Fanaroff and Martin's Neonatal-Perinatal Medicine: Diseases of the Fetus and Infant (2-Volume Set) [9 ed.] 0323065457, 9780323065450

Table of contents :
Cover......Page 1
Copyright......Page 5
Dedication......Page 6
Contributors......Page 8
Preface......Page 18
VOLUME ONE......Page 19
SECTION I......Page 24
Perinatal pioneers......Page 26
High-risk fetus and perinatal obstetrics......Page 27
Midwives and perinatal care......Page 28
Neonatal resuscitation: tales of heroism and desperation......Page 29
Saving infants to man the army......Page 30
Incubators, baby shows, and origins of neonatal intensive care units.......Page 31
Ventilatory care: “extended resuscitation”......Page 33
Tools and supplies for neonatal intensive care units......Page 35
Controlled clinical trials, evidence-based medicine, and research networks......Page 37
Some famous high-risk infants......Page 38
References......Page 39
Overview......Page 42
Infant mortality......Page 43
Preterm infants, infants with low birthweight, and infants who are small for gestational age......Page 44
Alarming increases in infants with low birthweight......Page 45
Legal and ethical issues......Page 46
Historical perspective......Page 47
Current recommendations on organization......Page 48
Regional perinatal center......Page 50
Problems......Page 52
Measurements of effectiveness of care organization......Page 53
References......Page 54
Medical ethics in neonatal care......Page 56
Beneficence......Page 57
Justice......Page 58
Communication with parents......Page 59
Family-centered neonatal intensive care......Page 60
Refusal of treatment during pregnancy......Page 61
To provide data on which decisions can be based......Page 62
Withholding and withdrawing life-sustaining medical treatment in the neonatal intensive care unit......Page 63
Collaborative, procedural framework for end-of-life decision making......Page 64
Palliative care in the neonatal intensive care unit......Page 65
Conflict resolution when consensus cannot be reached......Page 66
Ethics of research in the neonatal intensive care unit......Page 67
Ethical responsibilities of neonatal physicians......Page 68
References......Page 69
Disclaimer......Page 72
General structure of the federal and state court systems......Page 73
Residents and fellows......Page 74
Malpractice......Page 75
Prenatal consultation......Page 76
Role of the expert witness......Page 77
Causation......Page 78
Wrongful birth......Page 79
Strategies for avoiding tort litigation......Page 80
Live birth......Page 81
Handicapped newborns......Page 82
Baby jane doe......Page 83
Sun hudson......Page 84
Miller case......Page 85
Conclusions......Page 86
Summary......Page 87
References......Page 88
The case for improvement......Page 90
Data for improvement......Page 91
Identifiers......Page 92
Processes......Page 93
Risk adjusters......Page 94
Secondary data......Page 97
Costs and resources......Page 104
Role of networks......Page 105
Work of quality improvement......Page 106
Four key habits for improvement......Page 108
Patient safety......Page 109
References......Page 110
Teaching versus learning......Page 113
Simulation-based learning......Page 114
The evidence behind simulation......Page 116
The future of simulation in neonatal-perinatal medicine......Page 118
References......Page 119
Asking a focused clinical question......Page 122
Reviews......Page 123
Critically appraising evidence for its validity......Page 124
Applying the results to patient care......Page 125
Promoting evidence-based clinical practice......Page 126
References......Page 127
Alma-ata declaration and other global health initiatives......Page 130
Major causes of global neonatal mortality......Page 131
War conflicts and infant mortality......Page 133
Birth asphyxia......Page 135
Staffing......Page 136
Care of the mother and the newborn in the community......Page 137
Management of hypothermia......Page 138
Care of an infant not crying at birth......Page 139
Breast feeding......Page 140
Home visits by the health care worker......Page 141
Role of the traditional birth attendant and community health worker......Page 142
Social engineering......Page 143
Improvements in neonatal care practices......Page 144
Relevance of newer technologies in developing countries......Page 145
Use of nasal continuous positive airway pressure......Page 146
Establishment of tertiary care units in developing countries......Page 147
References......Page 148
SECTION II......Page 150
Advanced maternal age......Page 152
Aneuploidy......Page 153
Single-gene disorders......Page 154
Advanced paternal age......Page 155
Mitochondrial inheritance......Page 156
Epigenetics and uniparental disomy......Page 157
Teratogens......Page 158
Diagnostic imaging......Page 160
First-trimester screening......Page 161
Screening for hemoglobinopathies......Page 162
Diagnostic modalities......Page 163
Chorionic villus sampling......Page 164
Preimplantation genetic diagnosis......Page 165
Assisted reproductive technologies......Page 166
Genetic evaluation and counseling......Page 167
References......Page 168
Fetal imaging techniques......Page 170
Bioeffects and safety......Page 172
Ethical considerations......Page 173
Assisted reproduction......Page 174
Multiple gestation......Page 175
Pregnancy evaluation......Page 176
Placental location......Page 178
Cervical length and pelvic structures......Page 179
Doppler ultrasound......Page 180
Procedures......Page 181
Fetal ventriculomegaly......Page 182
Ventriculomegaly......Page 183
Dandy-walker cyst......Page 184
Head and neck......Page 185
Gastrointestinal tract......Page 186
Genitourinary tract......Page 187
Musculoskeletal system......Page 192
Summary......Page 194
References......Page 196
Physiologic basis for antenatal surveillance......Page 198
Contraction stress test......Page 199
Doppler flow velocimetry......Page 200
References......Page 202
Fetal heart rate variability......Page 204
Decelerations......Page 205
Category iii......Page 207
Evaluation and management of nonreassuring fetal heart rate patterns......Page 208
Summary......Page 209
References......Page 210
Hydronephrosis......Page 212
Fetal hydrothorax......Page 214
Miscellaneous procedures and other indications for shunts......Page 216
Fetal arrhythmias......Page 217
Diagnostic fetoscopy......Page 218
Diaphragmatic hernia (see chapter 44)......Page 219
Monochorionic twins discordant for anomaly......Page 221
Twin-twin transfusion syndrome......Page 222
Open fetal surgery......Page 224
Exit procedure......Page 228
Fetal anesthesia considerations......Page 229
References......Page 232
Exposures affecting the epigenome......Page 238
Paternal effects......Page 239
Lead......Page 241
Diet......Page 242
Heat......Page 244
Fetal distribution......Page 245
Fetal metabolism......Page 246
Spectrum of outcomes......Page 247
Conclusion......Page 248
References......Page 249
Additional reading......Page 251
Fetal origins of adult disease concept......Page 252
Postnatal and early childhood growth......Page 253
Mismatch concept......Page 254
Pathophysiology......Page 255
Cancer......Page 256
Psychosocial aspects......Page 257
Transgenerational persistence of phenotypic changes......Page 258
Epigenetics......Page 259
Translation to neonatology......Page 260
References......Page 261
SECTION III......Page 266
Fetal growth and body composition......Page 268
Fetal metabolism......Page 272
Epidemiology of low birthweight......Page 275
Maternal nutrition......Page 276
Chronic disease......Page 278
Drugs......Page 279
Placental determinants......Page 280
Fetal determinants......Page 282
Aberrant fetal growth patterns......Page 284
Diagnosis......Page 285
Antenatal management......Page 288
Approach to an infant who is small for gestational age......Page 289
Asphyxia......Page 291
Neonatal metabolism......Page 292
Temperature regulation......Page 293
Follow-up......Page 294
Growth......Page 295
References......Page 296
Superimposed preeclampsia......Page 300
Role of the renin-angiotensin system......Page 301
Pathophysiologic basis of clinical manifestations......Page 302
Uterine vasculature......Page 303
Predisposing factors......Page 304
Differential diagnosis......Page 305
Antihypertensives......Page 306
Prevention of preeclampsia......Page 307
Classification......Page 308
Other considerations......Page 309
Summary......Page 310
References......Page 311
Congenital malformations......Page 314
Fetal macrosomia......Page 315
Diabetic ketoacidosis......Page 316
Diagnosis......Page 317
Rationale and fundamental structure of intensified therapy......Page 318
Oral hypoglycemic and antihyperglycemic agents......Page 319
Antepartum fetal surveillance: what and how to test......Page 320
Timing and method of delivery......Page 321
References......Page 322
Prematurity......Page 326
Pathogenesis......Page 328
Demographics......Page 330
Cervical and uterine factors......Page 331
Infection......Page 332
Other risk factors......Page 334
Biochemical predictors......Page 335
Ultrasound predictors......Page 337
Prevention......Page 338
Bed rest......Page 339
Progesterone......Page 340
Cerclage......Page 341
Ritodrine......Page 342
Magnesium sulfate......Page 343
Indomethacin......Page 345
Calcium channel blockers......Page 346
Oxytocin antagonists......Page 347
Antibiotics......Page 348
Corticosteroids......Page 350
Summary......Page 351
References......Page 352
Immune thrombocytopenic purpura......Page 358
Management of a subsequent pregnancy......Page 360
Fetal-neonatal consequences of maternal antinuclear antibodies......Page 361
Fetal-neonatal consequences of maternal antiphospholipid antibodies......Page 362
Myasthenia gravis......Page 363
References......Page 364
Biology......Page 366
Fetal and neonatal consequences......Page 367
Malformations......Page 368
Twin-twin transfusion......Page 369
Mortality......Page 370
Outcome......Page 371
References......Page 372
Risks of post-term pregnancy......Page 374
Dysmaturity......Page 375
Assessment of fetal well-being......Page 376
How should labor be induced?......Page 377
References......Page 378
Historical background......Page 380
Sensitization......Page 381
Hydrops fetalis......Page 382
Rh immunoglobulin prophylaxis......Page 384
Screening......Page 385
Functional assays......Page 387
Rh isoimmunization......Page 388
Prediction of anemia......Page 389
Intravascular transfusion......Page 392
Intraperitoneal transfusion......Page 393
Short term......Page 394
Atypical antigens......Page 395
References......Page 396
Volume......Page 400
Production and regulation......Page 401
Amniotic fluid index measurement......Page 402
Comparing amniotic fluid assessment techniques......Page 403
Amniotic fluid assessment in multifetal pregnancies......Page 404
Increased perinatal morbidity and mortality......Page 405
Uteroplacental insufficiency......Page 406
Etiology......Page 407
Assess fetal growth......Page 408
Amnioinfusion......Page 409
Maternal hydration......Page 410
Etiology......Page 411
Assess fetal anatomy......Page 412
Nonimmune hydrops fetalis......Page 413
Etiology......Page 414
Laboratory testing......Page 416
Postnatal evaluation......Page 418
Fetal surgery and experimental treatment......Page 419
References......Page 420
Urinary tract infection......Page 422
Preterm premature rupture of membranes......Page 423
Diagnosis......Page 424
Bacterial vaginosis......Page 425
Diagnosis and treatment......Page 426
Varicella-zoster virus......Page 427
Prevention......Page 428
Epidemiology and clinical manifestations......Page 429
Epidemiology and clinical manifestations......Page 430
Fetal and neonatal infection......Page 431
Maternal-fetal transmission......Page 432
Hepatitis C......Page 433
Clinical features......Page 434
Group b streptococcal infection......Page 435
Prevention of perinatal disease......Page 436
Clinical manifestations, diagnosis, and treatment......Page 437
Maternal disease and transmission......Page 438
Effects on the newborn......Page 439
Transmission and neonatal disease......Page 440
References......Page 441
Overview......Page 446
Infection and inflammation......Page 449
Developmental and structural lesions......Page 450
Clinical correlation......Page 451
References......Page 452
SECTION IV......Page 454
Labor pain characteristics and challenges......Page 456
Hypnosis......Page 457
Systemic medications......Page 458
Pharmacokinetics, pharmacodynamics, and the fetus......Page 459
Opioids......Page 460
Opioid agonists......Page 461
Intrathecal injections (spinals)......Page 462
Maternal side effects......Page 463
Maternal temperature elevation......Page 464
Breast feeding......Page 465
Neuraxial anesthesia......Page 466
Ancillary medications......Page 467
Fetal distress......Page 468
References......Page 469
Fetus......Page 472
Transition at birth......Page 473
Causes of depression and asphyxia......Page 475
Response to asphyxia......Page 476
Anticipation......Page 477
Initial quick overview......Page 478
Thermal management......Page 479
References......Page 480
Role of Positive Pressure Ventilation in Neonatal Resuscitation......Page 481
When to initiate positive pressure ventilation......Page 482
Peak inflation pressures and tidal volumes......Page 483
Continuous distending pressure (positive end-expiratory pressure or continuous positive airway pressure)......Page 484
Pressure delivering device......Page 485
Facemasks......Page 486
Intubation procedure......Page 488
References......Page 489
Oxidative stress: pathophysiologic background......Page 491
Animal studies......Page 493
Clinical data......Page 494
Room air versus 100% oxygenor something else......Page 495
References......Page 496
Chest Compression, Medications, and Special Problems......Page 497
Chest compression......Page 498
Epinephrine......Page 499
Sodium bicarbonate......Page 501
Immediate care after establishing adequate ventilation and circulation......Page 502
Infants with very low birthweight and extremely premature infants......Page 503
Pneumothorax......Page 504
Physical examination......Page 505
References......Page 506
Order of the examination......Page 508
Eyes......Page 509
Femoral pulses......Page 510
Swollen eyelids......Page 511
Syndromes......Page 512
Extra digits......Page 513
Abnormalities of the genitalia......Page 514
Congenital heart disease......Page 515
Hearing and vision screening......Page 516
Inspection......Page 517
Cranial nerves......Page 519
Motor function......Page 520
The premature infant......Page 521
References......Page 522
Etiology......Page 524
Clinical manifestations......Page 525
Cephalhematoma......Page 526
Prognosis......Page 527
Differential diagnosis......Page 528
Differential diagnosis......Page 529
Clinical manifestations......Page 530
Fractures and dislocations of facial bones......Page 531
Subconjunctival hemorrhage......Page 532
Intraocular hemorrhage......Page 533
Treatment......Page 534
Mechanism of injury......Page 535
Etiology......Page 536
Treatment......Page 537
Prognosis......Page 538
Treatment......Page 539
Treatment......Page 540
Clinical manifestations......Page 541
Treatment......Page 542
Treatment......Page 543
Differential diagnosis......Page 544
Differential diagnosis......Page 545
Etiology......Page 546
Epiphyseal separations......Page 547
Management......Page 548
Scrotum and labia majora......Page 549
Injuries related to accidental and inflicted trauma incurred by the mother......Page 550
References......Page 551
General clinical approach......Page 554
Complex or multifactorial......Page 555
Environmental exposure and teratogens......Page 556
Minor anomalies and phenotypic variants......Page 558
Evaluation......Page 559
Skin......Page 560
Head......Page 561
Eyes......Page 562
Ears......Page 563
Mouth......Page 564
Anus......Page 566
Extremities......Page 567
General studies......Page 571
Genetic laboratory studies......Page 572
Educational resources and support organizations......Page 573
References......Page 574
SECTION V......Page 576
Water and heat exchange between the infant’s body surface and the environment......Page 578
Determination of water loss from the airway......Page 579
Heat exchange during the first hours after birth......Page 580
Transepidermal water loss during the first 4 postnatal weeks......Page 581
Heat exchange during the first weeks after birth......Page 582
Heat exchange between the infant’s skin and the environment during phototherapy......Page 583
Heat exchange between the infant’s skin and the environment during skin-to-skin care......Page 584
Water and heat exchange between the respiratory tract and environment......Page 586
Respiratory water and heat exchange during mechanical ventilation......Page 587
Neutral thermal environment......Page 588
The delivery room......Page 589
The cold infant......Page 590
References......Page 591
Touch and movement......Page 593
Summary......Page 594
Adjacencies......Page 595
Single-family rooms versus multiple-patient rooms......Page 596
The planning process......Page 597
References......Page 599
Biologic basis......Page 600
Leads and lead placement......Page 601
Electrocardiogram device and safety......Page 602
Blood pressure monitoring......Page 603
Indirect monitoring......Page 604
Surface and noninvasive monitoring......Page 605
Other monitoring techniques......Page 606
Capnography......Page 607
Pulse oximetry......Page 608
Principles......Page 609
Potential solutions......Page 610
Continuous blood gas monitoring......Page 611
Near-infrared spectroscopy......Page 612
Clinical applications of near-infrared spectroscopy......Page 613
Visible light spectroscopy......Page 614
References......Page 615
Transition phase and persistent pulmonary hypertension of the neonate......Page 620
Respiratory physiology: apnea, central control of ventilation, and respiratory distress syndrome......Page 621
Cardiac physiology and patent ductus arteriosus......Page 622
Oxygen therapy and retinopathy of prematurity and chronic lung disease......Page 623
Anesthetics and neurodevelopment......Page 624
Preoperative evaluation and preparation......Page 625
Transport......Page 626
Operating room equipment and monitoring......Page 630
Inhalational agents......Page 631
Intravenous agents......Page 632
Recovery from anesthesia......Page 634
References......Page 635
Provision for care......Page 638
Perception of the fetus as a separate individual......Page 639
Labor......Page 640
First hours after birth......Page 641
Sensitive period......Page 643
Interventions for premature or sick infants and their parents......Page 645
Kangaroo baby care......Page 646
Family-centered care in the neonatal intensive care unit......Page 647
Interventions for parents of malformed infants......Page 648
References......Page 649
Definition......Page 652
Epidemiology and trends......Page 653
Medical interventions and iatrogenic causes......Page 654
Gestational age assessment and obstetric practice guidelines......Page 655
Thermoregulation......Page 656
Role of fetal lung fluid clearance in neonatal transition......Page 657
Respiratory......Page 659
Temperature instability and hypoglycemia......Page 660
Hospitalizations and rehospitalizations after discharge......Page 661
Long-term outcomes and societal costs......Page 662
Discharge criteria......Page 663
References......Page 664
Incidence of postnatal growth failure......Page 666
Evidence supporting early nutritional support with parenteral amino acids......Page 667
Intravenous amino acid mixtures......Page 669
Intravenous carbohydrate requirements......Page 670
Intravenous lipid emulsions......Page 671
Energy requirements in the parenterally fed infant......Page 672
Complications of parenteral nutrition......Page 673
Protein requirements in the enterally fed infant......Page 674
Lipid requirements in the enterally fed infant......Page 676
Energy requirements in the enterally fed infant......Page 678
Minerals......Page 679
Vitamins......Page 680
Human milk and formula......Page 681
Preterm infants......Page 683
Initiation and advancement of enteral feeding......Page 686
Practical approach to administration of parenteral and enteral nutrition......Page 687
References......Page 688
Sodium balance in newborns......Page 692
Water balance in newborns......Page 693
Insensible losses......Page 694
Urinary losses......Page 695
Estimating pathogenic losses and deficit replacement......Page 696
Uncommon causes of hyponatremia in neonates......Page 697
Hypernatremia......Page 698
Chronic lung disease......Page 699
Long-term maintenance of acid-base balance......Page 700
Developmental aspects of acid-base physiology......Page 701
Diagnostic approach to disorders of acid-base balance......Page 702
Metabolic alkalosis......Page 703
Late metabolic acidosis of prematurity......Page 704
Neonatal bartter syndrome......Page 705
References......Page 706
Normal chest......Page 708
Respiratory distress syndrome......Page 709
Neonatal pneumonia......Page 710
Pulmonary hemorrhage......Page 711
Cystic adenomatoid malformation......Page 712
Esophageal atresia and tracheoesophageal fistula......Page 713
Heart......Page 714
Gastrointestinal tract......Page 715
Duodenal atresia......Page 716
Functional immaturity of the colon......Page 717
Necrotizing enterocolitis......Page 718
Biliary atresia......Page 720
Hepatic calcifications......Page 721
Urinary tract......Page 722
Multicystic dysplastic kidney......Page 723
Nephrocalcinosis......Page 724
Germinal matrix hemorrhage......Page 725
Periventricular leukomalacia......Page 726
Skeletal......Page 727
Future directions......Page 728
References......Page 729
Drug absorption......Page 732
Hepatic drug metabolism......Page 734
Placenta......Page 735
Drug transfer across the placenta......Page 736
Metabolic capability of the human placenta......Page 737
Drug disposition of the fetus......Page 738
Exposure of the fetus to drugs......Page 739
Effects of drugs on the fetus......Page 740
Genetic background......Page 741
Mechanisms of drug toxicity in the fetus......Page 744
Drug use, disposition, and metabolism in the newborn infant......Page 745
Absorption of drugs in the neonate......Page 746
Drug metabolism and disposition in the neonate......Page 747
Pharmacokinetic considerations, dosage guidelines, and therapeutic monitoring in neonatal drug therapy......Page 750
Passage of exogenous compounds from maternal blood to milk......Page 752
Antimicrobial drugs......Page 753
Drugs affecting the endocrine system......Page 754
Mercury......Page 755
References......Page 756
Clinician’s role in perinatal substance abuse......Page 758
Prevalence of fetal alcohol syndrome......Page 759
Neurobehavioral and developmental effects in infancy......Page 761
Recommendations for prevention and intervention......Page 763
Fetal effects......Page 764
Neuropsychological and behavioral effects in childhood and adolescence......Page 765
Recommendations for prevention and intervention......Page 766
Opioids......Page 767
Neonatal abstinence syndrome......Page 769
Central nervous system effects......Page 770
Pharmacologic effects of cocaine......Page 771
Intrauterine and postnatal growth......Page 772
Infant and toddler neurobehavior and mental and psychomotor development......Page 773
Child cognition, neurodevelopment, and behavior......Page 774
Amphetamines......Page 775
References......Page 776
VOLUME TWO......Page 20
SECTION VI......Page 782
Overview of hematopoiesis......Page 784
Kinetics of production and circulation......Page 785
Adhesion......Page 786
Microbicidal activity......Page 788
Production and differentiation......Page 789
Microbicidal activity......Page 790
Phenotypic and functional characteristics......Page 791
Overview of serum opsonins......Page 792
Ontogeny and analysis of the complement system in the neonatal period......Page 793
Structure and function......Page 794
Overview......Page 795
Tumor necrosis factor family......Page 796
Chemokines......Page 797
Coordinated inflammatory response in neonatal sepsis......Page 798
Overview......Page 800
Overview......Page 801
Role of cytokine receptor signaling in lymphocyte development......Page 802
Role of cytokines in t cell function......Page 803
T cell function in neonates......Page 804
Summary......Page 805
B cell development......Page 806
Role of cytokines in b cell function......Page 808
Immunoglobulin structure and function......Page 809
Antibody production in fetuses and neonates......Page 810
Immunologic properties of human breast milk......Page 811
Summary......Page 813
Evaluation of host defenses in neonates......Page 814
References......Page 815
Microbiology......Page 816
Maternal risk factors......Page 817
Other risk factors......Page 818
Differential diagnosis......Page 819
Leukocyte counts......Page 820
C-reactive protein......Page 821
Erythrocyte sedimentation rate......Page 822
Screening panels......Page 823
Empirical antimicrobial therapy......Page 824
Prevention......Page 828
Clinical manifestations......Page 829
Diagnosis......Page 830
Prognosis and outcome......Page 831
Clinical manifestations......Page 832
Etiology......Page 833
Pathogenesis......Page 834
Treatment......Page 835
Pathogenesis......Page 836
Etiology......Page 837
Diagnosis......Page 838
Etiology......Page 839
Treatment......Page 840
Mastitis (breast abscess)......Page 841
Diagnosis......Page 842
Treatment......Page 843
Diagnosis......Page 844
Diagnosis......Page 845
Follow-up......Page 847
Prevention......Page 848
Treatment......Page 849
Barrier nursing technique......Page 850
References......Page 851
Microbiology......Page 853
Clinical manifestations......Page 854
Treatment......Page 855
Incidence......Page 856
Clinical manifestations......Page 857
Treatment......Page 858
Clinical manifestations......Page 859
Diagnosis......Page 860
Clinical manifestations......Page 861
References......Page 862
Epidemiology and transmission......Page 864
Neonatal herpes simplex virus infection......Page 865
Diagnosis......Page 866
Therapy......Page 867
Cytomegalovirus......Page 868
Maternal clinical manifestations......Page 869
Prognosis......Page 870
Hearing and visual impairment......Page 871
Neonatal diagnosis......Page 872
Epidemiology and transmission......Page 873
Perinatal chickenpox......Page 874
Human herpesvirus 6 and 7......Page 875
Parainfluenza virus......Page 876
Effects on the fetus......Page 877
Transmission......Page 878
Subsequent respiratory morbidity......Page 879
Prophylaxis and therapy......Page 880
Epidemiology......Page 882
Pathogenesis......Page 883
Care of the mother and infant in developed countries......Page 884
Safety......Page 885
Care of the mother and infant in developing countries......Page 886
Clinical manifestations......Page 887
Epidemiology and transmission......Page 888
Laboratory diagnosis......Page 889
Hepatitis B virus......Page 890
Perinatal transmission......Page 891
Hepatitis C virus......Page 892
Hepatitis E virus......Page 893
Epidemiology and transmission......Page 894
Clinical manifestations......Page 895
Influenza a and b......Page 896
Novel h1n1 influenza a......Page 897
Rubella virus......Page 898
Congenital rubella syndrome......Page 899
Rhinovirus......Page 900
Severe acute respiratory syndrome......Page 901
Adenovirus......Page 902
Human papillomavirus......Page 903
References......Page 904
Neural induction and neurulation......Page 910
Neuronal proliferation......Page 911
Neuronal migration and cortical lamination......Page 913
Subplate neurons......Page 918
Axonal and dendritic growth......Page 919
Synaptogenesis......Page 922
Programmed cell death......Page 923
Astrocytes......Page 924
Oligodendrocytes and myelination......Page 925
Microglia and brain macrophages......Page 927
Craniorachischisis totalis, anencephaly, myeloschisis, encephalocele, myelomeningocele, and occult dysraphic states......Page 928
Holoprosencephaly and agenesis of the corpus callosum......Page 930
Microcephaly......Page 931
Micrencephaly vera......Page 932
Classic lissencephalies (type i lissencephalies)......Page 933
Tuba3 mutations......Page 934
Cobblestone lissencephalies (type ii lissencephalies)......Page 935
Disorders of central nervous system organization and maturation......Page 936
White Matter Damage and Encephalopathy of Prematurity......Page 937
Microscopic neuropathology......Page 941
Microglial activity......Page 942
Models of encephalopathy of prematurity: implications for pathogenesis......Page 943
Hyoperfusion and hypoxia-ischemia......Page 944
Infection and inflammation......Page 945
Excitotoxicity and oxidative stress......Page 946
Neonatal sonography......Page 947
Diffusion magnetic resonance imaging......Page 949
Magnetic resonance spectroscopy......Page 950
Impairment of brain growth and long-term development......Page 952
Intracranial Hemorrhage and Vascular Lesions......Page 956
Neuropathology......Page 959
Parenchymal hemorrhage......Page 960
Pathogenesis......Page 961
Cardiovascular factors......Page 962
Diagnosis......Page 963
Management......Page 964
Neurodevelopmental outcome......Page 965
Other hemorrhages......Page 966
Cerebral artery infarction......Page 967
Sinovenous thrombosis......Page 968
Diagnosis......Page 969
Risk factors......Page 970
References......Page 971
I. pathophysiology......Page 975
Vascular territories......Page 976
Types of hypoxic-ischemic insult......Page 977
Systemic adaptation to hypoxic-ischemic insult......Page 978
Secondary neuronal death......Page 979
Free radical formation......Page 980
Ii. assessment tools......Page 981
Cerebral edema......Page 990
Other neuroprotective strategies......Page 991
Hypoxic-ischemic encephalopathy......Page 993
Hypoxic-ischemic encephalopathy and less severe neurodisability......Page 994
Section iiassessment tools......Page 995
Seizures in Neonates......Page 997
Clinical seizure criteria......Page 999
Clonic seizures......Page 1000
Nonepileptic behaviors of neonates......Page 1002
Neonatal myoclonus without electrographic seizures......Page 1003
Seizure duration and topography......Page 1005
Incidence of neonatal seizures......Page 1006
Interictal electroencephalographic pattern abnormalities......Page 1007
Asphyxia-related events......Page 1008
Infection......Page 1010
Drug withdrawal and intoxication......Page 1011
Benign familial neonatal seizures......Page 1014
Treatment......Page 1015
Consequences of neonatal seizures......Page 1016
Prognosis......Page 1017
References......Page 1018
Neonatal neuromuscular disorders......Page 1020
Spinal muscular atrophies......Page 1021
Hereditary motor and sensory neuropathies......Page 1022
Neuromuscular junction disorders......Page 1023
Congenital muscular dystrophy......Page 1025
Congenital myotonic dystrophy......Page 1027
Nemaline myopathy......Page 1028
Myotubular myopathy......Page 1029
Debrancher enzyme deficiency (type iii glycogen storage disease)......Page 1030
References......Page 1031
The small head......Page 1033
Neurulation and cleavage anomalies......Page 1034
Migrational anomalies......Page 1035
Radial microbrain and micrencephaly vera......Page 1037
Evaluation and treatment of the small head......Page 1038
Macrencephaly and growth disorders......Page 1039
Degenerative disorders......Page 1040
Background......Page 1041
Posthemorrhagic hydrocephalus......Page 1042
External hydrocephalus......Page 1045
Intracranial cysts......Page 1046
Trauma......Page 1047
Craniosynostosis......Page 1048
Sagittal craniosynostosis......Page 1049
Unilateral coronal craniosynostosis......Page 1050
Metopic craniosynostosis......Page 1051
Crouzon syndrome......Page 1052
Apert syndrome......Page 1053
Saethre-chotzen syndrome......Page 1054
Treatment......Page 1055
References......Page 1056
Treatment......Page 1058
References......Page 1059
Follow-up for high-risk neonates......Page 1060
Physical growth......Page 1062
Major neurologic sequelae......Page 1063
Assessment of functional outcomes......Page 1064
Timing of follow-up visits......Page 1065
School-age outcome......Page 1066
Young adult outcomes......Page 1067
References......Page 1068
Normal hearing and hearing loss......Page 1072
Early intervention services......Page 1073
Risk factors......Page 1074
Hearing aids......Page 1076
Stress and impact on the family......Page 1077
References......Page 1078
A neurodevelopmental framework......Page 1080
Brain-environment interaction......Page 1081
A model for observing the preterm infant’s behavior......Page 1083
Behavioral language of the preterm infant......Page 1084
Testing the validity of nidcap......Page 1087
Direct assessment of preterm infants’ behavior......Page 1090
Overview of preterm newborn neurobehavioral assessments......Page 1091
The assessment of preterm infants’ behavior......Page 1092
Summary......Page 1094
References......Page 1095
Embryonic period......Page 1098
Saccular and alveolar stages......Page 1099
Pulmonary hypoplasia......Page 1102
Fetal lung fluid......Page 1103
Composition......Page 1104
Synthesis and secretion......Page 1105
Surfactant pool sizes......Page 1106
Pressure-volume curves......Page 1109
Surfactant appearance with development......Page 1111
Induced lung maturation......Page 1112
References......Page 1114
Respiratory rate......Page 1115
Cyanosis......Page 1116
Partial pressure of arterial carbon dioxide......Page 1117
Carbon dioxide......Page 1118
Airflow......Page 1119
Lung volume......Page 1120
Compliance......Page 1121
Resistance......Page 1122
Time constant......Page 1123
Limitations......Page 1125
Clinical applications......Page 1126
References......Page 1127
Pharmacologic acceleration of pulmonary maturation......Page 1129
Pathophysiology......Page 1130
Radiographic findings......Page 1131
Surfactant therapy......Page 1133
Inhaled nitric oxide therapy......Page 1135
Noninvasive nasal ventilation......Page 1155
Clinical associations......Page 1206
Resolution and consequences of apnea of prematurity......Page 1210
Pulmonary agenesis/hypoplasia......Page 1226
References......Page 1227
Cardiac Embryology......Page 1230
The tubular and looping heart......Page 1231
Endocardial development: formation of cushion tissue......Page 1234
Neural crest contribution to cardiogenesis......Page 1235
Epicardium......Page 1236
Therapy......Page 1237
References......Page 1238
Structural development......Page 1239
Pulmonary vascular transition......Page 1240
Persistent pulmonary hypertension of the newborn (see also part 8)......Page 1241
Congenital diaphragmatic hernia......Page 1242
Bronchopulmonary dysplasia......Page 1243
References......Page 1244
Chromosomal defects......Page 1245
Single gene defects......Page 1246
Major associated noncardiac defects......Page 1247
References......Page 1248
Fetal circulation......Page 1249
Indications for fetal cardiovascular assessment......Page 1250
Methods......Page 1251
Congenital heart defects......Page 1252
Fetal congestive heart failure and myocardial disease......Page 1253
Outcome......Page 1254
References......Page 1255
Pressure, flow, and resistance......Page 1256
Intracardiac shunting......Page 1257
Stroke volume and cardiac output......Page 1258
Diastolic function......Page 1259
Physical examination......Page 1260
Radiography......Page 1262
Echocardiography......Page 1263
Computed tomography and magnetic resonance imaging......Page 1264
Cyanosis......Page 1266
References......Page 1267
Clinical presentation......Page 1268
History......Page 1269
Laboratory evaluation......Page 1270
Laboratory evaluation......Page 1271
Management and prognosis......Page 1272
Anatomy and pathophysiology......Page 1273
Management and prognosis......Page 1274
Management and prognosis......Page 1275
Anatomy and pathophysiology......Page 1276
Clinical presentation......Page 1277
Anatomy and pathophysiology......Page 1278
Management and prognosis......Page 1279
Management and prognosis......Page 1280
Laboratory evaluation......Page 1282
Anatomy and pathophysiology......Page 1283
Associated defects......Page 1284
Management and prognosis......Page 1285
References......Page 1286
Cardiac malposition and abnormalities of abdominal situs......Page 1289
Diagnosis......Page 1290
Hypertension......Page 1291
Etiology......Page 1292
Treatment......Page 1293
Cardiac tumors......Page 1295
Neonatal marfan syndrome......Page 1297
References......Page 1298
Electrophysiologic testing......Page 1300
General mechanisms......Page 1301
Accessory pathway reentrant tachycardia......Page 1302
Atrioventricular node reentrant tachycardia......Page 1303
Junctional ectopic tachycardia......Page 1304
Ventricular tachycardia......Page 1305
Symptom-free or mildly affected infants......Page 1306
Rapid classification of abnormal tachycardia......Page 1307
Atrial flutter......Page 1308
Accessory pathway reentrant tachycardia......Page 1310
Treatment of bradyarrhythmias......Page 1311
General principles......Page 1312
Prostaglandin e1......Page 1313
Afterload......Page 1314
Prevention of bacterial endocarditis......Page 1315
Surgery......Page 1316
Neonatal heart transplantation......Page 1317
Balloon atrial septostomy......Page 1318
Pulmonary valve stenosis......Page 1319
Native coarctation of the aorta......Page 1320
Vessel occlusion......Page 1321
Fetal interventional catheterization......Page 1322
References......Page 1323
Anatomic and functional shifts in hematopoiesis......Page 1326
Hemoglobins and oxygen-carrying capacity......Page 1327
Red blood cell indices during prenatal and postnatal development......Page 1330
Definition......Page 1331
Evaluation......Page 1333
Anemia caused by blood loss......Page 1334
Twin-twin transfusion......Page 1335
Alloimmune hemolytic anemia......Page 1336
Minor blood group hemolytic diseases......Page 1337
Erythrocyte structural defects......Page 1338
Membrane lipid defects......Page 1339
Vitamin e deficiency......Page 1340
The thalassemias......Page 1341
Hemoglobin variants......Page 1342
Hemoglobin s and sickle cell anemia......Page 1343
Iron-deficiency anemia......Page 1344
Anemia of prematurity......Page 1345
Diamond-blackfan anemia......Page 1346
Methemoglobinemia......Page 1347
Polycythemia......Page 1348
Neutropenia......Page 1349
Autoimmune neutropenia of infancy......Page 1350
Shwachman-diamond syndrome......Page 1351
Reticular dysgenesis......Page 1353
Leukocyte adhesion deficiency......Page 1354
Chédiak-higashi syndrome......Page 1355
Chronic granulomatous disease......Page 1356
Platelets, endothelium, and von willebrand factor......Page 1357
Coagulation factors......Page 1358
Physiologic anticoagulant strategies and proteins......Page 1359
Physiologic alterations of coagulation and fibrinolysis in the neonate......Page 1360
Defects in the hemostatic system......Page 1361
Laboratory testing of the neonatal hemostatic system......Page 1363
Hemophilias......Page 1365
Factor vii deficiency......Page 1366
Hepatic disease......Page 1367
Extracorporeal membrane oxygenation......Page 1368
Neonatal thrombosis......Page 1369
Elevated factor viii levels......Page 1370
Low-molecular-weight heparin......Page 1371
Oral anticoagulation with warfarin (coumadin)......Page 1372
Fibrinolytic therapy......Page 1373
Platelet function......Page 1374
Neonatal alloimmune thrombocytopenia......Page 1375
Other conditions associated with increased platelet destruction......Page 1376
Thrombocytopenia with absent radii syndrome......Page 1377
Alport syndrome (hereditary nephritis) and epstein syndrome......Page 1378
Infant leukemia......Page 1379
Teratomas and other germ cell tumors......Page 1380
References......Page 1381
Risks of transfusion therapy......Page 1383
Hypocalcemia......Page 1384
Cytomegalovirus......Page 1385
Hepatitis b virus......Page 1386
Febrile nonhemolytic transfusion reactions......Page 1387
Autologous red blood cell transfusions......Page 1388
Indications......Page 1389
Red blood cell preparations......Page 1390
Granulocyte transfusion......Page 1391
Fresh-frozen plasma transfusion......Page 1392
Choice of blood components......Page 1393
References......Page 1394
Esophagus......Page 1398
Stomach......Page 1399
Duodenum, pancreas, and biliary system......Page 1401
Small intestine......Page 1402
References......Page 1403
Evaluation of the infant with chronic diarrhea......Page 1404
Management of intestinal failure......Page 1405
Prenatal diagnosis......Page 1406
Johanson-blizzard syndrome......Page 1408
Congenital sucrase-isomaltase deficiency......Page 1409
Maltase-glucoamylase deficiency......Page 1410
Laboratory investigations......Page 1411
Other causes of protein-losing enteropathy......Page 1412
Molecular pathophysiology......Page 1413
Clinical features......Page 1414
Clinical features......Page 1415
Clinical features......Page 1416
Allergic enteropathy......Page 1417
Short bowel syndrome......Page 1418
References......Page 1419
Esophageal atresia and tracheoesophageal fistula......Page 1423
Esophageal duplications......Page 1428
Congenital diaphragmatic hernia......Page 1429
Gastroesophageal reflux......Page 1430
Abdominal wall defects......Page 1431
Gastric volvulus......Page 1435
Gastric perforation......Page 1436
Pyloric atresia......Page 1437
Hypertrophic pyloric stenosis......Page 1438
Duodenal atresia and stenosis......Page 1440
Gastrointestinal duplications......Page 1441
Malrotation and midgut volvulus......Page 1442
Meconium ileus......Page 1444
Meconium plug syndrome......Page 1445
Jejunoileal atresia and stenosis......Page 1446
Hirschsprung disease......Page 1447
Anorectal anomalies......Page 1448
Colostomy......Page 1450
Long-term follow-up......Page 1451
References......Page 1452
Epidemiology......Page 1454
Outcome......Page 1455
Pathophysiology......Page 1456
Enteral feeding......Page 1457
Bacterial colonization......Page 1458
Host defense......Page 1459
Summary......Page 1461
References......Page 1462
Bilirubin biochemistry......Page 1466
Bilirubin production......Page 1468
Hepatic uptake of bilirubin......Page 1469
Excretion of bilirubin......Page 1470
Enterohepatic absorption of bilirubin......Page 1471
The ugt gene......Page 1472
Fetal bilirubin......Page 1473
Term neonate......Page 1474
Preterm neonate......Page 1475
Pathologic unconjugated hyperbilirubinemia......Page 1476
Disorders of production......Page 1477
Abo heterospecificity......Page 1478
Function of g6pd.......Page 1479
Erythrocyte structural defects......Page 1480
Sequestration......Page 1481
Crigler-najjar syndrome type i......Page 1482
Breast-feeding failure jaundice......Page 1483
Sequelae of unconjugated hyperbilirubinemia......Page 1484
Kernicterus......Page 1485
Total bilirubin (tb) measurements......Page 1488
Therapy for unconjugated hyperbilirubinemia......Page 1491
Mechanism of action......Page 1493
Technique......Page 1494
Complications......Page 1497
Phenobarbital......Page 1499
Miscellaneous agents......Page 1500
Complications......Page 1501
Prediction of hyperbilirubinemia and postdischarge follow-up......Page 1503
Conjugated hyperbilirubinemia......Page 1504
Causes of conjugated hyperbilirubinemia......Page 1506
Treatment of neonatal hepatitis......Page 1509
Treatment of extrahepatic biliary atresia......Page 1510
Known infectious causes......Page 1511
Cystic diseases......Page 1512
References......Page 1513
Other fetal substrates......Page 1520
Glucose metabolism after birth......Page 1521
Hypoglycemia......Page 1523
Intrapartum glucose administration......Page 1524
Diabetes in pregnancy: the infant of a diabetic mother......Page 1525
Hypoglycemia......Page 1527
Management of the diabetic mother......Page 1528
Prognosis......Page 1529
Intrauterine growth restriction and infants who are small for gestational age......Page 1530
Hepatic glucose-6-phosphatase and prematurity......Page 1531
Iatrogenic causes......Page 1532
Nesidioblastosis-adenoma spectrum......Page 1533
Beckwith-wiedemann syndrome......Page 1534
Clinical and laboratory manifestations......Page 1535
Fructose intolerance......Page 1536
Neurohypoglycemia (hypoglycorrhachia) caused by defective glucose transport......Page 1537
Intravenous glucose infusions......Page 1538
Somatostatin......Page 1539
Prognosis of neonatal hypoglycemia......Page 1540
Hyperglycemia......Page 1541
Diabetes mellitus in the newborn......Page 1542
Insulin therapy in the baby with low birthweight......Page 1543
Disorders of Calcium, Phosphorus, and Magnesium Metabolism......Page 1544
Serum calcium......Page 1546
Placental transport......Page 1547
Intestinal absorption......Page 1548
Serum phosphorus......Page 1550
Renal excretion......Page 1551
Intestinal absorption......Page 1553
Regulation of secretion......Page 1554
Synthesis and metabolism......Page 1555
Neonatal vitamin d: function and recommendations......Page 1556
Synthesis and metabolism......Page 1557
Ontogenesis......Page 1558
Evaluation of fetal mineral accretion......Page 1559
Infants with very low birth weight......Page 1561
Term infants......Page 1562
Maternal diabetes......Page 1563
Hypomagnesemia......Page 1564
Infantile osteopetrosis......Page 1565
Symptomatic hypocalcemia......Page 1566
Iatrogenic hypercalcemia......Page 1567
Primary hyperparathyroidism......Page 1568
Other causes of hypercalcemia......Page 1569
Treatment......Page 1570
Maternal diabetes......Page 1571
Other causes of hypomagnesemia......Page 1572
Osteopenia of prematurity......Page 1573
Etiology......Page 1574
Enteral nutrition......Page 1575
References......Page 1576
Thyroid Disorders......Page 1579
Cellular metabolism......Page 1580
Synthesis, release, transport, and use of thyroid hormones......Page 1581
Monodeiodination of thyroxine......Page 1582
Thyroxine......Page 1583
Free hormones......Page 1584
Thyroid autoantibodies......Page 1585
Thyroid-stimulating hormone surge test......Page 1586
Thyroid function: fetal-maternal relationship......Page 1587
Congenital hypothyroidism......Page 1589
Defective embryogenesis of the thyroid......Page 1591
Iodine deficiency......Page 1592
Neonatal screening for hypothyroidism......Page 1593
Clinical manifestations......Page 1597
Laboratory manifestations......Page 1598
Transient primary neonatal hypothyroidism......Page 1599
Transient hypothyroxinemia......Page 1600
Treatment......Page 1601
Etiology and pathogenesis......Page 1603
Prognosis......Page 1604
Increased transthyretin......Page 1605
References......Page 1606
Y chromosome and role of sry gene......Page 1607
Embryology and endocrinology......Page 1608
Development of the external genitalia......Page 1609
Fetal gonadal endocrine function......Page 1611
Control of genital differentiation: roles of antimÜllerian hormone and testosterone......Page 1612
Penis size......Page 1613
Clitoris size......Page 1614
Associated dysmorphology......Page 1615
Discussion with family and professional staff......Page 1616
Biochemistry......Page 1618
Nomenclature of disorders of sex development......Page 1619
45,x (turner syndrome and variants)......Page 1621
Ovotesticular disorder of sex development......Page 1623
Complete and partial gonadal dysgenesis......Page 1624
Mutation of sox9 (sry-like hmg box-related gene 9)......Page 1625
17,20-desmolase (17,20-lyase) deficiency......Page 1626
Leydig cell hypoplasia......Page 1627
Complete androgen insensitivity syndrome......Page 1628
Disorders of antimüllerian hormone and antimüllerian hormone receptor (persistent müllerian duct syndrome)......Page 1629
Congenital adrenal hyperplasia......Page 1630
Simple virilizing form.......Page 1632
Diagnosis.......Page 1633
Treatment.......Page 1634
3β-hydroxysteroid dehydrogenase deficiency......Page 1635
P-450 oxidoreductase deficiency......Page 1637
Maternally derived androgenic substances......Page 1638
Müllerian and vaginal dysgenesis......Page 1639
Cryptorchidism......Page 1640
Clitoromegaly......Page 1641
References......Page 1642
Misconception 1......Page 1644
Misconception 8......Page 1645
Principles of screening programs......Page 1646
Screening for disorders......Page 1647
Biotinidase deficiency......Page 1658
Homocystinuria......Page 1659
Medium-chain acyl-coa dehydrogenase deficiency......Page 1660
Phenylketonuria......Page 1661
Fetal diseases affecting the mother......Page 1662
The sick newborn infant......Page 1663
Pyridoxine-dependent seizures......Page 1664
Glycogen storage disorders......Page 1665
Fatty acid β-oxidation and mitochondrial respiratory chain defects......Page 1666
Hair and skin abnormalities......Page 1667
Tyrosinemia type i......Page 1668
Carbohydrate metabolism disorders......Page 1669
Hepatomegaly and splenomegaly......Page 1670
Sepsis......Page 1671
Congenital disorders of glycosylation......Page 1672
Cholesterol biosynthesis......Page 1673
Mevalonate kinase deficiency......Page 1674
Single-enzyme defects......Page 1675
Blood studies......Page 1676
Urine studies......Page 1677
Amino acid analysis......Page 1678
Lactate and pyruvate analysis......Page 1679
Tests for peroxisomal disorders......Page 1680
Postmortem evaluation......Page 1681
Differential diagnosis......Page 1682
Defects of ketolysis......Page 1683
Defects in branched-chain amino acid metabolism......Page 1684
Management......Page 1685
Differential diagnosis......Page 1686
Krebs cycle defects......Page 1688
Defects of the respiratory chain......Page 1689
Hypoglycemia......Page 1692
Differential diagnosis......Page 1693
Gluconeogenesis defects......Page 1694
Fatty acid oxidation disorders......Page 1695
Differential diagnosis......Page 1696
Urea cycle defects......Page 1697
Treatment......Page 1698
Treatment at the protein level......Page 1699
References......Page 1700
Development of the kidney......Page 1704
Development of the bladder and urethra......Page 1705
Physiology of the developing kidney......Page 1706
History......Page 1707
Physical examination......Page 1709
Urinalysis......Page 1710
Hematuria......Page 1711
Intrinsic (renal) acute kidney injury......Page 1712
Medical management......Page 1713
Definition and incidence......Page 1714
Causes......Page 1715
Treatment......Page 1716
Nephrocalcinosis......Page 1718
Congenital and inherited disorders of the kidney and urinary tract......Page 1719
Prenatal management......Page 1720
Posterior urethral valves......Page 1721
Exstrophy-epispadias complex......Page 1722
Autosomal recessive polycystic kidney disease......Page 1723
Tumors of the kidney......Page 1724
References......Page 1725
Development of the epidermis......Page 1728
Development of the appendages......Page 1730
Blood and lymphatic vessel development......Page 1731
Transient cutaneous lesion......Page 1732
Pigmentary lesions53......Page 1733
Transient neonatal pustular melanosis......Page 1734
Infantile acropustulosis......Page 1735
Collodion baby......Page 1736
Management of ichthyosis......Page 1738
Staphylococcal scalded skin syndrome......Page 1739
Viral lesions......Page 1740
Epidermolysis bullosa......Page 1741
Incontinentia pigmenti......Page 1743
Zinc deficiency......Page 1744
Congenital melanocytic nevi......Page 1745
Congenital giant melanocytic nevi......Page 1746
Partial albinism......Page 1747
Hemangiomas......Page 1748
Sturge-weber syndrome......Page 1750
Epidermal nevi......Page 1751
Atopic dermatitis......Page 1752
Langerhans cell histiocytosis......Page 1753
Ehlers-danlos syndrome......Page 1754
Porphyrias......Page 1755
Principles of newborn skin care......Page 1756
References......Page 1758
Neonatal eye examination......Page 1760
Normal ocular findings......Page 1764
Orbital abnormalities......Page 1765
Ocular hypotelorism or hypertelorism......Page 1766
Blepharoptosis......Page 1767
Watery eye......Page 1768
Cryptophthalmia......Page 1769
Cloudy cornea......Page 1770
Aniridia......Page 1771
Abnormal red reflex......Page 1772
Cataract......Page 1773
Persistent hyperplastic primary vitreous......Page 1774
Abnormal macula......Page 1775
Morning glory disc anomaly......Page 1776
Strabismus......Page 1777
Amblyopia......Page 1778
Craniosynostosis syndromes......Page 1779
Sturge-weber syndrome......Page 1780
Ocular trauma......Page 1781
Ocular tumors in infants......Page 1782
Rhabdomyosarcoma......Page 1783
References......Page 1784
Pathogenesis......Page 1787
Therapy......Page 1789
Recommendations for examination schedule......Page 1790
References......Page 1791
Spine......Page 1794
Multifactorial conditions......Page 1795
Spinal defects......Page 1796
Brachial plexus injury......Page 1797
Fractures......Page 1798
Bone and joint infections......Page 1800
Imaging......Page 1801
Diagnosis......Page 1803
Treatment......Page 1804
Radial hypoplasia and clubhand......Page 1805
Constriction bands......Page 1806
Polydactyly......Page 1807
Congenital scoliosis......Page 1808
Lower extremities......Page 1809
Anterolateral angulation......Page 1810
Diagnosis......Page 1811
Treatment......Page 1813
Hyperextension, subluxation, and dislocation of the knee......Page 1814
Diagnosis......Page 1815
Diagnosis......Page 1816
Congenital vertical talus......Page 1817
Syndromes......Page 1818
Osteogenesis imperfecta......Page 1819
Fibrodysplasia ossificans progressiva......Page 1820
References......Page 1822
Appendix......Page 1825
Therapeutic agents......Page 1826
Atomic weight and valence......Page 1836
Preterm infants with birthweight of 2000 grams or more......Page 1862
Preterm infants with birthweights less than 2000 grams......Page 1863
Z......Page 1864

Citation preview

3251 Riverport Lane St. Louis, Missouri 63043

FANAROFF AND MARTIN’S NEONATAL-PERINATAL MEDICINE ISBN: 978-0-323-06545-0 Copyright © 2011, 2006, 2002, 1997, 1992, 1987, 1983, 1977, 1973, by Mosby, Inc., an affiliate of Elsevier, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data Fanaroff and Martin’s neonatal-perinatal medicine : diseases of the fetus and infant / [edited by] Richard J. Martin, Avroy A. Fanaroff, Michele C. Walsh. — 9th ed. p. ; cm. Other title: Neonatal-perinatal medicine Includes bibliographical references and index. ISBN 978-0-323-06545-0 1. Newborn infants—Diseases. 2. Fetus—Diseases. I. Martin, Richard J. II. Fanaroff, Avroy A. III. Walsh, Michele C. IV. Title: Neonatal-perinatal medicine. [DNLM: 1. Fetal Diseases. 2. Infant, Newborn, Diseases. 3. Perinatal Care. 4. Pregnancy Complications. WS 420 F1985 2010] RJ254.N456 2010 618.92’01--dc22 2010009656

Acquisitions Editor: Judy Fletcher Developmental Editor: Arlene Chappelle Publishing Services Manager: Julie Eddy Senior Project Manager: Celeste Clingan Design Direction: Lou Forgione

Printed in United States Last digit is the print number:  9  8  7  6  5  4  3  2  1

To our spouses: Patricia, Roslyn, and Larry the Martin children and grandchild Scott & Molly; Sonya, Peter, & Mateo, the Fanaroff children and grandchildren Jonathan & Kristy; Jodi, Peter, Austin, & Morgan; and Amanda, Jason, & Jackson, and the Walsh children Sean and Ryan with love, admiration, and deep appreciation for their continued support and inspiration

Contributors Jalal M. Abu-Shaweesh, MBBS

Sundeep Arora, MD

Associate Professor of Pediatrics, Case Western Reserve University; Attending Neonatologist, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Attending Pediatric Gastroenterologist, Children’s Hospitals and Clinics of Minnesota; Minnesota Gastroenterology, St. Paul, Minnesota

Respiratory Disorders in Preterm and Term Infants

Disorders of Digestion

Veronica H. Accornero, PhD Assistant Professor of Clinical Pediatrics, University of Miami Miller School of Medicine, Miami, Florida

Infants of Substance-Abusing Mothers

Heidelise Als, PhD Associate Professor of Psychiatry (Psychology), Harvard Medical School; Director of Neurobehavioral Infant and Child Studies, and Senior Associate in Psychiatry, Children’s Hospital; Research Associate in Newborn Medicine, Brigham & Women’s Hospital; Consulting Staff in Pediatric Psychology, Spaulding Rehabilitation Hospital, Boston, Massachusetts

Neurobehavioral Development of the Preterm Infant

Brenna L. Anderson, MD, MSCR Assistant Professor of Obstetrics and Gynecology, Division of Maternal Fetal Medicine, Warren Alpert Medical School of Brown University, Providence, Rhode Island; Medical Staff, Women & Infant’s Hospital, Pittsburgh, Pennsylvania; Consulting Physician, South County Hospital, South Kingstown, Rhode Island; Consulting Physician, Charlton Hospital, Fall River, Massachusetts

Perinatal Infections

Jacob V. Aranda, MD, PhD, FRCP(C) Professor and Director, Neonatology, State University of New York Downstate Medical Center, Brooklyn, New York

Developmental Pharmacology

James E. Arnold, MS, FAAP The Julius W. McCall Professor and Chair, Department of Otolaryngology, Head and Neck Surgery, Case Western Reserve University; Chairman, Department of Otolaryngology, University Hospitals Case Medical Center, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Upper Airway Lesions

Komal Bajaj, MD Reproductive Genetics Fellow, MonteFiore Medical Center, Albert Einstein College of Medicine, Bronx, New York; Research Associate, Human Genetics Laboratory at Jacobi Medical Center, Bronx, New York

Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

Jill E. Baley, MD Professor of Pediatrics, Case Western Reserve University School of Medicine; Medical Director, Neonatal Transitional Unit, Division of Neonatology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Perinatal Viral Infections; Schedule for Immunization of Preterm Infants

Eduardo H. Bancalari, MD Professor of Pediatrics, University of Miami, Miller School of Medicine; Director, Division of Neonatology, University of Miami/Jackson Memorial Hospital, Miami, Florida

Bronchopulmonary Dysplasia

Emmalee S. Bandstra, MD Professor of Pediatrics and Obstetrics and Gynecology, University of Miami Miller School of Medicine; Attending Neonatologist, Jackson Memorial Medical Center, Holtz Children’s Hospital, Miami, Florida

Infants of Substance-Abusing Mothers

Edward M. Barksdale, Jr., MD Professor of Surgery, Case Western Reserve University; Chief, Division of Pediatric Surgery, Rainbow Babies and Children’s Hospital, Cleveland, Ohio Selected Gastrointestinal Anomalies

Cynthia F. Bearer, MD, PhD Mary Gray Cobey Professor of Neonatology, University of Maryland School of Medicine; Chief, Division of Neonatology, University of Maryland Medical System, Baltimore, Maryland

Occupational and Environmental Risks to the Fetus

vii

viii

Contributors

Isaac Blickstein, MD Professor, Hadassah-Hebrew University School of Medicine, Jerusalem, Israel; Senior Physician, Department of Obstetrics and Gynecology, Kaplan Medical Center, Rehovot, Israel

Fetal Effects of Autoimmune Disease; Obstetric Management of Multiple Gestation and Birth; Post-term Pregnancy

Jeffrey L. Blumer, MD, PhD Professor of Pediatrics and Pharmacology, Case School of Medicine; Department of Pediatrics, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Developmental Pharmacology

Samantha Butler, PhD Instructor, Harvard Medical School; Staff Psychologist, Children’s Hospital Boston, Boston, Massachusetts

Neurobehavioral Development of the Preterm Infant

Kara Calkins, MD Clinical Instructor, David Geffen School of Medicine at UCLA, University of California; Attending Physician, Mattel Children’s Hospital at UCLA, Department of Pediatrics, Division of Neonatology and Developmental Biology, Los Angeles, California

Developmental Origins of Adult Health and Disease

Michael S. Caplan, MD Clinical Professor of Pediatrics, University of Chicago, Pritzker School of Medicine, Chicago, Illinois; Chairman, Department of Pediatrics, North Shore University Health System, Evanston, Illinois

Neonatal Necrotizing Enterocolitis: Clinical Observations, Pathophysiology, and Prevention

Waldemar A. Carlo, MD Edwin M. Dixon Professor of Pediatrics; Director, Division of Neonatology, University of Alabama at Birmingham School of Medicine; Director, Newborn Nurseries, University of Alabama at Birmingham Medical Center, The Children’s Hospital of Alabama, Birmingham, Alabama

Assessment of Pulmonary Function

Gisela Chelimsky, MD Assistant Professor of Pediatrics, Case School of Medicine; Attending Physician, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Disorders of Digestion

Valerie Y. Chock, MD Instructor of Neonatology, Stanford University School of Medicine, Stanford, California

Biomedical Engineering Aspects of Neonatal Monitoring

Walter J. Chwals, MD Professor of Surgery and Pediatrics, Tufts University School of Medicine, Boston, Massachusetts

Development and Basic Physiology of the Neonatal Gastrointestinal Tract; Selected Gastrointestinal Anomalies

Alan R. Cohen, MD, FACS, FAAP Reinberger Chair in Pediatric Neurological Surgery; Professor of Neurological Surgery and Pediatrics, Case Western Reserve University School of Medicine; Surgeon-in-Chief and Chief of Pediatric Neurological Surgery, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Disorders in Head Shape and Size; Myelomeningocele

Daniel R. Cooperman, MD Professor of Orthopedic Surgery, Case Western Reserve University; Professor of Orthopedic Surgery, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Musculoskeletal Disorders; Bone and Joint Infections; Congenital Abnormalities of the Upper and Lower Extremities and Spine

Timothy M. Crombleholme, MD Richard G. and Geralyn Azizkhan Chair in Pediatric Surgery; Professor of Surgery, Pediatrics (Molecular and Developmental Biology), and Obstetrics and Gynecology, University of Cincinnati College of Medicine; Director, Fetal Care Center of Cincinnati, Division of Pediatric General, Thoracic, and Fetal Surgery, Cincinnati Children’s Hospital, Cincinnati, Ohio

Surgical Treatment of the Fetus

Mario De Curtis, MD Professor of Neonatology, Dipartimento di Scienze Ginecologiche, Perinatologia e Puericultura, University of Rome Faculty of Medicine; Director, Neonatology Unit, Patologia Neonatale e Terapia Intensiva Dipartimento di Scienze Ginecologiche, Perinatologia e Puericultura, Azienda Policlinico Umberto I, Rome, Italy

Disorders of Calcium, Phosphorus, and Magnesium Metabolism

Linda S. de Vries, MD, PhD Professor in Neonatal Neurology, Department of Neonatology, Wilhelmina Children’s Hospital, Utrecht, The Netherlands

Intracranial Hemorrhage and Vascular Lesions; Hypoxic-Ischemic Encephalopathy, Assessment Tools

Katherine MacRae Dell, MD Associate Professor of Pediatrics, Case Western Reserve University; Chief, Division of Pediatric Nephrology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Fluid and Electrolyte Management; Acid-Base Management; The Kidney and Urinary Tract

Scott Denne, MD Professor of Pediatrics, Indiana University School of Medicine, Riley Hospital for Children, Indianapolis, Indiana

Parenteral Nutrition; Enteral Nutrition

Sherin U. Devaskar, MD Professor of Pediatrics, David Geffen School of Medicine at UCLA; Director, Neonatal Research Center, Division of Neonatology and Developmental Biology, Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, California

Developmental Origins of Adult Health and Disease; Disorders of Carbohydrate Metabolism

Contributors

Juliann Di Fiore, BSEE Research Engineer, Department of Medicine, Case Western Reserve University, Division of Neonatology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Assessment of Pulmonary Function

Steven M. Donn, MD Professor of Pediatrics, Division of Neonatal-Perinatal Medicine; Faculty Associate, Center for Global Health, School of Public Health, University of Michigan Health System; Staff Neonato­logist, C.S. Mott Children’s Hospital, Ann Arbor, Michigan

Assisted Ventilation and Its Complications

Morven S. Edwards, MD Professor of Pediatrics, Baylor College of Medicine; Attending Physician, Texas Children’s Hospital, Houston, Texas

Postnatal Bacterial Infections; Fungal and Protozoal Infections

William H. Edwards, MD Professor of Pediatrics, Dartmouth Medical School, Hanover, New Hampshire; Vice Chairman, Department of Pediatrics, Neonatology Section Chief; Medical Director, Citad Nurseries, Dartmouth Hitchcock Medical Center, Lebanon, New Hampshire

Care of the Mother, Father, and Infant,

Francine Erenberg, MD Assistant Professor, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University; Staff, Pediatric Cardiologist, Cleveland Clinic Children’s Hospital, Department of Pediatric Cardiology, Cleveland, Ohio

Fetal Cardiac Physiology and Fetal Cardiovascular Assessment; Congenital Defects

Avroy A. Fanaroff, MB, BCh, FRCP[E], FRCPCN Professor, Pediatrics and Reproductive Biology, Case Western Reserve University School of Medicine; Eliza Henry Barnes Chair in Neonatology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Epidemiology; Perinatal Services; Obstetric Management of Prematurity

Jonathan M. Fanaroff, MD, JD Assistant Professor of Pediatrics, Case Western Reserve University School of Medicine; Associate Medical Director, NICU; Director, Rainbow Center for Pediatric Ethics, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Legal Issues in Neonatal-Perinatal Medicine

Ross Fasano, MD Pediatric Hematology/Oncology Fellow, Children’s National Medical Center, Washington, DC

Blood Component Therapy for the Neonate

Orna Flidel-Rimon, MD Lecturer in Pediatrics, Hebrew University, Jerusalem; Attending Physician, Kaplan Medical Center, Rehovot, Israel

Post-term Pregnancy

ix

Smadar Friedman, MD Senior Lecturer, Hadassah University Hospital, Hadassah School of Medicine; Director, Neonatology Unit, Department of Neonatology, Hadassah Ein Kerem Hospital, University Hospital, Jerusalem, Israel

Fetal Effects of Autoimmune Disease

Susan E. Gerber, MD, MPH Assistant Professor, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Northwestern University, Feinberg School of Medicine; Attending Physician, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Northwestern Memorial Hospital, Chicago, Illinois

Antepartum Fetal Surveillance; Evaluation of the Intrapartum Fetus

Jay P. Goldsmith, MD Clinical Professor of Pediatrics, Tulane University School of Medicine, New Orleans, Lousiana; Neonatologist, Women’s and Children’s Hospital, Lafayette, Louisiana

Delivery Room Resuscitation of the Newborn, Overview and Initial Management; Chest Compression, Medications, and Special Problems

Bernard Gonik, MD Professor and Fann Srere Chair of Perinatal Medicine, Department of Obstetrics and Gynecology, Wayne State University School of Medicine, Detroit, Michigan

Perinatal Infections

Jeffrey B. Gould, MD Robert L. Hess Professor in Pediatrics and Director, Perinatal Epidemiology and Health Outcomes Research Unit, Department of Pediatrics, Division of Neonatal and Developmental Medicine, Stanford University School of Medicine, Stanford; Attending Neonatologist, Lucile Packard Children’s Hospital at Stanford, Palo Alto, California

Evaluating and Improving the Quality and Safety of Neonatal Intensive Care

Pierre Gressens, MD, PhD Chief, Inserm-Paris 7 University, Hopital Robert Debré, Paris, France; Professor of Perinatal Neurology, Imperial College London, Hammersmith Hospital, London, United Kingdom

Normal and Abnormal Brain Development; White Matter Damage and Encephalopathy of Prematurity

Susan J. Gross, MD Professor, Department of Obstetrics and Gynecology, Women’s Health and Department of Pediatrics, Albert Einstein College of Medicine; Chairperson, Department of Obstetrics and Gynecology, North Bronx Healthcare Network, Bronx, New York

Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

Andrée M. Gruslin, MD, FRCS Associate Professor, Division of Maternal Fetal Medicine, Department of Obstetrics & Gynecology and Newborn Care, Cellular and Molecular Medicine, University of Ottawa, Ottawa, Ontario, Canada

Erythroblastosis Fetalis

x

Contributors

Balaji K. Gupta, MD Clinical Assistant Professor, Department of Ophthalmology and Visual Sciences, University of Chicago Pritzker School of Medicine, Chicago, Illinois

The Eye, Examination and Common Problems

Maureen Hack, MB, ChB Professor of Pediatrics, Case School of Medicine; Director, High Risk Follow-Up Program, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Follow-up for High-Risk Neonates

Louis P. Halamek, MD Associate Professor, Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University; Director, Center for Advanced Pediatric and Perinatal Education, Packard Children’s Hospital, Palo Alto, California; Attending Neonatologist, Packard Children’s Hospital, Palo Alto, California

Simulation in Neonatal-Perinatal Medicine

Aaron Hamvas, MD The James P. Keating, MD, Professor of Pediatrics, Washington University School of Medicine, St. Louis Children’s Hospital, St. Louis, Missouri

Pathophysiology and Management of Respiratory Distress Syndrome

Jonathan Hellmann, MBBCh, FCP(SA), FRCPC, MHSc

McCallum R. Hoyt, MD, MBA Assistant Professor of Anesthesiology, Harvard Medical School, Boston, Massachusetts; Director of Gynecologic and Ambulatory Anesthesia; Staff, Obstetric Anesthesia, Brigham and Women’s Hospital, Boston, Massachusetts

Anesthetic Options for Labor and Delivery

Petra S. Hüppi, MD Professor of Pediatrics, University of Geneva Faculty of Medicine; Director, Child Development Unit, Department of Pediatrics, Children’s Hospital, Geneva, Switzerland

White Matter Damage and Encephalopathy of Prematurity; Normal and Abnormal Brain Development

Lucky Jain, MD, MBA Richard W. Blumberg Professor and Executive Vice Chairman, Department of Pediatrics, Emory University School of Medicine; Medical Director, Emory Children’s Center, Children’s Healthcare of Atlanta at Eggleston in Atlanta, Georgia, Emory University Hospital Midtown, Atlanta, Georgia

The Late Preterm Infant

Alan H. Jobe, MD, PhD Professor of Pediatrics, University of Cincinnati School of Medicine; Director of Perinatology Research, Cincinnati Children’s Hospital, Cincinnati, Ohio

Lung Development and Maturation

Nancy E. Judge, MD, FACOG

Professor of Pediatrics, University of Toronto; Clinical Director, Neonatal Intensive Care Unit, Hospital for Sick Children, Toronto, Ontario, Canada

Associate Professor of Reproductive Biology, Albert Einstein Medical School of Yeshiva University; Director of Obstetrics, Montefiore North Division, Montefiore Hospitals, Bronx, New York

Medical Ethics in Neonatal Care

Perinatal Imaging

Susan R. Hintz, MD Associate Professor, Department of Pediatrics, Division of Neonatal and Perinatal Medicine, Stanford University School of Medicine; Attending Neonatologist, Lucile Packard Children’s Hospital, Stanford, California

Biomedical Engineering Aspects of Neonatal Monitoring

Steven B. Hoath, MD Professor of Pediatrics, University of Cincinnati College of Medicine, Division of Neonatology; Medical Director, Skin Sciences Program, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

The Skin

Jeffrey D. Horbar, MD Jerold F. Lucey Professor of Neonatal Medicine, University of Vermont College of Medicine; Chief Executive and Scientific Officer, Vermont Oxford Network, Burlington, Vermont

Evaluating and Improving the Quality and Safety of Neonatal Intensive Care

Michael Kaplan, MD Professor of Pediatrics, Faculty of Medicine of the Hebrew University; Director, Department of Neonatology, Shaare Zedek Medical Center, Jerusalem, Israel

Neonatal Jaundice and Liver Disease

Satish C. Kalhan, MBBS, FRCP, DCH Professor, Department of Molecular Medicine; Staff, Department of Pathobiology, Gastroenterology and Hepatology, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University Cleveland, Ohio

Disorders of Carbohydrate Metabolism

Reuben Kapur, PhD Professor of Pediatrics, Molecular Biology and Biochemistry, Medical and Molecular Genetics, Microbiology and Immunology, Indiana University School of Medicine, Department of Pediatrics; Director, Program in Hematologic Malignancies and Stem Cell Biology, Herman B. Wells Center for Pediatric Research, Cancer Research Institute, Indiana University School of Medicine, Indianapolis, Indiana

Developmental Immunology

Contributors

Ganga Karunamuni, MD Graduate Research Assistant, Case Western Reserve University,Cleveland, Ohio Cardiac Embryology

Lawrence M. Kaufman, MD Clinical Associate Professor of Ophthalmology, University of Illinois at Chicago; Director of Pediatric Ophthalmology, Advocate Illinois Masonic Medical Center, Chicago, Illinois Examination and Common Problems

Kathleen A. Kennedy, MD, MPH Professor of Pediatrics, Department of Pediatrics, University of Texas Health Science Center at Houston; Director, Division of Neonatal-Perinatal Medicine; Director, M.S. in Clinical Research Degree Program, University of Texas Health Science Center at Houston, Texas

Practicing Evidence-Based Neonatal-Perinatal Medicine

John H. Kennell, MD Professor of Pediatrics Emeritus, Department of Pediatrics, Case School of Medicine; Attending Physician, Division of Behavioral Pediatrics, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Care of the Mother, Father, and Infant

Joseph A. Kitterman, MD Professor of Pediatrics, University of California, San Francisco, School of Medicine, San Francisco, California Fibrodysplasia Ossificans Progressiva

Marshall H. Klaus, MD Professor of Pediatrics, Emeritus, University of California, San Francisco, School of Medicine, San Francisco, California

Care of the Mother, Father, and Infant,

Robert M. Kliegman, MD Professor and Chair of Pediatrics, Medical College of Wisconsin; Pediatrician-in-Chief/Pamela and Leslie Muma Chair in Pediatrics, Children’s Hospital of Wisconsin, Children’s Corporate Center, Milwaukee, Wisconsin

Intrauterine Growth Restriction

Oded Langer, MD, PhD Professor, Department of Obstetrics and Gynecology, Columbia University; Chairman, Department of Obstetrics and Gynecology, St. Luke’s-Roosevelt Hospital Center, New York, New York

Pregnancy Complicated by Diabetes Mellitus

Noam Lazebnik, MD Associate Professor of Reproductive Biology, Case Western University School of Medicine; Division of Maternal and Fetal Medicine, University MacDonald Women’s Hospital, Cleveland, Ohio

Perinatal Imaging

xi

Malcolm I. Levene, MD, FMedSe Professor of Paediatrics and Child Health, University of Leeds School of Medicine Department of Pediatrics; Consultant Neonatologist, Leeds Teaching Hospitals NHS Trust, Leeds, West York, United Kingdom

Hypoxic-Ischemic Encephalopathy; Pathophysiology; Management

Foong-Yen Lim, MD Assistant Professor of Surgery, Pediatrics and Obstetrics and Gynecology, University of Cincinnati College of Medicine; Pediatrics and Fetal Surgeon, Division of Pediatric General, Thoracic and Fetal Surgery, Cincinnati Children’s Hospital, Cincinnati, Ohio

Surgical Treatment of the Fetus

Tom Lissauer, MB, Bchir, FRCPCH Hon Consultant Neonatologist, Imperial College Healthcare Trust; Pediatric Program Director, Institute of Global Health, Imperial College, London, United Kingdom

Physical Examination of the Newborn

Suzanne M. Lopez, MD Associate Professor of Pediatrics, Department of Pediatrics, University of Texas Health Science Center at Houston; Associate Professor of Pediatrics; Director, Neonatal Perinatal Medicine Fellowship, University of Texas Health Science Center at Houston, Houston, Texas

Practicing Evidence-Based Neonatal-Perinatal Medicine

Timothy E. Lotze, MD Assistant Professor, Department of Pediatrics, Baylor College of Medicine; Section of Child Neurology, Texas Children’s Hospital, Houston, Texas

Hypotonia and Neuromuscular Disease

Naomi L. C. Luban, MD Division Chief, Children’s National Medical Center, Department of Laboratory Medicine, Children’s National Medical Center, Washington, DC

Blood Component Therapy for the Neonate

Lori Luchtman-Jones, MD Associate Professor, George Washington University Medical School; Division Chief, Hematology, Children’s National Medical Center, Washington, DC

Hematologic Problems in the Fetus and Neonate

David K. Magnuson, MD, FACS, FAAP Department Chair, Pediatric Surgery, Cleveland Clinic Children’s Hospital, Cleveland, Ohio

Development and Basic Physiology of the Neonatal Gastrointestinal Tract; Selected Gastrointestinal Anomalies

xii

Contributors

Henry H. Mangurten, MD

Mary L. Nock, MD

Professor of Pediatrics, Rosalind Franklin University of Medicine and Science/The Chicago Medical School, North Chicago, IL; Chairman Emeritus, Department of Pediatrics, Advocate Lutheran General Children’s Hospital, Park Ridge, Illinois

Assistant Professor, Department of Pediatrics, Case Western Reserve University School of Medicine; Attending Neonatologist, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Birth Injuries

Tables of Normal Values

Jacquelyn McClary, Pharm D Clinical Pharmacist Specialist, Neonatal Intensive Care Unit, Rainbow Babies and Children’s Hospital, Cleveland, Ohio Developmental Pharmacology

Geoffrey Miller, MA, MB, BCh, MPhil, MD, FRCP, FRACP Professor of Pediatrics and Neurology, Yale University School of Medicine; Clinical Director, Pediatric Neurology, Yale New Haven Hospital, New Haven, Connecticut

Hypotonia and Neuromuscular Disease

Marilyn T. Miller, MD Professor of Ophthalmology, Department of Ophthalmology and Visual Sciences, University of Illinois at Chicago College of Medicine; University of Illinois Hospital, Chicago, Illinois

The Eye, Examination and Common Problems

Mohamed W. Mohamed, MD Neonatology Fellow, The Hospital for Sick Children, Toronto, Ontario Disorders of Calcium, Phosphorus, and Magnesium Metabolism

Thomas R. Moore, MD Professor and Chairman, Department of Reproductive Medicine, University of California School of Medicine, San Diego, California

Erythroblastosis Fetalis; Amniotic Fluid and Nonimmune Hydrops Fetalis

Colin J. Morley MA, DCH, MD, FRCP, FRCPCH, FRACP Professor of Neonatal Medicine, The Royal Women’s Hospital, Melbourne, Australia, United Kingdom

Role of Positive Pressure Ventilation in Neonatal Resuscitation

Stuart C. Morrison, MD, ChB, FRCP Clinical Faculty, Cleveland Clinic Medical School; Staff Radiologist, Cleveland Clinic Foundation, Cleveland, Ohio

Perinatal Imaging

Anil Narang, MD Professor, Pediatrics (Neonatology), Department of Pediatrics, Postgraduate Institute of Medical Education and Research; Head, Department of Pediatrics, Advanced Pediatric Centre, Chandigarh, India

Perinatal and Neonatal Care in Developing Countries

Vivek Narendran, MD, MRCP, MBA Associate Professor of Pediatrics, University of Cincinnati College of Medicine; Medical Director, University Hospital NICU, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

The Skin

Mark R. Palmert, MD, PhD Associate Professor of Paediatrics, University of Toronto Faculty of Medicine; Head, Division of Endocrinology, The Hospital for Sick Children, Toronto, Ontario, Canada

Disorders of Sex Development

Aditi S. Parikh, BA, MD Clinical Instructor, Case Western Reserve University School of Medicine; Clinical Geneticist, University Hospital’s Case Medical Center, Cleveland, Ohio

Congenital Anomalies

Robert L. Parry, MD Associate Professor of Surgery and Pediatrics, Case Western Reserve School of Medicine; Pediatric Surgeon, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Development and Basic Physiology of the Neonatal Gastrointestinal Tract; Selected Gastrointestinal Anomalies

Dale L. Phelps, MD Professor of Pediatrics, University of Rochester School of Medicine, Rochester, New York

Retinopathy of Prematurity

Brenda Poindexter, MD, MS Associate Professor of Pediatrics, Indiana University School of Medicine, Riley Hospital for Children, Indianapolis, Indiana

Parenteral Nutrition; Enteral Nutrition

Richard A. Polin, MD Professor of Pediatrics, Department of Neonatology, Columbia University Medical Center; Director, Division of Neonatology, Morgan Stanley Children’s Hospital of New York, Columbia University Medical Center, New York, New York

Developmental Immunology

Bhagya L. Puppala, MD Assistant Professor of Pediatrics, Rosalind Franklin University of Medicine and Science, The Chicago Medical School, North Chicago, Illinois; Adjunct Professor of Pediatrics, Midwestern University, Downer’s Grove, Illinois; Director, Neonatal Perinatal Medicine-Fellowship, Director, Neonatal Perinatal Medicine Research, Advocate Lutheran General Children’s Hospital, Park Ridge, Illinois

Birth Injuries

Contributors

Tonse N.K. Raju, MD, DCH Adjunct Professor of Pediatrics, Georgetown University, Washington, DC; Program Scientist/Medical Officer, Pregnancy and Perinatalogy Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland

From Infant Hatcheries to Intensive Care: Highlights of the Century of Neonatal Medicine

Ashwin Ramachandrappa, MD Attending Neonatologist, Neonatology Associates, Ltd, Phoenix, Arizona The Late Preterm Infant

Raymond W. Redline, MD Professor of Pathology and Reproductive Biology, Case Western Reserve University School of Medicine; Pediatric Pathologist, University Hospital Case Medical Center, Cleveland, Ohio

Placental Pathology

Jacques Rigo, MD, PhD Professor of Neonatology and Pediatrics Nutrition, University of Liège; Head, Department of Neonatology, CHR Citadelle, Liège, Belgium

Disorders of Calcium, Phosphorus, and Magnesium Metabolism

Barrett K. Robinson, MD, MPH Clinical Fellow, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Northwestern University Feinberg School of Medicine, Chicago, Illinois

Evaluation of the Intrapartum Fetus

Susan R. Rose, MD Professor of Pediatrics and Endocrinology, University of Cincinnati; Professor of Pediatric Endocrinology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

Thyroid Disorders

Florence Rothenberg, MS, MD, FACC Assistance Professor, Department of Internal Medicine Staff; Cardiologist, VAMC Division of Cardiovascular Diseases, University of Cincinnati, Cincinnati, Ohio

Cardiac Embryology

Shaista Safder, MD Assistant Professor, University of Central Florida and Florida State University, Arnold Palmer Hospital for Children, Orlando, Florida Disorders of Digestion

Ola Didrik Saugstad, MD, PhD, FRCPE Professor of Pediatrics, Department of Pediatric Research, Oslo University, Oslo, Norway

Oxygen Therapy

xiii

Katherine S. Schaefer, PhD Associate Professor of Biology, Randolph College, Lynchburg, Virginia

Cardiac Embryology

Mark S. Scher, MD Professor of Pediatrics and Neurology, Case Western University School of Medicine; Division Chief, Pediatric Neurology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Seizures in Neonates

Gunnar Sedin, MD, PhD Professor, Perinatal-Neonatal Medicine, Uppsala University School of Medicine, Uppsala, Sweden

The Thermal Environment

Dinesh M. Shah, MD Professor, Department of Obstetrics/Gynecology; Director of Maternal-Fetal Medicine and Maternal-Fetal Medicine Fellowship, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin

Hypertensive Disorders of Pregnancy

Eric S. Shinwell, MD Clinical Associate Professor of Pediatrics, Hebrew University, Jerusalem, Israel; Director of Neonatology, Kaplan Medical Center, Rehovot, Israel

Obstetric Management of Multiple Gestation and Birth

Rayzel M. Shulman, MD, FRCPC Research Fellow, University of Toronto, The Hospital for Sick Children, Toronto, Ontario, Canada

Disorders of Sex Development

Eric Sibley, MD, PhD Associate Professor of Pediatrics, Division of Pediatric Gastroenterology, Stanford University School of Medicine, Stanford, California

Neonatal Jaundice and Liver Disease

Sunil K. Sinha, MD, PhD Professor of Pediatrics, University of Durham, Durham, United Kingdom; Consultant Neonatologist, The James Cook University Hospital, Middlesbroug, United Kingdom

Assisted Ventilation and Its Complications

Carlos J. Sivit, MD Professor of Radiology and Pediatrics, Case Western Reserve School of Medicine; Vice Chair, Clinical Operations, University Hospitals Case Medical Center, Cleveland, Ohio

Diagnostic Imaging

Ernest S. Siwik, MD Associate Professor of Clinical Pediatrics, Louisiana State University Health Sciences Center; Director, Cardiac Catheterization Program, Children’s Hospital of New Orleans, New Orleans, Louisiana

Principles of Medical and Surgical Management

xiv

Contributors

Robert C. Sprecher, MD, FACS, FAAP Chief, Pediatric Otolaryngology, Rainbow Babies and Children’s Hospital, Case Western Reserve University, Cleveland, Ohio

Upper Airway Lesions

Robin H. Steinhorn, MD Raymond and Hazel Speck Berry Professor of Pediatrics, Northwestern University Steinberg School of Medicine; Vice Chair of Pediatrics, Division Head of Neonatology, Children’s Memorial Hospital, Chicago, Illinois

Pulmonary Vascular Development

David K. Stevenson, MD Harold K. Faber Professor of Pediatrics; Vice Dean, Stanford University School of Medicine; Senior Associate Dean for Academic Affairs, Department of Pediatrics, Division of Neonatal and Developmental Medicine, Stanford University School of Medicine; Director, Charles B. and Ann L. Johnson Center for Pregnancy and Newborn Services, Lucile Packard Children’s Hospital, Stanford, California

Biomedical Engineering Aspects of Neonatal Monitoring; Neonatal Jaundice and Liver Disease

Eileen K. Stork, MD Professor of Pediatrics, Case Western Reserve University School of Medicine; Director, Neonatal ECMO Program, Rainbow Babies and Children’s Hospital, Division of Neonatology, Cleveland, Ohio

Therapy for Cardiorespiratory Failure

John E. Stork, MD Assistant Professor, Case Western Reserve University; Chief, Pediatric Cardiac Anesthesia, Rainbow Babies and Children’s Hospital, University Hospitals Case Medical Center, Cleveland, Ohio

Anesthesia in the Neonate

Arjan B. te Pas, MD, PhD Assistant Professor, Leiden University Medical Center; Paediatrician-Neonatologist, Department of Pediatrics, Division of Neonatology, Leiden University Medical Center, Leiden, The Netherlands

Role of Positive Pressure Ventilation in Neonatal Resuscitation

George H. Thompson, MD Professor, Orthopedic Surgery and Pediatrics, Case Western Reserve University; Director, Pediatric Orthopedics, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Musculoskeletal Disorders; Bone and Joint Infections; Congenital Abnormalities of the Upper and Lower Extremities and Spine

Philip Toltzis, MD Professor of Pediatrics, Case Western Reserve University School of Medicine; Attending Pediatrician, Division of Pharmacology and Critical Care, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Perinatal Viral Infections

Robert Turbow, MD, JD Adjunct Professor, Department of Biology, California Polytechnic State University, San Luis Obispo, California; Attending Neonatologist, Phoenix Children’s Hospital, Phoenix, Arizona

Legal Issues in Neonatal-Perinatal Medicine

Jon E. Tyson, MD, MPH UT Medical School Master’s Program—Co Director; Director, Center for Clinical Research and Evidence-Based Medicine; UT Medical School; Michelle Bain Distinguished Professor, Memorial Hermann Hospital, Lyndon Baines Johnson Hospital, Houston, Texas

Practicing Evidence-Based Neonatal-Perinatal Medicine

George F. Van Hare, MD Louis Larnck Ward Chair in Cardiology; Professor of Pediatrics; Director, Division of Pediatric Cardiology, Washington University School of Medicine, St. Louis Children’s Hospital, St. Louis, Missouri

Neonatal Arrhythmias

Maximo Vento, MD, PhD Professor of Pediatrics; Director, Neonatal Research Unit, University Hospital La Fe, Valencia, Spain

Oxygen Therapy

Dharmapuri Vidyasagar, MD, MSs, FAAP, FCCM, DHC Professor Emeritus, University of Illinois at Chicago, Department of Pediatrics, Chicago, Illinois

Perinatal and Neonatal Care in Developing Countries

Beth A. Vogt, MD Associate Professor of Pediatrics, Case Western Reserve University; Attending Physician, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

The Kidney and Urinary Tract

Betty Vohr, MD Professor of Pediatrics, Warren Alpert Medical School of Brown University; Director of Neonatal Follow-up Program; Medical Director, Rhode Island Hearing Assessment Program, Women and Infants Hospital, Providence, Rhode Island

Hearing Loss in the Newborn Infant

Michele C. Walsh, MD, MSE Professor of Pediatrics, Case Western Reserve University School of Medicine; Co-Director, Division of Neonatology; Medical Director, Neonatal Intensive Care Unit, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Epidemiology; Perinatal Services; Design Considerations; Myelomeningocele; Bronchopulmonary Dysplasia

Michiko Watanabe, PhD Associate Professor, Departments of Pediatrics, Genetics, and Anatomy, Case School of Medicine; Staff, Division of Pediatric Cardiology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio

Cardiac Embryology

Contributors

Diane K. Wherrett, MD, FRCPC Associate Professor, Division of Endocrinology, Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada

Disorders of Sex Development

Robert D. White, MD Adjunct Professor of Psychology, University of Notre Dame, Notre Dame, Indiana; Clinical Assistant Professor of Pediatrics, Indiana University School of Medicine, Indianapolis, Indiana; Director, Regional Newborn Program; Memorial Hospital, South Bend, Indiana

The Sensory Environment of the Intensive Care Nursery; Design Considerations

Georgia L. Wiesner, MS, MD Associate Professor of Genetics and Medicine, School of Medicine, Case Western Reserve University; Clinical Geneticist, Center for Human Genetics, University Hospitals Case Medical Center, Cleveland, Ohio

Congenital Anomalies

Jamie C. Wikenheiser, PhD Assistant Professor, Department of Pathology and Laboratory Medicine, Division of Integrative Anatomy, University of California, Los Angeles

Cardiac Embryology

David B. Wilson, MD, PhD Associate Professor, Department of Pediatrics, Washington University School of Medicine; Attending Physician, St Louis Children’s Hospital, St Louis, Missouri

Hematologic Problems in the Fetus and Neonate

Deanne Wilson-Costello, MD Professor of Pediatrics, Case Western Reserve University; Co-Director High Risk Follow-Up Program, Rainbow Babies and Children’s Hospital, Case Western Reserve University, Cleveland, Ohio

Follow-up for High-Risk Neonates

Richard B. Wolf, DO, FACOG Associate Clinical Professor, Department of Reproductive Medicine, University of California San Diego, School of Medicine, La Jolla, California; Attending Perinatologist, University of California San Diego Medical Center, San Diego, California

Amniotic Fluid and Nonimmune Hydrops Fetalis

xv

Ronald J. Wong, MD Senior Research Scientist, Department of Pediatrics, Division of Neonatal and Developmental Medicine, Stanford University School of Medicine, Stanford, California

Biomedical Engineering Aspects of Neonatal Monitoring; Neonatal Jaundice and Liver Disease

Mervin C. Yoder, MD Richard and Pauline Klingler Professor of Pediatrics; Professor of Biochemistry and Molecular Biology; Professor of Cellular and Integrative Physiology, Indiana University School of Medicine, Department of Pediatrics; Director, Herman B. Wells Center for Pediatric Research; Associate Chair for Pediatrics Research, Herman B. Wells Center for Pediatric Research, Cancer Research Institute, Indiana University School of Medicine, Indianapolis, Indiana

Developmental Immunology

Thomas Young, MD Professor of Pediatrics, School of Medicine, University of North Carolina Chapel Hill, Chapel Hill, North Carolina; Senior Neonatologist, WakeMed Faculty Physicians, Raleigh, North Carolina

Therapeutic Agents

Kenneth G. Zahka, MD Professor of Pediatrics, Cleveland Clinic Lerner College Medicine, Case Western Reserve School of Medicine; Pediatric Cardiology, Cleveland Clinic Children’s Hospital, Cleveland, Ohio

Genetic and Environmental Contributions to Congenital Heart Disease; Principles of Neonatal Cardiovascular Hemodynamics; Approach to the Neonate with Cardiovascular Disease; Congenital Defects; Cardiovascular Problems of the Neonate; Principles of Medical and Surgical Management

Arthur B. Zinn, MD, PhD Attending Physician, University Hospitals Case Medical Center; Associate Professor, Case Western Reserve University, Cleveland, Ohio

Inborn Errors of Metabolism

Preface The foundation for successful outcomes in neonatal-perinatal medicine has been the ability to apply knowledge of the fundamental pathophysiology of the various neonatal disorders to safe interventions. Molecular, biologic, and technologic advances have facilitated the diagnosis, monitoring, and therapy of these complex disorders. Advances at the bench have been transformed to the bedside, and survival statistics reveal steady improvements. Nonetheless, although the survival rates may give reason to rejoice, the high early morbidity and persistent neurodevelopmental problems remain cause for concern. Such problems include bronchopulmonary dysplasia, nosocomial infections, necrotizing enterocolitis, hypoxic-ischemic encep­h­ alopathy, cerebral palsy, and the inability to sustain the intrauterine rate of growth when the infants are born prematurely. These problems need to be addressed in addition to the complex birth defects and genetic disorders that now loom as major problems in the neonatal intensive care unit. The field of Neonatal Perinatal Medicine has transformed from anecdotal medicine to evidence-based medicine. The problem is that evidence-based medicine predicts outcomes for groups but not individuals. The next frontier, individualized, or personalized medicine, requires application of the human genome project to the individual patient. That frontier is gaining momentum amidst a dizzying proliferation of newly acquired scientific knowledge. The translation of bench research to bedside innovation is proceeding smoothly as is the understanding of the underlying mechanisms of many disorders. Advances in genetics have helped solve the etiology of many disorders, and many previously mysterious diseases can now be attributed to mitochondrial disorders accompanied by cellular energy failure. We have also attempted to address these advances. With the combination of print and electronic journals, the effort to stay current in a single subspecialty remains a daunting task. Indeed, presenting the current status of the field of

neonatal-perinatal medicine, even in a two-volume textbook, has become extremely challenging. It is a tribute to the contributors to Neonatal-Perinatal Medicine that this text has reached its ninth edition. We are profoundly grateful to both our loyal and our new contributors who give so freely of their time and knowledge. For this ninth edition, we have added several new sections ranging from problems of the late preterm infant to fetal origins of adult disease. We have, in addition, completely reorganized and rewritten a large number of chapters and significantly updated the rest. Our accomplished authors and careful editing continue to focus on the biologic basis of developmental disorders and evidence basis for their management. We have also increasingly sought to draw upon international leaders in the field of neonatal-perinatal medicine to provide a truly global perspective. This book would not exist without the remarkable clinical and intellectual environment that constitutes Rainbow Babies and Children’s Hospital in Cleveland. On a daily basis, we gain knowledge from our faculty colleagues and fellows, and wisdom from our nursing staff, who are so committed to their young patients. Once again, we have been blessed with an in-house editor, Bonnie Siner, to whom we cannot adequately express our thanks. She is the glue behind the binding in the book and has worked tirelessly with Elsevier staff members to bring this project to fruition. Elsevier has once again provided the resources to accomplish this mammoth task.

Richard J. Martin Avroy A. Fanaroff Michele C. Walsh

xvii

SECTION

I

The Field of Neonatal-Perinatal Medicine

1

CHAPTER

1

From Infant Hatcheries to Intensive Care: Highlights of the Century of Neonatal Medicine Tonse N. K. Raju

We trust we have been forgiven for coining the words, “neonatology” and “neonatologist.” We do not recall ever having seen them in print. The one designates the art and science of diagnosis and treatment of disorders of the newborn infant, the other the physician whose primary concern lies in the specialty. . . . We are not advocating now that a new subspecialty be lopped from pediatrics . . . yet such a subdivision . . . [has] as much merit as does pediatric hematology. — A. J. Schaffer, 196071 The American Board of Pediatrics offered the first examination in the neonatal-perinatal medicine subspecialty in 1975. By 2007, 4428 men and women were formally certified as neonatologists by the Board—a number more than the combined total of certified specialists in Pediatric Hematology and Oncology (2051), and Pediatric Cardiology (2016). The 1950 Index Medicus listed 218 publications that year under the subject heading “Infant, Newborn.” By 1967, the number had increased to 6365, and in 2008, more than 400,000 publications were listed in PubMed, the National Library of Medicine’s electronic database.59 Dr. Schaffer need not have apologized for his visionary understatement—the specialty and the physicians he christened proved him prophetic. Although it was not “lopped” from pediatrics, this final frontier carved a niche, bridging obstetrics with pediatrics and intensive care with primary care. The formal creation of neonatology may appear to be recent, but its roots extend into the 19th century, when systematic and organized care for premature infants began in earnest. This chapter traces the origins and growth of modern perinatal and

neonatal medicine, with a brief perspective on its promises and failures. The reader may consult scholarly monographs and review articles on specific topics for in-depth analyses.6,7,24,31,73,74

PERINATAL PIONEERS Many scientists played key roles in developing the basic concepts in neonatal-perinatal medicine that helped to formalize the scientific basis for neonatal clinical care. Their work and teachings inspired generations of further researchers advancing the field. For brevity’s sake, only a few are shown in Figure 1-1. Medicinal chemistry (later called biochemistry) and classic physiology gained popularity and acceptance toward the end of the 19th century, inaugurating studies on biochemical and physiologic problems in the fetus and the newborn. Some leading scientists in the early 20th century making fundamental contributions and training scores of scientists from around world included Barcroft8,34 and his mentee Dawes in England (gas exchange and nutritional transfer across the placenta and oxygen carrying in fetal and adult hemoglobin); Ylppö in Finland (neonatal nutrition, jaundice, and thermoregulation); Lind in Sweden (circulatory physiology); Smith in Boston76 (fetal and neonatal respiratory physiology); DeLee in Chicago26,27 (leading researcher on incubators and in high-risk obstetric topics, he also founded the first U.S. “incubator station” at the Chicago Lying-in Hospital); Day in New York (temperature regulation, retinopathy of prematurity, and jaundice); and Gordon38 in Denver (nutrition). Although the list appears long, the centers were scattered in location. Although it was not formal, the training

3

4

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

E

C

B

A

F

G

D

H

Figure 1–1.  Pioneers in perinatal and neonatal physiology and medicine. A, Joseph Barcroft. B, Arvo Ylppö. C, John Lind.

D, William Liley. E, Joseph DeLee. F, Richard Day. G, Clement Smith. H, Harry Gordon.  (A, From Barcroft J: Research on pre-natal life [vol 1], Oxford, 1977, Blackwell Scientific, courtesy of Blackwell Scientific; B-D, F-H, From Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: neonatal medicine [vol 1], Evansville, IN, 1984, Ross Publication, pp ix [B], xix [C], xxii [D], xvi [F], xii [G], xiv [H], courtesy of Mead Johnson Nutritional; E, Courtesy of Mrs. Nancy DeLee Frank, Chicago, IL.) was nevertheless rigorous. Smith once said, “If you were interested in babies and liked Boston, I was the only wheel in town!”60 Table 1-1 highlights some milestones in perinatal medicine.

HIGH-RISK FETUS AND PERINATAL OBSTETRICS Because so many deaths occurred in early infancy in times past, many cultures adopted remarkably innovative methods to deal with such tragedies. According to a Jewish tradition, full, year-long mourning is not required for infants who die before 30 days of age.40 In some Asian ethnic groups, infantnaming ceremonies are held only after several months, until which time the infant is simply called “it.” In India, an odd or coarse-sounding name is given to the first surviving infant after the death of a previous sibling; this is aimed at deflecting evil spirits. In her book on the history of the Middle Ages, Tuchman notes that infants were seldom depicted in medieval artworks.84 When they were drawn (e.g., the infant Jesus), women in the pictures looked away from the infant, ostensibly conveying respect, but perhaps because of fearful aloofness. Since antiquity, the care of pregnant women has been the purview of midwives, grandmothers, and experienced female elders in the community. Wet nurses helped when mothers were unavailable or unwilling to nurse their infants. Little or no assistance was needed for normal or uncomplicated labor and delivery. For complicated deliveries, male physicians had

to be summoned, but they could do little because many of them lacked expertise or interest in treating women. Disasters during labor and delivery were common, rendering this phase in their lives the most dreaded for women.43 In the early 1900s, unexpected intrapartum complications accounted for 50% to 70% of all maternal deaths in England and Wales.17,56 Because the immediate concern during most highrisk deliveries was saving the mother, sick newborns were not given substantial attention; their death rates remained very high. Occasionally, happy outcomes of high-risk deliveries did occur. In one of the oldest works of art depicting labor and delivery (Fig. 1-2A), a bearded man and his assistant are standing behind a woman in labor, holding devices remarkably similar to the modern obstetric forceps. The midwife has delivered an evidently live infant. In Figure 1-2B, three infants from a set of quadruplets, nicely swaddled, have been placed on the mother, as the unwrapped fourth infant is being handed to her for nursing. A divine figure in the background is blessing the newcomers. Cesarean sections were seldom performed on living women before the 13th century. Even subsequently, the procedure was performed only as a final act of desperation. Contrary to popular belief, Julius Caesar’s birth was not likely by cesarean section. Because Caesar’s mother was alive during his reign, historians believe that she probably delivered him vaginally. The term cesarean probably originated from lex caesarea, in turn from lex regia, the “royal law” prohibiting

Chapter 1  From Infant Hatcheries to Intensive Care

5

TABLE 1–1  Selected Milestones in Perinatal Medicine Category

Year(s)

Description

Antenatal aspects

1752

Queen Charlotte’s Hospital, the world’s first maternity hospital, is founded in London57

1915-1924

Campbell introduces outlines of regular prenatal visits, which become a standard

1923-1925

Estrogen and progesterone are discovered

1928

First pregnancy test is described, in which women’s urine is shown to cause changes in mouse ovaries

1543

Vesalius observes fetal breathing movements in pigs

1634

Paré teaches that absence of movement suggests a dead fetus

1819, 1821

Laënnec introduces the stethoscope in 1819, and his friend Kergaradec shows that fetal heart sounds can be heard using it

1866

Forceps are recommended when there is “weakening of the fetal heart rate”

1903

Einthoven publishes his work on the EKG

1906

The first recording of fetal heart EKG is made

1908

The term fetal distress is introduced

1948-1953

There are developments in the external tocodynamometer

1953

Apgar describes her scoring system3

1957-1963

Systematic studies are conducted on fetal heart rate monitoring

1970

Dawes reports studies on breathing movement in fetal lambs

1980

Fetal Doppler studies begin

1981

Nelson and Ellenberg report that Apgar scores are poor predictors of neurologic outcome

ca. 1000-500 bc

In Ayurveda, the ancient Hindu medical system, physicians describe obstetric instruments

98-138

Soranus develops the birthing stool and other instruments

1500s

There are isolated reports of cesarean sections on living women

1610

The first intentional cesarean section is documented

1700s

The Chamberlen forceps are kept as a family secret for three generations

1921

Lower uterine segment cesarean section is reported

1953

The modern vacuum extractor is introduced

1900-1950

Barcroft, Dawes, Lind, Liley, and others study physiologic principles of placental gas exchange and fetal circulation

Fetal assessment

Labor and delivery

Fetal physiology

EKG, electrocardiogram. See references 2, 41, 43, 61, 78 for primary citations.

burial of corpses of pregnant women without removal of their fetuses.11,89 The procedure allowed for baptism (or a similar blessing) if the child was alive or burial otherwise. Infants surviving the ordeal of cesarean birth were assumed to possess special powers, as supposedly did Shakespeare’s Macduff—“not of a woman born,” but of a corpse, and able to slay Macbeth.54 Soranus of Ephesus (circa 38-138 ad) influenced obstetric practice for 1400 years. His Gynecology can be regarded as the first formal “textbook” of perinatal medicine. Initially extant, it was rediscovered in 1870 and translated into English for the first time in 1956.83 Soranus wrote superbly about podalic version, obstructed labor, multiple gestations, fetal malformations, and numerous other maternal and fetal disorders. In an

age of belief in magic and the occult, he insisted that midwives should be educated and free from superstitions. He forbade wet nurses from drinking alcohol lest it render the infant “excessively sleepy.” His chapter, “How to Recognize the Newborn That Is Worth Rearing,” remains one of the earliest accounts on assessing viability of sick newborns—a topic of great concern even today.

MIDWIVES AND PERINATAL CARE Although occasionally caricatured (Fig. 1-3), midwives were responsible for delivering obstetric care for thousands of years. Men disliked obstetrics, and women were shy to let male physicians handle them. Good midwives were always in

6

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

A

B Figure 1–2.  High-risk deliveries. A, Marble relief of uncertain date depicting a high-risk delivery. The physician and his assistant in the background are holding devices similar to modern obstetric forceps. A midwife has just helped deliver a live infant while two people are looking through the window. B, Delivery of quadruplets. (From Graham H: Eternal Eve: the history of gynecology and obstetrics, New York, 1951, Doubleday, pp 68, 172.)

great demand, and many of them held important social and political positions in European courts.43,61,86 The emergence of man-midwives (Fig. 1-4) in England had a major effect on high-risk obstetric practice. Chamberlen the Elder (1575-1628) is usually credited for inventing the modern obstetric forceps.43,61,63 For 150 years, the instrument remained a trade secret through three generations of Chamberlens. By then, others had developed similar devices, and patients began associating good obstetric outcomes with male physicians—a key factor in transforming midwifery to a male-dominated craft.43 The shift from women-midwifery to men-midwifery might also have been due to changing social values and gender relationships in which women voluntarily began making choices about their bodies.86 Today’s increasing roles for female midwives and the higher proportion of women choosing specific birth practices (e.g., home versus hospital delivery, “underwater births,” cesarean delivery on request) offer interesting contrasts and perspectives to 18th century obstetrics.

NEONATAL RESUSCITATION: TALES OF HEROISM AND DESPERATION Figure 1–3.  On call. “A Midwife Going to a Labour,” carica-

ture by Thomas Rowlandson, 1811. (Courtesy of The British Museum, London.)

Popular artworks and ancient medical writings provide accounts of miraculous revivals of apparently dead adults and children.66 These are tales of successes only, for the failures were buried and rarely reported. Attempts to “stimulate” and revive apparently dead newborns included beating, shaking, yelling, fumigating, dipping in ice-cold water, and dilating and blowing smoke into the rectum.25,30,66 Oxygen administration

Chapter 1  From Infant Hatcheries to Intensive Care

Figure 1–4.  Man-midwife. (Courtesy of Clements C. Fry Print Collections, Harvey Cushing/John Hay Whitney Medical Library, Yale University, New Haven, CT.)

through an orogastric tube to revive asphyxiated infants persisted well into the mid-1950s, when James and Apgar showed conclusively that the therapy was useless.1,52

APGAR AND THE LANGUAGE OF ASPHYXIA Few scientists in the 20th century influenced the practice of neonatal resuscitation as profoundly as Apgar (1909-1974). A surgeon, she chose obstetric anesthesia for her career. Her simple scoring system inaugurated the modern era of assessing infants at birth on the basis of simple clinical examination.3 Right or wrong, the Apgar score became the language of asphyxia. It is often said that the first words heard by a newborn infant are, “What’s the Apgar score?” Although “giving an Apgar” has become a ritual, its profound effect has been on formalizing the process of observing, assessing, and communicating the infant status at birth in a consistent and uniform manner. This process eventually led to the formal steps of resuscitation at birth using the score. Few people know that it was also Apgar who was the first to catheterize the umbilical artery in a newborn.16 A woman of enormous energy, talent, and compassion, Apgar was honored with her depiction on a 1994 U.S. postage stamp (Fig. 1-5).

FOUNDLING ASYLUMS AND INFANT CARE In its early days, the Roman Empire experienced decreasing population growth. The emperors taxed bachelors and rewarded married couples to encourage procreation.76 In 315 ad, Emperor Constantine, hoping to curb infanticide

7

Figure 1–5.  Virginia Apgar, U.S. postage stamp. (Courtesy of the U.S. Postal Service.)

and encourage adopting of orphans, decreed that all “foundlings” would become slaves of those who adopted them. Similar humanitarian efforts by kings and the Council of the Roman Church led to the institutionalization of infant care by establishing foundling asylums for abandoned infants,77 also called “Hospitals for the Innocent,”— the first children’s hospitals. Parents of unwanted infants “dropped off” their infants in a revolving receptacle at the door of such asylums, rang the doorbells, and disappeared into the night (Fig. 1-6). Such accounts are poignant reminders of the contemporary problem of child abandonment because of which many states have programs to save such “dumpster babies,” or the abandoned infants.69 Foundling asylums adopted pragmatic techniques for fundraising. In 18th century France, lotteries were held, and souvenirs were sold. In May 1749, Handel gave a concert to support London’s “Hospital for the Maintenance and Education of Exposed and Deserted Young Children.” The final item of the program was the playing of “The Foundling Hymn.”77

SAVING INFANTS TO MAN THE ARMY During the French Revolution, France faced appalling rates of infant mortality. With rates greater than 50%, the Revolutionary Council in 1789 enacted a decree, proclaiming that working-class parents “have a right to the nation’s succors at all times.”77 The postrevolutionary euphoria about equality and fraternity among men stimulated reforms, heralding an idealistic welfare state, leading to collecting and maintaining valid statistics about children. The world’s first national databases began in France in the late 18th century.77

8

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

A

B

Figure 1–6.  Foundling homes. A, Le Tour—revolving receptacle. Mother ringing a bell to notify those within that she is leaving

her baby in the foundling home (watercolor by Herman Vogel, France, 1889). B, Remorce (“Remorse”)—parents after placing their infant in a foundling home (engraving and etching by Alberto Maso Gilli, France, 1875). (A and B, Courtesy of the Museum of the History of Medicine, Academy of Medicine, Toronto, Ontario, Canada; from Spaulding M, Welch P:Nurturing yesterday’s child: a portrayal of the Drake collection of pediatric history, Philadelphia, 1991, Decker, p 110 [A] and p 119 [B].) Over the next century, France faced a population problem similar to that of ancient Rome—a negative population growth. The birthrate had declined, and infant mortality remained high. Fearing future shortages of troops, the military leaders, deeply engaged in battles with Prussia, were naturally alarmed. Commissions were set up to study the depopulation problem and develop remedial actions. A series of measures began to improve maternal and neonatal care.6,7,22,24,77 Young parents were encouraged to uphold their patriotism and bear more children to “man the future armies.” It is the irony of our times that such noble intentions as saving infants were motivated by brutal needs for enhancing military might.

AN INGENIOUS CONTRIVANCE, THE COUVEUSE, AND PREMATURE BABY STATIONS A popular story of the origin of modern incubator technology is that on seeing the poultry section during a casual visit to the Paris Zoo in 1878, Tarnier (1828-1897), a renowned obstetrician, conceived the idea of “incubators” similar to the “brooding hen” or couveuse.6,7,22,24 He asked an instrument maker, Martin, to construct similar equipment for infants. With a “thermo-syphon” method to heat the outside with an alcohol lamp, Martin devised a sufficiently ventilated, 1 m3, double-walled metal cage, spacious enough to hold two premature infants. The first couveuses were installed at the Paris Maternity Hospital in 1880. Tarnier’s efforts led to dramatic improvements in survival rates for preterm infants. Although a few others had developed incubators before Tarnier,7 it was he and his students, Budin (1846-1907) and

Auvard, who are largely responsible for institutionalizing preterm infant care. They placed several incubators side by side, promoting the concept of caring for groups of sick preterm infants in geographically separate regions within their hospital.6,7,81 Budin and Auvard improved the original couveuse by replacing its walls with glass and using simpler methods for heating. Their efforts greatly influenced incubator technology during the first half of the 20th century in Europe and the United States (Fig. 1-7 and Table 1-2). In 1884, Tarnier made another important contribution; he invented a small, flexible rubber tube for introduction through the mouth into the stomach of preterm infants. With this tube, he could drip milk directly into the stomach. This method of nutritional support he called “gavage feeding.” Gavage feeding plus keeping infants in relatively constant and warm temperatures had a dramatic impact on improving survival rates.15,21 Tarnier also recommended that the legal definition of viability should be 180 days of gestation, which was opposed by contemporary obstetricians, who thought that the concept was “therapeutic nihilism.”7 Defining “viability” remains a highly emotional and contentious issue in contemporary neonatal/perinatal practice.

INCUBATORS, BABY SHOWS, AND ORIGINS OF NEONATAL INTENSIVE CARE UNITS. Almost two decades after its debut in France, incubator technology appeared in the United States, heralding organized newborn intensive care. As in France, it was an obstetrician who spearheaded the movement. In 1898, DeLee established the first “Premature Baby Incubator Station” at the Sara Morris Hospital in Chicago. During the early 1900s,

Chapter 1  From Infant Hatcheries to Intensive Care

A

B

Sponge

Bed

Hot-Water Tank

C

Air Exit

Air Entrance Filling Funnel Bunson Burner

Glass Cover

Figure 1–7.  Early incubators. A, Rotch incubator, circa 1893. B, Holt incubator. C, Schematics of the Holt incubator. (A, From Cone TE Jr: History of American pediatrics, Boston, 1979, Little Brown, pp 57 and 58, courtesy of Little Brown;B and C, From Holt LE: The diseases of infants and children, New York, 1897, Appleton, pp 12 and 13, courtesy of Appleton.)

as academic obstetricians and pediatricians were organizing specialized care for premature infants, an interesting, if bizarre, set of events led to the era of “premature baby shows,” which began in Europe and continued in the United States, lasting well into the 1940s.6,7,73,81 Couney, a Budin associate of doubtful medical credentials, wished to popularize the French technology abroad and show the value of “conserving” premature infants. (This account has been doubted.7) Couney obtained six incubators, probably from the French innovator Lion. Initially, Couney wanted to exhibit only the incubators as a technology of hope for saving infants. To add drama, however, he brought six preterm infants from Virchow’s maternity unit in Berlin and exhibited them inside the incubators at the 1896 Berlin Exposition. He coined a catchy phrase for

9

the show—kinderbrutanstalt or “child hatchery”—igniting the imagination of the public thirsty for sensational scientific breakthroughs. Couney’s Berlin exhibit was an astounding success. He continued to organize similar baby shows and brought one of them to Great Britain’s Victorian Era Exhibition in 1897. An editorial in Lancet praised the show. It recommended that large “incubator stations” be established similar to fire stations, where parents could borrow incubators.36 This was the origin of the phrase, “premature baby incubator stations,” which became part of the medical lexicon. Couney set sail to the United States and organized the first U.S. incubator show in 1898 at the Omaha Trans-Mississippi Exposition. He settled in New York and established a permanent annual premature infant exhibit at New York’s Coney Island boardwalk that lasted for 40 years. Couney also took his premature infant shows during summer months to state fairs, traveling circuses, science expositions, and recreational expositions all over the United States (Fig. 1-8). The shows were particularly large in San Francisco, Chicago, St. Louis, Buffalo, and Minneapolis. About 8000 “Couney babies” were cared for in all of his exhibits. He claimed not to have charged any parent for the care. He made a lot of money—all from the entrance fee for his show—but he died poor. The last infant show was held during the 1939-1940 season in Atlantic City. A commemorative bronze plaque has been placed on the wall next to the entrance to the Holiday Inn Hotel, where Couney exhibits were held.73 The infant shows evoke familiar contemporary themes. The news media’s desire for obtaining spectacular stories of medical breakthroughs and the public’s curiosity about “rare phenomena” (e.g., quintuplet, sextuplet, or octuplet births) recall the memory of preterm infant show spectacles. The shows are also symbolic of the difficulty in preserving a delicate balance between information, public education and awareness, and sensationalism. Traveling infant shows were similar to traveling moon rock exhibits, which exploited the popularity of man’s landing on the moon and helped increase NASA’s annual budgets. In 1914, Hess of Chicago started a Premature Infant Station at the Sarah Morris Children’s Hospital (of the Michael Reese Medical Center). With great attention to environmental control and aseptic practices and a regimental approach to feeding, Hess and his head nurse, Ms. Evelyn Lundeen (Fig. 1-9), achieved spectacular survival rates.47,67 Hess also developed an incubator built on the concept of a doublewalled metallic “cage” with warm water circulating between the walls. He used electric current for heating and devised a system to administer free-flow oxygen (Fig. 1-10). Only a few Hess incubators are known to have survived to this day. Hess’s premature unit outlasted the DeLee Premature Station. In December 2008, the Michael Reese Medical Center closed, however, declaring bankruptcy. The story of development of incubators and their impact on pediatrics is a tale of the success of technology and that of the perils technology might beget (see later section on relationship of improved incubator care and the retinopathy of prematurity [ROP] epidemic). That incubators were able to save infants became a powerful symbol of the might of the machine. In the heroic age of the mechanical revolution, the notion that machines could solve all human problems was all too appealing. The incubator stands as the most enduring

10

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 1–2  Evolution of Incubators Year(s)

Developer/Product

Comments

1835, ca. 1850

von Ruehl (1769-1846)

A physician to Czarina Feodorovna, wife of Czar Paul I, von Ruehl develops the first known incubator for the Imperial Foundling Hospital in St. Petersburg. About 40 of these “warming tubs” are installed in the Moscow Foundling Hospital in 1850

1857

Denucé (1824-1889)

The first published account of introducing an incubator is a 400-word report by Denucé. This is a “double-walled” cradle

1880-1883

Tarnier (1828-1897)

Tarnier incubator is developed by Martin and installed in 1880 at the Port-Royal Maternité

1884

Credé (1819-1892)

Credé reports the results of 647 infants treated over 20 years using an incubator similar to that of Denucé

1887

Bartlett

Bartlett reads a paper on a “warming-crib” based on Tarnier’s concept, but uses a “thermo-syphon”

1893

Budin (1846-1907)

Budin popularizes the Tarnier incubator and establishes the world’s first “special care unit for premature infants” at Maternité and Clinique Tarnier in Paris

1893

Rotch (1849-1914)

The first American incubator with a built-in scale, wheels, and fresh-air delivery system is developed; the equipment is very expensive and elaborate

1897

Holt Incubator

A simplified version of the Rotch incubator is developed. In this double-walled wooden box, hot water circulates between the walls

1897-1920s

Brown, Lyons, DeLee, Allin

Many modifications are made to the early American and European incubators by physicians. These are called baby-tents, baby boxes, warming beds, and other names

1922

Hess

Hess introduces his famous incubator with an electric heating system. For transportation, he develops special boxes that can be plugged into the cigarette lighters in Chicago’s taxicabs

1930-1950s

Large-scale commercial incubators

There is worldwide distribution of Air Shields and other commercial ventilators

1970-1980

Modern incubators

Transport incubators with built-in ventilators and monitoring equipment are developed—mobile intensive care units

See references 6, 7, 22-24, 73-75 for primary citations.

symbol of the spectacular success of modern intensive care and (paradoxically) some of its failures.74,75

SUPPORTIVE CARE AND OXYGEN THERAPY In a single-page note in 1891, Bonnaire referred to Tarnier’s use of oxygen in treating “debilitated” premature infants 2 years earlier14—this was the first published reference to the administration of supplemental oxygen in premature infants for a purpose other than resuscitation. The use of oxygen in premature infants did not become routine, however, until the 1920s. Initially, a mixture of oxygen and carbon dioxide—instead of oxygen alone—was employed to treat asphyxia-induced narcosis. It was argued that oxygen relieved hypoxia, whereas carbon dioxide stimulated the respiratory center.80 Oxygen alone was reserved for “pure asphyxia” (whatever that meant). The advent of mobile oxygen tanks and their easy availability in the mid-1940s enabled the use of oxygen for resuscitation.51,53,74 The success of incubator care brought new and unexpected challenges.68 Innovative methods had to be developed to feed the increasing number of premature infants who were surviving

for longer periods than ever before. Their growth needed to be monitored, and illnesses related to prematurity, such as sepsis, apnea, anemia, jaundice, and respiratory distress, had to be studied and treated. Another completely unexpected peril from “improved” incubator technology was the epidemic of blindness from ROP (then called retrolental fibroplasia), documented in vivid detail elsewhere.74,75 The apparent culprit in cases of ROP was the “leak proof” incubator that led to a great increase in the inspired oxygen concentrations (piped in freeflow manner), coupled with the belief that oxygen was innocuous, and if a little bit could save lives, a lot could save even more lives.74,75 Because more and more sick and small preterm infants began surviving with incubator care, providing ventilatory assistance became an urgent necessity.

VENTILATORY CARE: “EXTENDED RESUSCITATION” The first mechanical instrument used for intermittent positive pressure ventilation in newborns was the aerophore pulmonaire, a simple device developed by the French obstetrician

Chapter 1  From Infant Hatcheries to Intensive Care

11

A

B Figure 1–8.  Incubator baby shows. A, People lined up to see the Infant Incubator Show, Buffalo, New York. B, Interior of an incubator baby show, Buffalo, New York. (A and B, From Silverman WA: Incubator-baby side shows, Pediatrics 64:127, 1979, courtesy of the American Academy of Pediatrics, 1979.)

Gairal.65,66 It was a rubber bulb attached to a J-shaped tube. By placing the bent end of the tube into the infant’s upper airway, one could pump air into the lungs. Holt recommended its use for resuscitation in his influential 1897 book.48 Before starting mechanical ventilation, one needed to cannulate the airway, a task nearly impossible without a laryngoscope and an endotracheal tube. Blundell (1790-1878), a Scottish obstetrician, was the first to use a mechanical device

Figure 1–10.  A Hess incubator on display at the Spertus

Museum in Chicago.  (From the International Museum of Surgical Sciences, Chicago, IL.) for tracheal intubation in living newborns.13,32 Introducing two fingers of his left hand over the infant’s tongue, he would feel the epiglottis and then guide a silver pipe into the trachea with his right hand. His tracheal pipe had a blunt distal end and two side holes. By blowing air into the tube about 30 times a minute until the heartbeat began, Blundell saved hundreds of infants with birth asphyxia and infants with laryngeal diphtheria. His method of tracheal intubation is practiced in

Figure 1–9.  Hess and Lundeen medallions at the Michael Reese Hospital, Chicago. (Photo courtesy of Tonse N. K. Raju.)

12

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

many countries today.85 In the late 19th century, a wide array of instruments evolved to provide longer periods of augmented or extended ventilation for infants who had been resuscitated in the labor room. Most of the early instruments were designed for use in adults, however, and were used later in newborns and infants, particularly to treat paralytic polio and laryngeal diphtheria.39,45,79,80 The iron lung (or “man-can”) was one of the earliest mechanical ventilatory devices (Fig. 1-11), and a U.S. patent was issued for it in 1876.19,20,42 In other ventilatory equipment, varying methods for rhythmic inflation and deflation of the lungs were used for prolonged ventilation. Among those, the Fell-O’Dwyer apparatus used a unique foot-operated bellows system connected to an implement similar to the aerophore bulb.25,65,66 Between 1930 and 1950, there were sporadic but important reports of prolonged assisted ventilation provided to newborns.12,62,79,80 Beginning in the late 1950s and through the 1960s, more neonatal intensive care units (NICUs) began providing ventilatory assistance regularly (Table 1-3). Ventilatory care did not become predictably successful, however, until the early 1970s, when continuous positive pressure was incorporated into ventilatory devices.44,58,62,79

SUPPORTIVE CARE: INTRAVENOUS FLUID AND BLOOD TRANSFUSIONS When it comes to intravenous therapy, our legacy is one of bloodletting, not of transfusing. Blundel (of intubation fame) also made a major contribution to transfusion science. Believing that “only human blood should be employed for humans,” he developed instruments, syringes, and funnels for this purpose. In 1818, Blundel carried out the first direct transfusion from a healthy donor into a recipient; 5 of his first 10 patients survived. Human-to-human transfusions gradually became accepted, but physicians in the 19th century were puzzled about unexpected disasters among blood transfusion recipients. It took 15 years after Landsteiner’s discovery of blood groups in 1901 for the general acceptance and understanding of the scientific basis for blood group incompatibility.87

Figure 1–11.  The Man-Can, circa 1873 to 1875. A hand-held negative-pressure ventilatory device for which a patent was applied in 1876.19,20 (From DeBono E: Eureka: how and when the greatest inventions were made: an illustrated history of inventions from the wheel to the computer, New York, 1974, Holt, Rinehart & Winston, p 159.)

Adult transfusions were rare, but newborn transfusions were rarer still. On March 8, 1908, a 4-day-old term infant who had hemorrhagic disease of the newborn made history. “As the child’s skin became waxen white and mucous membranes without color, it was decided to attempt transfusion of blood obtained from the infant’s father,” wrote Lambert from New York.55 Carrel, a surgeon from Rockefeller University Hospital, performed an end-to-end anastomosis of the right popliteal vein of the infant with the left radial artery of the father. No anesthetic was given to either patient. “The amount of blood transfused could not be measured, but enough blood was allowed to flow into the baby to change her color from pale transparent whiteness to brilliant red . . . [and] as soon as the wound was sutured, the infant fed ravenously and immediately went to sleep,” according to Lambert. Incidentally, Carrel was the first surgeon to develop innovative methods of suturing blood vessels—a contribution for which he received the 1912 Nobel Prize. Despite Lambert’s dramatic report, direct father-to-infant transfusion did not become routine. Because of unexpected reactions among the recipients, blood transfusions continued to be risky, despite proper matching of the donors’ blood for major blood types. The mystery was understood only after the discovery of Rh subtypes by Landsteiner and Wiener in 1940.87,90 The discovery of the Rh blood types, leading to the conquest of erythroblastosis fetalis, remains an unparalleled triumph in pediatric medicine. This is an example of an orderly progression of accumulating knowledge leading to the near-eradication of a disease. First, there were the clinical descriptions of a disease (erythroblastosis); then there was a revolutionary, if symptomatic, therapy for it (exchange transfusion); and then there were efforts to prevent it (RhoGAM). The enthralling story of the conquest of erythroblastosis has been described superbly in many monographs and comprehensive review articles.28,29,87,90

TOOLS AND SUPPLIES FOR NEONATAL INTENSIVE CARE UNITS It may be impossible for us to realize the hardship of performing such simple and mundane chores as the collection of blood from or insertion of catheters into the veins of preterm infants before the advent of ultra-small needles, pumps, and tubing. These were not available until the 1930s. In 1912, Blackfan (1883-1941) developed an ingenious suction device for blood collection.9 Obtaining blood was done by puncturing the sagittal sinus or femoral or carotid veins; the latter sometimes led to accidental puncture of the nearby arteries. Well into the early 1970s, only a handful of laboratories could perform arterial blood gas analyses with less than 5 mL of blood. Using the intraperitoneal route for treating dehydration or hypovolemic shock was common. Electrolyte solutions and blood were being administered directly into the peritoneal cavity with the expectation that its large absorptive surface allowed for rapid absorption.10,91 In 1923, Sidbury introduced umbilical venous catheterization for neonatal blood transfusions,72 and in the 1950s, Diamond and colleagues began using this route for exchange transfusions.28,29 Indwelling polyethylene tubes were not introduced for gastric feeding until 1951.70

13

Chapter 1  From Infant Hatcheries to Intensive Care

TABLE 1–3  Ventilatory Care, Respiratory Disorders, and Intensive Care Category

Approximate Time Span

Procedures and Techniques

Resuscitation and oxygen

From antiquity to early 1970s

Mouth-to-mouth breathing (although it fell from favor in the late 18th century because many influential physicians declared it a “vulgar method” of revival)

1878

Tarnier uses oxygen in debilitated premature infants

1900-1930s

Schultz, Sylvester, and Laborde methods of resuscitation involve various forms of swinging infants (Schultz), traction of the tongue (Sylvester), and compression of the chest (Laborde)

1930-1960s

Oxygen administration to the oral cavity through a rubber catheter

1930s-1940s

Tight-fitting tracheal tube and direct tracheal oxygen administration

1913-1920s

Byrd-Dew method: immersion in warm water, with alternate flexing and extending of the pelvis to help the “lungs open”

1850-1930s

Dilation of the rectum

1930-1950s

Inhalation of oxygen and 7% CO2 mixture (for morphine-induced narcosis)

1940-1950s

Positive-pressure air-lock (Bloxsom method)

1940 to late 1950s

Concept that “air in the digestive tract is good for survival” is promoted— administration of oxygen to the stomach

1950 to late 1960s

Hyperbaric oxygen in Vickers pressure chamber

1950-1960s

Mouth-to-mouth or mouth-to-endotracheal tube breathing

1930s-1980s

Bell develops a negative-pressure jacket

1930-1950

Negative-pressure ventilators and iron lungs, used rarely in infants

1960s

Positive-pressure respirators used for prolonged ventilatory support

1971

Continuous positive airway pressure introduced for use in newborns

1973

Intermittent mandatory ventilator

1970-1980s

High-frequency ventilators; continuous monitoring of pulmonary function

1903

Hochheim reports “hyaline membranes” noted in the lungs of infants with RDS

1940-1950s

Clinical descriptions and pathology studied

1955-1956

Pattle discovers surfactant in pulmonary edema foam and lung extracts

1959

Avery and Mead show absence of surfactant in infants with hyaline membrane disease4

1971

Gluck introduces lecithin/sphingomyelin ratio

1973

Liggins suggests that antenatal steroids help mature the pulmonary surfactant system

1980

First effective clinical trial of postnatal surfactant therapy (bovine, Fujiwara)

1989-1991

Commercial surfactants become available

1995

Widespread antenatal steroid use leads to declines in rates for RDS and improves survival rates for infants with birthweight ,1000 g, heralding a new era of epidemics of bronchopulmonary dysplasia and retinopathy of prematurity

1950 and beyond

Era of controlled clinical trials in neonatal medicine begins

1970s

Regionalization of neonatal perinatal care

1990s

Evidence-based medicine, systematic reviews

Assisted ventilation

Surfactant

Education research and patient care

Large perinatal networks for research (self-funded and federally funded) 2005

Hypothermia for perinatal hypoxic-ischemic encephalopathy Continued

14

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 1–3  Ventilatory Care, Respiratory Disorders, and Intensive Care—cont’d Category

Approximate Time Span

Procedures and Techniques

Border of viability debates

2005

Improved survival rates for infants between 22 and 26 weeks’ gestation raise questions about the definition of border of viability, and ethics of intensive care for such infants. Debates and dilemma continue

RDS, respiratory distress syndrome. See references 5, 18, 25, 46, 49, 58, 61, 88 for primary citations.

PEDIATRIC SURGERY—NOT FOR RABBITS ANYMORE As the trend of specialization among surgical subspecialties became popular, generalist surgeons resisted the change. Churchill, a famous surgeon, once remarked that his surgical residents at Massachusetts General Hospital “were quite proficient at operating on rabbits,” and there was no need for a subspecialty in pediatric surgery.35 Despite those objections, Harvard Medical School founded the first department of pediatric surgery in 1941 and named Ladd as its chair. Ladd and his pupils (among others) went on to show that pediatric and neonatal surgery was not the same as operating on rabbits.

GLOBAL NEONATAL CARE By the middle of the 20th century, scores of neonatal units based on the Hess model were built in many European countries (Fig. 1-12) and the United Kingdom.33,61 During the final decades of the 20th century, in many Asian countries, indigenous instruments and devices were being developed to improve neonatal resuscitation and intensive care for sick newborns. The impact of these developments on global neonatal and infant outcomes has yet to be assessed.

Figure 1–12.  The first preterm infant unit in Athens, Greece, using incubators with oxygen flowing into them (circa 1947). (Courtesy of John Sofatzis, MD, Athens, Greece.)

CONTROLLED CLINICAL TRIALS, EVIDENCE-BASED MEDICINE, AND RESEARCH NETWORKS The modern era of controlled clinical trials evolved in the second half of the 20th century. The first randomized controlled clinical trial (RCT) in the United States was on a neonatal topic.74,75 In 1949, in their newly established “infant station” at the Babies Hospital, New York, Silverman and colleagues saw an infant with a severe stage of ROP. This 1200-g infant was the son of a biochemistry professor. Desperate to do “anything” to prevent the infant becoming totally blind, Silverman’s group decided to administer adrenocorticotropic hormone (ACTH). The infant made a dramatic recovery. This gave them the inspiration to “try” to design a placebocontrolled randomized trial using ACTH for ROP. The results were disappointing. In addition to causing steroid-related side effects, ACTH was no more effective than the placebo in reducing ROP severity. The experience of organizing this trial led Silverman to develop additional RCTs in neonatal medicine, however, the most famous of which is the multicenter trial of curtailed or liberal use of oxygen to prevent ROP.74,75 With RCT as the backbone, advances in statistical methods and computer sciences gave rise to the science of systematic analyses, culminating in the founding of Cochrane Collaboration in 1993, named after the British epidemiologist Cochrane. The Collaboration produces and disseminates systematic reviews on therapies in all of medicine. It has more recently added assessment of accuracy of test results to its review topics. Among the 51 Cochrane review groups, the Pregnancy and Childbirth and the Neonatal Review groups have produced the largest number of systematic reviews. Another welcome trend of the 1980s was the development of cooperative clinical and research networks. Examples of these include the self-funded Vermont Oxford Network and the federally funded Maternal Fetal Medicine Units and the Neonatal Research Network. Such cooperative networks have also been organized in the United Kingdom and European countries. These networks have conducted many highly successful collaborative clinical trials. With common protocols and shared resources, they have managed to enroll and study large numbers of research subjects within short time spans. The generic database maintained in these networks has become an invaluable resource for observational research. From the perspective of a historian, however, the long-term impact of these developments on the neonatal practice at the community (or practitioner) level and their overall impact on neonatal outcome need to be studied systematically.

Chapter 1  From Infant Hatcheries to Intensive Care

SOME FAMOUS HIGH-RISK INFANTS Shakespeare’s King Henry VI offers one of the most poignant musings on the burdens of disability and the difficult birth (owing to footling presentation) of his brother, the Duke of Gloucester, who later became Richard III. Henry says to the Duke,54 who was supposedly born premature (not confirmed by other historians), “Thy mother felt more than a mother’s pain, yet brought forth less than a mother’s hope.”

15

King Richard himself in a different Shakespeare play bemoans his misfortune54: “Deformed, unfinish’d, sent before my times/Into this breathing world scarce half made up.” Did King Richard have hemiplegic cerebral palsy as a consequence of prematurity? We cannot be sure. The list of leaders, celebrities, and famous individuals supposed to have been regarded as being at high risk at birth (Table 1-4) is impressive,64 although the authenticity of the stories is difficult to confirm because most of them were derived from anecdotal statements.

TABLE 1–4  Ominous Beginnings for Some Famous Personalities Category

Name

Description*

Religious

Moses

Jewish tradition holds that Moses was born “6 months and 1 day” after he was conceived; thus he could be hidden for 3 months from Pharoah’s soldiers who were looking to find and kill the liberator of Jews50,60

Historical personalities and characters

Duke of Gloucester (later Richard III) (1452-1485)

Footling presentation, possibly premature; might have had cerebral palsy (hemiplegia?)54

Macduff (Scottish nobleman in Shakespeare’s Macbeth)

Delivered by cesarean section after his mother’s death—”not of a woman born,” but of a corpse54

Jonathan Swift (1667-1745)

Mentioned by Cone24

Licetus Fortunio (1577-1657)

“A fetus no more than five and one-half inches” at birth

Pablo Picasso (1881-1973)

Left on the table as a stillbirth; his uncle, Don Salvador, a physician, resuscitated him

Voltaire (1694-1778)

Premature and asphyxiated, the “puny little boy” was not expected to live and was hurriedly baptized; he was raised in the attic to keep him warm

Samuel Johnson (1709-1784)

A huge baby, he was “strangely inert” at birth, required slapping and shaking. With persuasion, he made a few whimpers and lived

Johann Wolfgang von Goethe (1749-1823)

After 3 days of labor, his mother delivered him; he was “lifeless and miserable” and thought to be stillborn at birth

Anna Pavlova (1882-1931)

“A premature, so puny and weak,” she was wrapped in cotton wool for 3 months

Thomas Hardy (1840-1928)

He was thrown aside as dead at birth. “A good slapping” from the midwife revived him

Sidney Poitier (born 1927)

Being 3 months premature, he was so small that his father “could place him in a shoebox.” His grandmother said that despite prematurity, he would “walk with the kings.” He did, when he became Bahamian ambassador to Japan in the 1990s†

Johannes Kepler (1571-1630)

A “seven-month” baby; estimated IQ, 161

Christopher Wren (1632-1723)

Mentioned by Cone24

Isaac Newton (1642-1727)

Thought to be “as good as dead” at birth. He was such a “tiny mite” that he could be placed in a quart mug

Franklin D. Roosevelt (1882-1945)

Weighed 10 lb at birth, but was “blue and limp with a deathlike respiratory standstill” from too much chloroform given to his mother, Sara Roosevelt

Winston Churchill (1874-1965)

His early birth “upset the ball.” Later a duchess remarked that the baby had such a lusty “earth-shaking” cry as she had ever heard. Recent historians doubt his premature birth

Artists and writers

Scientists

Politicians

*The biographical notes are derived mostly from anecdotal statements of historians or family members or are from later recollection by the individuals themselves; we cannot be certain of the scientific validity of these stories. †  Quoted by Sidney Poitier in the television show Biography, CNN, Spring 2000.

16

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

FUTURE OF NEONATAL RESEARCH, EDUCATION, AND DATABASES IN THE INTERNET ERA Of all the advances in the 20th century, none has made a greater impact on our lives than computers, the Internet, and information technology. The Internet and information technology opened new avenues for dissemination of research information and helped the development of clinical research databases. These resources need to give way for creative means pooling our collective experiences in patient care in a prospective manner. Although today’s NICU is a technological marvel, conceptually it remains a miniaturized version of the adult intensive care unit. Knowledge about the environmental influences on growth and development may change the shape of future NICUs. While the immediate and long-term adverse effects of excessive noise, light, handling, and pain on infant growth are being studied, attempts are being made to transform the impersonal intensive care environment into an infant-friendly experience. Many ultramodern NICUs have now been built on the concepts of environmental care. Such “kind and gentle” NICUs may be what Tarnier, Budin, Hess, and others conceived of some 100 years ago. Despite incredible advances in the care of premature infants, today’s scientists are facing many unresolved issues, including limits of viability, cost of care, quality of life for intensive care “graduates” (into their adult lives), and an ever-increasing battle against opportunistic nosocomial microorganisms.74,75 A major concern is the definition of border of viability and the ethics of providing intensive or palliative care for infants born between 22 and 25 weeks’ gestation. To the long list of new problems, one might add the growing realization that many adult-onset disorders may have developmental origins. These and similar concerns also vexed the early pioneers of our subspecialty. Future historians may assess this century of neonatal medicine with the same sense of surprised wonder and awe that we now feel when remembering the days of infant hatcheries and baby incubator shows.

ACKNOWLEDGMENTS I sincerely thank Caroline Signore, MD, for helping during the preparation of the manuscript, and I thank all of the copyright holders for permission to reproduce the illustrations used in this chapter.

REFERENCES 1. Akerrén Y, Fürstenberg N: Gastrointestinal administration of oxygen in treatment of asphyxia in the newborn, J Obstet Gynaecol Br Emp 57:705, 1950. 2. Als H et al: Individualized developmental care for the very low-birth-weight preterm infant: medical and neurofunctional effects, JAMA 272:853, 1994. 3. Apgar V: A proposal for a new method of evaluation of the newborn infant, Anesth Analg 32:260, 1953. 4. Avery ME, Mead J: Surface properties in relation to atelectasis and hyaline membrane disease, Am J Dis Child 97:517, 1959.

5. Avery ME: Surfactant deficiency in hyaline membrane disease: the story of discovery, Am J Respir Crit Care Med 161:1074, 2000. 6. Baker JP: The incubator controversy: pediatricians and the origins of premature infant technology in the United States, 1890-1910, Pediatrics 87:654, 1991. 7. Baker JP: The machine in the nursery: incubator technology and the origins of newborn intensive care, Baltimore, Johns Hopkins University Press, 1996. 8. Barcroft J: Research on pre-natal life (vol 1), Oxford, Blackwell Scientific, 1977. 9. Blackfan KD: Apparatus for collecting infant’s blood for Wassermann reaction, Am J Dis Child 4:33, 1912. 10. Blackfan KD, Maxcy KF: The intraperitoneal injection of saline solution, Am J Dis Child 15:19, 1918. 11. Blumfeld-Kusinski R: Not of a woman born: representation of childbirth in medieval and renaissance culture, Ithaca, NY, Cornell University Press, 1990. 12. Bloxsom A: Resuscitation of the newborn infant: use of positive pressure oxygen-air lock, J Pediatr 37:311, 1950. 13. Blundell J: Principles and practice of obstetrics, London, E. Cox, 1834, p 246. 14. Bonnaire E: Inhalations of oxygen in the newborn, Arch Pediatr 8:769, 1891. 15. Budin P, Maloney WJ (translator): The nursling: the feeding and hygiene of premature and full-term infants, London, Caxton, 1907. 16. Butterfield LJ: Virginia Apgar, MD, MPhH, Neonatal Netw 13:81, 1994. 17. Campbell DJ et al: High maternal mortality in certain areas: reports on public health and medical subjects. London, Ministry of Health and Department of Health Publications, No. 68, 1932. 18. Clements JA, Avery ME: Lung surfactant and neonatal respiratory distress syndrome, Am J Respir Crit Care Med 157:S59, 1998. 19. Comroe JH Jr: Retrospectroscope: man-cans, Am Rev Resp Dis 116:945, 1977. 20. Comroe JH Jr: Man-cans (conclusion), Am Rev Resp Dis 116:1011, 1977. 21. Cone TE Jr: 200 Years of feeding infants in America, Columbus, OH, Ross Laboratories, 1976. 22. Cone TE Jr: History of American pediatrics, Boston, Little Brown, 1979, p 57. 23. Cone TE Jr: The first published report of an incubator for use in the care of the premature infants (1857), Am J Dis Child 135:658, 1981. 24. Cone TE Jr: Perspective in neonatology. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: neonatal medicine (vol 1), Evansville, IN, Ross Publication, 1984, p 9. 25. DeBard ML: The history of cardiopulmonary resuscitation, Ann Emerg Med 9:273, 1980. 26. DeLee J: A brief history of the Chicago Lying-In Hospital, Chicago, Alumni Association Lying-In Hospital and Dispensary Souvenir, 1895, p 1931. 27. DeLee JB: Infant incubation, with the presentation of a new incubator and a description of the system at the Chicago Lying-in Hospital, Chicago Medical Recorder 22:22, 1902. 28. Diamond LK et al: Erythroblastosis fetalis and its association with universal edema of the fetus, icterus gravis neonatorum, and anemia of the newborn, J Pediatr 1:269, 1932.

Chapter 1  From Infant Hatcheries to Intensive Care 29. Diamond LK et al: Erythroblastosis fetalis, VII: treatment with exchange transfusion, N Engl J Med 244:39, 1951. 30. Donald I, Lord J: Augmented respiration: studies in atelectasis neonatorum, Lancet 1:9, 1953. 31. Duncan RG: Neonatology on the web (website). http://neonatology. org/tour/history.html. Date accessed February 1, 2010. 32. Dunn PM: Dr. James Blundell (1790-1878) and neonatal resuscitation, Arch Dis Child Fetal Neonatal Ed 64:494, 1988. 33. Dunn PM: The development of newborn care in the UK since 1930, J Perinatol 18:471, 1998. 34. Dunn PM: Sir Joseph Barcroft of Cambridge (1872-1947) and prenatal research, Arch Dis Child Fetal Neonatal Ed 82:F75, 2000. 35. Easterbrook G: Surgeon Koop, Knoxville, TN, Whittle Direct Books, 1991. 36. Editorial. The Victorian Era exhibition at Earl’s Court, Lancet 2:161, 1897. 37. Editorial. The danger of making a public show of incubator for babies, Lancet 1:390, 1898. 38. Gartner LM: Dr. Harry Gordon. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: neonatal medicine (vol 1), Evansville, IN, Ross Publication, 1984, p xiv. 39. Gilmartin ME: Body ventilators: equipment and techniques, Respir Care Clin North Am 2:195, 1996. 40. Ginzberg L: The legend of the Jews, II: Bible times and characters from Joseph to the Exodus. (Translated from the German manuscript by Henrietta Szold.) Philadelphia, Jewish Publication Society of America, 1989, p 262. 41. Goodlin R: History of fetal monitoring, Am J Obstet Gynecol 133:323, 1979. 42. Gould D: Iron lung. In DeBono E, editor: Eureka! An illustrated history of invention from the wheel to the computer, New York, Holt, Rinehart & Winston, 1974, p 160. 43. Graham H: Eternal Eve: the history of gynecology and obstetrics, Garden City, NY, Doubleday, 1951. 44. Gregory GA et al: Treatment of idiopathic respiratory distress syndrome with continuous positive pressure, N Engl J Med 284:1333, 1971. 45. Henderson AR: Resuscitation experiments and breathing apparatus of Alexander Graham Bell, Chest 62:311, 1972. 46. Henderson Y: The inhalation method of resuscitation from asphyxia of the newborn, Am J Obstet Gynecol 21:542, 1931. 47. Hess JH: Premature and congenitally diseased infants, Philadelphia, Lea & Febiger, 1922. 48. Holt LE: The diseases of infants and children, New York, Appleton, 1897, p 12. 49. Hutchison JH et al: Controlled trials of hyperbaric oxygen and tracheal intubation in asphyxia neonatorum, Lancet 1:935, 1966. 50. Isaiah AB, Sharfman B: The Pentateuch and Rashi’s commentary: a linear translation into English, Exodus. New York, S. S. & R., 1960, p 9. 51. Jacobson RM, Feinstein AR: Oxygen as a cause of blindness in premature infants: “autopsy” of a decade of errors in clinical epidemiologic research, J Clin Epidemiol 11:1265, 1992. 52. James LS et al: Intragastric oxygen and resuscitation of the newborn, Acta Pediatr 52:245, 1963. 53. James LS, Lanman JT: History of oxygen therapy and retrolental fibroplasia, Pediatrics 57(suppl):59, 1976. 54. Kail AC: The medical mind of Shakespeare, Balgowhas, Australia, Williams & Wilkins, 1986, p 101.

17

55. Lambert SW: Melaena neonatorum with report of a case cured by transfusion, Med Rec 73:885, 1908. 56. Loudon I: Deaths in childbed from the 18th century to 1935, Med Hist 30:1, 1986. 57. Morton LT, Moore RJ: A chronology of the diseases and related sciences, Aldershot, UK, Scolar Press, 1997, p 84. 58. Murphy D et al: The Drinker respirator treatment of the immediate asphyxia of the newborn: with a report of 350 cases, Am J Obstet Gynecol 21:528, 1931. 59. National Library of Medicine (website). http://www.ncbi.nlm. nih.gov/sites/entrez?otool5nihlib. Accessed December 28, 2008. 60. Nelson NM: An appreciation of Clement Smith. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: neonatal medicine (vol 1), Evansville, IN, Ross Publication, 1984, p xii. 61. O’Dowd MJ, Phillipp AE: The history of obstetrics and gynecology, New York, Parthenon, 1994. 62. Papadopoulos MD, Swyer PR: Assisted ventilation in terminal hyaline membrane disease, Arch Dis Child 39:481, 1964. 63. Radcliffe W: The secret instrument, London, William Heinemann Medical Books, 1947. 64. Raju TNK: Some famous “high-risk” newborn babies. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: perinatal medicine (vol 2), Evansville, IN, Ross Publication, 1984, p 187. 65. Raju TNK: The principles of life: highlights from the history of pulmonary physiology. In Donn SM, editor: Neonatal and pediatric pulmonary graphics: principles and clinical applications, Armonk, NY, Futura, 1998, p 3. 66. Raju TNK: The history of neonatal respiration: tales of heroism and desperation, Clin Perinatol 26:629, 1999. 67. Rambar AC: Julius Hess, MD. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: perinatal medicine (vol 2), Evansville, IN, Ross Publication, 1984, p 161. 68. Ransom SW: The care of premature and feeble infants, Pediatrics 9:322, 1890. 69. Roche T: A refuge for throwaways: the spate of “dumpster babies” stirs a movement to provide a safe space for unwanted newborns, Time 155:50, 2000. 70. Royce S et al: Indwelling polyethylene nasogastric tube for feeding premature infants, Pediatrics 8:79, 1951. 71. Schaffer AJ: Diseases of the newborn, Philadelphia, Saunders, 1960, p 1. 72. Sidbury JB: Transfusion through the umbilical vein in hemorrhage of the newborn, Am J Dis Child 25:290, 1923. 73. Silverman WA: Incubator-baby side shows, Pediatrics 64:127, 1979. 74. Silverman WA: Retrolental fibroplasia: a modern parable, New York, Grune & Stratton, 1980. 75. Silverman WA: Personal reflections on lessons learned from randomized trials involving newborn infants from 1951 to 1967, Controlled Clin Trials 2004; 1:179-184 76. Smith CA: Physiology of the newborn infant, Springfield, Charles C. Thomas, 1945. 77. Spaulding M, Welch P: Nurturing yesterday’s child: a portrayal of the Drake collection of pediatric history, Philadelphia, Decker, 1991, p 110. 78. Speert H: Obstetrics and gynecology in America: a history, Baltimore, Waverly Press, 1980.

18

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

79. Stalhman MT: Assisted ventilation in newborn infants. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: perinatal medicine (vol 2), Evansville, IN, Ross Publication, 1984, p 21. 80. Stern L et al: Negative pressure artificial respiration: use in treatment of respiratory failure of the newborn, Can Med Assoc J 102:595, 1970. 81. Stern L: Thermoregulation in the newborn: historical, physiological, and clinical considerations. In Smith GF, Vidyasagar D, editors: Historical review and recent advances in neonatal and perinatal medicine: neonatal medicine (vol 1), Evansville, IN, Ross Publication, 1984, p 35. 82. Sterne L: The life and opinions of Tristram Shandy, gentleman, New York, Penguin, 1997, p 231. 83. Tempkin O: On the care of the newborn. In: Soranus’ gynecology, Baltimore, Johns Hopkins Press, 1956, p 79.

84. Tuchman BU: A distant mirror: the calamitous 14th century, New York, Ballantine Books, 1978, p 49. 85. Wijesundera CD: Digital intubation of the trachea, Ceylon Med J 35:81, 1990. 86. Wilson A: The making of man-midwifery: childbearing in England 1660-1770, Cambridge, MA, Harvard University Press, 1995, p 1. 87. Winthrobe MM: Blood: pure and eloquent, New York, McGraw-Hill, 1981. 88. Wrigley M, Nandi P: The Sparklet carbon dioxide resuscitator, Anaesthesia 49:148, 1994. 89. Young JH: Caesarean section: the history and development of the operation from earliest times, London, HK Lewis, 1994. 90. Zimerman DA: Rh, New York, MacMillan, 1973. 91. Zimmerman JJ, Strauss RH: History and current application of intravenous therapy in children, Pediatr Emerg Care 5:120, 1989.

CHAPTER

2

Epidemiology and Perinatal Services

PART 1

Epidemiology Michele C. Walsh and Avroy A. Fanaroff

OVERVIEW The neonatal-perinatal period is a time when the mother and fetus experience a period of rapid growth and development. At birth, the fetus makes an abrupt transition from the protective environment of the uterus to the outside world; the newborn infant must undergo extreme physiologic changes to survive this transition. The highest risk of infant death occurs during the first 24 hours after birth. Increased rates of mortality and morbidity continue during the neonatal period—from birth to the 28th day of life. Mortality and morbidity rates are high early in life because the fetus and infant are vulnerable to numerous metabolic, genetic, physiologic, social, economic, and environmental injuries. These factors influence the gestation, delivery, and neonatal period, and have a major impact on the health of the fetus and infant. The high incidence of mortality and morbidity during the perinatal period, which starts at the 28th week of pregnancy and extends to the 28th day after birth, makes it important to identify as early as possible mothers, fetuses, and infants who are at greatest risk. Of equal importance is the need to reduce the risk of morbidity, especially for handicapping conditions such as mental retardation. There is increasing evidence that early recognition of women with high-risk pregnancies and of high-risk infants followed by appropriate prenatal, intrapartum, and postpartum care can reduce the incidence of handicapping conditions and reduce the incidence of infant mortality.

Death of an infant is a source of anguish and grief to the parents and relatives. Infants who survive with disabilities and disease must endure personal suffering and may be a continuing source of pain, anguish, and loss of resources for their parents and society.3,7,8 They may also impose a biologic burden on future generations by increasing the frequency of maladaptive genes in the population. In addition to the human tragedy, the fiscal impact of these problems on society is estimated to be in the billions of dollars each year. Infants who are born before term gestation have the greatest risk of dying during infancy and of morbidity during childhood.5,6 Infants with low birthweight (LBW) (,2500 g) are 40 times more likely to die than infants with normal birthweight, and infants with very low birthweight (VLBW) (,1500 g) are 200 times more likely to die. Infants with LBW are at a much higher risk of being born with cerebral palsy, mental retardation, and other sensory and cognitive impairments compared with infants with normal birthweight. Surviving infants with LBW also have an increased incidence of disability for a broad range of conditions, including various neurodevelopmental handicaps, respiratory illness, and injuries acquired as a result of neonatal intensive care. These infants often have a diminished ability to adapt socially, psychologically, and physically to an increasingly complex environment. Box 2-1 lists the risk factors for high-risk pregnancies that are often associated with LBW. Important antecedents of many adult diseases, such as coronary artery disease, chronic renal and liver disease, and obesity, may have roots in early childhood, which implies that there may be very early opportunities for the prevention of adult chronic diseases.2 Further improvement in longevity and decreased morbidity are likely to result from a better understanding of the origins of adult disease in fetal life and infancy and from the prevention and early treatment of these diseases.

19

20

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

BOX 2–1 Factors Associated with High-Risk Pregnancy ECONOMIC Poverty Unemployment No health insurance; insufficient health insurance Poor access to prenatal care CULTURAL-BEHAVIORAL Low educational status Poor health care attitudes No or inadequate prenatal care Cigarette, alcohol, drug abuse Age ,16 or .35 years Unmarried Short time between pregnancies Lack of support group (husband, family, religion) Stress (physical, psychological) African-American race Abusive partner BIOLOGIC-GENETIC Previous infant with low birthweight Low maternal weight at mother’s birth Low weight for height Poor weight gain during pregnancy Short stature Poor nutrition Inbreeding (autosomal recessive?) Intergenerational effects Hereditary diseases (inborn error of metabolism) REPRODUCTIVE Previous cesarean section Previous infertility

Prolonged gestation Prolonged labor Previous infant with cerebral palsy, mental retardation, birth trauma, congenital anomalies Abnormal lie (breech) Abruption Multiple gestation Premature rupture of membranes Infections (systemic, amniotic, extra-amniotic, cervical) Preeclampsia or eclampsia Uterine bleeding (abruptio placentae, placenta previa) Parity (0 or .5) Uterine or cervical anomalies Fetal disease Abnormal fetal growth Idiopathic premature labor Iatrogenic prematurity High or low levels of maternal serum a-fetoprotein MEDICAL Diabetes mellitus Hypertension Congenital heart disease Autoimmune disease Sickle cell anemia TORCH infection Intercurrent surgery or trauma Sexually transmitted diseases Maternal hypercoagulable states

TORCH, toxoplasmosis, other agents, rubella, cytomegalovirus, herpes simplex. From Stoll JB, Kliegman RM: Section 1: noninfectious disorders. In Behrman RE et al, editors: Nelson textbook of pediatrics, 16th ed, Philadelphia, 2000, Saunders, p 461.

To decrease infant morbidity and mortality, pregnant women and infants at high risk should be identified as early as possible. Most high-risk infants can be identified before birth (see Box 2-1). Others are identified in the delivery room by abnormal growth status—small size for gestational age (weight ,3% for gestational age) or large size for gestational age (weight .90% for gestational age)—or congenital malformation. Examination of a fresh placenta, cord, and membranes can alert the physician to a newborn at high risk.

50 Deaths per 1000 live births

HIGH-RISK INFANTS

40 Infant

30

Neonatal

20

Postneonatal

10 0

Infant Mortality Infant mortality is a critical measure of the health and welfare of a population.5 In 2006, 4271 million infants were born in the United States, and 87,600 died before reaching age 1 year, resulting in an infant mortality rate of 6.7 deaths per 1000 live births. Rates of infant death in the United States have been declining steadily for at least 40 years and reached an all-time low in 2002 (Fig. 2-1).5 Despite the constant improvement in national infant mortality rates, the United States ranks only

1940

1950

1960

1970

1980

1990

2000 2007

NOTE: Rates are infant (under 1 year), neonatal (under 28 days), and postneonatal (28 days–11 months) deaths per 1000 live births in specified group.

Figure 2–1.  Infant mortality per 1000 live births in the

United States, 1940-2005. (From Kung HC et al: Deaths: final data for 2005, Natl Vital Stat Rep 56:1-120, 2008, and from Jiaquan XU et al: Deaths: Preliminary Data for 2007, Natl Vital Stat Rep 58:22, 2009.)

Chapter 2  Epidemiology and Perinatal Services

25th in the world in infant mortality, well behind Sweden, Japan, Singapore, and Hong Kong.5 Paradoxically, the birthweight-specific mortality in the United States is relatively low. That is, at each birthweight level, the infant mortality in the United States is very low. This low rate of birthweight-specific mortality is due to advances in neonatal care systems; most extremely tiny infants who may weigh 750 g at birth are now surviving.1 Most notable in the United States is the large disparity between African-American and white infant mortality. The mortality rate for African-American infants is more than double that for white infants (Fig. 2-2). In recent years, this ethnic and racial disparity has widened because the rate of decline in infant mortality has been higher among white infants than among African-American infants. Approximately half of all infant deaths were due to one of the four leading causes of infant death in 2005: congenital anomalies, short gestation and LBW, sudden infant death syndrome, and maternal complications of pregnancy.4 Of the four leading causes of infant deaths, African-American infants are much more likely to die from being born too soon or too small and from maternal complications of pregnancy. National policymakers have developed the following strategies for eliminating the health disparity. Infant deaths are divided into two categories according to age: neonatal (deaths of infants ,28 days old) and postneonatal (deaths of infants between the ages of 28 days and 1 year). A decline in infant mortality rates was observed for neonatal deaths (4.8 per 1000) and postneonatal deaths (2.4 per 1000). Infant mortality numbers have declined by more than 40% since 1980. Neonatal death rates declined more steeply in the 1980s, and postneonatal death rates declined more steeply in the 1990s. Neonatal deaths are generally attributable to factors that occur during pregnancy, such as congenital malformations, low birth weight, maternal toxic exposures (smoking or other forms of drug abuse), and lack of appropriate medical care. In contrast, postneonatal deaths are generally associated with the infant’s environmental circumstances, such as poverty, which

Rate

10

14.65 13.61

1995 2003

Infant Morbidity PRETERM INFANTS, INFANTS WITH LOW BIRTHWEIGHT, AND INFANTS WHO ARE SMALL FOR GESTATIONAL AGE Live-born infants born before 37 completed weeks of gestation (,259 days after the date of the mother’s last menstrual period) are defined as preterm (Fig. 2-3). Measures of live-born infant size include LBW (infants weighing ,2500 g) and two subgroups of LBW, moderately LBW (infants weighing 1500 to 2499 g) and VLBW (infants weighing ,1500 g). Other measures take into consideration gestational age and birthweight, such as small for gestational age, defined as live-born infants weighing less than the 10th percentile for gestational age; appropriate for gestational age, defined as infants weighing between the 10th and the 90th percentiles for gestational age; and large for gestational age, defined as infants weighing more than the 90th percentile for gestational age. 13

9.04 8.73 7.57

6.84

6.28

5.70

5

6.27

5.64 5.27

4.83

0 Non- American Hispanic Indian/ black Alaska Native

often results in inadequate food, housing, sanitation, and medical care. The decline in neonatal deaths among infants with LBW in the 1990s may be because of increased survival in neonatal intensive care units (NICUs), healthier infants with LBW, or both. It is estimated that two thirds of the decline in severity-adjusted mortality was due to increased survival in neonatal intensive care, and that one third was due to healthier infants with LBW. Increased survival in neonatal intensive care is attributed to the more aggressive use of respiratory and cardiovascular treatments. The improved health of infants with LBW is attributed to improvements in obstetric and delivery room care. Generally, for any given gestational age, the lower the infant birthweight, the higher the neonatal mortality; for any given birthweight, the younger the gestational age, the higher the neonatal mortality. Infant death rates decrease steeply with increasing infant birthweight. The lowest risk of infant death occurs among infants with birthweights of 3000 to 4000 g, and risk increases slightly for infants weighing more than 4000 g and for infants whose gestational age is older than 42 weeks.6 Infants who survive past the first 28 days of life have a vastly better prognosis.

Total

Non- Hispanic Asian/ Hispanic Pacific white Islander

Maternal race/ethnicity

Preterm births (% of live births)

15

21

12.5 12 11.5 11 10.5 10 9.5 1990 1993 1995 1997 1999 2000 2003 2004 2007

Figure 2–2.  Disparities in infant mortality by race and ethnic-

Figure 2–3.  Preterm births in the United States by year

ity in the United States, 1995 versus 2003. (From MacDorman MF et al: Fetal and perinatal mortality, United States, 2004, Natl Vital Stat Rep 56:1-19, 2007.)

expressed as percentage of total live births.  (Redrawn from National Center for Vital Health Statistics, Rep 56, 2007 and Natl Vital Stat Rep 57:14, 2009.)

22

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

The health of an infant with LBW is directly related to gestational age. An infant weighing 1800 g born at term is very different from an infant weighing 1800 g born at 32 weeks. An infant weighing 1800 g born at term is defined as having LBW and being small for gestational age, but an infant weighing 1800 g born at 32 weeks is defined as having LBW, being preterm, and being appropriate for gestational age. Two infants each weighing 1800 g of unequal gestational ages would require very different care (see Chapter 14). Infants with extremely low birth weight (ELBW) (weighing ,1000 g) have the greatest risk of death and neurodevelopmental impairment. Tyson and colleagues10 at the National Institute of Child Health and Human Development Neonatal Research Network developed an evidence-based calculator to inform physicians and families of the predicted risks for individual infants with ELBW born between 22 and 25 weeks using not only gestational age, but also exposure to antenatal steroids, sex, and multiple births. The calculator can be accessed at http://www. nichd.nih.gov/about/org/cdbpm/pp/prog%5Fepbo/. In the United States today, infants with LBW account for a relatively greater proportion of infant deaths than in the past. Early in the 20th century, two thirds of infant deaths occurred in the postneonatal period, primarily from infectious diseases. By 1950, 7.5% of live-born infants weighed less than 2500 g, and two thirds of all infant deaths occurred in the neonatal period. The causes of these deaths were related to antenatal and intrapartum events, such as birth injury, asphyxia, congenital malformations, and “immaturity.” Advances in the development of perinatal care in the 1950s and 1960s came as a result of the awareness of the increased morbidity in surviving infants with LBW coupled with a greater understanding of fetal and infant nutrition, pharmacology, and pathophysiology. The infant mortality rate decreased 47% from 1965 to 1980, primarily because of the increased survival of high-risk infants with LBW. Regionalization of neonatal intensive care has brought perinatal intensive care to most families in need and has contributed significantly to the increase in the fraction of infants with LBW and VLBW who are cared for in tertiary centers. Decreased rates of neonatal mortality also were observed after the introduction of regionalized care (see Part 2). Despite these impressive gains, the U.S. international ranking in infant mortality has continued to decline compared with other countries.6 In 1960, the United States ranked 12th in infant mortality; by 1990, the United States ranked 23rd; and by 2004, the United States ranked 29th, tied with Poland and Slovakia. In 2005, 36.5% of all infant deaths in the United States were due to a preterm-related cause of death. In addition, there has been an increase in the percentage of infants born preterm. Prematurity constitutes a major health problem. Surviving infants with LBW also remained three times as likely as infants of normal birthweight to have adverse neurologic sequelae. The risk of adverse sequelae increases with decreasing birthweight. Lower respiratory tract problems, particularly infections, are more common, as are complications of neonatal care. The risk of giving birth to an infant with LBW is increased among African-American women, women who smoke during pregnancy, unmarried women, women with low educational attainment, women who have no or inadequate prenatal care,

and women who have had a previous infant with LBW (see Box 2-1). There is a substantial and persistent difference between African-American and white infants in the risk of LBW and preterm delivery. African-American women are 2.4 times as likely to have an infant with LBW than white women and 2.6 times more likely than Chinese-American women. The higher LBW rate among African-American women has been observed for more than 20 years. AfricanAmerican infants are more likely to die of preventable causes than white infants. In addition, African-American infants have significantly higher rates of mortality for every cause of infant death except congenital anomalies and sudden infant death syndrome. Infant mortality rates were also elevated for Puerto Rican and American Indian women. Decades of research about the disparities in LBW rates and infant mortality between African-American and white infants have been unable to explain the racial disparities in birth outcomes. Scientists have studied the impact of education, maternal age, vaginal infection, exposure to cigarette smoke, use of alcohol, stress, socioeconomic status, and many other risk factors. None of these factors explain the racial disparities in death or LBW rates. Compounding the widening gap between African-American and white rates of LBW is the increasing number of women at high risk in the population. Increases during the 1980s in acquired immunodeficiency syndrome, poverty, use of illicit drugs such as crack cocaine, syphilis, and births to unmarried women are additional factors that are associated with further increases in infant mortality and LBW. Promising intervention strategies include modifying behaviors, lifestyles, and conditions that affect birth outcomes, such as smoking, substance abuse, poor nutrition, lack of prenatal care, medical problems, and chronic illness. Conditions that make the uterus unable to retain the fetus, interference with the course of pregnancy, premature separation of the placenta, or a stimulus to produce uterine contractions before term are generally associated with preterm infants with an appropriate weight for gestational age. Medical conditions that interfere with the circulation and efficiency of the placenta, the development or growth of the fetus, or the general health and nutrition of the mother are associated with infants who are small for gestational age.

ALARMING INCREASES IN INFANTS WITH LOW BIRTHWEIGHT In recent years, there has been alarming increase in the frequency of late preterm deliveries (see ch. 34) (defined as infants at 34 to ,37 weeks’ gestation). In part, the increase has been driven by parents and obstetricians who believe that neonatal morbidity at these gestations is equivalent to morbidity at term gestation. This perception is incorrect, however. Infants delivered by elective cesarean section at 38 weeks’ gestation are 1.3 to 2.1 times more likely to have adverse respiratory outcomes and intensive care unit admission, whereas infants delivered at 37 weeks’ gestation are 1.8 to 4.2 times as likely compared with infants delivered at 39 completed weeks of gestation (Fig. 2-4).9 The frequency of delivery by cesarean section has increased in parallel with the increasing LBW deliveries; together, these have ominous implications for the health of these infants and for the cost of health care. The March of Dimes and other national organizations, including the American Academy of Pediatrics, have

Chapter 2  Epidemiology and Perinatal Services 49.1

50

Proportion of deliveries Incidence of adverse outcome

Percent

40 29.5

30

19.5

20 10

23

15.3 11.0

8.0

6.3

10.4

11.3 7.3 3.8

0 37

38

39

40

41

0.9

≥42

Weeks of gestation at delivery

Figure 2–4.  Elective repeat cesarean delivery (N 5 13,258) and the incidence of adverse outcomes (composite of neonatal death and any of several adverse events, including respiratory complications, treated hypoglycemia, newborn sepsis, and admission to the neonatal intensive care unit) according to the number of completed weeks of gestation. (From Tita AT et al: Timing of elective repeat cesarean delivery at term and neonatal outcomes, N Engl J Med 360:111-120 2009.)

begun campaigns to educate the public about these increased risks and to encourage delays in delivery until 39 completed weeks of gestation.

FUTURE PROGRESS Prevention of Infant Mortality and Low Birthweight Although all the precise causes of infant mortality and LBW are unknown, prevention efforts can make substantial progress toward decreasing infant mortality and LBW.9,11 In the United States, much of the progress in reducing infant mortality has been a result of improvements in neonatal and perinatal care. As evidenced by the stable rates of LBW and preterm birth in the United States, there has been little or no progress to date in preventing LBW or preterm birth. Things can be done to prevent LBW, however. If smoking during pregnancy were eliminated, the infant mortality rate would decrease by 10%, and the LBW rate would decrease by 25%. Decreasing the incidence of unplanned pregnancies, decreasing the abuse of alcohol and illicit drugs, increasing enrollment in the Special Supplemental Food Program for Women, Infants, and Children (WIC), and providing prenatal care in a comprehensive and coordinated manner also could reduce infant mortality and LBW. Preventing pediatric mortality and morbidity associated with LBW requires a broad range of activities that involve the family, health care professionals, and community groups. These activities include identifying prepregnancy risks; counseling for risk reduction; implementing school health programs; increasing early, high-quality prenatal care services; expanding the content of prenatal care to meet the variable needs of individual women and selected high-risk groups, and developing, implementing, and supporting a long-term public information program with a few well-chosen messages targeted at LBW risks (e.g., smoking cessation).

Fetal Development, Physiology, and Biochemistry Improvements in our understanding of and ability to measure embryonic and fetal physiologic and biochemical homeostasis are likely to advance the practice of neonatalperinatal medicine. In addition, as we increase understanding of the mechanisms controlling labor, of the protective circulatory adjustments of the fetus to hypoxia, and of pharmacology, we can develop new means to detect high-risk pregnancies successfully, treat the fetus, and prevent preterm labor. Investigation of the cellular and molecular levels of development and pharmacology of the uterus, placenta, and fetus may be crucial to understanding the mechanisms that control labor. Such an understanding would enable physicians to decrease the incidence of preterm infants or infants with LBW. Prenatal diagnosis and biochemical and ultrasound fetal monitoring may be only the first steps toward the development of future treatments. These future treatments may include hybridization of cells in the early blastocyst or embryo to correct inborn errors of metabolism, stimulation of embryogenesis of organs, gene therapy, acceleration of organ maturation, and pharmacologic interventions to ameliorate hypoxic injury to the central nervous system.

Legal and Ethical Issues Advances in the field of neonatal-perinatal medicine have focused attention on numerous legal and ethical issues. There is continuing concern about life-and-death decision making in NICUs. New and complex societal relationships among the physician, patient, family, and nursing staff exist in these units, and this development has had an enormous impact on the process of making medical decisions. As a result of regionalization, we now have nearly universal access to care in NICUs, and this broad access has brought these ethical issues into a sharper and more demanding focus.

24

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Specific criteria must now be formulated for making decisions that previously were made on the basis of access to the health system, which was strongly influenced by the economic position of a family. Should a 500-g infant with a poor prognosis for intact survival be accepted to the NICU when it means one cannot accept a 2000-g infant who is at high risk but has a good prognosis if he or she survives, just because the referring physician in the case of the 500-g infant calls first, or because the 500-g infant is born in the same hospital where the intensive care unit is located? Ironically, although not surprisingly, as the technology of care increases in these units, the difficult choices for the physician are not the technical medical decisions, but rather the matters of judgment that require evaluating the complex human interests of the relatives, their friends and advisors, and the staff, and the various consequences for the people involved. These decisions have always been the most challenging and demanding ones for physicians because they cannot be delegated to others. The new elements are the frequency and complexity of these judgments in regional neonatal intensive care centers (see Chapters 3 and 4).

8. Sparks PJ: One size does not fit all: an examination of low birthweight disparities among a diverse set of racial/ethnic groups, Matern Child Health J 13:769, 2009. 9. Tita AT et al: Timing of elective repeat cesarean delivery at term and neonatal outcomes, N Engl J Med 360:111, 2009. 10. Tyson JE et al: Intensive care for extreme prematurity: moving beyond gestational age, N Engl J Med 358:1672, 2008. 11. US Department of Health and Human Services: Caring for our future: the content of prenatal care, Washington, DC: US Department of Health and Human Services, Public Health Service, Panel on the Content of Prenatal Care, 1988.

Quality of Care

HISTORICAL PERSPECTIVE

Whatever decision-making process is used to improve the quality of care, certain principles are important, but often not easy to apply. The fundamental responsibility of all concerned is to do no harm or, at least, no harm without a reasonable expectation of a compensating benefit for the patient. A corollary principle is that there must be continuous, scientific evaluation of the care provided and of proposed innovations. Clinicians should not initiate or continue activities that on balance do harm to the well-being of a newborn infant. The definition of well-being is a major problem because the varying ethical values, religious commitments, and life experiences of all the individuals who care for and about the infants and legal restraints must be taken into consideration. Generally, the minimal elements of well-being include a life prolonged beyond infancy, without excruciating pain and with the potential of participating in human experience to at least a minimal degree (see Chapter 5).

Before 1940, perinatal care services were delivered in the United States, Canada, and Europe without any particular organization. Most of the care was provided by an individual physician or midwife. In many areas, most deliveries occurred in the home. Larger urban areas often had numerous maternity hospitals, usually serving as teaching hospitals, with home delivery services and neighborhood clinics serving a geographic area. During the 1940s and early 1950s, many cities in the United States developed centers for the care of premature infants. Many European countries, particularly The Netherlands and the Scandinavian countries, developed systems of care for perinatal patients based on the development of primary prenatal care clinics staffed largely by midwives, with district and regional hospitals for the care of mothers with complications. From 1964 to 1968, studies were undertaken in Massachusetts, Wisconsin, and Arizona to analyze the causes of neonatal mortality and morbidity.4,8 These studies suggested that expert care of high-risk neonates could reduce infant mortality rates. In 1976, the March of Dimes Committee on Perinatal Health developed recommendations based on this research that supported a network of perinatal care providers that supplied care to a geographic region. The report, entitled “Toward Improving the Outcome of Pregnancy” (TIOP), spurred many states to regionalize care.31 By the end of the 1980s, 26 states had regionalized perinatal care, and studies documenting a shift in deliveries from level I to level II and level III centers emerged, together with data documenting reductions in neonatal mortality. Despite these data, in the early 1990s, state by state, the regionalized systems were weakened as factors put pressure on the system (Table 2-1). In 1993, the March of Dimes convened the second TIOP conference and reexamined the theory supporting regionalization.7 The conference participants concluded that regionalization continued to be the best option for reducing perinatal morbidity.

REFERENCES 1. Fanaroff AA et al: Trends in neonatal morbidity and mortality for very low birthweight infants, Am J Obstet Gynecol 196:147.e1, 2007. 2. Feigin RT: Future of child health through research, JAMA 294:1373, 2005. 3. Hack M et al: Self-perceived health, functioning and well-being of very low birth weight infants at age 20 years, J Pediatr 151:635, 2007. 4. Institute of Medicine: Prenatal care: reaching mothers, reaching infants, Washington, DC. National Academy Press, 1988. 5. Kung HC et al: Deaths: final data for 2005, Natl Vital Stat Rep 56:1-120, 2008. 6. Matthews TJ, MacDorman MF: Infant mortality statistics from the 2005 period linked birth/infant death data set, Natl Vital Stat Rep 57: 1-20, 2008. 7. Moster D et al: Long-term medical and social consequences of preterm birth, N Engl J Med 359:262, 2008.

PART 2

Perinatal Services Michele C. Walsh and Avroy A. Fanaroff

Chapter 2  Epidemiology and Perinatal Services

25

TABLE 2–1  Factors Threatening Regionalized Perinatal Care Physician Related

Institution Related

Oversupply and poor distribution of physicians

Use of high-tech procedures as a marketing tool

Desire to perform all services for which they were trained, regardless of level of care designated

Desire to perform all services to allow contractual arrangements with provider

Lack of incentive to improve care at other hospitals and potential decrease in patient referrals

Lack of incentive to improve care at other hospitals and potential decrease in patient referrals

Fear of loss of patients

Institutional ego

Failure to recognize problems warranting referral Modified from Institute of Medicine: Prenatal care: reaching mothers, reaching infants, Washington, DC, 1988, National Academy Press.

PRINCIPLES OF REGIONAL CARE General principles that form the basis for the development of regional health care services for perinatal patients include organization by geographic region, accountability for outcomes, and one standard of quality across care sites.1 A region denotes a geographic area or population with definable care needs. The regional center and the network of related institutions are accountable for the overall perinatal health care for the region. Data on mortality and morbidity, frequency of problems, and quality of care are assessed for the entire population in the area. The availability and quality of care in any given institution become the responsibility of all the institutions, including the perinatal center. Regionalization is based on the premise that there should be a single standard of quality across sites of care. Every mother and infant should have equal access to all the components of a functioning perinatal system (Table 2-2). Institutions operating within a region differ in their ability to provide perinatal care. Each institution is expected to deliver highquality care up to the level of its capability. When care requirements exceed the institution’s capability, the patient is referred to the closest facility that has the required capability. Of problems associated with increased risk for the mother, fetus, or newborn, 60% to 80% are detectable sufficiently in advance of the crisis either to permit the appropriate care resources to be made available locally or to transfer the patient to where appropriate resources are available.14 Even under ideal circumstances, some patients must move from one facility to another during the course of care. Institutions within a region must be effectively linked to permit ease of patient movement. Subspecialty centers are charged with outreach education to expand the care of basic and specialty units in their region. Regionalization permits optimal use of facilities and personnel. With a regional perinatal center, sufficient numbers of high-risk mothers and neonates are concentrated in a single location to justify economically the staffing and equipment necessary to meet care needs.44 In addition, personnel who have frequent opportunities to use their skills are able to maintain and enhance these skills. Institutions with smaller numbers of high-risk patients avoid the expense of developing services that are infrequently used. The health care needs of perinatal patients require a close working relationship among the various disciplines involved to prevent fragmentation and gaps in care.

CURRENT RECOMMENDATIONS ON ORGANIZATION Countries with better perinatal mortality and morbidity statistics have universal primary care for all pregnant women and have established mechanisms for referring mothers and infants from primary care institutions to district hospitals when needed. They have placed less emphasis on tertiary care because of the lower demand for such services. In contrast, perinatal health services of the United States and Canada place great emphasis on the specialized hospital services for perinatal patients, but primary care services available to mothers during pregnancy are disorganized and inconsistent. The number of perinatal patients requiring intensive in-hospital care is inversely related to the quality and availability of primary services and to the general health status of mothers before pregnancy. An organized system for perinatal care begins with a wellintegrated ambulatory care system that emphasizes early risk assessment. Prenatal care may be delivered by basic providers (family practice physicians or nurse practitioners), specialty providers (obstetricians or experienced family practitioners), or subspecialty providers (maternal-fetal medicine specialists) as appropriate to the level of risk. As the woman’s risk status changes during pregnancy, so may the level of care needed. No care system is effective if it is not used. TIOP II emphasized that “early and continuous prenatal care is an important and effective means to improve the outcome of pregnancy.”7 In the 1980s, virtually no progress was made, however, in increasing the number of women receiving early prenatal care, and the number of women receiving late or no care increased. More work needs to be done to understand better the barriers to prenatal care. The Institute of Medicine described the following six barriers to adequate prenatal care: financial problems, inadequate capacity in care systems used by low-income women, services that are difficult to navigate (unfriendly), lack of awareness and acceptance of unintended pregnancy, personal beliefs about prenatal care, and social isolation.3,18 TIOP II proposed solutions to ameliorate each of these barriers (Table 2-3).7 An organized system for providing care to a pregnant woman and her infant after birth consists of the following types of facilities: physicians’ offices and clinics, basic perinatal facilities, specialty perinatal facilities, subspecialty perinatal facilities, regional perinatal centers, and specialized units such as children’s hospitals and cardiac centers.

26

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 2–2  Definitions of Care at Perinatal Centers Service

Basic

Specialty

Subspecialty

X

X

X

X

X

Care Provided

Basic inpatient care for women and newborns without complications High-risk pregnancies with moderate complications Neonatal intensive care including ventilation

X

Inpatient care for critically ill neonates

X

Follow-up medical care of infants released from NICU

X

Follow-up developmental assessment

X

Consultation and referral arrangements

X

Transport service

X

Personnel

Physician and nursing staff to care for uncomplicated pregnancy

X

X

Obstetrician

X

X

X

Pediatrician

X

X

Obstetric anesthesia

X

Neonatologist

X

Perinatal social worker

X

Genetic counselor

X

Pediatric subspecialists

X

Pediatric Surgery, Subspecialties, and Support Services

Laboratories to assess fetal well-being and maturity

X

Level III ultrasound capability

X X

Laboratories with microspecimen capability

X

Blood gases available on 24-hr basis

X X

NICU, neonatal intensive care unit. From March of Dimes Birth Defects Foundation: Toward improving the outcome of pregnancy: the 1990s and beyond, White Plains, NY, 1993, March of Dimes.

TABLE 2–3  Reducing Barriers to Prenatal Care through System Change Barriers

Related COPH Recommendations

Financing

Health care coverage for all pregnant women Mechanisms to ensure adequate provider payment

Capacity

More efficient use of existing providers Improved linkages between public and private and ambulatory and inpatient providers

Lack of user-friendly services

Matching provider capabilities and expertise to individual need and risks Risk assessment to identify medical, personal, and cultural barriers

Unintended pregnancy

Reproductive awareness among all women Greater emphasis on preconception care and care during pregnancy, including family planning

Personal beliefs and attitudes

Health promotion and health education for all children Reproductive awareness among all women

Social isolation

Outreach programs

COPH, Committee on Perinatal Health. From March of Dimes Birth Defects Foundation: Toward improving the outcome of pregnancy: the 1990s and beyond, White Plains, NY, 1993, March of Dimes.

Chapter 2  Epidemiology and Perinatal Services

Physicians’ Offices and Clinics The basic units for care during pregnancy and for care of the mother and infant after delivery are physicians’ offices and general clinics. These units must be able to obtain a complete health history, careful physical assessment of the mother or infant, systematic risk assessment (using one of the available risk-scoring systems), and laboratory resources for determining hematocrit or hemoglobin concentration and urinalysis.

Basic Perinatal Facilities (Level I) Basic perinatal facilities are designed primarily for the care of maternal and neonatal patients who have no complications. Because complications can arise in previously uncomplicated cases, basic units must have the resources to provide competent emergency services when the need arises. The necessary services for a basic perinatal facility are shown in Table 2-2 and include a normal newborn nursery. Capabilities for different levels of neonatal care are summarized in Table 2-4.

Specialty Perinatal Facilities (Level II) Specialty perinatal facilities are hospitals that have larger maternity and newborn services. These hospitals are located in urban and suburban areas serving larger communities. In addition to providing a full range of maternal and newborn services for perinatal patients who have no complications, they provide services for obstetric and neonatal patients who have one or more complications. The range of obstetric and neonatal complications an institution can treat depends on its resources. The services available in specialty units are presented in Table 2-2 and include a normal newborn nursery and a transitional-care nursery. Care of high-risk neonates should be provided by appropriately qualified physicians. A board-certified pediatrician with special interest, experience, and in some situations subspecialty certification in neonatal-perinatal medicine, should be chief of the neonatal care services.

Subspecialty Perinatal Facilities (Level III) In addition to the resources and capabilities of a specialty unit, subspecialty facilities can provide a full range of maternal complications and newborn intensive care. The neonatal intensive care unit (NICU) must have a full range of services for neonates, with the possible exception of infants with congenital heart disease or other complex congenital anomalies who require a specialized unit. The services offered by subspecialty facilities are listed in Table 2-2 and include intensive care for mothers and infants. The director of a subspecialty unit should be a full-time, board-certified pediatrician with subspecialty certification in neonatal-perinatal medicine.

Regional Perinatal Center A regional perinatal center is a subspecialty facility that also is responsible for coordinating and managing special services, including transportation, that are needed for the region. In areas where there is only one subspecialty facility, the facility is expected to function as the regional perinatal center. In areas

27

where there is more than one unit with subspecialty capabilities, one of the units would serve as the regional perinatal center. The regional perinatal center must offer outpatient and inpatient consultation and diagnostic services for basic and specialty facilities within the region, including ultrasonography, laboratory analysis of amniotic fluid, assessment of gestational age, genetic studies, and other studies of fetal health. It also should provide specialized nursing services and consultation in nutrition, social services, respiratory therapy, and laboratory and radiology services. The center is responsible for conducting an active outreach education program for the institutions, health professionals, and public within the region. A regional perinatal center has unique personnel needs. It should be directed by a full-time physician with extensive training and experience in perinatal medicine and administration. There also should be a director of obstetric services and a director of neonatal services. The director of obstetric services should be a full-time physician with training and experience in fetal-maternal medicine, including maternal intensive care. The director of neonatal services should be a full-time neonatologist with training and experience in neonatal care, including newborn intensive care. The perinatal nursing services should be directed by a clinical nurse specialist with advanced experience in maternal and neonatal nursing and in administration. The center also may require a full-time director of the outreach education program to coordinate the active participation of physicians in obstetrics and newborn care, nurses in obstetrics and newborn care, nutritionists, social workers, and other specialized personnel. The obstetrics and newborn care units, including the newborn intensive care unit, should have clinical nurse specialists in obstetrics and neonatal care responsible for organizing the nursing program and coordinating the patient care needs. It is important to estimate the number of pregnant women who might need specialized obstetric and neonatal services within the area of a given regional perinatal center. The percentage of pregnancies at increased risk may vary from 10% for general populations, such as an entire state or country, to more than 90% in some urban hospitals. In an Ontario, Canada, perinatal study, 32% of pregnancies had some increased risk factor that resulted in 60% of the neonatal problems. In the Nova Scotia Fetal Risk Project, 11% of 9483 patients accounted for 50% of the stillbirths and 75% of the neonatal deaths. The number of neonatal intensive care beds and neonatal intensive care days needed for a given population are most influenced by the frequency of premature birth and low birthweight (LBW). There are great differences in these frequencies among countries and among populations within a country. Infants with LBW account for less than 5% of the births in the Scandinavian countries and 6% to 9% of infants in some states in the United States; in some institutions, the percentage may be 15% to 20%.28 Swyer and associates39 calculated a need for neonatal intensive care beds at 0.7 beds per 1000 live births on the basis of a 7% LBW rate. Transitional or intermediate and convalescent bed needs were approximately 4 per 1000 live births. The Wisconsin Perinatal Care Program predicted a need for 12 intensive care beds for 6000 live births (7% LBW rate). Data from Utah indicate a need for 2 beds per 1000 annual live births in special care facilities. The estimated breakdown

28

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 2–4  P  roposed Uniform Definitions for Capabilities Associated with Highest Level of Neonatal Care within an Institution Level I Neonatal Care (Basic)

Well-newborn nursery: has the capabilities to Provide neonatal resuscitation at every delivery Evaluate and provide postnatal care to healthy newborns Stabilize and provide care for infants born at 35-37 wk gestation who remain physiologically stable Stabilize newborns who are ill and infants born at ,35 wk gestation until transfer to a facility that can provide the appropriate level of neonatal care Level II Neonatal Care (Specialty)

Special care nursery: level II units are subdivided into 2 categories on the basis of their ability to provide assisted ventilation including continuous positive airway pressure Level IIA: has the capabilities to Resuscitate and stabilize preterm or ill infants before transfer to a facility at which newborn intensive care is provided Provide care for infants born at .32 wk gestation and weighing 1500 g (1) who have physiologic immaturity such as apnea of prematurity, inability to maintain body temperature, or inability to take oral feedings or (2) who are moderately ill with problems that are anticipated to resolve rapidly and are not anticipated to need subspecialty services on an urgent basis Provide care for infants who are convalescing after intensive care Level IIB has the capabilities of a level IIA nursery and the additional capability to provide mechanical ventilation for brief durations (,24 hr) or continuous positive airway pressure Level III (Subspecialty) NICU

Level III NICUs are subdivided into 3 categories Level IIIA: has the capabilities to Provide comprehensive care for infants born at .28 wk gestation and weighing .1000 g Provide sustained life support limited to conventional mechanical ventilation Perform minor surgical procedures such as placement of central venous catheter or inguinal hernia repair Level IIIB NICU: has the capabilities to provide Comprehensive care for extremely low birthweight infants (#1000 g and #28 wk gestation) Advanced respiratory support such as high-frequency ventilation and inhaled nitric oxide for as long as required Prompt and on-site access to a full range of pediatric medical subspecialists Advanced imaging, with interpretation on an urgent basis, including computed tomography, magnetic resonance imaging, and echocardiography Pediatric surgical specialists and pediatric anesthesiologists on site or at a closely related institution to perform major surgery such as ligation of patent ductus arteriosus and repair of abdominal wall defects, necrotizing enterocolitis with bowel perforation, tracheoesophageal fistula or esophageal atresia, and myelomeningocele Level IIIC NICU: has the capabilities of a level IIIB NICU and is located within an institution that has the capability to provide ECMO and surgical repair of complex congenital cardiac malformations that require cardiopulmonary bypass ECMO, extracorporeal membrane oxygenation; NICU, neonatal intensive care unit. From Stark AR, Couto J; American Academy of Pediatrics Committee on Fetus and Newborn: Levels of neonatal care, Pediatrics 114:1341, 2004.

Chapter 2  Epidemiology and Perinatal Services

was 0.5 level I beds, 0.5 level II beds, and 1 level III bed per 1000 annual live births.20 In the United Kingdom, Field and colleagues12 reported that the demand for neonatal intensive care was 1.1 beds per 1000 deliveries. This was a minimum estimate, and factors such as increased survival of extremely immature infants would increase the demand for beds.

REGIONALIZATION Impact on Neonatal Morbidity and Mortality Paneth and associates27 analyzed all singleton births and deaths with known birthweight and gestational age in New York City from 1976 through 1978. Mortality rates for fullterm, appropriately grown infants were not influenced by the hospital of birth. Preterm infants and infants with LBW were at a 24% higher risk of death, however, if birth occurred at either a level I or a level II unit. These small infants constituted only a small percentage of the births, but accounted for 70% of the deaths.35 Fanaroff’s group compared the outcomes of neonates delivered between 24 and 28 weeks’ gestation at National Institute of Child Health and Human Development network level III units with neonates of similar gestational age who were transported to the centers after birth.10 Outborn infants had significantly more respiratory distress syndrome (88% versus 81%), more grade III or IV intraventricular hemorrhages (24% versus 17%; odds ratio 1.61, 1.12 to 2.16), and greater mortality (32% versus 22%; odds ratio 1.63, 1.15 to 2.30). Yeast and colleagues43 compared neonatal mortality in two 5-year periods (1982-1986 versus 1990-1994) in Missouri. They found that in both periods the relative risk of neonatal mortality in level II centers was 2.28 compared with level III centers, and that no substantial improvement had occurred in those 10 years. Similar data have been reported from California.6,29 There are compelling reasons for preterm deliveries to occur at tertiary centers.

Maternal and Neonatal Transport Services It is estimated that antenatal maternal transfer is impossible in 50% of high-risk pregnancies. In these situations, neonatal transport must be performed by specially trained teams skilled in adequate stabilization before and effective management during transport. Hood and associates17 documented a 60% greater mortality rate when neonates were transferred by an untrained versus a trained neonatal transport team. Hypothermia and acidosis in particular were more common after transfer by an untrained team. National groups have recommended standards for personnel configuration, training, and accreditation.42 The transport team from the referral hospital also serves an important educational function and can influence and improve methods of stabilization at the referring center, which may influence mortality statistics (see Chapter 5). An effective means of managing overcrowding at a perinatal referral center is to encourage reverse transport of previously ill neonates to levels I and II nurseries for intermediate care.19 Transfer can include not only infants with resolved acute medical problems, but also infants with chronic problems such as bronchopulmonary dysplasia. The tertiary center must be familiar, however, with the capabilities, facilities, and

29

resources of the hospitals to which reverse transport is occurring so that quality of care can be ensured. This evaluation can be combined with an outreach education program. Apart from the cost-effectiveness of reverse transport, it can encourage family bonding and greater involvement of the pediatrician who will be offering continuing care to the infant. The outcomes of infants transported back from tertiary centers to community hospitals (level II units staffed by skilled personnel) were compared with outcomes of infants convalescing in the tertiary center. Lynch and colleagues24 documented that the transported infants received appropriate care, were less likely to need readmission to the intensive care unit (7% versus 14%), and required fewer transfusions. Major new health problems developed in 27% of the patients during convalescence. The overall complication rate was lower, however, for the reverse transfers. The current medical economic climate mandates that tertiary units establish criteria for reverse transfer, and many health maintenance organizations and preferred provider organizations are demanding early reverse transport. Tertiary units are obliged to ensure that there are appropriately trained personnel and adequate facilities at their community hospitals so as not to compromise the medical needs or care of the neonates requiring reverse transfer.

Problems One of the most common problems of regionalization is centralization in place of regionalization; this is the end product of a regional center that operates with no outreach education program or other mechanisms for continuing the development and improvement of services in the other hospitals of the region. With such a system, the central hospital continues to receive the referrals of high-risk mothers or high-risk neonates, but makes no effort to help the referring hospital develop programs for preventing the problems. This may be particularly true in university medical centers, where outreach education and service are not considered a regular academic or hospital activity. This problem improved with increased use of reverse transports.5 This progress is threatened, however, because some third-party and government payers refuse to reimburse the expense of the reverse transport, even when the result is to relocate care to a less expensive institution. A second common problem of regionalization is unnecessary duplication of units36; this includes the unnecessary duplication of basic and specialty units within rural or urban areas and competing subspecialty units, particularly in urban areas. The duplication results in difficulties in recruiting and maintaining the necessary personnel for such care units and an increased cost per patient for such care. Such duplication invariably is a result of competing institutions and competing medical staffs, who view maternal intensive care and NICUs as important to their business, income, or institutional image. Some regional centers are staffed with inadequate or inappropriately trained personnel. In the past, a common problem was the staffing of infant intensive care nurseries with inadequately supervised resident staff as the primary responsible physicians, particularly during night hours and weekends. Data indicate, however, that 23% to 75% of subspecialty centers have increasingly moved to 24-hour in-house specialty coverage by attending neonatologists.9,36

30

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

NECESSITY OF REGIONALIZATION OF PERINATAL CARE In the 1970s, a spirit of cooperation drove the development of regionalized systems. The realities of health care reform, competition, and cost constraints have dimmed that cooperative spirit. More hospitals are merging and forming networks, which jointly contract with payers to supply complete health services to a population of “covered lives.” These forces lead every network to wish to provide all levels of service in a given area so that covered lives stay within that system. This situation leads to duplication of services that runs counter to the principles of regionalization.33 The availability of a large supply of highly trained neonatologists, together with financial incentives, led to the creation of new levels of care termed level IIB, level II plus, or community NICUs.26,38 The development of these centers shifted the location of high-risk births away from level III centers to community NICUs in California and other states.16 Gould and coworkers15 evaluated the impact of this deregionalization on neonatal mortality and documented that by 1997 regional NICUs lost 5.2% of California’s live births. Only 20% of the reduction occurred because of a shift of deliveries to community NICUs, however; 80% of the shift was to level I and level II (intermediate) NICUs. Neonatal mortality was similar at the regional NICUs and community NICUs, but remained elevated at intermediate NICUs (odds ratio 1.54), self-designated intermediate NICUs that were not certified by the state (odds ratio 1.33), and primary care units (odds ratio 1.56). Cifuentes and coworkers,6 also working with health statistics from California, similarly identified a mortality disadvantage for neonates with very LBW at level II NICUs and identified an impact of average daily census: centers with a census less than 15 had increased mortality. Rogowski and colleagues34 were unable to confirm this association in the Vermont Oxford Network. Kamanth and associates21 analyzed births in Colorado and showed decreased mortality risk among infants weighing less than 750 g born in a level III center versus those born in a non–level III center.

Financial Impact The facilities within a regional network, including a regional perinatal center, financially depend on a combination of patient revenue and public support to care for neonatal patients. With carefully coordinated use of resources and facilities, the cost of care can be contained.13,22,23,44 It is essential that health insurance programs provide adequate coverage for obstetric and neonatal conditions, and that charges reflect the cost of delivering the services. Medical assistance and other forms of payment and direct support of state and county hospitals must be adequate for the care needs. There must be adequate nursing staff, physician coverage, equipment, and supplies to achieve an acceptable quality of care. The reimbursement schemes should provide for patient transfer between institutions without multiplication of deductibles or major financial hardship to the patient. The cost of emergency transport of high-risk maternal and newborn patients should be covered. Financial incentives should promote, rather than discourage, the use of resources appropriate to patient care needs.

Diagnosis-related grouping has been implemented as the basis for federal reimbursement, and other third-party payers have followed suit. The limited initial database used to determine the reimbursement level did not take into account many variables that influence length of stay, particularly at a tertiary center.2,23,30 If the guidelines are not modified, major tertiary centers will be at a distinct financial disadvantage, particularly when providing care for extremely complicated problems and infants with LBW. The system could also discourage transport of infants from primary and specialty units if the referring hospitals perceive financial gain; the quality of care may be compromised.

MEASUREMENTS OF EFFECTIVENESS OF CARE ORGANIZATION When the effectiveness of any care system and its individual components is measured, it is essential to analyze data for the entire region. A hospital can show dramatic changes in mortality or morbidity statistics in the course of a single year, not as a result of improvement of care, but as a result of movement of patients with particular problems to another institution. If the geographic boundaries of a region are well designed, there should be limited patient movement from region to region. This permits consistent year-to-year evaluation of the care within the region. Maternal mortality has declined to the point that it is no longer a satisfactory index of quality of care in developed countries. Fetal and neonatal mortalities are still reasonable indicators of perinatal care.11 In evaluating a region, the data must link the hospital in which the death is recorded with the institution and the community in which the birth occurred or the care was initiated. Fetal and neonatal mortalities should also be divided by weight groups. The weight groups should be in 500-g increments or less, beginning with 500 g. If possible, the gestational age distribution and cause of death for the fetal or neonatal deaths should be established. Sex and transport status (maternal or neonatal) have been shown to affect mortality.10,37 Such practices make it possible to identify areas or institutions within a region in which there are major problems with care. Any comparison of mortality is most useful if a risk adjustment tool is applied, such as the Clinical Risk Index for Babies (CRIB) or Score for Neonatal Acute Physiology (SNAP) (see Chapter 5).32,40 Such scores correct for the inherent bias of subspecialty centers that receive only the sickest neonates and experience the highest mortality. When maternal, fetal, or neonatal mortality and morbidity rates are used, certain other information is necessary to permit useful interpretation. For maternal mortalities, it is essential to separate the maternal deaths that occurred during or as a result of pregnancy associated with other maternal disease, such as severe cardiac disease or malignancy, from deaths associated primarily with the pregnancy. This separation requires analysis of each death and assignment to preventable or unpreventable categories. For evaluation of neonatal programs and regionalization, the frequency of 1-minute Apgar scores of 3 or less and 5-minute Apgar scores of 5 or less should be recorded. Also helpful in assessing effectiveness of care are the frequency of sepsis, traumatic delivery,

Chapter 2  Epidemiology and Perinatal Services

respiratory distress syndrome, and neurologic problems; the number of intensive care days; and the number of patient transfers from community institutions to institutions of greater care capability. All high-risk mothers and neonates should have systematic follow-up care. Infants with a birthweight of less than 1500 g are at high risk for developmental, neurologic, or learning problems, and should be followed into school age with careful neurologic and educational testing (see Chapter 41).25,41 The incidence of child abuse and failure to thrive may also reflect parenting disorders having antecedents in the perinatal period.

FUTURE CONSIDERATIONS Centers for newborn care have expanded at specialty (level II) institutions with large delivery services so that many have developed capabilities similar to those at subspecialty centers. In many geographic areas, the subspecialty and specialty units compete; in others, the two centers work hand in hand. In all communities, it is essential that mutually productive relationships be reestablished between the academic and community hospitals. Most board-certified neonatologists in the United States now practice outside traditional academic settings. These personnel needs are being met by the university-based training programs. With fewer training programs accredited and more trainees completing a third fellowship year, however, it is conceivable that in the near future there will not be enough trained neonatologists to staff the ever-expanding specialty units. At present, the major strategies are to provide costeffective medical care, to comply with regulations that dictate standards required for reimbursement, and, above all, to prevent LBW and prematurity.

REFERENCES 1. Aubrey RH et al: High-risk obstetrics, I: perinatal outcome in relation to a broadened approach to obstetrical care for patients at special risk, Am J Obstet Gynecol 105:241, 1969. 2. Beeby PJ: How well do diagnosis-related groups perform in the case of extremely low birthweight neonates? J Paediatr Child Health 39:602, 2003. 3. Brown SS et al: Barriers to access to prenatal care. In Kotch JB et al, editors: A pound of prevention: the case for universal maternity care in the United States, Washington, DC, American Public Health Association, 1992. 4. Callon HF: Regionalizing perinatal care in Wisconsin, Nurs Clin North Am 10:263, 1975. 5. Chiu T et al: University neonatal centers and level II centers capability: the Jacksonville experience, J Fla Med Assoc 79:464, 1992. 6. Cifuentes J et al: Mortality in low birth weight infants according to level of neonatal care at hospital of birth, Pediatrics 109:745, 2002. 7. Committee on Perinatal Health: Toward improving the outcomes of pregnancy: the 1990s and beyond, White Plains, NY, March of Dimes, 1993. 8. Committee on Perinatal Welfare: Report on perinatal and infant mortality in Massachusetts, 1967 and 1968, Boston, Massachusetts Medical Society, 1971.

31

9. Denson SE et al: Twenty-four hour in-house coverage for NICUs in academic centers: who, how and why? J Perinatol 10:257, 1990. 10. Fanaroff AA et al: Very-low-birth-weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, May 1991 through December 1992, Am J Obstet Gynecol 173:1423, 1995. 11. Fanaroff AA et al: The NICHD neonatal research network: changes in practice and outcomes during the first 15 years, Semin Perinatol 27:281, 2003. 12. Field D et al: The demand for neonatal intensive care, BMJ 299:1305, 1989. 13. Friedman B et al: The use of expensive health technologies in the era of managed care: the remarkable case of neonatal intensive care, J Health Polit Policy Law 27:441, 2002. 14. Goodwin JW et al: Antepartum identification of the fetus at risk, Can Med Assoc J 101:458, 1969. 15. Gould JB et al: Expansion of community-based perinatal care in California, J Perinatol 22:630, 2002. 16. Hernandez JA et al: Impact of infants born at the threshold of viability on the neonatal mortality rate in Colorado, J Perinatol 1:21, 2000. 17. Hood JL et al: Effectiveness of the neonatal transport team, Crit Care Med 11:419, 1983. 18. Institute of Medicine: Prenatal care: reaching mothers, reaching infants, Washington, DC, National Academy Press, 1988. 19. Jung AL, Bose CL: Back transport of neonates: improved efficiency of tertiary nursery bed utilization, Pediatrics 71:918, 1983. 20. Jung AL, Streeter NS: Total population estimate of newborn special-care bed needs, Pediatrics 75:993, 1985. 21. Kamanth BD et al: Infants born at the threshold of viability in relation to neonatal mortality: Colorado, 1991 to 2003, J Perinatol 28:354, 2008. 22. Kitchen WH et al; the Victorian Infant Collaborative Study Group: The cost of improving the outcome for infants of birthweight 500-999 g in Victoria, J Paediatr Child Health 29:56, 1993. 23. Lagoe TJ et al: Impact of selected diagnosis-related groups on regional neonatal care, Pediatrics 77:627, 1993. 24. Lynch T et al: Neonatal back transport: clinical outcomes, Pediatrics 82:845, 1998. 25. McCormick MC et al: The health and developmental status of very-low-birth-weight children at school age, JAMA 267:2204, 1992. 26. Merenstein CB et al: Personnel in neonatal pediatrics: assessment of numbers and distribution, Pediatrics 76:454, 1985. 27. Paneth N et al: The choice of place of delivery: effect of hospital level on mortality in all singleton births in New York City, Am J Dis Child 141:60, 1987. 28. Paneth NS: The problem of low birthweight, Future Child 5:19, 1995. 29. Phibbs CS et al: The effects of patient volume and level of care at the hospital of birth on neonatal mortality, JAMA 276:1054, 1996. 30. Poland RL et al: Analysis of the effects of applying federal diagnosis-related grouping (DRG) guidelines to a population of high-risk newborn infants, Pediatrics 76:104, 1985. 31. Report of the Committee on Perinatal Health of the American Medical Association, American College of Obstetricians and Gynecologists, American Academy of Pediatrics, and American Academy of Family Physicians: Toward improving the outcome

32

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

of pregnancy, New York, March of Dimes National foundation, 1975. 32. Richardson DK et al: Score for Neonatal Acute Physiology: a physiologic severity index for neonatal intensive care, Pediatrics 91:617, 1993. 33. Richardson DK et al: Perinatal regionalization versus hospital competition: the Hartford example, Pediatrics 96:417, 1995. 34. Rogowski JA et al: Indirect vs direct hospital quality indicators for very low-birth-weight infants, JAMA 291:202, 2004. 35. Roth J et al: Changes in survival patterns of VLBW infants from 1980 to 1993, Arch Pediatr Adolesc Med 149:1311, 1995. 36. Schwartz RM, Kellogg R, Muri JH: Specialty newborn care: trends and issues, J Perinatol 20:520, 2000. 37. Stevenson DK et al: Very low birthweight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1993 through December 1994, Am J Obstet Gynecol 179:1632, 1998. 38. Stoddard JJ et al: Providing pediatric subspecialty care: a workforce analysis. AAP Committee on Pediatric Workforce Subcommittee on subspecialty Workforce, Pediatrics 106:1325, 2000.

39. Swyer PR et al, editors: Regional services in reproductive medicine, Toronto, Joint Committee of the Society of Obstetricians and Gynaecologists of Canada and the Canadian Paediatric Society, 1973. 40. Tarnow-Mordi W et al: The CRIB (Clinical Risk Index for Babies) score: a tool for assessing initial neonatal risk and comparing performance of NICUs, Lancet 342:193, 1993. 41. Taylor HG et al: Middle-school-age outcomes in children with very low birthweight, Child Dev 71:1495, 2000. 42. Woodward GA et al: The state of pediatric interfacility transport: consensus of the second National Pediatric and Neonatal Interfacility Transport Medicine Leadership Conference, Pediatr Emerg Care 18:38, 2002. 43. Yeast JD et al: Changing patterns in regionalization of perinatal care and the impact on neonatal mortality, Am J Obstet Gynecol 178:131, 1998. 44. Zupancic JA et al: Economics of prematurity in the era of managed care, Clin Perinatol 27:483, 2000.

CHAPTER

3

Medical Ethics in Neonatal Care Jonathan Hellmann

This chapter explores the complexity of moral problem solving in neonatal medicine. First, principles of medical ethics and key terms and concepts are defined, followed by the application of these concepts in specific moral problems that arise (1) when a pregnant patient refuses treatment, (2) in the prenatal consultation at the limits of viability, and (3) when withholding and withdrawing life-sustaining medical treatment in the neonatal intensive care unit (NICU) is undertaken. A collaborative, procedural framework for consensual end-of-life decision making is described, and specific ethical issues that may arise in end-of-life care are discussed, including the use of analgesic agents, brain death and organ donation, palliative care, and the withdrawal or withholding of artificial nutrition and hydration. This chapter includes guidelines for ethical conflict resolution and an approach to the conduct of clinical research. Finally, a brief summary of ethical responsibilities of neonatal physicians is presented. Three elements characterize the practice of medicine that are timeless, universal, and irrefutably true: (1) the fact of illness and the vulnerability it creates, (2) the act of the profession (the use of medical skills for the benefit of the patient), and (3) the practice of medicine itself—that which physicians and patients do together in the clinical encounter, characterized by mutual intentionality.61 These three elements are well exemplified in the care of sick newborn infants, where the vulnerability of anxious parents is manifest, where the clinical competence and moral discretion of health professionals are used for the benefit of newborn patients and their parents, and where the practice is carried out in a patient-parentphysician relationship characterized by mutual trust and pursuit of the neonatal patient’s and parents’ good. Medical ethics involves the systematic, reasoned evaluation and justification of the “right” action in pursuit of human good or well-being in the context of medical practice. It involves a critical examination of the concepts and assumptions underlying medical and moral decision making, and it may include a critical examination of the kind of person a physician should be.61 Medical ethics has become a

central focus in the practice of neonatal intensive care because it contains all the elements of moral discourse but without clear-cut, “right” answers. Numerous ethical issues arise as physicians attempt to determine what is right and good for their vulnerable patients, when physicians and parents deliberate on what constitutes the “best interests of the infant,” and when NICUs use their technological capability in the quest for medical progress. These issues are compounded by the fact that the treatment occurs in a fast-paced environment that precludes or severely limits opportunities for in-depth exploration of parents’ and physicians’ views and values. It is also complicated by the medical uncertainty surrounding accurate outcome prediction of potentially adverse findings, and the very nature of many decisions in which quality of life and life itself may be under consideration. All these issues make the NICU a challenging, highly scrutinized environment in how ethical issues are examined, reasoned, and justified, and raises fundamental questions about what the responsibilities of neonatal health care providers should be. Ethical issues may be experienced as moral dilemmas, moral uncertainty, or moral distress. A moral dilemma is present when the physician believes there is an obligation to pursue two (or more) conflicting courses of action. As only one of these courses can be pursued, the physician has to make a value-based choice that compromises one of these obligations. A classic example in neonatal care is a conflict between action required by the principle of respect for autonomy and action required by the principle of beneficence. Moral uncertainty typically arises when the presenting issue is unclear. Parents whose fetus has a major congenital anomaly are presented with the option of terminating that pregnancy or anticipating major surgery for their infant at the time of birth. They are unwilling to end the pregnancy, but are uncertain about the varying prognoses given to them by respective physicians. They, and possibly their care providers, remain unclear and uncertain about the principles and values that are in conflict.

33

34

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Moral distress arises when the decision maker feels certain about the morally right thing to do, but this perceived “right” course of action is precluded for numerous reasons, including the caregiver’s lack of decision-making authority or institutional or financial constraints. Neonatal nurses who have intense hands-on contact with newborns may feel powerless and experience moral distress when treatment decisions are made by others.32 Ulrich and colleagues74 have elucidated the feelings of powerlessness, frustration, and fatigue experienced by nurses and social workers in relation to ethical issues in their workplace. The number of surveys of neonatologists’ preferences and practices at the limits of viability is also likely a reflection of a degree of moral distress within that group, as is the more personal, plaintive, and painful expression of the neonatal fellow in dealing with an infant “abandoned in the NICU” when faced with parental insistence on “doing everything” possible for a severely compromised and socially vulnerable infant.10

PRINCIPLES IN MEDICAL ETHICS Ethical reasoning requires an understanding of the principles that define a domain (Box 3-1); the principles of autonomy, beneficence, nonmaleficence, and justice in the care of newborns are briefly discussed to suggest ways in which these principles can be interpreted. Although reference to principles serves as a useful checklist of moral concerns, appealing to principles per se in the clinical context can be of limited value when the problem revolves around a conflict in the principles themselves. In many cases, a “principlist” paradigm may not do justice to the nature or complexity of the moral problem. In addition, in a purely principle-based approach, the role of the neonatal health care provider as moral agent (i.e., someone with personal virtues and values and the opportunity to take action that comports with that judgment of morality) is not highlighted. This chapter does not prescribe adherence to any particular ethical theory, but aims to enhance appreciation of the language of ethical discourse to enable nuanced reflection, reasoning and ethical decision making, and to underline the role of the health care provider in these situations.

Autonomy Autonomy grounds the right of competent patients to make their own health care choices. An individual who makes an autonomous choice acts intentionally, with understanding

BOX 3–1 Key Concepts Underlying Ethical Care in the Neonatal Intensive Care Unit Respecting parental authority/autonomy Applying the best interests of the infant standard of judgment Minimizing harm to the newborn Developing sound parent-physician relationships Empowering and informing parents Applying family-centered care principles Respecting parents’ values and cultural and religious beliefs Sharing decision making Developing respectful interprofessional (moral) teamwork

and without external controlling influences.8 An idealized expression of respect for autonomy with adult patients is one in which mentally and emotionally capable patients choose voluntarily and intelligently from among various options whose relative risks and benefits have been fully explained to them by their physicians (i.e., via a truly informed consent process). The term parental autonomy is often used in the context of neonatal care, without making a distinction between that term and parental authority, although most argue that autonomy can strictly be used only when making decisions for oneself, and that parental authority is the more correct term when referring to the role of parents in decision making for their newborn infants. The use of parental autonomy may be appropriate, however, when a parent has a vision of what it means to be a good parent, and it is necessary for that parent to choose to be allowed to parent in that self-directed way. There are at least two primary reasons why neonatal health care professionals should show respect for parental authority/autonomy: (1) because of the rightful presumption that parents will make decisions that are in their child’s best interests, and (2) as a means of equalizing the parental role in decisions with physicians who shape parental views, even at times without consciously intending to do so. Physicians need to be conscious of the impact of their authority and the power that derives from expertise or that follows from information given in a convincing manner. They also have to recognize that parents’ ability to reason and to act in a selfdirected way is often diminished in the presence of serious illness in their newborn infants. In addition, there is a body of evidence and opinion that families may not want the decision-making authority that the prevailing bioethics emphasis on autonomy and selfdetermination advocates.58,82 Numerous studies show that respect for mothers’ or parents’ autonomy does not require physician adherence to a strict informed consent interaction; a more nuanced application of this respect can be satisfactorily achieved by including parents in decision making within a trusting parent-physician relationship.37,57 This inclusion may be more important than the parental “right” to make the final decision per se. Parents’ insistence on their right and role as final decision maker may become their stance only after a breakdown in communication, or where respect for their role in a shared decision-making process has been disallowed or denigrated. Too rigid or simplistic an interpretation and application of respect for parental autonomy by physicians is also problematic, such as when agreeing to a parental statement that “we want everything done.” Without exploring what underlies this declaration, physicians may neglect the complexity of the situation and an unperceived parental fear of “abandonment.”26 Finding the right balance between respect for parental autonomy and the physician’s role and responsibility in any decision-making process requires insight, empathy, and great analytical and communication skills.

Beneficence Beneficence is the obligation of others to promote the best interests of patients who are unable to make autonomous health care choices. In newborns, this obligation is embodied in the concept of the “best interests of the newborn.” This is

Chapter 3  Medical Ethics in Neonatal Care

a moral and legal standard of judgment that helps to establish the primacy of duties to infants, ensuring they be regarded as fully human individuals with interests, even when clearly unable to express their own value system. Pursuing a course of action in the best interests of an infant implies determining what treatment course has a more favorable benefit-to-harm ratio than other possible options—whether the treatment is likely to improve the newborn’s condition or well-being, prevent it from deteriorating, or reduce the extent or rate at which it is likely to deteriorate, and whether the newborn’s condition is likely to improve, remain the same, or deteriorate without the treatment. It also requires consideration of whether a less restrictive or less intrusive treatment course would be as beneficial as the treatment that is being proposed. Determination of best interests requires an assessment of the child’s potential quality of life. Quality-of-life considerations encompass the predicted cognitive and neurodevelopmental outcome, the potential for motor disability or other physical handicap (e.g., vision, hearing), and longer term concerns such as behavioral and learning difficulties or school problems. It also considers the requirements for repeated or prolonged hospitalization, surgery or medication, and the potential for pain and suffering to be endured. Quality-of-life considerations may also include less concrete medical states, such as the capacity for meaningful and potentially enjoyable interaction with other people and the environment. The standard of best interests of the newborn acts as a threshold for judgment, yet the subjective nature of this assessment and the fact that the assessment is being done by surrogate decision makers must always be recognized. The procedural question of who are the rightful decision makers is of pivotal importance when parents, physicians, and nurses have different assessments of what is in the infant’s best interests. Consider parents who believe in the preservation of life regardless of their infant’s neurodevelopmental outcome; their perception of the benefits and harms of aggressive intervention may be quite different from the perception of a physician who places greater value on quality of life and relief of suffering. The best interests standard is not only a standard of judgment, but also a standard for possible intervention, if it is perceived by rightful decision makers that the infant’s best interests are not being served by a particular course of action. In most jurisdictions, the state confers the responsibility for health care decision making on the parents of young children because parents have a unique relationship with their children—one of concern, obligation, responsibility, and intimacy. Recognition of parents’ moral authority in decision making presumes that parents would normally act to promote their children’s best interests and endeavor to make health care decisions in that light. The concept of best interests engenders much debate, not only because of the subjectivity of its assessment, but also because some authorities believe the concept is too complex: Defining “best” may be unattainable; it is culture specific, it is too individualistic, and it ignores the interests of others. The concept does imply, however, that the child’s best interests should be pursued, even sometimes to the exclusion of family interests. Other authorities argue that it is legitimate for physicians and parents to consider family interests in

35

making health care decisions for a sick newborn.30 It has been shown that most neonatologists ascribe to an incorporation of family interests into decision making for their incompetent patients.29 Despite these definitional difficulties, “best interests of the newborn” is accepted as a guiding principle for decision makers to use because it unites under one standard different meanings and exhibits reasonableness, given the prevailing conditions.41 Properly understood, the concept can serve as a powerful tool in settling disputes about how to make good decisions for individuals who cannot decide for themselves.

Nonmaleficence The principle of nonmaleficence implies an obligation not to inflict harm on others. It has been closely associated with the maxim primum non nocere (first do no harm).8 Although beneficence incorporates preventing and removing harm as part of promoting “the good” of a patient, the injunction not to inflict harm remains a distinct principle and requires intentionally refraining from actions that cause harm. Such harm is generally interpreted as physical harm, especially pain, disability, or death. Nonmaleficence requires that no initiation or continuation of treatment be considered without regard to the infant’s pain, suffering, and discomfort; in situations in which a treatment is perceived as overly burdensome or harmful without foreseeable benefit, it should not be undertaken.

Justice Justice is the dominant principle relating to social cooperation: It is expected to determine how social benefits, such as health care, are distributed via justified norms that are the product of human choices and values. Although justice is regarded as a guiding principle for any health care system, no single, simple, ideal implementation of it can suffice or satisfy all who wish to receive resources and be treated in ways that safeguard their rights. Generally, concepts of justice range from the broader utilitarian calculus to promote the greatest good for the greatest number in underwriting the distribution of resources (i.e., macroallocation) to a more narrowly focused equality of opportunity for each individual (microallocation). Justice in the distribution of resources would require (1) that patients in similar situations have access to the same health care, and (2) that the level of health care available for one set of patients takes into account the effect of such a use of resources on other patients. Reasoning in the context of allocation of resources for the individual patient allows that fairness is an important consideration, but is not the fundamental standard in this interaction. Microallocation issues are seldom appropriate at the bedside, although some consistency between macrodistribution and microdistribution must exist if the two systems are to operate smoothly, because the two are inevitably linked. In every system, the provision of neonatal services requires guidelines to be developed for determining appropriate use of these costly and often limited resources. Priorities must be set and justified by authorities in decision-making positions. Any changes in the way neonatal care is delivered also need to be fair and accessible, and open and challengeable. Disputes should be worked out based on the acceptance of objective

36

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

measures and trust in the integrity and fairness of the process and the individuals responsible for management of the process. The fact that the principle of justice has not been at the forefront of bioethical discourse at the neonatal bedside may change with increasing recognition of this deficiency, particularly in times of financial constraint; how resources for sick newborns are used and justified may become more critically and ethically questioned.

KEY TERMS AND CONCEPTS Communication with Parents Parents require complete and truthful information about their infant—the diagnosis and prognosis, the available treatment options (including, where relevant, the option of no treatment), the benefits and harms associated with each option, and the limits of available technology. The manner in which this information is communicated influences parents’ understanding of the situation, their ability to discuss moral issues and values openly, and their ability to participate effectively in a decision-making process (Box 3-2). Information communicated in an honest and respectful manner is likely to foster trust; information that is confusing, incomplete, evasive, or conveyed in a hurried or dismissive way is unlikely to do so. Transparency in communication is crucial: It emphasizes the physician’s reasoning, builds an understanding of the illness, makes the connection between data and their implications, and tempers unrealistic parental expectations.38 As medicine advances, physicians and parents must sometimes struggle with information that is at the limits of medical knowledge and where the implications of findings are uncertain. The manner in which medical uncertainty, specifically prognostic uncertainty, is (or is not) communicated is extremely important and influences subsequent decision making. Neonatal physicians have been described as dealing with prognostic uncertainty via one or more of three strategies: (1) a statistical approach, (2) a wait-until-certainty approach, or (3) an individualized prognostic approach.65 With a statistical approach, decisions are made on the basis of an analysis of the known outcomes of infants with that condition—be it extremely low gestational age or a major chromosomal or congenital anomaly. The primary objective is to avoid enhancing the survival of infants with conditions where the known outcome data show profoundly poor prognoses, even if this means that some potentially viable infants do not survive. The wait-until-certainty approach begins with treatment for almost every infant with any chance of survival. It establishes a momentum in favor of continuing treatment until a major medical event does or does not occur. The possibility of withdrawal or discontinuation of life-sustaining medical treatment is considered only when severe, adverse medical findings become unequivocally evident. A problem with this strategy is that frequently, even with the passage of time, uncertainty persists. The issue is not certainty per se, but rather the degree of certainty required by physicians and parents to facilitate decision making. This approach denies the ethical complexity of situations by failing to address the moral and the medical uncertainty and risks relegating

BOX 3–2 Guidelines for Respectful Communication with Parents Create an environment for communication that encourages parents’ participation and their becoming as fully informed as possible Identify and remove barriers that limit parents’ role in communication (e.g., language, physical distance) Communicate with parents: at the time of admission, at any crisis point in their child’s NICU course, via periodic reviews of longer stay patients, and other unstructured opportunities Encourage parents to seek clarification of information at any point by requesting an appointment with the child’s responsible physician Provide open, truthful communication at all times Provide information as accurately as possible and with as much certainty of diagnosis and prognosis as is possible in each clinical situation Identify areas of medical uncertainty Use easily understandable language Assess family communication preferences, and attempt to communicate within those parameters Be preemptive in communication (i.e., foresee what problems or issues may arise in the child’s course) Be proactive in communication in any clinical situation in which a poor outcome is predicted Convene meetings with both parents when important decisions need to be made Keep parents informed of any special investigations/tests that are planned in the course of management of their child Recognize the need for time for processing and absorption of information Ensure consistency and continuity of communication in the face of medical staff changes and handovers Practice open, honest, and timely disclosure regarding medical error NICU, neonatal intensive care unit.

parents and other members of the health care team to the role of bystanders as the medical course unfolds. With an individualized prognostic strategy, treatment is initiated for all infants who have a reasonable chance of survival. Decision making solely on the basis of biomedical factors is avoided, and the ongoing moral responsibility of the decision makers is emphasized in an effort to involve parents and the health care team in navigating the ensuing prognostic uncertainty. This approach avoids the extremes of either withholding treatment from all infants who fall below a minimum threshold or treating all infants until the outcome is absolutely certain. The timing of any communication with parents is crucial. Ideally, parents’ readiness to receive information and their coping resources should be ascertained, so that appropriate information is shared with consideration of their adaptation to the medical setting. In acute situations, if an urgent decision is required, the physician needs to move the relationship

Chapter 3  Medical Ethics in Neonatal Care

rapidly, however, from one of “moral strangers” to one in which moral issues can be openly discussed. Obtaining parental consent for each planned intervention for the infant is often the prompt for communicating with parents in the NICU. Some authorities may regard that obtaining formal, informed consent from parents for virtually every neonatal test or procedure is a means of maintaining a high standard of ethical care. Ensuring that parents are kept up to date and advised of treatment plans is important; however, information gathered from an ethnographic study showed that parents did not want to be asked to consent to every procedure.1 They often felt overwhelmed when asked to consent for routine and minor procedures and felt they were given the illusion that they could or should say stop when there was no real choice and no time to learn more about each procedure. It was also clear that the more the staff offered information and time to listen to parents when not driven by a consent process, the easier it was for parents to discuss questions and dilemmas on fairly equal terms. When consent was required, parents emphasized the need for a twoway informed agreement between fairly equal partners with established mutual trust and respect. A sound patient-physician relationship is a sine qua non of good medicine, for it is within this relationship that physicians exercise their humanity, understanding, and respect for the values of others. Every communication interaction with parents is an opportunity for relationship building. The ideal model in adult patient-physician relationships is considered to be the deliberative-interactive model, wherein physicians not only help the patient with clarification of his or her values, but also strive to make their own reasoning transparent for the patient to appreciate the many factors that inform their professional recommendation.21 In neonatal medicine, the physician’s communication relationship is with the parents (or legal guardians) of the newborn, and the optimal parent-physician relationship aims to mirror the deliberative-interactive model, wherein physicians provide parents with accurate and timely information (with as much medical certainty as possible), and encourage and empower them to identify their values and treatment preferences. This model of relationship respects parental authority, encourages the physician’s expression of his or her own clinical judgment, and, in so doing, promotes the best interests of the newborn and the family. Environmental and contextual factors challenge the achievement of this parent-physician relationship, including the fast NICU pace and the demands of time and scheduling changes. There may also be personal physician-related factors that need to be overcome, such as physicians’ reluctance to express their views, a focus on short-term goals, an inherent avoidance of prognostication and discussion about outcomes, and a fear of damaging the relationship by being the bearer of bad news. Nevertheless, it is incumbent on physicians to attempt to develop a therapeutic alliance with parents and to engage in relationship building with them. The responsibility for a parent-physician relationship always rests with the physician.

Teams and Communication Although it is important that the responsible physician attempt to integrate all of the important information and maintain a consistent relationship and pattern of communication with

37

parents, neonatal intensive care is provided by many clinicians with expertise in different fields. This situation requires an understanding among the team members of the differences in responsibility in communication with parents, such as when conveying day-to-day quantifiable, objective information, or when communicating severe diagnoses or the more speculative, prognostic significance of specific findings. Although the responsibilities of each discipline are generally known, in certain situations communication boundaries may need to be defined to minimize fragmented and inconsistent information. Interdisciplinary meetings, in which the contribution of all the practitioners involved with the family are combined, are crucial in ensuring that consistent patterns of communication are maintained.

FAMILY-CENTERED NEONATAL INTENSIVE CARE Family-centered care is a philosophy that acknowledges the sick newborn infant’s place within the social unit of the family, and that cultural, emotional, and social support by the family is an integral component of the infant’s care. This concept also considers the effects of a decision on all family members, their responsibilities toward one another, and the burdens and benefits for each member of the family. Family-centered care shapes policies, programs, facility design, and day-to-day interactions, but should reflect values and attitudes more than protocols. The potential benefits of a family-oriented approach include improved parental satisfaction with care and decision making, decreased parental stress, greater parental ability to cope with their infant’s appearance and behavior, improved success with breast feeding, and increased parental comfort and competence for postdischarge care.19 Recognition of the child’s place within the family highlights the cultural, religious, and spiritual dimensions of families’ lives and may bring to light significant diversity within these domains. In a pluralistic society, parents and physicians are unlikely to share the same values, cultural systems, personal histories, and experiences, and at times of stress these differences may present health care providers with significant challenges. The first challenge is to recognize, understand, and respect the cultural, religious, and spiritual views and values of parents and families. Meeting this first challenge is crucial because misperceptions caused by a lack of sensitivity can lead to inappropriate care or poor clinical outcomes. Cultural competence is more than sensitivity to cultural norms different from one’s own. Cultural competence implies an ability to interact effectively with people of different cultures and comprises four components: (1) the individual’s awareness of his or her own cultural worldview, (2) the individual’s attitude toward cultural differences, (3) the individual’s knowledge of different cultural practices and worldviews, and (4) the individual’s cross-cultural skills.50 Developing cultural competence results in an ability to understand, communicate with, and interact effectively with people across cultures. The second challenge for neonatal health care providers involves the limits of tolerance—where to draw the line between accepting patterns of decision making between couples or within families that contrast markedly with the

38

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

prevailing cultural norm of shared parental responsibility for decision making. A third challenge arises because not only are parents and physicians products of their own respective cultures, but also their interactions occur within a further “culture”—that of medicine and intensive care itself—with its own values, assumptions, and understanding of what should be done. Although the health care team should recognize how families’ interests are shaped by social, cultural, and other contexts, it is important not to stereotype the members of specific social, cultural, ethnic, or religious groups. Individuals’ affiliations may not be predictive of their beliefs and values in the care of their infant, and the health care team should regard each patient and family as unique and attend carefully to their specific views and values. When attentiveness to the views and values of families is difficult because of language barriers, professional interpreters should be used. Use of an interpreter is advisable for three reasons: (1) it ensures that parents’ views are available to the health care team, (2) it removes the burden on family members or friends for the transfer of information, and (3) it limits the potential for miscommunication. In certain situations, a cultural interpreter not only can facilitate language comprehension, but can also provide useful information about cultural norms and traditions that are unfamiliar to the health care team. Access to high-quality interpretation services is an essential component of ethically and culturally sensitive care. Religion and the more general concept of spirituality as a major determinant of culture, tradition, and family values often needs to be addressed with parents, particularly when end-of-life decision making is undertaken. A qualitative questionnaire study completed by parents after their child’s death revealed the emergence of four explicitly spiritual/ religious themes: prayer, faith, access to and care from clergy, and belief in the transcendent quality of the parent-child relationship that endures beyond death.66 Other significant themes with a religious/spiritual dynamic included finding meaning, hope, trust, and love. Similar findings were obtained in a study focused on the delivery room consultation: mothers stated that religion, spirituality, and hope were the major factors that guided their decision making.9 The implication of these studies is that health care teams need to consider whether they have or need to create an environment that is hospitable to, and supportive of, religious or spiritual practice, that clinical staff recognize parents’ spiritual needs and provide access to hospital chaplains and community clergy, and, on a deeper level, appreciate parents’ religious and spiritual perspectives in prenatal consultations and end-of-life discussions.

Consensual Decision Making Consensual decision making implies that the parents and the physician/health care team are involved in the decisionmaking process—that they share relevant information with each other, they express their treatment preferences, and when a final decision is made, all parties are in agreement. The ideal consensual decision is one in which neither party feels individually responsible for that decision. Parents often feel burdened by what they perceive as their responsibility for

the decision. When the process of consensual decision making is handled well by the physician/health care team, the burden of the consequences of the decision is shared.

Team Consensus in Ethical Issues How teams that normally function synergistically in terms of purely medical matters operate when dealing with ethical issues may be very challenging, not only because of the difficulty in defining roles and responsibilities in matters of ethical deliberation, but also from a more fundamental aspect in that the two major professions, medicine and nursing, have differed over time in their preparation for this practice.71 Medical professionals tend to view their role as one in which the best science is in the patient’s best interest, with less attention to deeper questions such as the goals and limits of medicine and the moral core of the profession. Seeing things from the patient’s point of view, relating to the patient’s family in a nonjudgmental way, emphasizing the healing potential of communication, and having respect for patients’ informed choices seem to be more intrinsic to the nursing tradition. In interactions concerning the end of life of a neonate, physicians see the focus of their moral obligations on decision making with parents, whereas neonatal nurses see their moral obligations focused on the process and moment immediately surrounding death.22 What is needed, according to Storch and Kenny,71 is “shared moral work”: Interprofessional practice implies that individual health care practitioners are aware of their own professional values and the need to work collaboratively, build understanding, and work toward resolution with other professionals, particularly when different perspectives threaten team function. The beneficial effects of multidisciplinary participation and perspectives have been well shown in teams developing guidelines for decision making, such as those at the limits of viability.6,36

CLINICAL APPLICATIONS IN SPECIFIC MORAL PROBLEMS Refusal of Treatment during Pregnancy When a pregnant woman acts in such a way as potentially to create a serious risk of harm for her developing fetus, some authorities argue that state intervention is morally justified in an effort to promote fetal health and well-being. Other authorities maintain that such intervention not only violates the pregnant woman’s autonomy, integrity, and privacy, but also undermines the principle of reproductive freedom.24 Authorities who advocate state intervention consider the principle of reproductive freedom to be morally unsound: Whatever rights a pregnant woman may have to direct the course of her pregnancy, they do not include the right to harm the fetus. This perspective ignores the contested moral status of the fetus; the fetus-to-be-born, in contrast to the infant or the pregnant woman, is not uniformly recognized as a person with full moral standing. This perspective also suggests that the pregnant woman intentionally seeks to harm her fetus, which is generally false. Other participants in this debate readily acknowledge that the fetus does not have the same rights as a person, and that the pregnant woman may not intend to harm her fetus. They

Chapter 3  Medical Ethics in Neonatal Care

maintain, however, that the woman has obligations to the fetus that include care. In choosing to continue her pregnancy, the woman is said to have incurred an obligation to do what is necessary (within reasonable limits) to ensure that the fetus is born healthy.51 If she violates this obligation, the state is entitled or obliged to intervene to protect the fetus. This argument misunderstands the context in which women continue their pregnancy, however, and ignores the overlapping interests of the fetus and the pregnant woman. The relationship between the pregnant woman and her fetus is falsely characterized as adversarial—the woman is cast in the role of aggressor with the fetus in the role of innocent victim—when in fact the situation is far more complex. Significantly, professional bodies have commonly resolved the moral dilemma between the pregnant woman’s autonomy and the fetus’s well-being in favor of respecting the principle of autonomy. The American College of Obstetricians and Gynecologists has stipulated, “Every reasonable effort should be made to protect the fetus, but the pregnant woman’s autonomy should be respected. . . . The use of courts to resolve these conflicts is almost never warranted.”3 Similarly, the Society of Obstetricians and Gynecologists of Canada Ethics Committee “opposes involuntary intervention in the lives of pregnant women. . . . The primary objective of physicians who work with pregnant women should be to promote women’s health and well-being while respecting their autonomy.”70 Many reasons explain the prevailing ethical view of not condoning state intervention in the lives of pregnant women. First, there is the principled commitment to respect personal autonomy. Second, there is the belief that state intervention harms the pregnant woman without any benefit accruing to her fetus (e.g., when the court intervenes after fetal harm has occurred). Third, there is the pragmatic concern that a policy of state intervention may discourage women whose fetuses are most at risk from seeking appropriate care, for fear of being prosecuted.49 Fourth, there are concerns about oppression and gender discrimination. State intervention is disproportionately oppressive toward poor and minority women. It also typically ignores paternal actions that are hazardous to the fetus.47 Fifth, state intervention in pregnancy is an intrusion into the lives of pregnant women in excess of anything that would be tolerated to protect nonfetal lives.24 For these compelling reasons, caring, compassionate health care providers confronted with a refusal of treatment during pregnancy are well advised to educate and attempt to persuade, but not to coerce, pregnant women.

Prenatal Consultation at the Limits of Viability With advances in medical technology, neonatology teams have developed the capacity to maintain physiologic signs of life at extremely low gestational ages, with survival possible after a gestation of 22 weeks. Use of a physiologic definition of viability, such as the point at which life can be maintained outside of the uterus, might suggest that every neonate born at such gestational ages should be given an opportunity for extrauterine life and be actively supported. Other definitions of viability do not focus exclusively on the likelihood of survival, but rather include quality-of-life considerations, in which there is an explicit value judgment regarding the degree

39

of morbidity that is acceptable if such a life were to be maintained. Viability, so understood, might suggest that full support should be provided only at a gestational age at which a “good enough” quality of life is foreseen. In clinical practice, there is no universally accepted definition of a viable fetus; however, guidelines from professional societies in several countries have defined gestational age ranges at which the benefit-to-burden ratio of aggressive obstetric or neonatal care becomes questionable.46,63,68 This “gray area” comprises 23 to 25 weeks of gestation. There seems to be consensus among most perinatal care providers that “the full gamut of intensive care should be provided at 26 weeks’ gestation and beyond.”23 The challenge facing the life-defining “resuscitate or not” issue for parents and the neonatal team in the “gray zone” is compounded by the “moral strangeness” between the participants, the vulnerability and unfamiliar physical and emotional context for parents, constraints of time, and the great degree of medical uncertainty in prognosticating outcome for individual infants. Nevertheless, the urgent context exerts its own demands, and the opportunity must be optimized because it is considered more conducive to an exploration of parental and physician/team views than if this conversation were to occur in the delivery room, or after delivery in the NICU in the first few hours of the infant’s life. A discussion with the pregnant patient about the limits of viability even earlier in pregnancy, free of the threat of imminent delivery and the lack of time for absorption of information and ability to participate objectively in such difficult decisions, has been proposed.14,67 Generating an advance directive for managing early delivery would obviate much uncertainty for the medical team; this approach may gain momentum, despite the potential of producing anxiety in most pregnancies that would continue to term. The prenatal consultation at the limits of viability has two primary aims: (1) to provide data on which decisions can be based, and (2) to explore the values of the pregnant patient and her partner and negotiate a shared decision between them and the medical team.

TO PROVIDE DATA ON WHICH DECISIONS CAN BE BASED It is important to have the best data available45—data that are current, based on gestational age rather than birthweight, and specific to the unit where the patient is being cared for (or the geographic region, where appropriate). The data need to include survival statistics and information about the quality of that survival, with quantitative and qualitative outcome measures (functional abilities, learning, behavior, impact on family). Fully informing parents requires more than simple disclosure of the potential harms associated with outcomes at each gestational age, because there is a tendency for parents to “use hope and denial to interpret the limits imposed by statistics.”17 Is there an optimal way of conveying this information? In an in-depth qualitative study on the discussion between neonatologists and parents at risk of premature delivery between 23 and 25 weeks, two divergent models were used by neonatologists regarding resuscitation decisions at this threshold.60 In a neutral information model, information was given, and parents ultimately decided—after information on mortality

40

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

statistics, risk estimates of complications, and sequelae had been provided. Parents were expected to manage the probabilities and uncertainties via their own decision-making process. In an assent model, the neonatologist’s preferences were clearly expressed, and a decision was sought during the consultation. Prognostic statistics were used as information to justify and reflect on the suggested course of action. From the parents’ perspective, neither the neutral information model nor the assent model fully addressed their expectations. In the first model, the neonatologist was described as a “messenger of uncertainty,” and parents felt essentially on their own to make decisions. Parents expressed the need for a more individualized and humane relationship, which could not be addressed by a clinical focus on “objective/neutral facts.” In the assent model, when decisions between parents and neonatologists were in accord, parents felt included, but when the neonatologist’s recommendation did not fit with the parents’ expectations, parents tended to feel abandoned and left on their own to confront the event. This study advocated that between the “liberal autonomous decision” and a “paternalistic decision-making process,”57 there is the need for an intermediate “shared relational space”—for a more caring relationship between the decision makers, and more time to allow exploration of the facts, expectations, and values that are inherent in this interaction.

TO EXPLORE THE VALUES OF THE PREGNANT PATIENT AND HER PARTNER AND NEGOTIATE A SHARED DECISION BETWEEN THEM AND THE MEDICAL TEAM It is crucial to explore with prospective parents their preferences in relation to the data provided. Some parents may regard 20% as a fair chance of a “good enough” outcome, whereas others may regard 80% as not enough of a guarantee.28 Elucidating parents’ views raises the question of how much moral weight should be given to these views for delivery room decisions. Leuthner43 described different weightings ascribed to parental views in terms of models of best interests: in a “medical expertise” model, outcome data are used in more directive counseling in pursuit of the physician’s judgment of the best possible outcome, with less parental input. In the “negotiated” model, parental input is maximized, and the decision attends to the moral values of the physician and the parents. The guidelines derived by Kaempf and associates36 seem to conform to these conceptual models in that they move from a more directive approach in not recommending NICU care at 23 to 24 weeks of gestation to a more negotiated model of counseling from 25 weeks of gestation, when most medical staff members do recommend NICU care. An important feature of this approach is the recommendation that parents be provided with the majority opinion of the unit’s staff and the responsible physician’s recommendation based on their own best professional judgment. It is worth reflecting on the reasons for the lack of congruency in decisions between physicians and parents, where differences seem to arise from three sources:60

interruption and fracture of the family narrative and a threat to their anticipated parenthood. 2. How information is used—physicians attempt to be as objective as possible with facts and figures, whereas parents reformulate those chances. 3. What constitutes the “right” decision—for physicians, it is after they are convinced the parents have fully understood the information; whereas for parents, it is after they have had their experience taken into account and it is supported by a scientifically competent and humane medical team. By focusing first on the parents, sharing two-way information, recognizing the importance of “relational space,” and overcoming to any degree these different starting points, physicians may achieve greater consensus and more acceptable decisions for all parties in this complex and contentious arena. The challenge in moving forward from a purely information-sharing interaction cannot be overemphasized; most neonatologists in New England who responded to a questionnaire viewed their primary role as providing factual information almost exclusively, with few considering their role being to assist in weighing risks and benefits, and only 2% regarding their primary role to discuss potential differences in views between parents and the medical team.5 Numerous guidelines and frameworks for decision making at the limits of viability have been developed and are helpful in setting out parameters of practice.36,63,68,79 They are generally not intended to be prescriptive, however, and decisions still need to be made specific for the individual patient, parent, physician, and team. The reminder that “decisions should be made on a case-by-case basis”12,23,48 requires that in each interaction, attention be paid to the many factors intrinsic to that unique situation. In addition, the individuals undertaking these decisions need to make their reasoning behind their approach explicit, because this has a major impact on the way they present information— how mortality and survival statistics are framed and options are discussed.31

Withholding and Withdrawing Life-Sustaining Medical Treatment in the Neonatal Intensive Care Unit Withholding life-sustaining medical treatment involves a choice to omit a form of treatment that is not considered beneficial, whereas withdrawal involves a choice to remove treatment that has not achieved its beneficial intent. From a moral perspective, there is no difference: If it is morally right (or wrong) to withhold treatment deemed to be ineffective, it is equally right (or wrong) to withdraw this same treatment after it is started, should it later become clear that the treatment is ineffective. Criteria for decisions to withhold or withdraw lifesustaining medical treatment are usually based on one of three general criteria:

1. Different starting points—physicians’ intent is to give

1. Inevitability of death. This is a situation where it is likely that

the facts, whereas families’ reaction is to the premature

the infant would die whether intensive care is continued or

Chapter 3  Medical Ethics in Neonatal Care

not. At minimum, the infant would likely not survive to discharge from intensive care. 2. Ineffective treatment. Treatment that is not meeting or would not meet the goals set for that treatment is considered ineffective. 3. Poor quality of life. The criteria involved in assessment of quality of life have been described earlier. Despite the difficulty in determining the quality of a life with limited cognitive or relational capacity, mobility, or self-awareness, or a life of continued pain and suffering, a poor quality of life is a valid consideration as to whether treatment should be initiated or continued in the face of an extremely poor predicted outcome. Some authors cite futility of medical treatment as a criterion for withholding or withdrawing medical treatment. It is difficult to know what follows from such claims, however, because there is considerable debate and diversity of opinion as to the meaning of futility.33 In some instances, continuing care is deemed futile because death is inevitable; in other instances, continuing care is considered futile because death is likely but uncertain, and in the event of survival, an extremely poor quality of life is inevitable. In addition to the debate about the meaning of the term futility, arguments have emerged about the authority of the physician to determine when an intervention is futile. Some authors contend that futility is a medical decision to be made by the physician alone, whereas others believe that the decision is value laden, and that parents should be involved in this determination. When futility is determined solely on the basis of medical or physiologic factors (a rare occurrence unless death is imminent), unilateral decision making by the physician based on sound medical knowledge and expertise may be appropriate. When subjective elements form part of the determination, however, the physician has no unique claim to moral expertise. Instead of attempting to center decision making using references to futility, physicians should state their reasons for considering withdrawing or withholding medical treatment. Use of the above-listed criteria—the inevitability of death, the low probability of successful treatment, or a poor predicted quality of life—should rather be cited. Although there is some hesitancy to use quality-of-life considerations, in the study by Wall and Partridge,76 23% of deaths resulting from a decision to withhold or withdraw medical treatment did include quality-of-life considerations.

COLLABORATIVE, PROCEDURAL FRAMEWORK FOR END-OF-LIFE DECISION MAKING Following a structured decision-making process helps to ensure that appropriate views and preferences are made explicit. Harmonious decision making is promoted, as all of the participants in the process may come to understand the reasons and values underlying a particular choice. The following procedural framework is suggested: 1. Create an optimal environment for discussion. It is important

to create a quiet and uninterrupted environment in which ethical issues and values can be thoroughly explored, despite the demands on the time and energy of parents and staff.

41

2. Establish that the presenting issue is an ethical problem, one

in which moral values conflict or moral uncertainty exists. Ethical deliberation is often complicated by communication problems and psychological issues. These need to be disentangled from the ethical issues. 3. Identify the rightful decision makers. Many individuals may legitimately be involved in the decision-making process, including, at least, the parents, the physician with primary responsibility, and other members of the health care team directly involved in the care of the patient. More generally, individuals who bear the greatest burden of care and conscience; individuals with special knowledge; and health care professionals with the most continuous, committed, and trusting relationship with the patient or parents should be involved in decision making.53 4. Establish the relevant facts. “Good ethics begins with good facts.” Medical facts include the diagnosis, the prognosis (and the estimated certainty of outcomes), past experience on the unit, relevant institutional policies, and relevant professional guidelines. Nonmedical facts include information about family relationships, language barriers, cultural and religious beliefs, and past experiences with the health care system. It is also important to ascertain parents’ understanding of the medical facts, parents’ expectations of the technology involved, the quality of communication between the parents themselves, and the degree of trust in physicians and the medical system. The willingness of the physician to discuss personal views and beliefs may enhance gathering of such information. 5. Explore the options. Explicit discussion of treatment options and their known potential short-term and long-term consequences should occur. For physicians, a troubling element at this stage is whether to describe all possible options or only the options they consider beneficial. Different physicians perceive their obligations differently. Physicians who feel morally obliged to inform the parents of all possible options should not hesitate, at the same time, to offer a professional recommendation on the course of action considered most appropriate.7 6. Develop consensus. All decision makers should be in agreement with the plan of action proposed at the time, even though, occasionally, agreement may be a temporizing measure. Open, honest discussion of the goals and consequences of treatment allows parents, physicians, and other legitimate decision makers to consider carefully a range of professional and personal beliefs, values, and preferences, and to explore reasoned arguments for and against various options meaningfully. This process holds the promise of harmonious, consensual decision making. The final responsibility does rest with the parents, however, and no decision to withdraw or withhold life-sustaining treatment would be made without their agreement. 7. Implement the decision. When a consensual decision has been reached, additional issues may need to be addressed with parents to ensure effective implementation. When the decision is made to withhold or withdraw lifesustaining treatment, there should be an open discussion about the manner of death, the option for parents

42

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

to be present, the performance of religious rituals, the anticipated grieving process, and the supports available for bereaved parents.80

SPECIFIC ISSUES IN END-OF-LIFE CARE Brain Death and Organ Donation Brain death is well defined for older age groups and has been accepted in most jurisdictions for more than 20 years as the threshold criterion for the removal of organs from heart-beating donors (without violating the dead donor rule). The diagnosis of brain death, particularly in infants less than 7 days old, is difficult, however, and is very rarely made. Numerous clinical issues limit the usefulness of the concept in this patient population, including the difficulty of establishing the exact cause of the coma, the clinical assessment of brain death (particularly the determination and reliability of the absence of brainstem reflexes), and the uncertainty regarding the validity of adjunctive laboratory tests. Aside from the clinical difficulties in diagnosing neonatal brain death, from an ethical perspective there are concerns that the process for determining brain death for the purpose of solid organ procurement may affect parental decision making and patient management. Introducing parents to the concept of brain death moves the conversation from a discussion of the inevitability of death or a predicted poor quality of life to a discussion about the criteria for whole brain silence and its confirmation. A possible negative consequence of this shift in conversation is that parents may mistakenly believe that withdrawal of life-sustaining medical treatment should be undertaken only when electrocerebral silence is confirmed. This belief could result in an infant being kept alive longer than might otherwise be the case, solely to determine if the criteria for brain death can be met. This situation raises the potential for the infant’s condition to change during the evaluation period, so that an infant with severe hypoxic-ischemic encephalopathy who would have died after withdrawal of assisted ventilation may survive in a severely compromised state. Even when both parents and caregivers wish to salvage some good from a tragic neonatal death, extreme caution should be exercised in discussions with grieving parents about neonatal brain death and potential organ donation.

Use of Analgesic Agents at the Time of Withdrawing Life-Sustaining Medical Treatment If we accept that an infant should not have to experience pain or discomfort at the end of life, choosing medications with the goal of achieving adequate pain relief or sedation is considered appropriate.54 Partridge and Wall59 showed that in most cases of withholding or withdrawing life support from critically ill infants, neonatologists provided opioid analgesia to these infants, despite the potential respiratory depression of these agents, the so-called double effect. The intent of the action—to alleviate pain and promote comfort—distinguishes the use of analgesics from the use of other agents, such as paralyzing agents, whose intent is to ensure death. The introduction of neuromuscular blocking agents at the time of

withdrawal of life-sustaining medical treatment is considered ethically inappropriate, as is any form of active euthanasia.72 Although the expressed intention of alleviating suffering and discomfort is usually cited as justification for the “double effect” of analgesics and sedatives, the distinction from hastening death is sometimes difficult. To some authors, such as April and Parker,4 the ban on active euthanasia (i.e., the use of drugs with the specific intent to end life) exists largely because professional guidelines in most jurisdictions call for the distinction, and as such is a reflection of current consensus. They contend that relief of symptoms and concern for a newborn’s suffering should be the primary goal, and the ban on physicians’ role in active euthanasia is in their view not a moral argument, but an appeal to public opinion and the moral primacy of the status quo. This issue has been stirred by the Groningen protocol in The Netherlands where, in rare cases, in a newborn with an extremely poor prognosis and intractable suffering, and after all measures to alleviate the suffering have been unsuccessful, the deliberate ending of life may be considered legal.75 Many authorities consider such active neonatal euthanasia to be unsupportable and believe that the Groningen protocol should be abandoned.35,40 Most authorities believe, similar to Costeloe,18 that, despite the conceptual difficulty in defining the difference between drugs used to alleviate pain and drugs given with the intention of ending life, in practice, “experienced neonatologists and neonatal nurses feel comfortable with this distinction. They can discuss it openly with families and help them to understand the acceptability of infusing opiates at a dose that controls pain and distress, but the unacceptability of increasing the dose further with the primary intention of hastening death.18

PALLIATIVE CARE IN THE NEONATAL INTENSIVE CARE UNIT Until more recently, withdrawal or withholding of lifesustaining treatment in the NICU tended to be considered only when life-prolonging treatments were determined to be ineffectual and burdensome, and death seemed imminent. Many factors may account for the slow adoption of palliative care in the NICU, such as the uncertainty in determining an infant’s terminal prognosis, the difficulty in moving away from an interventionist approach when therapeutic options still seem possible, divergent caregiver perceptions of the “right” action, a degree of moral and legal ambiguity in pursuing such a course, fear of a “lingering death,”52 a tendency for the compartmentalization of medical care into specialist teams with differing agendas, and a lack of formal training and experience in this aspect of care. The dilemma of continuing the intensive care of a critically ill newborn when it is likely to result in a prolonged dying process or survival with profoundly limited capabilities and long-term suffering for the infant and his or her family has highlighted the need for a change in this approach. This has led to the incorporation of palliative care practices into the NICU environment, where a cure-oriented approach is replaced by an entire spectrum of management to prevent and relieve suffering and improve the transition to dying. This approach has been supported in three general clinical

Chapter 3  Medical Ethics in Neonatal Care

situations: (1) congenital anomalies incompatible with life, (2) newborns born at the limits of viability, and (3) infants who have overwhelming illness not responding to lifesustaining intervention.55 This approach is also being increasingly considered when a diagnosis is made prenatally in conditions of certain lethality. Palliative care principles can be used to guide supportive conversations during the pregnancy and enable the preparation for a perinatal death in facilities supportive of such an approach. Greater adoption of palliative care at the time of birth for fetally diagnosed lethal anomalies has had an impact on decisions of pregnant women regarding termination during the pregnancy.11 Palliative care involves a team approach to the prevention and relief of physical, psychological, social, and spiritual suffering for the dying infant and the family.15 Protocols aim to ensure continuity of care; symptom management and comfort care for the infant; family-centered decision making; practical, emotional, and spiritual support for the parents; and organizational, emotional, and spiritual support for the intensive care clinicians.16 It is crucial that the neonatal staff fully explain their reasoning and solicit the agreement of parents when considering the move to a “comfort care” approach, and that the care plan be developed with input from the parents, appropriate hospital and community resources, and occasionally the district coroner, in situations in which there is the strong likelihood of the infant dying at home. Within the spectrum of palliative care, after other, more obviously invasive forms of life-sustaining medical treatment have been withheld or withdrawn, the withdrawal of artificial hydration and nutrition may become a focus of consideration. Justification for this practice (after clearly showing the inability of an infant to tolerate oral feeds safely) revolves around the question of whether providing hydration and nutrition via other routes is a medical treatment or an obligatory part of simple humane care. All forms of maintaining artificial hydration and nutrition (the passage of nasogastric tubes, the insertion of intravenous needles, or more invasive interventions such as the surgical placement of a gastrostomy tube) are invasive to some extent and carry medical risks of dislodgment, migration, error, infiltration, and infection.34 There seems to be general agreement, at least over the past two decades, that artificial nutrition and hydration is a medical treatment, on a par with mechanical ventilation and other life-sustaining technologies,34,73 and that it should not be held to a higher standard than other forms of life-sustaining treatment.13 In terms of the best interests standard of judgment in which the benefits and burdens of withdrawing or withholding hydration and nutrition are considered, studies in adult patients have shown that death is due to dehydration, not starvation, and that dehydration leads to a decrease in nausea, vomiting, diarrhea, and urine output, with little, if any, discomfort, perhaps because of the release of endogenous opioids with fasting and ketosis.81 In addition, patients experience little, if any, hunger or thirst if appropriate mouth care is provided.20 Justification is in accord with the application of the respect for parents in their role as decision makers for the care of their newborn infants: Their decision to refuse artificial nutrition and hydration as medical treatment is respected if based on their child’s best interests, unless the parents are deemed incompetent or are disqualified on other grounds.

43

Withdrawal of artificial hydration and nutrition in newborns is considered an ethically legitimate option when it is clear that cure is no longer possible, and that an interventionist approach does not serve the child’s (or family’s) best interests.44,56 Despite the acceptance by many authors of withdrawal of hydration and nutrition as a moral way of responding to severe terminal suffering,64 the literature has highlighted difficulties, particularly for nurses with “ethical proximity” to the patient, when this has been implemented.42,62 These types of issues require significant attention if this practice is to become more clinically accepted.

CONFLICT RESOLUTION WHEN CONSENSUS CANNOT BE REACHED In most instances, participants in a decision-making process can arrive at a morally sound decision regarding the best course of action in a particular situation. Attempts at consensual decision making are sometimes unsuccessful, however, and may result in conflict and intractability. Occasionally, this conflict is between parents who disagree with each other regarding what is in the best interests of their newborn. More frequently, the conflict is between the parents of a newborn and the health care providers who have different perceptions of the child’s best interests. The following guidelines are proposed to promote continuing negotiation and resolution of conflict in the NICU: 1. Allow time for further clinical observation. In the specific

case of parental objections to forgoing treatment that the physician believes is not beneficial, it may be unrealistic to expect agreement from the parents the first time this option is raised. Prudence suggests moving as fast as the slowest member of the decision-making group, provided that the infant is not compromised further. 2. Ensure full parental comprehension of the medical information. Early expressions of treatment preference by parents, such as “do everything possible,” need to be examined carefully. There is a tendency for busy medical teams to reduce parental expressions into simple one-line statements and not to explore their meaning. Such statements may be an expression of parental love or an expression of frustration with the medical team, and not really an informed choice about what is in the best interests of the newborn. In addition, in view of parents’ potential denial of the severity of their infant’s condition, it is important to check parents’ understanding repeatedly. If it seems that inconsistent information has been provided by different individuals or teams, it may be advisable to convene a formal interdisciplinary case conference to explicate these differing viewpoints. 3. Continue to discuss, explore, and challenge the underlying reasons for the differences in choice. It is important for health care providers to try to understand the parents’ views, beliefs, and preferences. Physicians must recognize that the parents’ beliefs and values are informed by ethnic and cultural traditions, customs, and institutions, and that these influences may be significantly divergent from their own. 4. Continue to negotiate toward consensus. Parents may have great difficulty in making an unassisted decision. The

44

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

burden and potential guilt of decision making experienced by parents is often immense and may be underappreciated by the medical team. The consensual nature of a joint decision-making process and the shared burden of the decision must be reinforced. In addition, there may be medically imposed obstacles to achieving consensus, such as frequent changes in the responsible physician from one neonatologist to the next, the failure of a physician to establish a therapeutic alliance with the parents, the persistence of communication barriers between the physician and the parents, or the development of a “contest of wills” between the physician and parents as to who will sway the other. All these factors need to be overcome when negotiating with parents toward consensus. 5. Broaden the parents’ moral community. Identifying the locus of decision-making authority within each family’s moral community may involve the inclusion of additional family members, grandparents, significant others, and religious or spiritual advisors in meetings with parents and physician/ team or in less structured communication opportunities. 6. Share the attending physician’s moral load by actively seeking opinions from colleagues. Although the physician responsible for the infant bears the final responsibility for the approach taken with the parents, for personal and legal reasons, it is important to establish that the physician is supported by a responsible body of medical opinion. 7. Involve a bioethicist or ethics consultation team, as appropriate. It may be beneficial to involve an institutional ethics committee with experience in case consultation and review, a clinical ethics consultation team, or a clinical ethicist, depending on available resources. Most hospital ethics committees include experts in medicine, nursing, philosophy, law, religion, and social services. The function of these committees varies widely. In some institutions, the ethics committee as a whole reviews cases; more commonly, a smaller subcommittee, an infant care review team, or an individual ethics consultant undertakes this responsibility. It is important to appreciate the philosophy and mandate of the institution’s ethics committee because this may vary from a decision-making body via adjudication of conflicting rights and claims to one in which decisions are facilitated via an exercise in values clarification (i.e., a process to ensure that facts are confirmed, relevant parties are involved, and decision makers are empowered, but not to be the decisionmaking group per se). In a multidisciplinary ethics committee, team members may disagree among themselves, some members may dominate others, the committee or consultation team may not be qualified to deal with the subject matter, or the team may be overly concerned with the institutional impact of a decision rather than with the specifics of the case. Experience suggests that consultation with individual bioethicists and smaller ethics consultation teams may be of greater benefit to decision makers than the practice of “ethics by committee.” 8. Consider transferring responsibility of care for the infant to another, accepting physician. When differences of opinion remain, and the degree of physician moral compromise is significant, it may be advisable to involve another staff member with whom the parents have formed a therapeutic alliance.

Despite these efforts, the moral problem may not be amenable to consensual resolution. Rational people of good will may hold views that are irreconcilable. Individuals involved in a failed attempt at deriving consensus may experience what Webster and Baylis77 term moral residue—“that which each of us carries with us from those times in our lives when in the face of moral distress we have seriously compromised ourselves or allowed ourselves to be compromised.” Until consensus can be achieved, withdrawal of lifesustaining medical treatment should not be undertaken. Rarely, physicians seek authority to make unilateral decisions via institutional or legal redress. When ethical conflict seems intractable, an institutional decision may be made to seek legal recourse. This course of action is generally unsatisfactory: It increases the anguish for patients and families, it destroys the parent-physician relationship, it creates (or increases) conflict between members of the health care team, and it invariably results in a significant drain on staff time and morale. It can also be extremely costly and timeconsuming for all parties involved. Use of the judicial system can be very damaging to the institution’s reputation. Ideally, institutional policies developed by the hospital ethics committee and staff should be in place to minimize the need for judicial intervention.

ETHICS OF RESEARCH IN THE NEONATAL INTENSIVE CARE UNIT There is no lack of justification for the conduct of research in sick newborn infants; improvements in the practice of clinical neonatology would not occur without such research. Research in children, let alone newborn infants, has not always been easily justified, however. Until more recently, a very protectionist view prevailed, such that studies in children were justified only if they were unsuitable in adults. This perspective seems to have changed, reflecting the move from the dominance of beneficence or protectionism toward vulnerable groups, such as infants, to a stance based more on the principle of justice as the important consideration whereby individuals who are the subject of a treatment should have an equal opportunity to share the benefits and risks of human research.25 This move is also no doubt a reflection of the fact that newborns have benefited, but also have been harmed, by the adaptation of results of treatment in other groups being applied to them without adequate research. The view that neonates should have an equal opportunity to participate in research accepts a certain amount of justified risk. The current view is that research can be approved if (1) the risk is only a minor increase over minimal risk, (2) the research is likely to yield important results, and (3) the research presents an important opportunity to gain knowledge. This view leads to the important debate over what constitutes a “minor increase over minimal risk” and “more than a minor increase over minimal risk,” such that every research proposal needs to categorize each risk potential carefully and accurately.78 Despite the imperative for research, the NICU is a difficult context for its ethical requirements. There is a great challenge to obtaining “authentic,” morally valid, informed consent from parents, because this requires surrogate decision making by

Chapter 3  Medical Ethics in Neonatal Care

anxious and stressed decision makers, usually following an unanticipated, acute emergency. In addition, parents are often young, healthy members of society with little prior medical exposure and familiarity with the concept of medical research. Language and cultural and religious diversity add further complexity, and the act of soliciting consent itself often further exacerbates parental stress. Parents or guardians feel beholden to the caregivers of their vulnerable infants, and in cases in which the relationship between caregiver and researcher is unclear there is the potential for “therapeutic misconception” (i.e., that the researcher is acting in the best interests of the infant). It may also be disturbing for parents to learn that there is much uncertainty about neonatal practice, such that their confidence in NICU caregivers may be diminished. Before discussing informed consent in neonatal research, other more general issues facing the ethical conduct of research need to be ensured, including the scientific value and validity of the study proposal, the existence of clinical equipoise in randomized trials, the distinction between therapeutic and nontherapeutic research (where the study would not lead to direct benefit for that infant), the existence of any potential conflicts of interest or financial incentives, and the overall risk/ benefit analysis of the research proposal. Informed consent is enshrined as a foundational cornerstone of the ethical practice of protecting human subjects from research risk.39 The four domains within informed consent are (1) disclosure of information, (2) understanding, (3) competence or capacity, and (4) voluntariness or freedom to choose. Decisions that adults make on their own are morally robust, but decisions made by others cannot have the same degree of authenticity and are necessarily less valid in children.39 Despite these challenges, it is widely accepted that neonatal research investigators have the obligation to approximate truly informed consent to the greatest extent possible. Golec and colleagues27 described various models of consent in neonatal research: the standard model where parents are solicited when their infant becomes eligible for a study, and they are given written and verbal information, encouraged to ask questions, and required to sign a consent form. Other means of obtaining consent have included a steplike process of consent2; advanced consent, in which parents are approached in anticipation that their infant may meet inclusion criteria at a later date; emergency consent; and randomization without consent, whereby randomization occurs before potential participants are approached, and only participants allocated to experimental therapies are informed of the trial and invited to give or withhold consent (Zelen randomization).69 The individuals allocated to continue with standard therapies are not informed that they are trial participants at that stage. The optimal model for consent for neonatal research is one that protects and promotes parental autonomy, is sensitive to the vulnerability and stress of the parents, and is beneficent to the infants. Golec and colleagues27 suggested that a “morally optimizing approach to research recruitment” is one in which: . Parents are approached one study at a time. 1 2. Studies have relevance to the current clinical status of the

neonate.

. Researchers minimize information overload. 3 4. Researchers promote respect for parental autonomy.

45

5. Researchers inform parents about research in a continuing

process, not as an event.

. Adequate time is allowed to make decisions. 6 7. There are limits on subsequent solicitation approaches.

No model of soliciting parental permission is without some difficulty, and the concept of informed consent cannot be the sole safeguard protecting the welfare of the neonate in research studies. The integrity of the researcher, the role of the institutional review board, and safety and monitoring by all individuals involved help to ensure the safety and protection of neonates. Another important resource to ensure ethical conduct of neonatal research is the role of the NICU nurse. Not only are nurses in a key position to promote parental understanding and improve the likelihood that the conditions of informed consent are met, but nurses also can help set priorities in neonatal research by playing an active role in the formulation of research policy and being “at the table” where priorities are set.25

ETHICAL RESPONSIBILITIES OF NEONATAL PHYSICIANS How neonatal medicine is practiced will certainly change, but the fundamental ethical obligations of neonatal physicians and health care practitioners will remain unchanged (Box 3-3). The physician’s responsibility to the competent pregnant patient is defined and well established—to respect the patient’s wishes regarding treatment even when this may be contrary to fetal best interests. So too, the physician’s responsibility to the neonatal patient is well defined and established—right and good action within the best interests standard of judgment. In the broadest terms, these responsibilities are best discharged in the context of a constructive and mutually respectful parent-physician relationship that recognizes that patient and family values and beliefs are integral to the decision-making process. Physicians’ responsibilities include the necessity of challenging parental views that they consider contrary to the patient’s best interests. In the context of team medicine, the attending physician is responsible for developing and maintaining positive relationships with all members of the health care team to

BOX 3–3 Ethical Responsibilities of Neonatal Physicians To neonatal patients—”right and good” action in their best interests To parents—constructive, respectful relationship To NICU team—leadership, direction with open, questioning culture To trainees—educational experience To institution—in accord with mission, maintenance of data, review of practice To society—trust in profession, technical competence, and moral discretion in resource use To self—moral conscience NICU, neonatal intensive care unit.

46

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

promote better, open discussion of moral problems and to ensure consensus among the individuals directly involved in the patient’s care. The physician also has an obligation to foster the ethical experience and education of the interdisciplinary team and that of junior staff and trainees. The perinatal high-risk unit or NICU team should be more than a group of physicians, nurses, and many other professionals working in an isolated area of the hospital trying to master new technology and break new ground. Ideally, it should be an open, analytic, self-critical, and responsive group providing ethically responsible care to pregnant women, newborn infants, and their families. By setting a standard of ethical responsibility for the care of newborns and families, physicians working in neonatal care send a message to society that promotes public confidence and trust in their professional practice and responsible use of expensive resources. Finally, all health care practitioners in neonatal care need to consider their own moral conscience and agency. Ethics is essentially a reflective task that requires participants to be explicit about what they believe and why, and what they value and on what grounds.

REFERENCES 1. Alderson P et al: Parents’ experiences of sharing neonatal information and decisions: consent, cost and risk, Soc Sci Med 62:1319, 2006. 2. Allmark P et al: Obtaining consent for neonatal research, Arch Dis Child Fetal Neonatal Ed 88:F166, 2003. 3. American College of Obstetricians and Gynecologists, Committee on Ethics: Patient choice: maternal-fetal conflict. Washington, DC, 1987, American College of Obstetricians and Gynecologists, Committee Opinion No. 55. 4. April C, Parker M: End of life decision-making in neonatal care, J Med Ethics 33:126, 2007. 5. Bastek TK et al: Prenatal consultation practices at the border of viability: a regional survey, Pediatrics 116:407, 2005. 6. Baumann-Hölzle R et al: A framework for ethical decision making in neonatal care, Acta Paediatr 94:1777, 2005. 7. Baylis F, Downie J: Professional recommendations: disclosing facts and values, J Med Ethics 27:20, 2001. 8. Beauchamp TL, Childress JF: Principles of biomedical ethics, New York, Oxford University Press, 2001. 9. Boss RD et al: Values parents apply to decision-making regarding delivery room resuscitation for high-risk newborns, Pediatrics 122:583, 2008. 10. Boss RD: End-of-life decision-making for infants abandoned in the neonatal intensive care unit, J Palliat Med 11:109, 2008. 11. Breeze ACG et al: Palliative care for prenatally diagnosed lethal fetal abnormality, Arch Dis Child Fetal Neonatal Ed 92:F56, 2007. 12. Canadian Paediatric Society, Maternal-Fetal Medicine Committee, Society of Obstetricians and Gynaecologists of Canada: Management of the woman with threatened birth of an infant of extremely low gestational age, Can Med Assoc J 151:547, 1994. 13. Casarett D et al: Appropriate use of artificial nutrition and hydration—fundamental principles and recommendations, N Engl J Med 353:2607, 2005. 14. Catlin A: Thinking outside the box—prenatal care and the call for a prenatal advance directive, J Perinat Neonatal Nurs 19:169, 2005.

15. Catlin A, Carter B: Creation of a neonatal end of life palliative care protocol, J Perinatol 22:184, 2002. 16. Clarke EB et al: Quality indicators for end-of-life care in the intensive care unit, Crit Care Med 31:2255, 2003. 17. Cole FS: Extremely preterm birth—defining the limits of hope, N Engl J Med 343:429, 2000. 18. Costeloe K: Euthanasia in neonates, BMJ 334:912, 2007. 19. Dudek-Shriber L: Parent stress in neonatal intensive care unit and the influence of parent and infant characteristics, Am J Occup Ther 58:509, 2004. 20. Ellershaw JE et al: Dehydration and the dying, J Pain Symptom 10:192, 1995. 21. Emanuel EJ, Emanuel LL: Four models of the physician-patient relationship, JAMA 267:2221, 1992. 22. Epstein EG: End-of-life experiences of nurses and physicians in the newborn intensive care unit, J Perinatol 28:771, 2008. 23. Fanaroff AA: Extremely low birthweight infants—the interplay between outcomes and ethics, Acta Paediatr 97:144, 2008. 24. Flagler E et al: Bioethics for clinicians, XII: ethical dilemmas that arise in the care of pregnant women: rethinking “maternal-fetal conflicts,” Can Med Assoc J 156:1729, 1997. 25. Franck LS: Research with newborn participants: doing the right research and doing it right, J Perinat Neonatal Nurs 19:177, 2005. 26. Gillis J: We want everything done, Arch Dis Child 93:192, 2008. 27. Golec L et al: Informed consent in the NICU setting: an ethically optimal model for research solicitation, J Perinatol 24:783, 2004. 28. Greisen G: Meaningful care for babies born after 22, 23 or 24 weeks, Acta Paediatr 93:153, 2004. 29. Hardart GE, Truog RD: Attitudes and preferences of intensivists regarding the role of family interests in medical decision making for incompetent patients, Crit Care Med 31:1895, 2003. 30. Hardwig J: What about the family? Hastings Cent Rep 20:5, 1990. 31. Haward MF et al: Message framing and perinatal decisions, Pediatrics 122:109, 2008. 32. Hefferman P, Heilig S: Giving “moral distress” a voice: ethical concerns among neonatal intensive care unit personnel, Camb Q Healthc Ethics 8:173, 1999. 33. Helft P et al: The rise and fall of the futility movement, N Engl J Med 343:293, 2000. 34. Johnson J, Mitchell C: Responding to parental request to forego pediatric nutrition and hydration, J Clin Ethics 11:128, 2000. 35. Jotkowitz AB, Glick S: The Groningen protocol: another perspective, J Med Ethics 32:157, 2006. 36. Kaempf JW et al: Medical staff guidelines for periviability pregnancy counseling and medical treatment of extremely premature infants, Pediatrics 117:22, 2006. 37. Keenan HT et al: Comparison of mothers’ and counselors’ perceptions of predelivery counseling for extremely premature infants, Pediatrics 116:104, 2005. 38. King NM: Transparency in neonatal intensive care, Hastings Cent Rep 22:18, 1992. 39. Kodish E: Informed consent for pediatric research: is it really possible? J Pediatr 142:89, 2003. 40. Kon AA: Neonatal euthanasia is unsupportable: the Groningen protocol should be abandoned, Theor Med Bioeth 28:453, 2007.

Chapter 3  Medical Ethics in Neonatal Care 41. Kopelman LM, Kopelman AE: Using a new analysis of the best interests standard to address cultural disputes: whose data, which values? Theor Med Bioeth 28:373, 2007. 42. Kuczewski MG et al: Providing comfort or prolonging death for a baby with “dead gut syndrome”? Camb Q Healthc Ethic 8:538, 1999. 43. Leuthner SR: Decisions regarding resuscitation of the extremely premature infant and models of best interest, J Perinatol 21:193, 2001. 44. Leuthner SR, Carter BS: Artificial hydration and nutrition in the neonate: ethical issues. In Bhatia J, editor: Perinatal nutrition: optimizing infant health and development, New York, 2005, Marcel Dekker, pp 347-362. 45. Levene M: Is intensive care for immature babies justified? Acta Paediatr 93:149, 2004. 46. Lorenz JM: Ethical dilemmas in the care of the most premature infants: the waters are murkier than ever, Curr Opin Pediatr 17:186, 2005. 47. Losco J, Shublak M: Paternal-fetal conflict: an examination of paternal responsibilities to the fetus, Polit Life Sci 13:63, 1994. 48. MacDonald H: Perinatal care at the threshold of viability, Pediatrics 110:1024, 2002. 49. Mahowald MB: Women and children in health care: an unequal majority, New York, 1993, Oxford University Press. 50. Martin M, Vaughn B: Cultural competence: The nuts and bolts of diversity and inclusion. Strategic Diversity and Inclusion Management Magazine 2004. Available at URL: http:// diversityofficemagazine.com/magazine/?page_id5391. 51. McCullough LB, Chervenak FA: Ethics in obstetrics and gynecology, New York, 1994, Oxford University Press. 52. McHaffie HE et al: Lingering death after treatment withdrawal in the neonatal intensive care unit, Arch Dis Child Fetal Neonatal Ed 85:F8, 2001. 53. Mitchell C: Care of severely impaired infant raises ethical issues, Am Nurse 16:9, 1984. 54. Munson D: Withdrawal of mechanical ventilation in pediatric and neonatal intensive care units, Pediatr Clin North Am 54:773, 2007. 55. Munson D, Leuthner SR: Palliative care for the family carrying a fetus with a life-limited diagnosis, Pediatr Clin North Am 54:787, 2007. 56. Nelson LJ et al: Forgoing medically provided nutrition and hydration in pediatric patients, J Law Med Ethics 23:33, 1995. 57. Orfali K: Parental role in medical decision making: fact or fiction? A comparative study of ethical dilemmas in French and American neonatal intensive care units, Soc Sci Med 58:2009, 2004. 58. Paris JJ et al: Has the emphasis on autonomy gone too far? Insights from Dostoevsky on parental decision-making in the NICU, Camb Q Healthc Ethic 15:147, 2006. 59. Partridge JC, Wall SN: Analgesia for dying infants whose life support is withdrawn or withheld, Pediatrics 99:76, 1997. 60. Payot A et al: Deciding to resuscitate extremely premature babies: how do parents and neonatologists engage in the decision? Soc Sci Med 64:1487, 2007. 61. Pellegrino E: Physician and philosopher: The Philosophical Foundation of Medicine: Essays by Dr. Edmund Pellegrino. In Bulger RJ, McGovern JP, editors: Physician and philosopher, Charlottesville, VA, 2001, Carden Jennings. 62. Penticuff J: Nursing perspectives on withholding food and fluids in pediatrics. In Frankel LR et al, editors: Ethical dilemmas

47

in pediatrics: cases and commentaries, New York, 2005, Cambridge University Press. 63. Pignotti MS, Donzelli G: Perinatal care at the threshold of viability: an international comparison of practical guidelines for the treatment of extremely preterm births, Pediatrics 121:e193, 2008. 64. Quill TE, Byock IR: Responding to intractable terminal suffering: the role of terminal sedation and voluntary refusal of food and fluids, Ann Intern Med 132:408, 2000. 65. Rhoden NK: Treating Baby Doe: the ethics of uncertainty, Hastings Cent Rep 16:34, 1986. 66. Robinson MR et al: Matters of spirituality at the end of life in the pediatric intensive care unit, Pediatrics 118:e719, 2006. 67. Schroeder J: Ethical issues for parents of extremely premature infants, J Pediatr Child Health 44:302, 2008. 68. Seri I, Evans J: Limits of viability: definition of the gray zone, J Perinatol 28(suppl 1):S4, 2008. 69. Snowdon C et al: Making sense of randomization; responses of parents of critically ill babies to random allocation of treatment in a critical trial, Soc Sci Med 45:1337, 1997. 70. Society of Obstetricians and Gynaecologists of Canada (SOGC): SOGC Clinical Practice Guidelines, Policy Statement No. 67: Involuntary intervention in the lives of pregnant women, J Soc Obstet Gynaecol Can 19:1200, 1997. 71. Storch KJ, Kenny N: Shared moral work of nurses and physicians, Nurs Ethics 14:478, 2007. 72. Truog RD et al: Pharmacologic paralysis and withdrawal of mechanical ventilation at the end of life, N Engl J Med 342:508, 2000. 73. Truog RD, Cochrane TI: Refusal of hydration and nutrition irrelevance of the “artificial” vs “natural” distinction, Arch Intern Med 165:2574, 2005. 74. Ulrich C et al: Ethical climate, ethics stress, and the job satisfaction of nurses and social workers in the United States, Soc Sci Med 65:1708, 2007. 75. Verhagen E: End of life decisions in newborns in the Netherlands: medical and legal aspects of the Groningen protocol, Med Law 25:399, 2006. 76. Wall SN, Partridge JC: Death in the intensive care nursery: physician practice of withdrawing and withholding life support, Pediatrics 99:64, 1997. 77. Webster G, Baylis F: Moral residue. In Rubin S, Zoloth L, editors: Margin of error: the ethics of mistakes in the practice of medicine, Hagerstown, MD, 2000, University Publishing Group. 78. Wendler D, Emanuel EJ: What is a “minor” increase over minimal risk, J Pediatr 147:575, 2005. 79. Wilkinson AR et al: Management of babies born extremely preterm at less than 26 weeks of gestation: a framework for clinical practice at the time of birth, Arch Dis Child Fetal Neonatal Ed 94:F2, 2009. 80. Williams C et al: Supporting bereaved parents: practical steps in providing compassionate perinatal and neonatal end-of-life care—a North American perspective, Semin Fetal Neonatal Med 113:335, 2008. 81. Winter S: Terminal nutrition: framing the debate for the withdrawal of nutritional support in terminally ill patients, Am J Med 109:723, 2000. 82. Zupancic JA et al: Characterizing doctor-parent communication in counseling for impending preterm delivery, Arch Dis Child Fetal Neonatal Ed 87:F113, 2002.

CHAPTER

4

Legal Issues in Neonatal-Perinatal Medicine Jonathan M. Fanaroff and Robert Turbow

Clinicians, especially clinicians working in the intensive care unit environment, are accustomed to a modicum of predictability. Treatment plans are generally based on years of clinical experience coupled with robust dialogue with one’s colleagues. How should physicians react to the $21 million “wrongful birth” verdict against a geneticist who missed the diagnosis of Smith-Lemli-Opitz syndrome? How does one prepare for the resuscitation of a 23-week gestation infant knowing that physicians have been sued for resuscitating such infants, and others have been sued for failing to resuscitate them? Does the Born-Alive Infants Protection Act require that 22-week fetuses be given a trial of an endotracheal tube? The medical profession tends to view the legal system with mistrust. The unfamiliar concepts and vocabulary coupled with the seemingly unpredictable nature of legal decision making can create an environment of confusion and apprehension. For various reasons, these concerns are particularly acute for neonatal-perinatal practitioners. Tremendous clinical and ethical uncertainty can surround the decision to resuscitate an extremely premature infant. Neonatologists are often asked to attend deliveries for premature infants at the limits of viability. Must the physician honor the parents’ requests? What if the parents request that their extremely premature infant not be resuscitated? What are the roles and duties of the perinatologist, the neonatologist, and the hospital administration? These questions are not obscure or theoretical. A $60 million verdict by one jury, subsequently overturned, accentuates the importance of a clinician’s familiarity with the laws that affect clinical practice. In these cases, the legal system can seem capricious and arbitrary. When the stakes are so high, and there is a lack of applicable case law, it is understandable that clinicians are left in a quandary. During residency and fellowship training and after training is completed, clinicians interact with the legal system. This interaction may be in the form of a contract with a new employer, a lease for office space, or as a defendant in a

medical malpractice suit. This chapter focuses on more recent legal developments in perinatal-neonatal medicine that can affect the daily professional lives of individuals who work in high-risk maternal units, delivery rooms, and neonatal intensive care units (NICUs). Several complex issues are addressed: What are the legal ramifications of a neonatologist disregarding a parent’s request to forgo delivery room resuscitation? What are a physician’s liabilities when providing phone supervision of an ambulance transfer of a critically ill patient? What are the elements of a medical malpractice case? Practicing clinicians must understand their rights, duties, and liabilities as physicians. They must understand the legal relationship that they have with their employers, the hospital, referring physicians, consultants, and neonatal nurse practitioners (NNPs) and physician assistants (PAs) they supervise. This chapter also assists the clinician in understanding basic terms and concepts of medical law. A certain baseline vocabulary is necessary to discuss the relevant issues adequately. Terms and definitions are introduced throughout the chapter. Additionally, certain landmark cases are discussed. This chapter provides legal background so that the clinician has a more complete understanding of the legal principles, cases, and statutes that affect the daily practice of perinatalneonatal medicine.

DISCLAIMER The authors of this chapter have attempted to provide a background or framework of law for the purpose of educating clinicians. Nothing contained in this chapter should be viewed as substantive legal advice. This chapter does not create an attorney-client relationship between the authors and any readers. Laws generally vary from state to state. Federal laws may represent a separate body of rules that can affect a given practitioner.

49

50

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Legal cases often hinge on very specific facts. Courts and juries make determinations based on the facts of a given case. A slight variation in circumstances, or the same facts argued by a different lawyer or in front of a different jury, can result in a completely different legal outcome. A practitioner should never assume that his or her situation is identical to the parties in another situation. Seasoned practitioners recognize that not all 27-week gestation infants with respiratory distress syndrome have identical courses. Likewise, each legal situation has its own nuances that can determine a distinct outcome. Courts expend considerable effort to distinguish the facts when comparing one case with another. The authors neither advocate for nor reject the judicial decisions and legislative actions described in this chapter. Readers should become familiar with the current laws that affect practice in their state. If readers have specific questions, they should consult with a qualified attorney.

GENERAL LEGAL PRINCIPLES Legislative Law and Case Law Individuals unfamiliar with the U.S. legal system may have difficulty understanding the distinction between case law and statutes. Generally, a significant portion of U.S. law is based on the common law. These laws have roots in English law from the last few hundred years. In many ways, the common law provides the foundation for the U.S. perspective on contracts, property, torts, criminal law, evidence, and many other legal disciplines. The common law was created by judges who were generally evaluating disputes between parties. More recently, well-known cases, such as Brown v Board of Education or Roe v Wade are examples of judicial decisions that became U.S. law. These were cases that involved defined parties. The U.S. Supreme Court made a determination, and the law was established. In addition, many laws are created by elected legislative bodies, such as the U.S. Congress or a state legislature. Often, court opinions state that it is not the role of the judiciary to redefine or change the definition of laws that were created by a legislative body; rather, the legislature is generally responsible for changing the law. The case of Vo v Superior Court79 is illustrative. This Arizona case involved a woman who was shot in the head during a drive-by shooting on the freeway. The woman and her 23-week fetus died as a result of the shooting. The prosecutor subsequently charged Nghia Hugh Vo with two counts of murder. The Arizona Court of Appeals considered the propriety and legality of charging Vo with two counts of murder. The court stated that when the legislature created the murder statutes, they did not intend to include a fetus in the definition of a person or human being. The court concluded that the unlawful killing of a fetus could not be murder. Then the court stated that if the legislature intended to include a fetus in the definition of a person, it was the responsibility of the legislature to change the homicide statute. Shortly after the Vo opinion, the Arizona legislature amended the manslaughter statute to include the intentional or reckless killing of an

“unborn child at any stage of its development by any physical injury to the mother of such child which would be murder if the death of the mother had occurred.”12 The Vo case serves as an example of the dynamic balance between the two branches of government that create law. In this case, the judges stated that it was the responsibility of the legislature to change the definition of manslaughter. The legislature responded to this case by expanding the definition of manslaughter to include unlawful killing of a fetus. Although the Vo case has also been included in the acrimonious debate of fetal rights, it is presented here to elucidate the concept of legislative law as opposed to judicial law.

State Law and Federal Law Another area of potential confusion is the differences between state laws and federal laws. Generally, clinicians find themselves in state court subjected to state laws. Significant restrictions exist that keep most cases out of federal court. A civil rights case, a dispute involving the Americans with Disabilities Act (ADA), or a malpractice case that occurred at a military hospital are three examples of cases that could be adjudicated in federal court. Unless a case meets narrow criteria to qualify for federal adjudication, most legal disputes involving perinatal or neonatal practitioners are tried in state court. How do laws in one state affect clinical practice in another state? Practitioners may wonder how a Michigan court decision would affect a practitioner in Ohio or Nevada. Generally, state court decisions are binding only in that state. If the Texas Supreme Court has ruled on an issue, the courts’ findings are viewed as state law in Texas, and the legislatures and courts of California, North Carolina, or Wyoming are not bound by the Texas court ruling. A state court’s ruling could be persuasive in other states, but the conclusions of one state court are not generally viewed as binding on courts in other states. This concept of one state’s laws affecting another state also holds true for laws passed by state legislatures. If the California legislature passes a law concerning access to prenatal care, the law would have essentially no effect on citizens of Connecticut or Virginia. Issues such as the definition of “live birth” are treated differently by different state governments. Clinicians should be familiar with their applicable state laws before relying on case law or statutes cited in this chapter.

General Structure of the Federal and State Court Systems Several of the cases cited in this chapter mention the holdings of various state and federal appellate courts, and several U.S. Supreme Court decisions are also discussed. How does a case get to an appellate court or to the U.S. Supreme Court? Various rules determine which court hears a dispute and which appellate court has the jurisdiction to review the decisions of the lower courts. Most of the cases discussed in this chapter would be adjudicated in the state court system. The general hierarchies of the federal and state court systems are depicted in Figure 4-1.

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine THE FEDERAL COURT SYSTEM Supreme Court United States Supreme Court

Appellate Courts United States Court of Appeals 12 regional Circuit Courts of Appeals 1 U.S. Court of Appeals–Federal Circuit

Trial Courts United States District Court 94 judicial districts (several per circuit) U.S. Bankruptcy Court

THE STATE COURT SYSTEM State Supreme Court

State appellate courts

State trial courts

Figure 4–1.  Hierarchy of the federal and state court systems.

SUPERVISION OF OTHERS Theories of Liability for Attending Physicians Attending neonatologists carry substantial responsibility. Generally, they bear ultimate medical responsibility for the neonates under their care. Practically speaking, it is impossible for one physician to provide all of the care for a sick newborn. Depending on the clinical setting, nurses, NNPs, PAs, respiratory therapists, social workers, residents, fellows, consultants, and many others all contribute greatly to patient care. The exact demarcation of responsibility and liability borne by attending physicians for these alternate providers is often difficult to determine. In holding attending physicians liable for the acts of others, courts tend to rely on three different theories of liability. An early theory of attending liability was known as the “captain of the ship doctrine.” Physicians, particularly surgeons, were assumed to be similar to naval captains and have complete control over the operating room (the “ship”) and all the medical personnel (the “crew”) within. With this control came responsibility for all negligent actions performed by anyone under the surgeon’s “command.”19 Most courts now recognize the increasing complexity of health care provision and have rejected the captain of the ship doctrine as “an antiquated doctrine that fails to reflect the emergence of hospitals as modern health care facilities.”46 Respondeat superior is a more accepted doctrine of physician liability for negligence of others under his or her control.

51

Respondeat superior literally means “let the master answer.” Because an attending physician has a right of control over an NNP or resident, the negligence of that provider is imputed to the attending physician in certain circumstances. Under respondeat superior, the attending physician would be responsible if a resident negligently places umbilical lines or an endotracheal tube. The attending physician would not be liable, however, under respondeat superior, if a bedside nurse was negligent in the observation or reporting of a significant intravenous infiltrate. Under the earlier “captain of the ship” doctrine, the attending physician may have been deemed to be responsible for the intravenous infiltrate. The attending physician can also be held liable for providing “negligent supervision.” For supervisees under his or her charge, the attending physician is responsible for providing adequate training and supervision. The attending physician must be readily available and promptly respond to requests for assistance. This responsibility was underscored in a 2004 obstetric malpractice case in which the attending anesthesiologist was not immediately available for an emergency cesarean section, and the fetus allegedly suffered as a result. The case was settled for $35 million.13 Finally, in many cases, an attending physician becomes liable for the actions of supervisees by the creation of a physicianpatient relationship that flows from the patient through the supervisee to the attending physician. In a case in New York, a patient was seen by a nurse practitioner in an emergency department, and the nurse practitioner misdiagnosed the condition. The attending physician discussed the patient with the nurse practitioner and signed the chart, but did not personally examine or speak with the patient. The court, interpreting New York’s law regarding nurse practitioners, held that “the ultimate responsibility for diagnosis and treatment rests with the physician.”61

Residents and Fellows During their postgraduate training, residents and fellows gain increasing experience and clinical skills under the supervision of attending physicians. Under the doctrine of respondeat superior, the educational institution and the attending physician are generally responsible for the medical care provided by residents and fellows. Residents who have completed their first year of training are eligible to be licensed to practice medicine without supervision. Because of this fact and the expectation of appropriate supervision, most states treat residents as physicians rather than students and hold them to “the same standard of care as physicians who have completed their residency in the same field of medicine.”14 Nevertheless, because trainees are thought to be agents of the hospitals in which they work, often have limited financial resources from which to pay a judgment against them, and always have an attending physician assigned to the patients they are caring for, the institution and the attending physician are almost always named in the lawsuit as well. Neonatologists must be very careful about appropriately supervising residents and fellows. In some cases, inexperienced trainees are responsible for caring for some of the sicker patients in the NICU. From a legal standpoint, the

52

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

supervising neonatologist must remain involved in the care of these patients and provide an appropriate level of oversight. The level of supervision would vary based on a variety of factors, including the condition of the patient, the likelihood of major changes in that condition, and the experience and skill of the resident providing the care. Failure to provide appropriate supervision can result in liability for negligent supervision.65

Physician Assistants Many NICUs now employ PAs. A PA is a health care professional who may practice medicine only with physician supervision. A certified PA is a PA who has completed training and passed a national certification examination. The scope of practice of PAs is governed by state law and varies from state to state. In many states, PAs are authorized to prescribe medications. PAs can contribute greatly to the care of neonates, and some centers that have found it difficult to recruit an adequate number of NNPs have found PAs to be an “untapped resource for the NICU.”62 It is important, however, for physicians to determine the scope of practice for PAs in their individual state for two reasons. First, there is increased potential malpractice liability for the supervising physician when PAs exceed their scope of practice. Second, physicians may risk loss of their license if they are “found to have condoned the unauthorized practice of medicine by a nurse or other health care professional for whose conduct [they are] responsible.”17

American Academy of Pediatrics Policy Statement on Advanced Practice in Neonatal Nursing released in June 2003 recommends the following: n A

neonatologist should supervise an APNN in the NICU. APNN should collaborate and consult with other health care professionals. n The APNN should be certified by a nationally recognized organization and should maintain that certification. n The APNN should participate in continuing education. n The APNN should comply with hospital policy regarding credentialing and recredentialing.11 n The

One critical issue concerning APNNs is the liability of the supervising physician. In some states, NNPs are licensed to practice independently and require no supervision under the law. If an NNP is hired by a hospital as an independent contractor with his or her own privileges, the physician does not employ the NNP. In these situations, when the NNP and physician are interacting in the management of an infant, the physician is acting the same as when consulting with any other provider, with similar liability. In most instances, however, and in almost all NICU environments, APNNs are not hired as independent contractors, but rather as employees of the physician or hospital. In these cases, the employer is vicariously liable for the acts of the employee. The NNP is often supervised by an attending neonatologist who bears ultimate responsibility for the patient as discussed earlier.

Advanced Practice Neonatal Nurses

MALPRACTICE

There has been a rapid increase in recent years in the number of advanced practice neonatal nurses (APNNs) in the United States. An APNN is a registered nurse who has completed a master’s degree in advanced nursing practice and, in most cases, has passed a national certifying examination. Advanced practice nurses are regulated at the state level, and educational requirements can vary. Although 43 states require a graduate degree, generally a master’s, for authorization to practice at the advanced practice level, 7 states do not have this educational requirement. Additionally, the National Association of Neonatal Nurses has proposed that all APNs be prepared through a Doctor of Nursing Practice degree by 2015.52 APNNs have played an invaluable role in improving health care for neonates in various settings ranging from urban academic centers to small rural hospitals. Under certain circumstances, there can also be additional liability for the physician. In neonatology, there are generally two recognized types of APNNs: clinical nurse specialist and NNP. The clinical nurse specialist is a registered nurse with a master’s degree who has expertise in neonatal nursing. The NNP is a registered nurse with experience in neonatal nursing (many have completed a master’s degree) and additional clinical training in the management of newborns. NNPs are allowed to assess, diagnose, and treat newborns independently or under the supervision of a physician. The licensure and scope of practice of APNNs vary considerably from state to state. Each institution must have policies and procedures for granting privileges for APNNs. The

Medical malpractice litigation can be contentious and acrimonious. It is a pervasive issue that appears with great regularity in the popular press. Many books have been dedicated to the subject of medical malpractice. In his 2008 State of the Union Address, President Bush reiterated the need to reform the current medical malpractice system.72 This section is largely limited to a discussion of malpractice in perinatalneonatal medicine (Box 4-1). Malpractice is part of a broader area of law known as torts. Tort law largely deals with the duties and responsibilities that individuals have toward one another. Torts are generally divided into two groups: intentional torts and unintentional torts. Defamation, invasion of privacy, civil battery, and professional malpractice are all torts, but malpractice is a type of unintentional tort. Negligence means that an individual’s behavior has deviated from a standard of “due care.” Malpractice is considered a specific type of negligence. By some interpretations, malpractice is also considered a type of breach of contract with the patient, so the defendant is technically being accused of committing a tort and violating contract law. Lawyers, accountants, physicians, and other professionals are held to a certain level of conduct. If one’s professional conduct is substandard, and a client, customer, or patient is harmed by this substandard conduct, a plaintiff may attempt to show that the practitioner has committed malpractice. To win a malpractice case, the plaintiff must show four critical elements: duty, breach, causation, and damages.

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

BOX 4–1 Common Malpractice Suits in Neonatology DELIVERY ROOM MANAGEMENT OR RESUSCITATION Poor neurologic outcome Cerebral palsy: neonatologist named as codefendant with obstetrician-perinatologist Asphyxia: plaintiff alleges some component of injury occurred postnatally LINE COMPLICATIONS Vascular accidents related to central venous lines Loss of fingers or toes associated with central lines Thrombus and complications from thrombus DELAY IN DIAGNOSIS OR TREATMENT Poor blood gases, prolonged hypotension (see also poor neurologic outcome) Delay in antibiotic administration Congenital hip dislocation Congenital heart disease TRANSPORT TEAM Medications or care provided by transport team (e.g., excessive heparin given) FAILURE TO MONITOR ADEQUATELY Blood glucose Blood oxygen: either hypoxia (brain damage) or hyperoxia (retinopathy of prematurity) Seizure

Duty “The duty of care owed to an individual, for purposes of a claim of medical malpractice, is based primarily on the existence of the physician-patient relationship.”69 To proceed with a negligence case, the plaintiff must show that the defendant had a duty to the plaintiff. This has been described as a “threshold issue.” Does the defendant owe a duty to the injured party? If there is no duty, no claim of negligence can be sustained. If a neonatologist has privileges only at hospital A, and the physician is called and refuses to attend a high-risk delivery at hospital B, the physician likely would have no professional relationship with the pregnant woman or her infant at hospital B. The neonatologist cannot breach his or her duty if no duty to the defendant exists. This concept of duty is separate from the moral or ethical obligation to provide care. A physician cannot be liable to a patient if there is no legal duty. Likewise, if NNP Smith is on call for the evening, and NNP Jones has left town with his family for a scheduled vacation, it would be difficult for an injured plaintiff to show that NNP Jones had a duty to attend a high-risk delivery while he was out of town. Does a physician caring for a pregnant woman have a duty to the newborn even after the infant is born and being cared for by another physician? In Nold v Binyon,54 a woman tested positive for hepatitis B. Her newborn did not receive hepatitis B immunoglobulin or the hepatitis B vaccine, and the infant subsequently became a chronic carrier for hepatitis B. The trial court ruled, and the Kansas Supreme Court agreed, that the delivering physician had a duty to inform the woman of her hepatitis B status. The Supreme Court stated, “a physician

53

who has a doctor-patient relationship with a pregnant woman who intends to carry her fetus to term and deliver a healthy baby also has a doctor-patient relationship with the fetus.”

SUPERVISING OTHERS As discussed earlier, neonatologists are often asked to supervise the care provided by others. This supervisory role generally establishes a physician-patient relationship with any patient who is cared for by the supervised NNP or PA. In these cases, the physician has a duty to the patient even if the supervised NNP is providing all of the bedside care.

TELEPHONE ADVICE Is “duty” established when one physician consults with another over the phone? Telephone advice and transport present an interesting legal challenge. On many transports, the responsible physician at the receiving facility is not physically present with the transport team. The receiving physician often begins to offer clinical advice, however, when first contact is initiated by the referring facility. Generally, this can be a situation of shared duty. The referring physician and the receiving physician may have a duty to the patient. The receiving physician may have no duty to the patient, however, if the receiving physician is acting more in the role of a consultant. In Sterling v Johns Hopkins,69 a woman was admitted at approximately 32 to 33 weeks’ gestation to a hospital, and she developed severe preeclampsia and suspected HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome. The treating physician called the emergency transport service to arrange transport to Johns Hopkins because the receiving hospital had an NICU that could potentially care for a premature infant. The receiving physician spoke with the referring physician on the telephone. The woman became unresponsive during the transport. She experienced an intracranial bleed and later died. The husband sued the receiving hospital alleging negligent advice given over the phone. The court determined that there was no physician-patient relationship between the receiving physician and the pregnant woman. Because the receiving physician was acting more in the role of a consultant, and the referring physician was free to make his own management decisions, the court ruled that the receiving physician did not have a duty to the pregnant woman. “Where the treating physician exercises his or her own independent judgment in determining whether to accept or reject [a consultant’s] advice, . . . the consultative physician should not be regarded as a joint provider of medical services with respect to the patient.” In this case, the court determined that the treating physician maintained decision-making power, and that the physician at the receiving facility was acting more as a consultant than comanaging the patient. The court determined that no duty existed between the receiving facility and the patient.

PRENATAL CONSULTATION Prenatal consultations might or might not give rise to a duty between the neonatologist and the pregnant woman and her child. Among the determining factors, courts seem to evaluate

54

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

the formality of the consultation and the presence or absence of contact between the parties. In Hill v Kokosky,35 an obstetrician informally consulted with a neonatologist. The plaintiff had been admitted to the hospital at 22 weeks’ gestation with a diagnosis of incompetent cervix. The two physicians informally discussed the case over the telephone. The obstetrician discussed the case in the abstract, and the neonatologist tended to agree with the obstetrician’s management. There was no referral or formal consultation. From the record, it seemed that the neonatologist did not review the chart, speak with the mother, or even know the mother’s name. The infant was born 2 weeks later and developed severe cerebral palsy. In her malpractice action against the neonatologist, the mother maintained that the neonatologist gave substandard advice concerning birthing options, and that this substandard advice contributed to her infant’s injuries. Given the facts, the court concluded that no physician-patient relationship existed between the neonatologist and the family. It is important to understand why the court ruled for the defendant in this case. Casual telephone advice was given to a colleague. The neonatologist did not know the name of the patient and never spoke with her. The court concluded that the neonatologist did not prescribe a course of treatment, but rather gave recommendations that could be accepted or rejected by the obstetrician. In contrast to this case, if a neonatologist is formally consulted and speaks with a family and makes recommendations concerning management, there may be a duty to the mother and her infant. Judicial decisions also seem to hinge on whether or not a consulting physician is recommending a specific course of therapy or merely making suggestions that the original physician either can follow or can ignore. Generally, it is not difficult for a plaintiff to establish that a clinician had a duty to the patient. Usually, the physician has provided care to the patient, and the plaintiff easily establishes that the duty requirement has been met. Especially in the case of hospital-based physicians, such as neonatologists, the element of “duty” is generally established.

Breach STANDARD OF CARE What is the duty that is owed? The duty is to provide reasonable care under the circumstances. Although generalists are held to a standard of “same or similar community,” specialists and subspecialists, such as neonatologists, NNPs, and perinatologists, are generally held to the higher national standard of care.55 An expert witness from a different state can testify about the national standard of care for a neonatologist. In these cases, the neonatologist’s care is not being compared with the care provided in a similar community; the care is evaluated in light of national standards. In many malpractice suits involving obstetricians, perinatologists, and neonatologists, considerable emphasis is placed on this element. Often the defense vigorously maintains that “the doctor did nothing wrong.” The defense often takes the position that there has been an unfortunate outcome, but the defendant practiced within the standard of care. If the defense can prove that the physician acted within the

standard of care, the plaintiff cannot successfully maintain a malpractice action. A typical case is a brachial plexus injury after birth. In Knapp v Northeastern Ohio Obstetricians,43 a mother alleged that her infant’s brachial plexus injury was the result of excessive traction applied by the obstetrician. The trial court found, and the appellate court affirmed, that the evidence did not support the mother’s allegations. The court concluded that the obstetrician had not breached the standard of care. Retinopathy of prematurity (ROP) is another common source of negligence cases involving newborns. In Brownsville Pediatric Associates v Reyes,21 a pediatrician was found liable for substandard ventilator management. The expert witness testified that the child’s resulting brain damage and ROP were related to hyperventilation and hyperoxia. The plaintiff was awarded $8 million, and the defendant’s appeal was denied. Another ROP case32 dealt with the responsibility of the neonatologist, the pediatricians, the pediatric ophthalmologist, and the parents when 29-week gestation twins missed their follow-up appointments for ROP evaluation. The twins became legally blind. In this case, the neonatologist apparently provided all necessary referrals and documentation, and she was not named in the resulting suit.

ROLE OF THE EXPERT WITNESS In tort proceedings other than medical malpractice, a person with common knowledge generally knows and understands “due care.” If someone is walking down the street with closed eyes and bumps into someone else and injures the other person, a lay juror does not need an expert witness. The lay juror understands that one should not walk down the street with closed eyes because someone else could be injured as a result. In medical malpractice cases, however, the lay juror generally does not have a grasp of “reasonable care under the circumstances.” This is one of the roles of the expert witness. To serve as an expert witness, an individual must have specific knowledge and training that qualifies him or her to serve in this capacity. The Rhode Island statute,63 for example, states that “only those persons who by knowledge, skill, experience, training, or education qualify as experts in the field of the alleged malpractice.” If a fetal monitoring strip shows severe, repetitive, late decelerations, should the obstetrician perform a cesarean section? It is the role of the expert witnesses to educate the jury so that the jurors have a grasp of what is (and is not) reasonable care under the circumstances. Both sides (plaintiff and defendant) usually hire their own expert witnesses. The plaintiff’s expert generally maintains that the physician practiced outside of the standard of care. The defense expert maintains just the opposite. After the expert witnesses are examined and cross-examined, it is up to the jury (or arbitrators) to decide whether the clinician committed a breach in the standard of care. Expert witness testimony is often required in perinatalneonatal malpractice cases. Expert witnesses were used in the 2000 Pennsylvania case Sonlin v Abington Memorial Hospital.67 In this case, a premature infant girl who was born at approximately 34 weeks’ gestation had an umbilical line. The infant developed vascular compromise in her left leg, which

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

resulted in a thrombus that required amputation of the extremity. The plaintiff maintained that the neonatologist did not recognize the thrombus, and he did not institute corrective action in a timely fashion. Expert witnesses would have to explain to a jury why umbilical lines are placed, how long they are left in place, and the potential complications from indwelling arterial catheters. In this case, an expert would have to explain the effect of prematurity on lung development and the subsequent necessity for monitoring blood oxygen levels. Basically, the expert must explain the standard of care, the indications for the procedures, and the potential complications. In the Louisiana case Hubbard v State,37 a full-term newborn was admitted with meconium aspiration and hypoglycemia. A peripheral intravenous infusion of 10% dextrose in water was ordered. After a change was made in the intravenous fluid, it was noted that the infant’s hand became red and swollen, and the infant became lethargic. His blood glucose was 450 mg/dL. Because of an error, the infant had received 50% dextrose in water instead of the 10% dextrose in water that was ordered. The infant sustained third-degree burns that left permanent disfigurement of the hand, and a computed tomography scan showed a “possible venous thrombosis of the transverse and sagittal sinus.” By the time this case reached the Louisiana Appellate Court, the child was almost 8 years old. In the interim, the child had been found to have Russell-Silver syndrome, a condition known to be associated with developmental impairment. The expert witnesses were extensively questioned about whether the dehydration and possible venous thrombosis contributed to his observed mental delays. These cases represent a potential breach in the standard of care. Mistakes were made, infants were harmed, and the families attempted to hold the caregivers responsible for the damages that occurred. In both of these cases, expert witnesses were needed to assist in delineating the standard of care for the legal decision makers.

RES IPSA LOQUITUR Res ipsa loquitur is a legal doctrine that means “the thing speaks for itself.” Common medical examples of this doctrine include a retained surgical sponge, removal of the wrong kidney, or operation on the wrong patient. These three examples are factually simple. A juror’s common knowledge would guide him or her to a reasonable conclusion. In practical terms, res ipsa loquitur generally means that the plaintiff does not need an expert to show that there was a deviation in the standard of care. The doctrine means that certain things do not just happen. Historically, the res ipsa loquitur doctrine can be traced to an English case22 from 1863 in which a passerby was injured when a barrel fell from a window. The doctrine was developed to explain that barrels do not fall out of windows unless someone has acted negligently. A 2003 case from New York relied on the res ipsa loquitur doctrine. In Rosales-Rosario v Brookdale,64 a woman was hospitalized to give birth. An epidural line was placed, and she was partially anesthetized. It was subsequently noted that she had sustained a burn to her leg. She had no idea how she sustained the injury. Her leg possibly was burned by an examination

55

light, but because of her anesthetized state, she did not recall how the injury occurred. She relied on the doctrine of res ipsa loquitur. The trial court dismissed the case, but the appellate court reversed the lower court’s decision. The appellate court stated the rule of res ipsa loquitur: “To rely on the doctrine of res ipsa loquitur, a plaintiff must submit sufficient proof that: (1) the injury is of a kind that does not occur in the absence of someone’s negligence; (2) the injury is caused by an agency or instrumentality within the exclusive control of the defendants; and (3) the injury is not due to any voluntary action on the part of the injured plaintiff.”64 In the Rosales-Rosario case, the court concluded that leg burns do not just occur, so the first element of res ipsa was satisfied. In addition, the physicians and hospital personnel were in exclusive control of all equipment in the delivery room, including the examination light. Finally, the patient did not engage in a voluntary act that resulted in her injuries. Showing res ipsa loquitur means that the plaintiff has overcome the burden for breach. The plaintiff still has the burden of showing the other elements of the tort suit, but the deviation in standard of care has been proved if the court accepts the doctrine of res ipsa loquitur.

Causation Of the four essential elements of a tort suit, causation is perhaps the most challenging to understand. Regardless of whether the dispute is malpractice or another civil complaint, causation is not always intuitive. In brief, the defendant’s breach must be the cause of the plaintiff’s injury; mere correlation is insufficient. The plaintiff must show a reasonable inference that the deviation in care resulted in the injury. Stated differently, assuming the plaintiff has suffered an injury, did the deviation in standard of care cause this injury? Expert testimony is often required to answer these questions. Did the obstetrician’s decision to allow a vaginal birth in the face of severe decelerations result in the newborn’s neurologic damage? Did the neonatologist’s “delay” in the decision to perform a double-volume exchange transfusion result in kernicterus that would have otherwise been avoided? In many malpractice cases, the issue of causation can be complex. Causation can be particularly perplexing, however, in the context of a perinatal-neonatal medical malpractice case. In the Hubbard case37 described earlier, multiple expert witnesses disputed whether the severe dehydration contributed to the child’s mental delay. What is the impact of the diagnosis of Russell-Silver syndrome? Was this the cause of the mental delay? The child had also had at least two documented falls during early development. Did the head injuries, which led to evaluation in an emergency department, cause the findings? This case elucidates the particular challenge of causation. Board-certified neonatologists and pediatric neurologists could have well-substantiated, yet differing, opinions on the etiology of this child’s mental deficits. How is a jury to rule on this complex issue? As in essentially all medical malpractice cases, expert witnesses must testify on the issue of causation. The concepts are usually too specialized for a nonexpert juror. Absent the testimony of an expert witness, a juror’s common knowledge is often inadequate to decide the issue of causation.

56

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Damages The final element that a plaintiff must show is damages. For plaintiffs to recover any type of award, they must show that they were harmed. They experienced either pain and suffering or a loss of some kind, such as loss of wages or loss of consortium. In some cases, damages are presumed. If a surgeon leaves a surgical sponge in the patient, the patient requires an additional surgery to remove the sponge. In this case, the patient would be able to show that the second surgery led to discomfort, time away from home, time away from work, and lost wages. To understand the concept of damages, one should be familiar with the common distinction between economic and noneconomic damages. Economic damages include medical expenses, costs of burial, lost earnings, and loss of employment. These damages are often less complicated to calculate. Alternatively, some plaintiffs claim noneconomic damages, which are more subjective. Among other claims, noneconomic damages can include pain and suffering, emotional distress, mental anguish, or destruction of parentchild relationship. Many state legislatures have attempted to cap awards for pain and suffering or for wrongful death. The California legislature23 has placed a $250,000 cap on noneconomic damages. This generally means that any litigation involving the death of a newborn has a maximum award of $250,000. If older patients die, their estate can seek damages for lost wages, lost consortium, or other losses. The award for wrongful death of a newborn is essentially capped in California, however. In addition, several state legislatures have placed caps on other noneconomic damages such as pain and suffering. In Louisiana, the Malpractice Liability for State Services Act has capped state hospitals’ liability at $500,000 for “pain and suffering.”44 A great deal of tort reform litigation deals with the issue of damages. Specifically, some state legislatures have attempted to limit large jury awards by placing these caps on noneconomic damages. These disputes generally do not involve issues such as lost wages, hospitalization, or loss of property. Perceived excesses of some jury awards coupled with increasing malpractice rates are likely to fuel the national debate concerning caps on noneconomic damages. If a negligent act is committed, and someone is harmed, damages are usually limited to compensatory awards. Punitive damages are awarded if the defendant is found to have committed a particularly egregious act. These damages are rarely awarded in medical malpractice cases. If a physician were to alter a chart or make another attempt to change the medical record, however, a jury may award punitive, or punishment, damages. “Spoliation of evidence” can lead to punitive damages.

Burden of Proof How does a plaintiff win a case? Generally, each element must be proved by a preponderance of the data. This is also called the “51% test.” The attorneys often frame their questions in terms of “is it more likely than not.” Whether the attorneys are questioning expert witnesses or making statements to the jury, they often refer to the burden of proof.

Generally, this burden is on the plaintiff. The plaintiff must show that the defendant had a duty, that the duty was breached, that the plaintiff sustained damages, and that the breach was the legal cause of the damages. The plaintiff must prove each of these elements by a preponderance of the data. Burden of proof can be contrasted with the common criminal standard of proof known as “beyond a reasonable doubt.” In criminal cases, the state must show the defendant’s guilt beyond a reasonable doubt. Another standard that may be familiar to readers is the “clear and convincing” standard. This standard is sometimes used in cases involving the withdrawal of care when the patient cannot communicate his or her wishes.39 In civil cases, and particularly in malpractice cases, the standard is generally a preponderance of the data.

Protected (Nondiscoverable) Proceedings Many hospitals have committees that review the care provided at that institution. Although the precise operations of these committees might differ, committees generally review cases that have had an untoward or unexpected result. In some institutions, all deaths are reviewed. Many state legislatures have provided protection for the proceedings from these committee meetings. The Georgia statute specifically shields the specified proceedings from being used by a plaintiff in a medical malpractice case.29 This is an issue of public policy. It is believed that care would be improved if caregivers could discuss challenging cases openly in a protected environment. Clinicians must recognize that these proceedings do not limit the plaintiff’s ability to bring suit against the physicians or the institution. A meritorious plaintiff can still subpoena hospital records and successfully bring suit against the caregiver. The protected proceedings would be unavailable as evidence in the case, however. From a legal standpoint, the proceedings are “nondiscoverable.” Before a physician discusses a case with an untoward outcome at a medical staff proceeding, the physician might wish to confirm that the proceedings of that meeting are nondiscoverable.

Other Tort Actions In the context of neonatal-perinatal medicine, there are two unique causes of action. These actions are wrongful life and wrongful birth. There are distinctions between the two, and these causes of actions are not allowed in some states.

WRONGFUL BIRTH Wrongful birth cases are brought by parents who have given birth. The cause of action is maintained on behalf of the parents. The parents contend that the child should never have been born, and they seek recovery based on the birth. These cases are often seen after the failure of a sterilization procedure or after a physician has assured a patient that he or she is not fertile. The parents generally seek economic damages related to the cost incurred in raising the child. In 2007, a Florida family was successful in suing a geneticist on a wrongful birth cause of action. The case involved the alleged misdiagnosis of a child with Smith-Lemli-Opitz syndrome.24 The family had one child with multiple anomalies,

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

including microcephaly, micrognathia, cleft palate, syndactyly, hypospadias, and cryptorchidism. This child also had severe developmental delay. The family alleged that they brought the child to a geneticist who failed to make the diagnosis. The family was allegedly told that their chances of having a normal child were the same as anyone else’s. Based on this information, the family conceived and did not seek genetic diagnosis of the fetus. Within hours of birth, this child was diagnosed with Smith-Lemli-Opitz syndrome. Subsequently, the first child was also diagnosed with Smith-Lemli-Opitz syndrome. The family contended that they would have terminated the pregnancy if they had been aware of the diagnosis of the second fetus. Expert witnesses testified that failure to diagnose the oldest child with Smith-Lemli-Opitz syndrome was below the standard of care. A plaintiff’s expert also testified that the conduct of the original geneticist was egregious. The plaintiff expert testified that the family should never have been told that they had no increased risk of having a child with birth defects. The family sued, and a jury awarded more than $21 million.71

WRONGFUL LIFE Wrongful life cases are maintained on behalf of a newborn. The plaintiff usually maintains there was negligence in the diagnosis or treatment of the mother, and that the infant should not have been born. In essence, the parent is claiming that the infant would be better off if the infant had not been born. Historically, these cases tended to be brought on behalf of newborns with severe congenital anomalies. A family maintains that an obstetrician or ultrasonographer missed certain important findings. The family claims that they would have terminated the pregnancy if they had known the child’s diagnosis. Ethically and philosophically, this tort raises many more questions. What is the value of human life? Can a person actually sustain a cause of action simply because that life exists? Would any person actually be better off if he or she had not been born?31 As ill-equipped as many clinicians are to deal with these questions, the courts are at an even greater disadvantage. Courts tend to rely on facts and evidence. How does one compare a damaged existence with no existence at all? Courts have grappled with this issue. Some courts have stated that there is sanctity in an impaired existence, but that this sanctity does not preclude a child’s recovery for wrongful birth.77 Courts have said that there is an almost insurmountable challenge in attempting to compare what judicial opinions characterize as the “utter void of nonexistence” with an impaired existence; however, some states do allow recovery. The Estrada case (noted earlier in “Wrongful Birth”) addresses some of the issues associated with genetic testing.30 If genetic testing is indicated and is not offered to a family, certain states allow a wrongful life cause of action. Most states do not allow this cause of action.

WRONGFUL DEATH A wrongful death action is seen in other civil suits as well as in medical malpractice cases. If a person dies as the result of another’s negligence, the estate may pursue a wrongful death claim. If a pregnant woman is involved in a car crash, and she miscarries as a result, the woman may be able to sue for wrongful death of her fetus.

57

BOX 4–2 Strategies to Avoid Tort Litigation Stay current by reading journals and textbooks, and by attending continuing medical education conferences Maintain professional ties with a large medical center When facing a difficult situation, consider consulting with a colleague Maintain open communication with parents and families Practice timely documentation of procedures, communication, complications, and persons present Document telephone advice Be aware of state laws that affect your practice

Although wrongful death claims can be part of other negligence suits, the claim is common in malpractice actions. Although some states allow for the wrongful death of a fetus, other states recognize this claim only after a live birth.18 A case in Connecticut seems to allow for a wrongful death if the fetus is viable.28 Other states, such as West Virginia, allow recovery for the wrongful death of a fetus of any gestation.27

Strategies for Avoiding Tort Litigation Physicians maintain a vested interest in avoiding malpractice litigation. When untoward medical complications arise, physicians may experience deep empathy for the patient and personal feelings of doubt or inadequacy. The allegation of medical malpractice may lead the physician to experience numerous physical and emotional symptoms, collectively referred to as the medical malpractice stress syndrome.66 Besides the mental anguish that one can experience as a result of being named as a defendant in a malpractice case, these proceedings are often time-consuming and expensive. Physicians can adopt certain personal guidelines to minimize their chances of being named in a malpractice suit (Box 4-2). Clinicians should make all necessary efforts to stay current in their discipline. Attending conferences and reading journals and textbooks assist physicians in their clinical practice. In the 1990s, dexamethasone was widely used to wean premature infants with chronic lung disease from ventilators. Because of long-term neurologic concerns, this therapy is now normally reserved for only the sickest infants. A neonatologist who does not keep up with current practices could be administering a medication or offering a form of therapy well after its use has been widely abandoned or curtailed. NICUs are often part of a regional system of perinatalneonatal services. Small and large community NICUs can maintain professional ties with one another and with larger medical centers. These ties allow professional interchange of ideas and recent developments. Clinicians would be wise to view their professional development as an ongoing process. A clinician who is faced with a rare or particularly challenging case can consider calling a colleague. Attending physicians often maintain contact with their former trainees. These contacts with former mentors can provide great benefit for one’s patients. Perhaps a former attending physician can shed some light on a difficult situation. At large academic centers, attending physicians generally make it a point to discuss regularly the most difficult clinical cases at fetal

58

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

boards, morning report, grand rounds, or other venues. By phoning a colleague, a community-based neonatologist can maintain a similar professional network. In addition to maintaining good communication with colleagues, few issues are more vital than optimal communication with parents. It is not always easy to maintain good communication with parents, but the implications of poor communication are generally unacceptable. If parents believe a clinician is hiding something from them, they often become frustrated and angry. Parents are already dealing with having a sick infant in the NICU. If they feel mistrust, the physician’s relationship with a family can quickly deteriorate. A suboptimal relationship with the family of a sick newborn can be a harbinger of a pending malpractice suit. Besides optimal communication, the clinician’s other major defense to a tort suit is documentation. Entries in the chart should be punctual, legible, and accurate. If circumstances necessitate that a late entry be made, this should be clearly documented as such. If a particularly important event has occurred, whether it be a complication with the infant or a comprehensive family conference, it should be documented in the chart. With respect to family conferences, it is important to document who was present and what was discussed. One area where clinicians often fail to document is the advice that they give over the phone. Because this can be time-consuming and logistically difficult, some physicians open themselves to liability by giving advice and failing to document that advice. Referring physicians or parents may ignore a physician’s advice. In these situations, documentation of the phone discussion can save a tremendous amount of money, time, and frustration at a later date. Finally, clinicians should be aware of state laws that affect their practice. This chapter presents cases and statutes from several states. Which ones affect a particular physician’s practice? Many issues, such as resuscitation of extremely premature infants or accreditation of NICUs, are largely controlled by state law. Physicians should know the laws that affect their practice.

Medical Association has expressed its support for MICRAtype reforms.48 In 2006, the National Conference of State Legislatures reported that 36 states were considering medical malpractice legislation.53 Generally, state legislatures evaluated the merits of the medical malpractice system and passed bills designed to curb certain monetary awards. In a move that alarmed many physicians and hospitals, Florida voters passed a constitutional amendment that has come to be known as the “three strikes malpractice law.”74 Article X, Section 26 of the Florida Constitution15 automatically revokes the license of any physician with three malpractice judgments against him or her. To avoid a “strike” from losing a malpractice case, there is strong incentive for physicians to settle all malpractice claims. It has been predicted that essentially all physicians in Florida would be sued within 10 years.58 In contrast, legislation placing limits on damages in Texas, authorized by a recently passed constitutional amendment, preceded a 50% increase in physicians applying to practice in that state. An environment in Texas that provides some protection for physicians has helped to relieve specialty shortages in rural areas.73 There are efforts to overturn this tort reform, however. In 2008, a class action lawsuit was filed alleging that limits on damages violate constitutionally guaranteed rights, including equal protection and due process. The plaintiffs in this class action suit claim that the state-mandated limits on medical malpractice awards do not provide adequate compensation for their injuries. Among the plaintiffs in this class action suit are at least two children that allege injuries they sustained as newborns. One plaintiff in this suit allegedly sustained permanent brain damage because fetal distress was not diagnosed quickly enough. Another infant plaintiff allegedly sustained injury because of inappropriate respiratory management and because hypoglycemia was not promptly diagnosed and treated.75

Trends in Malpractice Legislation

Generally, a plaintiff must show that all four elements (duty, breach, causation, and damages) are present. Showing only that the physician made a decision outside an acceptable standard of care is insufficient. Likewise, a patient cannot sustain a cause of action simply because of sustaining harm. The physician may have no duty to the patient, or the damages sustained may have no relation to the alleged breach in the standard of care. Of the four elements, breach and causation are the main elements that are commonly disputed in a malpractice case. The plaintiffs argue that the physician’s care deviated from an acceptable standard, and that this deviation harmed the patient. The defendant maintains that the care was acceptable, and any damages were not the result of faulty care (or decision making) by the defendant.

For the last several decades, insurance companies, large corporations, and attorneys have actively participated in the tort reform debate. Tort reforms may include limits on noneconomic damages, limits on attorney fees, expert witness standards, and inadmissibility of apology statements by health care providers.53 In the context of malpractice law, consumer advocates argue that meritorious plaintiffs are entitled to compensation when they are injured by medical negligence. Insurance companies and providers counterargue that juries should be restricted from awarding massive monetary verdicts because the costs of these verdicts are reflected in increasing health care bills, physicians’ unwillingness to practice in certain locales, and the increase of “defensive medicine.” California enacted the Medical Injury Compensation Reform Act of 1975 (MICRA) in an effort to address increasing malpractice premiums. MICRA addressed various issues, including caps on noneconomic damages (e.g., pain and suffering, loss of consortium) and mandated that attorneys’ contingency fees be based on a sliding scale. The American

Conclusions

LIVE BIRTH As the fetus descends through the birth canal and emerges as a living infant, many critical transitions occur. Readers of this textbook are familiar with the physiologic adaptations that accompany a live birth. In a legal sense, the fetus generally acquires full personhood when there has been a declaration

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

of a live birth. Before birth, the cases cited earlier determine the fetus’s legal status. Although the rules differ from state to state, there is consistent interpretation that at the moment of live birth, the infant has all of the associated rights, privileges, and consequences of personhood in civil and criminal matters.76 It would seem intuitive how to define “live birth.” Each state has a definition for “live birth,” and there is considerable overlap in these definitions, but some states have placed additional clarification in the statutory language. The Utah statute states that live birth “means the birth of a child who shows evidence of life after it is entirely outside of the mother.”78 This statute can be contrasted with the Alabama statute, which states, “When used with regard to a human being, [live birth] means that the human being was completely expelled or extracted from his or her mother and after such separation, breathed or showed evidence of any of the following: beating of the heart, pulsation of the umbilical cord, definite movement of voluntary muscles, or any brainwave activity.”9 A common thread in the states’ statutes is to include some physiologic sign of life, whether it is a beating heart, pulsation of the umbilical cord, spontaneous respiratory activity, or spontaneous movement. Beyond these findings, some states clarify further that an infant is considered to be alive whether or not the placenta is still attached, and that an infant of any gestation can be a live birth.49 Some statutes26 specifically differentiate between heartbeats and “transient cardiac contractions.” There is also an effort to distinguish breathing from “fleeting respiratory efforts or gasps.” The Alaska10 statute seems to contemplate all of these variables. It states that “live birth means the complete expulsion or extraction from its mother of a product of human conception, irrespective of the duration of pregnancy, which, after the expulsion or extraction, breathes or shows any other evidence of life such as beating of the heart, pulsation of the umbilical cord, or definite movement of voluntary muscles, whether or not the umbilical cord has been cut or the placenta is attached. Heartbeats shall be distinguished from transient cardiac contractions; respirations shall be distinguished from fleeting respiratory efforts or gasps.” Additional language is added to the Maine statute, which specifically states that “each product of such a birth is considered live born and fully recognized as a human person under Maine law.”47 In the context of criminal law, the American Law Reporter proposes a different test for “life.” This is known as a showing of a “separate and independent existence.”6 For purposes of homicide, a newborn is considered to have been born alive if it ever showed a separate and independent existence from its mother.

Born-Alive Infants Protection Act In 2002, Congress passed by an overwhelming majority the Born-Alive Infants Protection Act (BAIPA). The Act states that when the terms individual, person, human being, and child are used in any law, it should be interpreted to include every infant who is born alive at any stage of development. The law defines born alive to mean “the complete expulsion or extraction from his or her mother of that member, at any stage of

59

development, who after such expulsion or extraction breathes or has a beating heart, pulsation of the umbilical cord, or definite movement of voluntary muscles, regardless of whether the umbilical cord has been cut, and regardless of whether the expulsion or extraction occurs as a result of natural or induced labor, cesarean section, or induced abortion.”60 BAIPA8 was passed with the notion that legal protections should be provided to infants born alive after failed abortions. According to the U.S. Congress, it was not relevant if the parents intended to have a live infant. If the infant was born alive, the infant must receive a “medical screening,” and appropriate care must be provided. Various parties have since raised concerns that previable fetuses that were born alive would have to be resuscitated in light of BAIPA. In 2003, the American Academy of Pediatrics Neonatal Resuscitation Program Steering Committee responded to this concern: “It is the opinion of the American Academy of Pediatrics Neonatal Resuscitation Program (NRP) Steering Committee that the Born-Alive Infants Protection Act of 2001 should not in any way affect the approach that physicians currently follow with respect to the extremely premature infant.”56 The full impact of BAIPA has not yet been determined. At least one commentator has expressed concern that BAIPA could be interpreted to require resuscitation of extremely premature infants.16 At a minimum, however, the federal law requires that all live-born infants require the same level of medical screening regardless of the reasons for the birth.

HANDICAPPED NEWBORNS Federal and state governments have created protections for the most vulnerable members of society. This section deals largely with federal issues, but the states have also adopted guidelines to protect handicapped individuals. Individuals with disabilities are protected by ADA.4 All children are protected by child abuse and child neglect statutes. The government generally places a high emphasis on protecting the lives of fragile children. Parents can lose custody of their children if they violate laws related to abuse or neglect, and they can be incarcerated for criminal endangerment50 if their behavior is particularly egregious. With the growing expertise in prenatal diagnosis, it is increasingly rare that the family and the health care team are surprised by the birth of an infant with congenital anomalies. Maternal serum markers, prenatal ultrasound, amniocentesis, prenatal percutaneous umbilical blood sampling, and other procedures provide the practitioner with a considerable armamentarium to diagnose anomalies. In the case of “lethal” anomalies, this advance notice gives the family time to consider how they wish to proceed. There is often ample opportunity for the practitioners and the family to discuss the diagnosis, the implications of the diagnosis, and the care options. In the case of lethal anomalies, care options selected by the parents may include comfort care only, aggressive resuscitation, or any level of care in between. Federal law prohibits discrimination on the basis of handicap. Under this law, nourishment and medically beneficial treatment (as determined with respect for reasonable medical judgments) should not be withheld from handicapped infants

60

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

solely on the basis of their present or anticipated mental or physical impairments.4

Baby Doe Handicapped newborns and the care that they receive became a mainstream issue in the 1980s. The controversy created by the Baby Doe case still governs many decisions made by neonatologists. Baby Doe lived for less than 1 week, but his legacy remains more than a quarter century later. Many neonatal-perinatal practitioners are familiar with the basic facts of Baby Doe.40 The infant was born with Down syndrome in Bloomington, Indiana, on April 9, 1982. As is the case with many infants with this disorder, he had a gastrointestinal tract atresia. Practitioners recognize that gastrointestinal atresias are usually surgically correctable. Duodenal atresia is more common in Down syndrome, but Baby Doe had esophageal atresia. Because of the atresia, the infant could not be fed. Baby Doe’s parents, on the advice of their obstetrician, elected to forgo surgery. After discussions with the family, food and water were not provided, and the infant died at 6 days of age. Down syndrome is not considered a lethal anomaly. Had Baby Doe not had Down syndrome, deferring surgery would not have been considered an acceptable option. Because Baby Doe did not have lethal anomalies, it was assumed that medical and surgical care were withheld because of the mental deficits associated with Down syndrome. A hospital nurse had heard about the baby who was being “starved,” and she filed a complaint. This resulted in a judicial opinion that stated that the parents should “have the right to choose a medically recommended course of treatment.” Higher courts68 refused to hear the case. Baby Doe died while a stay was being sought in the U.S. Supreme Court.20 The U.S. Supreme Court38 later refused to hear the case. The decision to forgo care was largely viewed as unacceptable. Advocates for the handicapped were particularly concerned about this case. They proposed that Baby Doe had been discriminated against on the basis of the infant’s handicap, and there had been a violation of Section 504 of the Rehabilitation Act of 1973.1 For various reasons, the Baby Doe case was a landmark decision. In the 3 years following Baby Doe’s death, the executive, the legislative, and the judicial branches of the federal government became involved. The American Medical Association, the American Academy of Pediatrics, the American College of Obstetrics and Gynecology, and other professional organizations also became involved. The Reagan administration’s position was that all handicapped newborns must be treated aggressively unless the care is obviously futile. The Reagan administration believed that physicians should be liable for neglect and discrimination if they did not comply with these rules. The Reagan administration attempted to force hospitals to treat all severely handicapped newborns, regardless of the parents’ wishes. The Department of Health and Human Services (HHS) promulgated “Baby Doe regulations” that required federally funded hospitals to post certain rules5 in the hospital. Notices were to be prominently posted in delivery wards, maternity wards, pediatric wards, and each nursery. These signs encouraged concerned parties to call the HHS toll-free

number to report suspected cases of discriminatory withholding of care from handicapped newborns. The rules also encouraged the creation of infant care review committees to assist in decision making for difficult cases. Many of these regulations were ultimately disallowed by the courts. The current incarnation of the Baby Doe regulations is found in the Child Abuse Protection and Treatments Act (CAPTA), which prevents the “withholding of medically indicated treatment,” which is defined as “failure to respond to the infant’s life-threatening conditions by providing treatment . . . which, in the treating physician’s . . . reasonable medical judgment, will be . . . effective in . . . correcting all such conditions.”2 Although infant care review committees are encouraged, the courts have ruled that parents, in conjunction with their physicians, should have the right to make health care decisions for their handicapped children. When the U.S. Supreme Court decided against hearing the Baby Doe case, the matter was deferred only for a few years, until the Baby Jane Doe case.

Baby Jane Doe In 1983, the year after Baby Doe died, a child was born in New York with meningomyelocele, hydrocephalus, microcephaly, bilateral upper extremity spasticity, a prolapsed rectum, and a malformed brainstem. She has been immortalized in the neonatal literature as “Baby Jane Doe.”81 Her parents were presented with two options to treat the meningomyelocele: primary skin healing or surgical repair. The parents refused consent for surgical repair of the defect and for the placement of a shunt for hydrocephalus. Instead, the parents requested that the infant be treated with antibiotics and nutritional support. An attorney who was not related to the family thought that the parents’ request was an inappropriate medical decision. This attorney requested that the trial court appoint an independent guardian for the infant so that consent could be given to perform the surgeries. The trial court granted the attorney’s request, but the appellate court overturned that decision the following day. The appellate court found that the parents had chosen an acceptable medical option and had acted in the best interest of their child. While this issue was being dealt with in the state court system, HHS received a complaint from a “private citizen.” This complaint stated that Baby Jane Doe was being discriminated against because of her handicap. HHS referred the case to Child Protective Services, which concluded that there was no cause for state intervention. During this time, HHS also made repeated requests of the hospital to produce the infant’s medical records. The hospital refused on the grounds that the parents had not consented to release the records. The federal government sought to compel access to the medical chart, and filed a suit in federal district court under section 504 of the 1973 Rehabilitation Act.3 The courts found that the hospital had not violated any of the pertinent statutes because they were willing to perform the surgery if the parents would consent. The case is significant because it represents a rare case in which a judge has allowed withdrawal of life-sustaining treatment against parents’ wishes. As the case proceeded through the federal court system, various judges asserted that the infant was not being

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

discriminated against on the basis of her handicap. Ultimately, the case reached the U.S. Supreme Court.20 The U.S. Supreme Court took the opportunity to review the Baby Doe case, the Baby Doe regulations, the care of handicapped newborns, the role of the federal and state government in these cases, the rights of parents, and the rights and duties of caregivers. The court found that there was no violation of Section 504 because the withholding of treatment was secondary to lack of parental consent, not secondary to discriminatory withholding based on the infant’s handicap. Among other conclusions, the U.S. Supreme Court found that the parents had made reasonable decisions that were consistent with the best interests of their child. The court found no discrimination. The court also stated, “A hospital’s withholding of treatment when no parental consent has been given cannot violate Section 504, for without the consent of the parents or a surrogate decision maker the infant is neither ‘otherwise qualified’ for treatment nor has he been denied care ‘solely by reason of his handicap.’ Indeed, it would almost certainly be a tort as a matter of state law to operate on an infant without parental consent.” The final sentence of this quotation raises interesting questions in light of the Miller case, which is discussed later.

Baby K Baby K was found prenatally to have anencephaly.41 Her mother declined termination of the pregnancy, and the infant was delivered by cesarean section on October 13, 1992. The infant was initially placed on mechanical ventilation so that the diagnosis could be confirmed. After confirmation of the diagnosis, the caregivers approached the infant’s mother to request permission to withdraw the ventilator. Based on the mother’s religious beliefs that all life is sacred and must be protected, she insisted that the ventilator support be continued. When the infant was 9 days old, the hospital ethics committee met with the physicians and concluded that the care was futile. Attempts to transfer the infant to another facility were not successful, and the infant was eventually transferred to an extended care facility. Baby K required three subsequent hospitalizations secondary to respiratory distress. Each time, the mother insisted that the infant be reintubated. The caregivers and the infant’s father believed that the treatment was futile and inappropriate. The hospital sought a federal court ruling that would allow them to withhold the ventilator from Baby K in the future. The hospital sought a declaratory judgment that by withholding the ventilator that they would not violate the Emergency Medical Treatment and Active Labor Act (EMTALA), the state Child Abuse Amendments, the state Malpractice Act, or ADA. The court ruled (and the 4th Circuit of the U.S. Court of Appeals42 upheld) that the hospital was not entitled to such a declaration. The court reasoned that the infant’s anencephaly qualified her as “handicapped” and “disabled” for the purposes of ADA. Because of procedural concerns and federal and state issues, the court did not rule on the topics of malpractice or child abuse. In the court’s interpretation of EMTALA, they reasoned that because “stabilization” included establishing and securing an

61

airway, the refusal to intubate Baby K would be a violation of EMTALA. Because Baby K was handicapped and disabled, and the hospital received federal funds (e.g., Medicare payments), the hospital could not deny the requests of Baby K’s mother. The EMTALA legislation was initially intended to prevent hospitals from “dumping” nonpaying patients to other facilities. In this case, the hospital was not contending that the issue was payment for the treatment. The court found that EMTALA applied, however, and Baby K must be intubated and ventilated as long as that was the mother’s wish. The court stated that “absent finding of neglect or abuse, parents retain plenary authority to seek medical care for their children, even when the decision might impinge on a liberty interest of the child.” The 4th Circuit of the U.S. Court of Appeals stated that it was beyond their judicial limits to address the moral and ethical implications of providing emergency care to anencephalic infants. They stated that Congress did not want a case-by-case analysis, but rather desired that hospitals and physicians provide stabilizing care to all patients who present with an emergency condition. Many ethicists and clinicians view anencephaly as “a paradigm case of medical futility.”45 For them, the Baby K decision was considered inappropriate, mandating treatment that would not be considered in the best interests of the patient. Four years after the original decision, the 4th Circuit placed some limits on the reach of the Baby K decision. In Bryan v Rectors of the University of Virginia,7 the court limited the scope of EMTALA to emergency stabilizing treatment and declined to apply EMTALA to a hospital that entered a “do not resuscitate” order against the wishes of the family 12 days after the patient had been admitted.

Sun Hudson Sun Hudson was born in Houston, Texas, in September 2004 to his mother Wanda and an unknown father. He was intubated shortly after birth and subsequently diagnosed with thanatophoric dysplasia, a typically fatal short-limbed skeletal dysplasia. The neonatologists and bioethicists at Texas Children’s hospital believed that “it would be unethical to continue with care that is futile and prolongs Sun’s suffering.”36 His mother disputed the hospital’s diagnosis, however, and fought to keep Sun on the ventilator. Under the Advanced Directives Act,70 signed into Texas state law by then Governor Bush, physicians and hospitals can unilaterally withdraw life support against the families wishes as long as the hospital’s ethics committee agrees with the decision. The hospital is also required to allow at least 10 days so that the family can find another facility to accept the patient. In Sun’s case, Texas Children’s contacted more than 40 hospitals, but was unable to find another facility to accept him. In addition, although court approval is not required by the statute, the hospital offered to pay Ms. Hudson’s attorney fees and the case was heard in court. After several months, a judge supported the position of the hospital. Sun was extubated and died in February 2005. The case is significant because it represents a rare case in which a judge has allowed withdrawal of life-sustaining treatment against a parent’s wishes.

62

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Conclusions The preceding cases present the challenges faced when damaged newborns receive a level of care that is viewed to be inappropriate by some observers. If a child has “lethal” anomalies or a “terminal” condition, the courts generally find that parents are the primary decision makers concerning the level of care. In the Baby Doe case, the infant did not have a lethal underlying condition. Treatment would not have been “futile,” and, in retrospect, many observers believed that the decision to forgo surgery was not “reasonable.” The resulting national outcry was a measure of people’s dissatisfaction with the decision made by the physicians and the parents. In contrast, Baby Jane Doe had a course of therapy that was selected by her parents. Although it was not the most aggressive course of treatment, some experts believed that allowing the skin to grow over the meningomyelocele was an acceptable option. Other experts take issue with this course of therapy. The court ruled, however, that the parents selected an acceptable treatment option for their daughter. In the Baby K case, the child had what is largely viewed as a “lethal” anomaly. Despite the heartfelt objections of the physicians and staff, the courts ruled that care could not be withheld simply because the caregivers considered it to be morally and ethically inappropriate. The case of Sun Hudson shows that parents’ rights are not completely unlimited. In this case, however, the medical community and bioethicists agreed that Sun was suffering, and Texas had a law in place specifically allowing unilateral withdrawal of support. Can these four cases be reconciled? In the Baby Doe case, the parents made decisions that could be viewed as neglect. Baby Doe did not receive life-saving surgery because he was handicapped. It would seem that such withholding of treatment would currently be viewed as illegal. Baby Jane Doe’s parents made a choice that the court ruled was a reasonable one. There was no discriminatory withholding of care. In Baby K’s case, the courts supported a parental decision to ventilate a child with no cerebral cortex, a decision that many clinicians find untenable. Sun Hudson’s condition, thanatophoric dysplasia, was believed to lead to his prolonged suffering, and this may have contributed to the judge’s decision to override the mother’s wishes. If conclusions can be drawn, they would seem to indicate that the courts support parents’ interests as long as decisions are not being made solely on the basis of handicap, and as long as the infant is not believed to be suffering. The personal morals and ethics of the caregivers are fundamentally irrelevant in the legal context of parental decision making. If the parents’ decisions are reasonable, and there is no evidence of neglect or suffering, parents seem to have substantial decision-making power.

PROVIDING CARE AGAINST PARENTS’ WISHES One of the most challenging aspects of neonatology is caring for extremely premature infants. A substantial portion of this textbook addresses the medical issues involved in caring for these infants. Although the care of extremely premature

neonates largely defines the parameters of the specialty, it also creates many of the legal and ethical quandaries for caregivers and families. These tiny individuals existing on the cusp of viability have the same legal rights as all citizens. They are entitled to equal protection, due process, and all other constitutionally guaranteed rights and privileges of the citizens of the United States. By virtue of a heartbeat or spontaneous respiratory effort, an extremely premature infant is transformed into a “person.” As mentioned in the earlier discussion of live birth, in many states, the gestational age and whether the placenta is attached are legally irrelevant. The pending delivery of a 23-week gestation fetus generally carries a wide variety of concerns. What are the family’s wishes? Will the resuscitation go smoothly? Will all the equipment function properly? How will the infant respond to the resuscitation? Will stabilization be difficult? Will the infant require significant ventilator support, volume boluses, or an infusion of catecholamines? More recently, the overriding concern is whether the infant must be resuscitated if it seems to be previable. What if the parents request no resuscitation? Are the caregivers liable if they overrule the parents and proceed with resuscitation of a 23-week gestation infant? What if the parents request no resuscitation for a 24-week or 25-week infant? This extraordinarily difficult situation is exacerbated by a relative lack of statutory and case law. The existing case law seems to conflict with itself. This section addresses the unique challenge associated with the delivery of extremely premature infants. To gain insight into this convoluted area of law, the reader should review the material on perinatal issues (maternal-fetal conflict), live birth, informed consent, and limiting care. The reader should have a grasp of the Cruzan25 decision and the associated liberty interest in keeping one’s body free from unwanted medical intervention. Likewise, one must appreciate that courts have found that parents generally have the right to refuse certain unwanted medical intervention for their children as long as this refusal is not neglect or abuse. Without a familiarity with the rules governing informed consent, the reader lacks the necessary foundation to appreciate the following discussion.

Miller Case In 1990, Karla Miller went into preterm labor at approximately 23 weeks’ gestation. After the neonatologist explained the grim prognosis for infants born at this gestation, Mark and Karla Miller requested that the infant not receive heroic measures. Initially, the obstetrician and neonatologist agreed to honor the parent’s wishes. A hospital administrator claimed, however, that the hospital had a policy requiring that the infant be resuscitated if she weighed more than 500 g. This policy was explained to the parents, who again requested that the infant not be resuscitated. Sidney Miller was born later that night, several hours after Mrs. Miller had been admitted to the hospital. The infant was resuscitated, and she survived with severe impairment. The parents sued the hospital, asserting claims in battery and negligence. They did not sue the physicians, although they were involved in the trial, because they believed that the physicians were following orders from the administration. The main allegation was that the hospital was liable for mandating

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

Sidney’s resuscitation without parental consent. The hospital maintained that the parents had no right to refuse life-saving intervention. The jury awarded the family approximately $60 million. The appellate court33 overturned this verdict. In their analysis, the appellate court stated that this was a situation of the emergency exception to the informed consent rule. The court also relied on the Advanced Directives Act. This act protects caregivers and hospitals who withhold care from terminally ill patients. The court reasoned that because Sidney Miller’s condition did not fit the definition of terminal, the parents had no right to refuse life-saving therapy. The court stated that although parents have a right to determine health care decisions for their children, this is not an absolute right, and the state also has an interest in the health of children. “Having recognized, as a general rule, that parents have no right to refuse urgently-needed life-sustaining medical treatment to their non-terminally ill children, a compelling argument can be made to carve out an exception for infants born so prematurely and in such poor condition that sustaining their life, even if medically possible, cannot be justified.” The appellate court concluded that perhaps the legislature should address the issue of defining “terminal” with respect to some premature infants who are born so small and so sick. A dissenting judge on the appellate court believed that the parents’ course of action was lawful. This judge supported his opinion by quoting the U.S. Supreme Court’s decision in the Baby Jane Doe case. This dissenting judge stated that no emergency existed. The emergency exception to the informed consent rule was not available to the caregivers. According to the dissent, it was the caregivers’ delay and indecision that led to the urgency. Thirteen years after Sidney Miller was born, the Texas Supreme Court34 ruled on the case. This court found that the hospital was not liable for resuscitating the infant. Although the Texas Supreme Court analyzed the case differently than the appellate court had, the decision was the same: The hospital had no liability. The high court reasoned that because it was impossible to predict how sick the infant would be at birth, the emergency did not exist until after Sidney was born. The physician could not evaluate the situation until after the infant was born.

Messenger Case Gregory Messenger and Traci Messenger were faced with a situation similar to the Miller family. Traci Messenger went into labor at 26 weeks. Gregory Messenger was a physician (dermatologist). After discussing their options with their caregivers, the Messengers requested that the infant not be resuscitated. The request was not honored. The infant was resuscitated and brought to the NICU. Dr. Messenger went into the NICU, extubated his son, and placed him in Traci Messenger’s arms. The baby died shortly thereafter. The neonatologist listed the cause of death as “homicide.” Approximately 1 year later, Dr. Messenger was acquitted of manslaughter.57 Among other conclusions, the jury believed that Dr. Messenger was acting in the “best interest”80 of his son. The Messenger case was a criminal trial with a higher standard of proof. Dr. Messenger was acquitted of a crime. He

63

removed his own extremely premature son from a ventilator, and he was found to be not guilty of intentional killing.

Montalvo Case Baby Emanuel (Montalvo) Vila was born in Wisconsin in November 1996 at 23 weeks’ gestation weighing approximately 679 g. His mother had presented earlier with preterm labor, and after attempts to delay his birth were unsuccessful, a cesarean section was performed after informed consent was obtained. Three years after his birth, the family sued the obstetrician, the neonatologist, and the hospital alleging that they had resuscitated Emanuel without advising the parents of the risks or potential consequences of an extremely premature infant. The trial court found for the physicians and the hospital, and the family appealed. The Wisconsin State Court of Appeals agreed with the trial court for two reasons. First, the court noted that under Wisconsin law, in the absence of a persistent vegetative state, a parent does not have the right to withhold lifesustaining medical treatment. Second, the court looked at the current incarnation of the Baby Doe regulations, known as CAPTA regulations (discussed earlier under “Handicapped Newborns”). The court believed that the parents did not have a choice to refuse resuscitation since “[t]he implied choice of withholding treatment . . . is exactly what CAPTA prohibits.”51 The Trial Court and the Appellate Court addressed the public policy issues associated with resuscitation of extremely premature infants, and both courts pointed out that the decision-making authority in these cases cannot be placed “wholly” in the hands of the parents. This portion of the opinion would seem to imply that parents and caregivers have decision-making responsibility in these cases. The potential implications of the Montalvo case are considerable. Because neonates are rarely in a persistent vegetative state, the ruling does not seem to give the option to withhold or withdrawal treatment for most critically ill neonates. The decision applies only in Milwaukee, which is the area covered by District 1. Also, the court stated that parents were not “wholly” responsible for these decisions. This language would imply some role for parental decision making in concert with the attending physicians. The long-term impact of the case remains to be seen.

Conclusions The Miller case was one decision made in one state. The decision was made by the Texas Supreme Court, so the decision is the current law only in Texas. The court’s decision in the Miller case may or may not have any effect in any other state. Currently, if one practices in Texas, Miller v HCA controls one’s practice. Clinicians should be familiar with the facts and the judicial conclusions in this case. It seems that clinicians and hospitals in Texas would not be liable for resuscitating 23-week gestation infants against the parents’ wishes. Despite the parents’ clear request to forgo resuscitation, a lapse of several hours between the time of the request and the time of birth, and no effort on the part of the clinicians or the hospital to transfer care to another provider, the court found no liability. Unless the U.S. Supreme Court ultimately rules

64

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

on Miller v HCA, or the Texas legislature passes contrary statutes, this case is binding in Texas. In Wisconsin, the Montalvo case seems go a step further and suggests that neonatologists may be required to resuscitate all extremely premature infants. In stark contrast to the American Academy of Pediatrics emphasis on shared decision making, the court claims that parents have no rights to decline treatment under the current Baby Doe regulations. Yet this interpretation is controversial. The Institute of Medicine report on preterm birth notes that the Baby Doe regulations “were not originally intended to apply to premature infants; rather, they were intended to apply to disabled fullterm infants.”59 As a state appellate court ruling, however, this interpretation applies only in the Milwaukee area. In the Baby Jane Doe case, the U.S. Supreme Court did comment on decision making being taken away from parents. In defending the hospital’s decision to honor the request of Baby Jane Doe’s parents, the U.S. Supreme Court stated, “it would almost certainly be a tort as a matter of state law to operate on an infant without parental consent.” What of the remainder of the United States? Perhaps some courts in other states would follow the trial court or the appellate court dissent in Miller. The trial court in Miller allowed a $60 million verdict. The dissent in the appellate court quoted the U.S. Supreme Court decision in Baby Jane Doe, to question the propriety of imposing on parents the consequences of resuscitation of 23-week gestation infants. The Texas Supreme Court and the Wisconsin Appellate Court decided that parents have no right to refuse resuscitation of their extremely premature infants, using different rationales. The Texas court relied heavily on their interpretation that the emergency did not exist until Sidney Miller was born. The hospital and caregivers would basically always be protected by the emergency exception to the informed consent rule. The Montalvo court relied on a State (Wisconsin) Supreme Court ruling and the federal Baby Doe regulations. The court believes that CAPTA prohibits withholding of resuscitation no matter how small or immature the infant. This is a drastic change from how many clinicians currently practice and removes physician judgment and the parents’ view of what is best for their child from the equation. How does one reconcile Miller, Montalvo, and Messenger? It is intellectually dissatisfying to conclude that the major difference between these cases is that they were adjudicated in different states. The inherent conflict between the cases must have more substantial legal underpinnings than a simple difference in jurisdiction. The fact that the Messenger case was a criminal trial with a higher burden of proof on the prosecution has some effect on the legal comparisons, but the inherent inconsistencies exist. In the Miller case, an extremely premature infant is resuscitated against the parent’s wishes, and the hospital has no liability. In the Montalvo case, an extremely premature infant is resuscitated, and when the parents state they were not adequately informed, the court states that informed consent is irrelevant because the Baby Doe regulations do not allow parents to refuse resuscitation for even the smallest and most immature patients. In the Messenger case, an extremely premature infant was resuscitated against the parent’s wishes, and the father had no criminal liability for disconnecting the ventilator with the intent to hasten the infant’s death. If the

BOX 4–3 When Parents and Caregivers Disagree on the Care of Critically Ill Newborns: Lessons from Existing Case Law PARENTS REQUEST FULL SUPPORTIVE CARE AND CAREGIVERS DISAGREE General rule: provide full supportive care Anencephaly: Baby K case likely controls (federal case) PARENTS WANT NO RESUSCITATION AND CAREGIVERS DISAGREE Miller (Texas): no liability for caregivers Montalvo (opinion covers Milwaukee, Wisconsin): no liability for caregivers Messenger (Michigan): no criminal liability for father who disconnects the ventilator PARENTS AND CAREGIVERS AGREE TO FORGO AGGRESSIVE TREATMENT Possibly covered by Baby Jane Doe case: parents have right to make reasonable choice Baby Doe case: cannot make decision solely based on present or future handicap Montalvo (opinion covers Milwaukee, Wisconsin): parents may not have right to make reasonable choice.

infant was lawfully being cared for, how could the father be acquitted of manslaughter? If caring for the infant against the parent’s wishes was not lawful, why did the Miller and Montalvo courts find for the hospital? These cases leave in question the exact legal status of extremely premature infants. The infant is “alive” by statute. In some jurisdictions, the infant can recover for wrongful life. In some jurisdictions, a wrongful life claim would be denied because the parents have no right to refuse the care. In some jurisdictions, there is no criminal liability for a parent who overrules the physicians and takes matters into his or her own hands. So what should a clinician do when called to the birth of a previable or periviable newborn? One should understand the issues, understand the rights of the newborn, understand the rights and the duties of the parents, and understand one’s own rights and duties. Box 4-3 provides some general lessons from existing case law concerning situations in which parents and caregivers disagree on the care of critically ill newborns. What would happen if Sidney Miller had been born in a state other than Texas? Did state politics play a role in this case? Is it a coincidence that Gregory Messenger was acquitted in Michigan? These are not trivial questions, but they are unanswerable given the current case law on this issue. To the family of Sidney Miller, the Supreme Court of Texas determined that the child and the family should recover nothing.

SUMMARY A physician facing a legal issue can experience various difficulties. Physicians are often unfamiliar with the legal process, and the new terminology can be daunting. Compared with science, legal results can seem to be unpredictable. In law, there are no double-blind, randomized, controlled trials.

Chapter 4  Legal Issues in Neonatal-Perinatal Medicine

Additionally, legal events often elicit strong emotions. With respect to malpractice, families may have suffered tremendous losses, and juries in some states can award tens of millions of dollars. Generally, practicing strong clinical medicine requires the clinician to stay current in the specialty, to maintain excellent communication with families and the hospital staff, to strive consistently to make ethical decisions, and never to violate the law knowingly. Veteran neonatologists know the value of good documentation and effective communication with colleagues, staff, and families. Many other nuances exist, however. Practitioners benefit from understanding the elements of a tort suit; recognizing the importance of informed consent; and knowing the law concerning the care of handicapped newborns, anencephalic infants, and extremely premature infants. With respect to extremely premature newborns, it is particularly difficult to discern clear legal principles from the various court decisions in this arena. This issue speaks to the importance of knowing the law in the state in which one practices. In Wisconsin, clinicians must know the appellate court district in which they are practicing. Although the law may seem to be arbitrary, there are substantial underlying principles that courts and legislatures have honored for hundreds of years. Fundamentally, the legal system is not unpredictable. It can be daunting to a physician, but the process can be demystified. If one is experiencing a significant health problem, one seeks out medical advice. Likewise, the legal profession can provide meaningful insight to a physician who is in need of counsel.

REFERENCES 1. 29 USCA §794. 2. 42 USC 5101 et seq; 42 USC 5116 et seq. 3. 45 CFR §84.61 (1985). 4. 45 CFR §84.55. 5. 48 Fed Reg 9630. 6. 65 ALR3d 413 (1975). 7. 95 F3d 349 (4th Cir 1996). 8. Administration for Children & Families (website). http://www. acf.hhs.gov/programs/cb/laws_policies/policy/pi/pi0501a.htm. Accessed February 9, 2010. 9. Ala Code 1975 §§26-21-2, 26-22-2 (2001). 10. Alaska Stat §18.50.950 (2001). 11. American Academy of Pediatrics Committee on Fetus and Newborn: Advanced practice in neonatal nursing, Pediatrics 111:1453, 2003. 12. Ariz Rev Stat §13-1103 (A)(5) (West 1989 & Supp 1998). 13. Arkebauer v Lojeski, No. 99 L 005157D (Ill 2004). 14. Arpin v US, 521 F.3d 769 (7th Cir. 2008). 15. Article X, Section 26 (website). http://www.leg.state.fl.us/ statutes/index. 16. Baby Doe Redux. Pediatrics 116:e576, 2005. 17. Becker S: Health care law: a practical guide, 6th ed. New York, LexisNexis Matthew Bender, 1999, §17.02. 18. Bolin v Wingert, 764 NE2d 201 (Ind 2002). 19. Boumil M et al: Medical liability, 2nd ed. St. Paul, MN, Thompson West Publishers, 2003, p 186. 20. Bowen v American Hospital Association, 476 US 610 (1986). 21. Brownsville Pediatric Associates v Reyes, 68 SW3d 184 (Tex App 2002).

65

22. Byrne v Boadle, 2 H&C 722, 159 EngRep 299 (Court of Exchequer 1863). 23. Cal Civil Code §3333.2 24. CNN.com (website). http://www.cnn.com/2007/US/law/07/24/ wrongful.birth.ap/index.html. Accessed February 9, 2010. 25. Cruzan v Director, MDH, 497 US 261 (1990). 26. Del Stat Title 16 §3101. 27. Farley v Sartin, 466 SE2d 522,528 (WVa, 1995). 28. Florence v Plainfield, 849A.2d7 (Conn, Sup Ct, Jan 16, 2004). 29. Ga Code Ann, §31-7-130. 30. Genetic Engineering & Biotechnology News (website). http:// www.genengnews.com/articles/chitem.aspx?aid52132. Accessed February 9, 2010. 31. Greco v US, 111 Nev 405, 893 P2d 345 (1995). 32. Gross v Burt, 149 SW3d 213 (Tex. App, 2004). 33. HCA Inc v Miller ex rel Miller, 36 SW3d 187 (2000). 34. HCA Inc v Miller ex rel Miller, 118 SW3d 758 (Tex 2003). 35. Hill by Burton v Kokosky, 463 NW2d 265 (Mich App, 1990). 36. Hopper L. Ruling keeps baby on life support / Mom given time to find alternative after hospital says case is hopeless. Houston Chronicle January 26, 2005, section A, page 01. 37. Hubbard v State 852 So2d 1097 (La App 4th Circ, 2003). 38. Infant Doe v Bloomington Hospital, 464 US 961, 104 SCt 394 (1983). 39. In re Fiori 673 A2d 905 (1996). 40. In re Infant Doe, No. GU8204-004A (Ind Ct App Apr 12, 1982). 41. In the Matter of Baby K, 832 FSupp1022 (EDVa 1993). 42. In the Matter of Baby K, 16 F3d 590 (4th Cir 1994). 43. Knapp v Northeastern Ohio Obstetricians, Ohio App 11 Dist (2003). 44. La Rev Stat Ann §40:1299.39. 45. Lantos JD, Meadow WL. Neonatal bioethics: the moral challenges of medical innovation. Baltimore, Johns Hopkins, 2006. 46. Lewis v Physicians Insurance Co of Wisconsin, 243 Wis2d 648 (2001). 47. Maine Rev Stat Ann tit 22, §1596. 48. Medical News Today (website). http://www.medicalnewstoday. com/articles/75438.php. Accessed February 9, 2010. 49. Mo Rev Stat §193.015. 50. Mo Rev Stat §568.045 (VAMS 2003). 51. Montalvo v Borkovec, 647 NW 2d 413 (Wis App 2002). 52. National Association of Neonatal Nurse Practitioners. Position statement on the Doctor of Nursing Practice (DNP) degree. Glenview, IL: National Association of Neonatal Nurses; 2008. http://www.nann.org/nannp/index.html. Accessed November 20, 2008. 53. NCSL (website). http://www.ncsl.org/standcomm/sclaw/ medmaloverview.htm. Accessed February 9, 2010. 54. Nold v Binyon, 31 P3d 274 (2001). 55. O’Neil v Great Plains Women’s Clinic, 759 F2d 787 (1985). 56. Pediatrics (website). http://pediatrics.aappublications.org/cgi/ reprint/111/3/680-a.pdf. Accessed February 9, 2010. 57. People of the State of Michigan v Gregory Messenger, Ingham County Circuit Court, Lansing, Mich, Docket 94-67694FH, February 2, 1995. 58. Point of Law (website). http://www.pointoflaw.com/ archives/000747.php. Accessed February 9, 2010. 59. Preterm birth: causes, consequences, and prevention. Institute of Medicine, 2007. 60. Pub L No. 107-207, 116 Stat 926 (2002).

66

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

1. Quirk v Zuckerman, 196 Misc 2d 496 (NY, 2003). 6 62. Reynolds EW, Bricker JT. Nonphysician clinicians in the neonatal intensive care unit: meeting the needs of our smallest patients, Pediatrics 2007;119:361-369. 63. RI Gen Laws §9-19-41. 64. Rosario v Brookdale University Hospital, 767 NYS2d 122 (2003). 65. Rouse v Pitt Co Memorial Hosp, Inc, 343 NC 186, 470 SE2d 44 (1996). 66. Sanbar SS, Firestone MH: Medical malpractice stress syndrome. In Sanbar SS, editor: The medical malpractice survival handbook. Philadelphia, Mosby Elsevier, 2007. 67. Sonlin v Abington Memorial Hospital 748A2d 213 (2000). 68. State ex rel Infant Doe v Baker, No. 482 §140, May 27, 1982. 69. Sterling v Johns Hopkins Hospital, 802 A2d 440, 2002. 70. Texas Health & Safety Code Ann. § 166.046(a). 71. The Briefing Room (website). http://www.flsenate.gov/data/ session/2008/senate/bills/billtext/pdf/s0072.pdf. Accessed February 9, 2010. 72. The Briefing Room (website). http://www.whitehouse.gov/news/ releases/2008/01/20080128-13.html. Accessed February 9, 2010.

73. The Dallas News (website). http://www.dallasnews.com/ sharedcontent/dws/bus/stories/DN-medmal_17bus.ART0.State. Edition2.43983f4.html. Accessed February 9, 2010. 74. The New York Times (website). http://www.nytimes. com/2004/11/26/national/26malpractice.html. Accessed February 9, 2010. 75. The Southeast Texas Record (website). http://www.setexasrecord. com/news/208641-class-action-challenges-texas-cap-onmedical-malpractice-damages. Accessed February 9, 2010. 76. Turbow R, Fanaroff J: Legal issues in newborn intensive care. In Sanbar SS, editor: Legal medicine, 7th ed. Philadelphia, Mosby Elsevier, 2007. 77. Turpin v Sortini, 32 Cal3d 220, 643 P2d 954 (1982). 78. Utah Code Ann §26-2-2. 79. Vo v Superior Court, 172 Ariz 195,198, 836 P2d 408,411 (App 1992). 80. Walters S: Life-sustaining medical decisions involving children: father knows best, 15 Thomas Cooley Law Review 115:143, 1998. 81. Weber v Stony Brook Hospital, 60 NY2d 208.

CHAPTER

5

Quality and Safety of Neonatal Intensive Care Medicine PART 1

Evaluating and Improving the Quality and Safety of Neonatal Intensive Care Jeffrey D. Horbar and Jeffrey B. Gould

In this chapter, we review the ways information can be collected, evaluated, and applied to improve the quality and safety of medical care for newborn infants and their families. We discuss the available sources of such data for neonatology and describe how these data can be used to evaluate and improve the processes, outcomes, and costs of medical care for newborn infants. We begin by focusing on the case that improvement is necessary in neonatology.

THE CASE FOR IMPROVEMENT The nation’s health care system lacks . . . the capabilities to ensure that services are safe, effective, patient-centered, timely, efficient and equitable. . . . Between the health care we have and the care we could have lies not just a gap but a chasm.13 More people die in a given year as a result of medical errors than from motor vehicle accidents, breast cancer, or AIDS.48

Systematic evaluation of the quality, safety, and efficiency of clinical care has become an integral part of medical practice. Physicians, hospitals, and large health care organizations are under increasing pressure to monitor, report on, and continuously improve the quality, safety, and cost-effectiveness of their services. Public release of hospital performance data and report cards are becoming increasingly common.51 Relman has described this as the “era of assessment and accountability.”69 In this new era, health professionals in neonatology must learn how to evaluate themselves and learn how they will be evaluated by others, including policy makers, hospital administrators, regulators, payers, and the families and public they serve. Evaluation is not an end in itself. Health professionals must learn how to use available information to improve the quality and safety of medical care continuously. This is not just an option. It is a responsibility. The American Board of Pediatrics’ 2010 maintenance of certification requirement that neonatologists be actively engaged in quality improvement emphasizes this professional responsibility (www.ABP.org).

The Institute of Medicine of the National Academy of Sciences has issued two landmark reports: Crossing the Quality Chasm: A New Health System for the 21st Century, and To Err is Human: Building a Safer Health Care System. These reports present a clear and compelling challenge to all health care professionals to improve the quality and safety of the medical care for the patients and families they serve.13,48 Despite overwhelming evidence that deficiencies in quality and safety are widespread throughout the American health care system, many health care professionals in neonatology may feel that these problems do not apply to our clinical specialty. This is not the case. In neonatology, as in other clinical fields, opportunities for improving the quality and safety of medical care are substantial. First, infants receiving neonatal intensive care remain at high risk for mortality and acute as well as long-term morbidity. Although mortality rates for high-risk preterm infants declined steadily over the past several decades, this trend appears to have now reached a plateau.19,39,94 Furthermore, extremely premature infants who do survive are at high risk for neurodevelopmental and sensory disabilities as well as educational disadvantages in young adulthood.30 Second, variation among neonatal intensive care units in the processes and outcomes of care is dramatic. This variation cannot be explained by differences in case mix, suggesting that it is at least in part due to differences in the quality

67

68

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

of care.96 Nosocomial infection provides a striking example.10 The 682 units participating in the Vermont Oxford Network database in 2007 reported that 19.8% of the more than 49,000 very low birthweight infants had a nosocomial bacterial infection during their hospitalization.94 The rates varied dramatically among units, with 25% of units having rates under 9.8% and 25% having rates over 25%. This dramatic variation could not be explained by case mix or type of neonatal intensive care unit (NICU), indicating a significant opportunity for improvement. Third, inappropriate care—defined as underuse, overuse, and misuse of interventions24—is common in neonatal intensive care. Examples include the underuse of hand hygiene by NICU personnel,9 the overuse of antibiotics, and the misuse of medications because of medical errors. Medical errors occur frequently in the neonatal intensive care unit, leading to adverse events. A study at two Boston hospitals has documented that errors in the process of ordering, dispensing, or monitoring medications occurred for more than 90% of the infants cared for in the NICU.47 Using a voluntary, anonymous, Internet-based error-reporting system established by the Vermont Oxford Network, Suresh and colleagues have documented a broad range of errors and near errors at 54 neonatal intensive care units.89 Only about half of the reported events involved medications; the remainder involved a wide variety of errors in multiple domains of care. A recent study from eight Dutch NICUs found that incidents concerning mechanical ventilation, blood products, intravascular lines, parenteral nutrition, and medication dosing errors pose the highest risk to patients in the NICU.86 These observations are consistent with the findings of the Institute of Medicine reports. We have many opportunities to improve the quality and safety of neonatal intensive care.

Data for Improvement It is essential that quality improvement be data driven. As a first step to quality improvement, data are used to determine baseline performance relative to a set of standards or peerderived benchmarks. As changes are implemented, data are used to track improvement. Data regarding neonatal patients are available from a variety of sources (Box 5-1). There are two potential sources of data: primary and secondary databases. Primary databases are designed specifically for evaluating neonatal care; secondary databases were originally designed for other purposes.24 Examples of secondary data sets that have been used to evaluate quality of care are the hospital discharge database and files that link birth certificates and death certificates. Although these secondary databases were not designed for evaluating perinatal care, they contain data elements that have made it possible to examine risk-adjusted perinatal complication rates,21 cesarean delivery,26 neonatal morbidity and mortality rates,64 and neonatal readmission rates.15 Although secondary data sets can provide useful information, the accuracy and completeness of reporting poses a potentially important limitation.53,54 Strategies to improve the collection and recording of key quality indicators on secondary data sets may be a workable alternative to building a primary data collection system. The California Maternal Quality Care Collaborative (www.cmqcc.org) is evaluating the feasibility of applying this

BOX 5–1 Data Sources for Evaluating Neonatal Care VITAL STATISTICS Federal and state data Birth certificates Death certificates Hospital information systems Clinical information systems Administrative information systems Decision support systems CLAIMS DATA Medicaid Other insurers NEONATAL DATABASES Locally designed databases Commercial databases NATIONAL AND INTERNATIONAL NEONATAL NETWORKS Australian and New Zealand Neonatal Network British Association of Perinatal Medicine Canadian NICU Network EuroNeoNet Indian Neonatal Forum Network International Neonatal Network Ireland-Northern Ireland Network Israel Neonatal Network Italian Neonatal Network Japanese Neonatal Network National Collaborative Perinatal Neonatal Network of Greater Beirut NICHD Neonatal Research Network SEN1500 (Spain) Swiss Neonatal Network Vermont Oxford Network NICHD, National Institute of Child Health and Human Development; NICU, neonatal intensive care unit.

approach to hospital discharge billing data on maternal hemorrhage. Primary databases are specifically created to address perinatal issues. There are two basic approaches to primary database design. One approach attempts to create a virtual patient. This approach is exhaustive because it contains a detailed description of all aspects of a patient’s hospital stay that would be included in the medical record. The advantage of the virtual approach is that it allows detailed and broadbased analyses that even include factors whose importance will be identified only in the future. The disadvantage is that multiple-item primary data collection, entry, and quality control are resource intensive. In the future, as clinical medicine moves to the electronic medical record and as the various electronic repositories of medical information (e.g., laboratory, consultant, imaging, outpatient, resource use, and costs) are integrated,57 it will be possible to construct a multidimensional virtual patient based entirely on routinely collected secondary data. Today, the minimal data set is a practical and popular approach to collecting primary perinatal data.

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

Minimal Data Set The minimal data set approach begins with the notion of creating an information base. An information base is founded on the premise that information is data that promotes action by informing decision-making and quality-improvement efforts.98 Therefore, the first task in constructing a minimal database is to determine the areas of decision making and quality improvement to be addressed and the information needed to address these specific issues. Only data that are needed to inform these issues qualify for inclusion in a minimal database. Achieving the appropriate granularity—that is, the proper balance between detail and simplicity—is difficult but crucial. It involves making tradeoffs between the depth and breadth of the information to be collected and the costs of collecting the data. These costs are not trivial. As more and more emphasis is placed on documenting and improving outcomes, data abstraction, entry, and analysis have become essential components of clinical care. Unfortunately, administrative budgets often fall short of what is required to support these activities adequately. Regardless of how many data elements are included in a database, the elements must be clearly defined. Database users should have access to standardized definitions in a printed manual of operations to facilitate both the coding and the interpretation of data. Access to standardized definitions is particularly important with regard to diagnostic information. Because physicians and nurses rarely use precise definitions in their daily notes, the medical record might not provide a reliable source of clinical information. The importance of uniform definitions cannot be overstated. Without them, valid comparisons and inferences cannot be made from a database. A few well-defined data items are far more valuable than an extensive list of poorly defined items. There are currently three methods for transferring these items to a database. The most efficient technique is to download information that has already been collected via an electronic medical record. Although this will doubtless be the standard method in the future, it is still in the early stages of development. The second and most commonly used method is the paper form. Data are abstracted from the medical record to the paper form, which is sent to the data center and either scanned or hand entered into a master database. The paper forms that are used to record data must be simple and easy to use. However, even the clearest of paper forms has the major disadvantage of quality-control lag. Thus, to identify an error on the form (e.g., omission of an item, an out-of-range entry, or a logically inconsistent value), one must first expend the effort to enter the form into the database and run an error check. Having found the error, one must then communicate back to the NICU. The NICU then has to request the record from the medical records repository, reabstract the data item, and send back the correction, which is then used to update and correct the initial error. This lag in quality control results in a detection-correction process that requires a great deal of effort by the NICU and the data center. To avoid this labor-intensive process, there has been a recent movement toward on-site computer-based data entry systems using local or Web-based data entry interfaces. For

69

example, the Vermont Oxford Network provides members with dedicated software, eNICQ, to collect, manage, and submit standardized data. The California Perinatal Quality Care Collaborative (CPQCC) supports data submission and management with a dedicated on-line tool (www.cpqcc.org). The great advantage of these systems is that they detect errors in completeness, range, and consistency as the data are being abstracted and entered. When the data entry program detects an error, it requests a valid value and does not allow data entry to proceed until it receives the correct value. The result is a clean record at the point of data entry that avoids the costly off-site error detection and correction cycles. However, even with this system it is important to use methods to minimize data entry errors, such as visual verification.37 Creating a clean data set is only the first step in using data to inform quality improvement. To be effective agents for change, reports generated from the data must be available in a timely manner. Data that are too old cannot inform decision making reliably. Because of the lengthy hospitalization of the very premature infant, it is often useful to enter data at the end of the first 28 days as well as at discharge. This strategy facilitates the timely analysis of neonatal outcomes. The format in which data reports are produced and distributed is evolving from the yearly paper report to more flexible electronic media such as the compact disk and the confidential Internet-based report such as the Vermont Oxford Network Nightingale Internet Reporting System. Nightingale provides users with secure real-time access to all data submitted since 1990, allows users to track trends over time, and enables them to compare their unit’s performance with the Network as a whole and with predefined subgroups of units similar to their own (Fig. 5-1). The CPQCC also provides its members with web-based reports (see www. cpqccreport.org, use logon “0000” and password “test”). The CPQCC web-based system has an integrated data entry and report structure that generates confidential individual member reports and network-wide benchmark metrics in real time. Advantages of electronic media reports are the ease of performing local secondary data analysis as well as the ability to easily incorporate the report’s tables and figures into local customized presentations. The presentation of a hospital’s data in comparison with that from other NICUs in a network plays an important role in motivating quality improvement activities.99 When designing a primary database or deciding which variables to extract from a secondary database, it is useful to consider the type of variables to be collected. Clinical databases consist of a series of records, each record containing data on an individual patient. A perinatal database contains four types of data: identifiers, processes, outcomes, and risk adjusters.

DATABASE ELEMENTS Identifiers Although identifiers are usually thought of as being normative (e.g., name, Social Security number, case record number), they can also be virtual. A virtual identifier is a combination of characteristics, such as date of birth, race, residential zip code, and birthweight, that are unique to a specific infant.

70

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE Figure 5–1.  Sample screen from

the Vermont Oxford Network Nightingale Internet reporting system showing data for a fictitious center 999 compared with all NICUs in the Vermont Oxford Network in 2003. Network data include the overall mean value and the quartiles for hospital values. Users may generate a wide variety of tables and figures. (Courtesy of the Vermont Oxford Network, Burlington, Vt.)

Although one might favor normative identifiers, it is not uncommon for an infant’s name to be changed during the first several months of life, and Social Security numbers are not always assigned to infants. NICU patients are frequently at risk for misidentification errors as a result of similarities in standard identifiers, such as names and medical record numbers.28 Because virtual identifiers consist of a constellation of factors that are constant (e.g., birth date, sex, birthweight, mother’s age) and tend to be unique to a specific person, they are extremely powerful, and it is possible to effectively link birth, death, and discharge data sets using virtual identifiers.32 The Health Insurance Portability and Accountability Act of 1996 (HIPAA) has required increasing attention to the use of personal identifiers in patient databases.49

Processes Having identified the patient, it is important to record certain processes of care. In a minimal data set, only processes that have been strongly linked to the quality of outcomes should be included. The goal of using process analysis for quality improvement is to detect the underuse of processes that have been demonstrated to improve outcomes, such as the use of antenatal steroids in mothers facing premature delivery. Process analysis can also detect the overuse of processes that have been demonstrated to be detrimental, such as the use of postnatal steroids. In contrast to the minimal data set, the virtual patient approach attempts to capture all activities and the diagnoses

they address. Detailed electronic medical record, billing, and cost-accounting systems can be used to create a virtual patient based on secondary data. Such databases, developed as components of administrative hospital information systems, are relatively new to pediatrics and neonatology. Because these databases were designed primarily for financial analysis, coding of clinical data and procedures might not be accurate. Another drawback is that physician procedures often are not included in hospital billing systems. To be effective for evaluating clinical care, cost-accounting data sets must be customized to the local environment. This process requires clinical input and provides the neonatologist with an opportunity to build in appropriate safeguards to reflect more accurately clinical status and procedures. Administrative information systems promise important possibilities for the future, but today primary neonatal databases offer the best approach to recording what doctors and nurses do for their patients. As a primary database, the minimal data set can be expanded readily to collect more detailed information on the processes of care. For example, one can incorporate the Neonatal Therapeutic Intervention Scoring System (NTISS) to obtain a measure of neonatal therapeutic intensity.27 This score, based on a modification of the adult Therapeutic Intervention Scoring System (TISS), assigns 1 to 4 points for each of 62 intensive care therapies in 8 categories (respiratory, cardiovascular, drug therapy, monitoring, metabolic/nutrition, transfusion, procedural, vascular access). Data for the score are abstracted from the medical record. The

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

NTISS score is highly correlated with markers of illness severity and is a measure of nursing acuity, and it is predictive of NICU length of stay and total hospital charges for survivors. If scores such as NTISS can be simplified and validated in a large number of NICUs, they will be valuable tools for measuring the scope and intensity of therapeutic interventions. Ideally, the data items needed for measuring therapeutic intensity should be available in either the primary or the secondary data sets already being collected. Future neonatal database systems should attempt to incorporate measures of therapeutic intensity so that chart review is not necessary for collecting the required data items. Patients differ inherently in the therapeutic intensity that they require, so a perinatal service’s outcomes must always be measured within the context of its case mix.

Outcomes Outcomes are the third constituent of the perinatal database and are essential to assessing quality of care. Pragmatically, outcomes are negative events such as death and morbidity, and quality is inferred on the basis of a lower-than-expected negative event rate. It is important to record a wide spectrum of outcomes, but for an outcome to serve as an effective quality indicator, evidence must be strong that variations in process or structure can decrease its incidence. Nosocomial infection is an important quality indicator because it is a significant source of morbidity, and incorporating certain processes of patient care (e.g., hand hygiene and intravenous line care) has been shown to reduce its incidence.90 Characterizing the incidence of outcomes at an individual NICU and across a network of NICUs plays several key roles in the quality improvement cycle. It allows identification of problem outcomes on the basis of their high incidence. It also allows assessment of the extent of variability in outcome across a network of reporting NICUs and the identification of NICUs in which the incidence of the problem is lower or higher than expected. NICUs with lower adverse outcomes can be assessed to identify factors and practices that have promoted the superior outcomes, with the goal being to transfer these benchmark approaches to units with high adverse outcomes. Finally, the database can be used longitudinally to track the extent to which quality improvement efforts have been effective. In addition to the rate of adverse events (e.g., death, pneumothorax, chronic lung disease), indicators of resource use such as number of days on the ventilator and length of stay are easily captured in a database and are considered important indicators of quality, especially by payers.

Risk Adjusters In the previous discussion we touched on expectation and the concept of an expected rate of negative outcomes. Expectation is pivotal to the notion of quality because quality can be defined as the ratio of observed to expected outcome.24 In the context of clinical medicine, expected outcome includes the extent of risk features and comorbidities that are not under the control of the clinician. Does hospital A with an observed neonatal mortality rate (NMR) of 5 provide better care than hospital B with an observed NMR of 10? Without knowing the expected rate of mortality at these two hospitals,

71

it is impossible to assess their relative quality of care. For example, hospital B might be a regional center whose expected NMR is much higher than the observed rate of 10. Hospital A could be a primary care hospital whose observed NMR of 5 is much higher than its expected NMR. A key challenge to quality assessment is how to control for differences in case mix to estimate their expected morbidity and mortality rates fairly and accurately. To estimate expected rate, we use risk adjusters, the fourth type of variable contained in a perinatal database. Risk adjusters have two essential requirements. They must predict adverse outcome and they must not be under the control of the entity being evaluated. Birthweight, gestational age, plurality, intrauterine growth, birth defects, and sex are risk factors that are commonly used as risk adjusters to control for institutional differences in case mix in order to make fair comparisons across NICUs.24,72 They are highly predictive of morbidity and mortality and are not under the control of the clinician. Mode of delivery and antenatal steroid use are also important predictors of mortality.101 Although not under the control of the neonatologist, they are under the control of the obstetrician. If neonatal mortality is used to compare the quality of care across NICUs, one can include these two factors in the risk adjustment. However, if the goal is to use neonatal mortality to compare the quality of care across perinatal services, these factors cannot be included because they are under the control of the obstetrician. The documented outcomes of an individual NICU must always be considered in the context of the severity and complexity of that NICU’s case mix, using an appropriate analytic approach to measure and adjust for differences in risk. A simple approach is to compare outcomes for relatively homogeneous categories of patients, such as individual diagnosisrelated groups (DRGs), discharge categories, or birthweight groupings. The outcomes and interventions for infants at a given NICU in a particular category are then compared with those for infants in that category at all other NICUs in the network. This approach has been used by both the Healthy People 2010 report and the American Congress of Obstetricians and Gynecologists’ (ACOG) guidelines for assessing overuse of cesarean delivery.1,92 The recommendation is to compare cesarean rates in a defined section of a hospital’s population: primiparas with full-term, singleton infants in the vertex position. Although this subset of women may seem homogeneous, across hospitals this subset may have marked differences in the distribution of significant risk factors. For example, a study by Gould and coworkers demonstrated that even when a stratum was restricted to women without a previous cesarean delivery, in active labor at term, with a singleton infant and no evidence of maternal complications or neonatal anomalies, significant differences in maternal age, parity, and ethnicity were identified across the study hospitals. After further adjustment for these differences, the results changed, demonstrating the importance of a multivariate approach even within a seemingly homogeneous stratum.26 More sophisticated multivariate methods can be used to adjust for differences in case mix, illness severity, and patient risk.31,71,72 This type of approach assesses the objective outcomes of fetal, neonatal, perinatal, postneonatal, and infant

72

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

mortality. The basic paradigm is that to make valid interpretations of the observed mortality rate, one must account for the components that affect its variability: risk, chance, and quality of care. The risk component reflects differences in observed outcome that are due solely to differences in case mix. Williams considered the primary risk factors to be birthweight, race, ethnicity, sex, and plurality, because these factors have been shown to be important predictors of neonatal mortality and are available from birth and death certificates.98a His strategy employs indirect standardization. Using all California births, the mortality rate for each combination of the four risk factors is calculated, producing a risk matrix of 190 cells. Each cell represents the average statewide mortality for an infant with that set of characteristics. In 2001, for example, neonates weighing 500 to 750 grams born to white, non-Hispanic single women had a state-wide NMR of 389.5 per 1000.64 For a given hospital, the overall observed mortality is compared with an expected mortality calculated by applying the overall California mortality rates to each of the hospital’s neonates and summing the results. Gould and coworkers revised this calculation using the Poisson statistic to account for the chance component in evaluating the significance between a hospital’s observed and expected mortality rates.64 A hospital with poor performance would have a higher observed mortality rate than would be expected on the basis of its birthweight, race, ethnicity, sex, and plurality case mix. Although the use of the risk matrix approach is technically sound, techniques based on regression analysis allow greater flexibility with respect to the number and types of variables and a more precise understanding of the characteristics of the adjustment model. Regression techniques make it possible to adjust for a wider variety of risk factors that are not under the control of the neonatologist. A mortality prediction model based on variables measured before NICU admission that are routine components of a minimal neonatal database was developed by the Vermont Oxford Network and has been used in routine annual reporting to members since 1991. The logistic regression model on which the predictions are now based includes terms for gestational age (birthweight had been used in some years), gestational age squared, race (African American, Hispanic, white, other), sex, location of birth (inborn or

outborn), multiple birth (yes or no), 1-minute Apgar score, small size for gestational age (lowest 10th percentile), major birth defect (yes or no, added in 1994), severity of birth defect (added in 2001), and mode of delivery (vaginal or cesarean). The model is recalibrated each year. A similar approach that takes into account gestational age (linear and quadratic terms), small for gestational age, birth defects, multiple gestation, 5-minute Apgar score, race/ethnicity, sex, transfer status, and prenatal care is used by the CPQCC to assess morbidity and mortality.11 Using this approach, the expected number of deaths (or adverse outcomes) at each NICU can be determined based on the characteristics of infants treated at that NICU. The ratio of the observed number of deaths (or adverse outcomes) to the expected number of deaths (or adverse outcomes), called the standardized mortality (or morbidity) ratio (SMR), can then be calculated. Rather than reporting the SMR as a measure of performance, an alternative approach is to calculate the difference between the observed and expected number of cases (O-E) with a particular finding, where the expected number can be determined using the same regression approach described earlier. The Vermont Oxford Network routinely reports the O-E for mortality and major morbidities to its members using a method that accounts for risk using regression models as described earlier and accounts for chance variation using an empirical bayesian shrinkage method.20,84 The shrunken O-E values are displayed on a funnel plot, with larger hospitals plotted to the left and smaller hospitals to the right. Figure 5-2 shows a plot for the O-E values for severe retinopathy of prematurity in 25 hospitals from the Australia and New Zealand Neonatal Network in 1998-1999 compared with those in 2000-2001.16 Figure 5-3 shows a similar O-E plot for nosocomial infection at Vermont Oxford Network hospitals in 2007. In this plot, for hospitals above or below the 95% control limits indicated by the parabolic curve, there is evidence that the number of observed cases of infection is either greater than or less than the number expected based on case mix. An advantage of O-E as a measure of performance is that it gives an easily interpretable estimate of the number of excess cases seen at hospital. It is important to recognize that all estimates of risk-adjusted performance must be interpreted carefully and that these

Excess number of severe ROP cases

30 25 20 15

Excess 1998–99 Excess 2000–01 95% limits 1998–99 95% limits 2000–01

10 5 0 –5 –10 –15 NICU ordered by decreasing number of eligible infants in 1998–99 cohort

Figure 5–2.  Observed minus expected number of cases (O-E) of retinopathy of prematurity (ROP) at 25 NICUs in the Australia and New Zealand Neonatal Network for 1998-1999 compared with cases for 2000-2001. Shown are 95% control limits. NICU, neonatal intensive care unit.  (Reprinted with permission of Br J Opthalmol.)

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine 80 60 40

O–E

20 0 –20 –40 –60 –80

Figure 5–3.  Observed minus expected number of cases

(O-E) of nosocomial infection at 682 neonatal intensive care units in the Vermont Oxford Network in 2007. Shown are 95% control limits. Hospitals with O-E above the limit, having more cases than expected, are shown in dark gray and those below the limit, having fewer cases than expected, are shown in light gray. (Courtesy of the Vermont Oxford Network, Burlington, Vt.)

estimates are only the first step in assessing the quality of care.82,88 Because the statistical power to detect quality of care outliers using multivariate risk-adjustment methods based on a single year may be low, it is very useful to also perform the analysis on data combined from several years.64 Even employing very accurate predictive models and combining several years of data might not be able to overcome the problem of a small sample size in some NICUs.33 In this case, multivariate risk models may be useful for identifying individual infants who died despite having a low predicted probability of death. The medical records of such infants can then be chosen for detailed review and audit. Recognizing that risk adjustments based on infant characteristics are imperfect, there is value in reporting comparative performance data stratified by the type of NICU. The Vermont Oxford Network provides such stratified comparisons to its members. The Committee on the Fetus and Newborn of the American Academy of Pediatrics has proposed a NICU classification system that may be useful for this purpose.12 In addition to reporting standardized ratios of observed-to-predicted rates for mortality, the Vermont Oxford Network and the CPQCC provide their members with similar data for a range of important morbidities, with the goal of using these data to identify opportunities for improvement. Although multivariate prediction models that are based on admission variables perform well for infants with very low birthweights for whom gestational age or birthweight is highly predictive of mortality, physiologic measures of disease severity may be necessary to achieve similar predictive performance for larger, more mature infants. In addition to stratification and multivariate modeling based on patient

73

characteristics that are present before therapy is initiated, it is also possible to perform case mix adjustment based on comparable severity of illness. Severity of illness scores for both adult and pediatric intensive care patients have been developed and validated.31,67 These scores, based on multivariate modeling techniques, can be used to adjust for case mix differences among intensive care units when comparing patient outcomes. Similar physiology-based severity scores have been developed for use in neonatal intensive care.71 The Clinical Risk Index for Babies (CRIB) was developed by the International Neonatal Network under the leadership of W. O. TarnowMordi to predict mortality risk for infants with birthweights of less than 1500 grams or gestational ages younger than 31 weeks.44,62 The score has been recalibrated with data from 1998 to 1999, using the variables sex, birthweight, gestational age, temperature on admission, and maximum base excess during the first hour. The potential for early treatment bias has been reduced by obtaining measurements in the first hour after admission. The CRIB score correlates with mortality risk or the risk for major cerebral abnormality on cranial ultrasound with a receiver operating characteristic (ROC) curve area under the curve of 0.82. A major strength of the CRIB II score (a 5-item version of the CRIB score) is its simplicity; a limitation is that it was designed specifically for infants younger than 32 gestational weeks.62 The Score for Neonatal Acute Physiology (SNAP), developed by Richardson and coworkers, is a physiology-based illness severity score originally based on measurements of 26 routine clinical tests and vital signs.70,74 Birthweight and SNAP are independent predictors of mortality. An additive score that is based on birthweight, 5-minute Apgar score, size for gestational age, and SNAP, called the SNAP-PE (SNAPPerinatal Extension), has been shown to be superior to either birthweight or SNAP alone. The more recent version of the score, SNAP II, uses only six laboratory and clinical parameters (lowest mean blood pressure, lowest temperature, lowest pH level, lowest Pao2-to-Fio2 ratio, urine output, and seizures) that are collected during the first 12 hours after admission. It has been shown to be compatible with SNAP I, is valid for infants of all birthweights, and takes only 5 minutes to collect.75 A study of more than 10,000 infants at 58 sites in the Vermont Oxford Network showed that the current performance of SNAP II and SNAP-PE II is similar to that observed in the original validation report, and addition of congenital anomalies as defined by the Vermont Oxford Network to SNAP-PE II significantly improves discrimination to a level consistent with the Vermont Oxford Network risk-adjustment algorithm described earlier.103 The CRIB score and the SNAP score are potentially useful for comparing mortality rates and other outcomes at different NICUs. The limited number of data elements required for both CRIB II and SNAP II makes them compatible with a minimal data set approach. One drawback of both scores is their use of variables that are measured during the first 1 to 12 hours after NICU admission. This raises two potential problems. The first problem relates to the 1- to 12-hour period of observation. Richardson and associates state that the longer the period of observation, “the more contaminated it becomes with the effects of successful (or unsuccessful) treatment and thus no

74

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

longer reflects admission severity.”72 Because their values may be influenced by treatments provided after admission, these illness severity scores are not truly independent of the effectiveness or quality of care. The second problem is that the observed severity of illness in the first hours of life may differ from the observed severity of illness in the very same infant in the first 6 hours following transfer and admission to another unit. Further studies are required to determine the extent to which these potential problems limit the usefulness of CRIB II and SNAP II for adjusting case mix. The Canadian Transport Risk Index of Physiologic Stability (TRIPS) is a scoring system that was developed to assess infant transport care. Based on the collection of only four variables (temperature, respiratory status, systolic blood pressure, and response to noxious stimuli), this approach had an area under the curve prediction of .83 for 7-day survival and .74 for severe intraventricular hemorrhage in a Canadian population.53 This scoring system has also been used to evaluate the effectiveness of different Canadian transport systems.54 An advantage of this score is that it assesses infant condition in a time frame that is not limited to the first 24 hours of life with very high prediction characteristics. However, this score has not been validated outside of Canada. Possible limitations of this approach in other settings include respiratory severity being scored maximum with intubation and there being no consideration of pressor use for blood pressure support. Tyson and coworkers have recently used the National Institute of Child Health and Human Development (NICHD) Neonatal Network database to develop a multivariable model for prediction of survival and neurodevelopmental outcome for preterm infants of 22 to 25 weeks’ gestation based on birthweight, gestational age, gender, multiplicity, and antenatal steroid status.91 An online calculator is available to determine the model predictions for specific value of the five variables in the model (http://www.nichd.nih.gov/about/org/ cdbpm/pp/prog_epbo/epbo_case.cfm). Both survival and survival free of handicap are estimates. Further research is required to identify the best models for predicting neonatal risk and to determine their precision in identifying individual cases or institutions with poor quality of care.61 However, even without a firm foundation in research, risk-adjusted comparisons of NICUs will become more common. The public release of risk-adjusted comparisons of mortality rates for U.S. hospitals by the Centers for Medicare and Medicaid Services and the publication of hospital-specific and surgeon-specific mortality rates for cardiovascular surgery by the New York State Department of Health are two examples of this phenomenon. An example in the field of perinatology is the Pacific Business Group on Health’s publication of risk-adjusted primary cesarean section rates for California hospitals (www.healthscope.org). This analysis is unique in that technical oversight was provided by neonatologists, perinatologists, and researchers who are members of the CPQCC. The public release of comparative performance data and the use of such data for contracting and performance-based reimbursement will become increasingly common over the next few years. The Leapfrog Group, a consortium of Fortune 500 companies, now provides hospital-specific performance data to the public on the Internet. NICU ratings are currently

based on caring for more than 50 very low birthweight infants per year and antenatal steroid use.51 Neonatologists must understand the strengths and weaknesses of different methods for making risk-adjusted comparisons of neonatal outcomes as they attempt to assist the public in understanding these data and to use the data themselves to monitor, evaluate, and improve the quality of care that they provide. In recent years, interest in quality assessment has rapidly increased, and there has been a proliferation of perinatal quality measures. An important development in ensuring that proposed measures can actually reflect the quality of care has been the work of the National Quality Forum (NQF) (http://qualityforum.org). The NQF serves as a clearing house for the evaluation of proposed quality measures. In considering measures, several important areas are assessed, including the importance of the measure to making significant gains in health care quality, the measure’s reliability and validity, the extent to which the results can be understood and be useful for decision making, and the feasibility of collection. Table 5-1 shows the perinatal measures recently endorsed by the NQF (http://qualityforum.org/pdf/ projects/perinatal/tbAppA-Specs-%2010-20-08.pdf). When considering a potential measure for widespread use, it is important to have evidence that the aspect of care that the measure reflects has been shown to be malleable, that is, shown to be improvable, in at least one multi-institutional quality improvement initiative.

Secondary Data Although secondary data have been briefly described, thus far the discussion has focused on primary data, that is, data specifically collected to address perinatal issues. Secondary data are data collected for other purposes that can also be used to assess and improve perinatal care. Secondary data sources that are frequently used are the birth certificate, the hospital discharge abstract, and hospital billing data. The major advantage of using secondary data is that someone else maintains the data system. The major disadvantages are that the secondary data sources might not have all of the necessary data items, the definitions might not be appropriate or the same as those used in other NICUs, and the accuracy might not be adequate. For example, demographic information, prenatal care, mode of delivery, and birthweight tend to be fairly reliable on birth certificates.23 However, the presence of congenital anomalies, an important item because of the high degree of complexity and mortality in this group, is markedly under-reported on both birth and death certificates.85 A recent review describes both the advantages and disadvantages of using vital records for quality improvement.23 Linked birth and death certificate files are an important source of population-based studies of factors that affect perinatal outcomes. These studies span a wide range of areas (e.g., the effect of the increase in multiple births on infant mortality,7 factors associated with the birth of infants with very low birthweight at non-NICU hospitals,25 and quality assessment of perinatal regionalization17). However, to maximize the potential usefulness of vital records, clinicians must become actively involved in ensuring their uniformity of definition, accuracy, and completeness. (Text continues on page 80)

75

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

TABLE 5–1A  N  ational Voluntary Consensus Standards for Perinatal Care: Selected Abridged Neonatal Measures* Measure

Numerator

Denominator

Data Source

Appropriate use of antenatal steroids

Number of mothers from the denominator receiving antenatal steroids (corticosteroids administered intramuscularly) during pregnancy at any time before delivery

Total number of mothers who delivered preterm infants (24-32 weeks with preterm premature rupture of membranes or 24-34 weeks with intact membranes)

Medical record, clinical database, electronic health record

Infants under 1500 g delivered at appropriate site

Liveborn infants from the denominator with birthweight ,1500 g at the given birth hospital

All live births older than 24 weeks of gestation at the given birth hospital. Is this hospital a level III or equivalent neonatal intensive care unit as defined by AAP? Yes _____ No _____

Birth records

Nosocomial bloodstream infections in neonates

Any diagnosis code for specific bacterial or fungal infection, OR patients with one of the following diagnosis codes: • Septicemia (sepsis) of newborn (771.81) OR • Bacteremia of newborn (771.83)

All inborn and outborn infants (admitted at 0-28 days) with a birthweight between 500 and 1499 g OR a gestational age between 24 and 30 weeks AND all inborn and outborn infants with a birthweight $1500 g if the infant experienced death, major surgery, mechanical ventilation, or transfer in or out from/to an acute care facility. Inborn refers to neonates born within that institution; outborn refers to neonates born elsewhere but transferred within the first 2 days of life.

Claims/discharge abstract data

Exclusive breast feeding at hospital discharge

Proportion of the denominator that were fed by “breast only” since birth

Live births not discharged from the neonatal intensive care unit who had newborn genetic screening performed

Newborn screening data

*http://qualityforum.org/pdf/projects/perinatal/tbAppASpecs-%2010-20-08.pdf

TABLE 5–1B  National Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications* Measure #, Title, IP Owner† PN-007-07 Elective delivery before 39 completed weeks’ gestation IP Owner Hospital Corporation of American (HCA)/St. Marks Perinatal Center ‡,§

Numerator

Denominator

Exclusions

Data Source

Babies from the denominator electively delivered before 39 completed weeks’ gestation

All singletons delivered at $37 completed weeks’ gestation

Post-dates (ICD-9 code 645), IUGR (656.5), oligohydramnios (658.0), hypertension (642), diabetes (648.0), maternal cardiac disease (648.8), previous stillbirth (648.5), placental abruption (648.6), placenta previa (641), unspecified antenatal hemorrhage (646.2), maternal renal disease (646.7), acute fatty liver of pregnancy (651), multiple gestation (652), malpresentation (656.1), isoimmunization (656.2), maternal coagulopathy (656.4), fetal demise (657), hydramnios (658.1), and ruptured membranes (649.3), V27.1

Medical records

Continued

76

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 5–1B  N  ational Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications—cont’d Measure #, Title, IP Owner†

Numerator

Denominator

Exclusions

Data Source

PN-013-07 Incidence of episiotomy IP Owner Christiana Care Health Service/NPIC

Number of patients from the denominator with episiotomy procedures (CPT code: 59410 or ICD-9 codes 72.1, 72.21, 72.31, 72.71, 73.6, 75.6) performed

Number of vaginal deliveries (CPT 59410 or by DRG)

Vaginal deliveries complicated by shoulder dystocia (ICD-9 660.41 or 660.42)

Claims, medical records, electronic health records

PN-010-07 Cesarean rate for low-risk first birth women IP Owner California Maternal Quality Care Collaborative

Proportion of patients from the denominator that had a cesarean birth

Livebirths at or beyond 37.0 weeks’ gestation that are having their first delivery and are singleton (no twins or beyond) and vertex presentation (no breech or transverse positions)

Patients with abnormal presentation, preterm, fetal death, multiple gestation diagnosis codes, or breech procedure codes

Claims data and vital records (birth certificate)

PN-011-07 Prophylactic antibiotic in cesarean section IP Owner Massachusetts General Hospital

Number of patients who received prophylactic antibiotics within 1 hour before surgical incision or at the time of delivery

All patients undergoing cesarean section without evidence of prior infection or already receiving prophylactic antibiotics for other reasons

Patients who had a principal ICD-9 diagnosis code suggestive of preoperative infectious disease (as defined in Appendix A, Table 5.09 of the Specification Manual for National Hospital Quality Measures, Version 2.2, and future updates). Patients who were receiving antibiotics within 24 hours before surgery, except that prophylaxis with penicillin or ampicillin for group B Streptococcus (GBS) is not a reason for exclusion. Patients with physician/advanced practice nurse/physician assistant/certified nurse midwife– documented infection or prophylaxis for infection, except that prophylaxis for GBS is not a reason for exclusion. Patients who undergo other surgeries within 3 days before or after the cesarean section

Administrative, medical records, clinician survey, paper medical record, and electronic health record

PN-006-07‡ Appropriate DVT prophylaxis in women undergoing cesarean delivery IP Owner HCA /St. Marks Perinatal Center

Patients from the denominator who receive either fractionated or unfractionated heparin or pneumatic compression devices before surgery

All women undergoing cesarean delivery

None

Medical records

‡

77

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

TABLE 5–1B  N  ational Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications—cont’d Measure #, Title, IP Owner†

Numerator

Denominator

Exclusions

Data Source

PN-002 & 019-07‡ Birth trauma rate measures (harmonized) IP Owners Agency for Healthcare Research and Quality (AHRQ)/ National Perinatal Information Center (NPIC)

Discharges from the denominator with one of the following birth outcomes: 1. ICD-9-CM code 7670: Subdural and cerebral hemorrhage due to trauma, intrapartum anoxia, or hypoxia 2. 76711: Epicranial subaponeurotic hemorrhage (massive) 3. 7673: Injuries to skeleton (excludes clavicle) 4. 7674: Injury to spine and spinal cord 5. 7675: Facial nerve injury 6. 7677: Other cranial and peripheral nerve injuries 7. 7678: Other specified birth trauma 8. 767.8: Other specified birth trauma, eye damage, hematoma of liver, testes, vulva, rupture of liver, spleen, scalpel wound, traumatic glaucoma

All neonates within a hospital. A neonate is any newborn aged 0 to 28 days (inclusive) at discharge with: 1. An ICD-9-CM code for in-hospital liveborn birth OR 2. An admission type of newborn, age in days at admission equal to 0, and no code for an out-of hospital birth OR 3. Any DRG in MDC 15 (if age in days is missing)

Infants with a birthweight of less than 2000 g (ICD-9-CM codes 765.00-07, 765.11-17). Infants with any diagnosis code of osteogenesis imperfecta (756.51). Infants with injury to the brachial plexus, palsy or paralysis, Erb’s palsy (767.6)

Claims/ discharge abstract data

PN-001-07‡ Hepatitis B vaccine administration to all newborns before discharge IP Owner Centers for Disease Control and Prevention

Number of newborns from the denominator administered hepatitis B vaccine (CPT for hepatitis B vaccine 90744, CPT for immunization administration 90471, diagnosis code V05.3 for hepatitis B vaccination) before discharge

Number of live newborns discharged from the hospital

Parental refusal

Claims, medical records, clinical database, pharmacy data, and electronic health record data

PN-016-07‡ Appropriate use of antenatal steroids IP Owner Providence St. Vincent Hospital/ Council of Women and Infants Specialty Hospitals (CWISH)

Number of mothers from the denominator receiving antenatal steroids (corticosteroids administered IM) during pregnancy at any time before delivery

Total number of mothers who delivered preterm infants (24-32 weeks with preterm premature rupture of membranes or 24-34 weeks with intact membranes)

None

Medical record, clinical database, electronic health record

Continued

TABLE 5–1B  N  ational Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications—cont’d Measure #, Title, IP Owner†

Numerator

Denominator

Exclusions

Data Source

PN-022-07 Infants ,1500 g delivered at appropriate site IP Owner California Maternal Quality Care Collaborative

Liveborn infants from the denominator with birthweight ,1500 g at the given birth hospital

All live births .24 weeks’ gestation at the given birth hospital Is this hospital a level III¶ or equivalent neonatal intensive care unit as defined by AAP¶? Yes _____ No____

None

Birth records

PN-003-07‡ Nosocomial blood stream infections in neonates IP Owner AHRQ

Any diagnosis code for the following: • Staphylococcal septicemia, unspecified [038.10] • Staphylococcus aureus septicemia [038.11] • Other staphylococcal septicemia [038.19] • Gram-negative organism NOS [038.40] • Septicemia due to other gram-negative organisms, Escherichia coli [038.42] • Septicemia due to other gram-negative organisms, Pseudomonas [038.43] • Septicemia due to other gram-negative organisms, Serratia [038.44] • Septicemia due to other gram-negative organisms, Other [038.49] • Disseminated candidiasis/ systemic candidiasis [112.5] OR Patients with one of the following diagnosis codes: • Septicemia [sepsis] of newborn [771.81] OR • Bacteremia of newborn [771.83] OR • Bacteremia [790.7] AND one of the following diagnosis codes: • Streptococcus Group D (Enterococcus) [041.04] • Staphylococcus, unspecified [041.10] • Staphylococcus aureus [041.11] • Other Staphylococcus [041.19] • Friedländer’s bacillus • (Klebsiella pneumoniae) [041.3] • Escherichia coli [041.4] • Pseudomonas [041.7]

All inborn and outborn infants (admitted at 0-28 days) with a birthweight between 500 and 1499 g OR a gestational age between 24 and 30 weeks AND all inborn and outborn infants with a birthweight greater than or equal to 1500 g, if the infant experienced death, major surgery, mechanical ventilation, or transfer into or out from an acute care facility. Inborn refers to neonates born within that institution; outborn refers to neonates born elsewhere but transferred within the first 2 days of life

Patients with a principal diagnosis of sepsis or bacterial infection Patients with a length of stay of less than 2 days

Claims/ discharge abstract data

79

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

TABLE 5–1B  N  ational Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications—cont’d Measure #, Title, IP Owner†

Numerator

Denominator

Exclusions

Data Source

PN-025-07 Birth dose of hepatitis B vaccine (HBV) and hepatitis immune globulin (HBIG) for newborns of mothers with chronic hepatitis B IP Owner Asian Liver Center at Stanford University

Number of newborns from the denominator who receive birth doses of HBV vaccine and HBIG within 12 hours of delivery

Number of newborns delivered from mothers who tested positive for the hepatitis B surface antigen (HBsAg) during pregnancy

Stillbirths

Medical records, clinical database, laboratory data, and electronic health record data

PN-021-07 Exclusive breast feeding at hospital discharge IP Owner California Maternal Quality Care Collaborative

Proportion of the denominator that were fed by “breast only” since birth

Live births not discharged from the NICU who had newborn genetic screening performed

Infants in the NICU at time of newborn screen and infants who received TPN or other nutrition supplements

Newborn screening data

Paired Measures: PN-029-07a First temperature within 1 hour of admission to NICU AND PN-029-07b First NICU temperature ,36°C IP Owner Vermont Oxford Network

Patients from the denominator with a first temperature taken within 1 hour of NICU admission Patients from the denominator whose first temperature was ,36°C

All NICU admissions with a birthweight of 501-1500 g All NICU admissions with a birthweight of 501-1500 g whose first temperature was measured within 1 hour of admission to the NICU

Outborn infants admitted more than 28 days after birth Outborn infants that had been home before admission Infants without temperature taken within 1 hour of NICU admission. (PN-029-07b only)

Medical records, registries, the Vermont Oxford Network Database (when applicable), and the eNICQ data collection instrument

PN-030-07 Retinopathy of prematurity (ROP) screening IP Owner Vermont Oxford Network

Number of infants from the denominator receiving a retinal examination for ROP

Number of infants aged 22-29 weeks’ gestation hospitalized at the postnatal age at which a retinal examination is recommended by the AAP

Outborn infants admitted more than 28 days after birth. Outborn infants that had been home before admission

Medical records, registries, the Vermont Oxford Network database (when applicable), and the eNICQ data collection instrument

‡

Continued

80

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 5–1B  N  ational Voluntary Consensus Standards for Perinatal Care: Performance Measure Specifications—cont’d Measure #, Title, IP Owner†

Numerator

Denominator

Exclusions

Data Source

PN-031-07 Timely surfactant administration to premature neonates IP Owner Vermont Oxford Network

Patients from the denominator treated with surfactant within 2 hours of birth

Number of infants born at 22-29 weeks’ gestation treated with surfactant at any time

Outborn infants admitted .28 days after birth Outborn infants that had been home before admission

Medical records, registries, and the Vermont Oxford Network database (when applicable)

PN-032-07** Neonatal immunization IP Owner Child Health Corporation of America

Patients from the denominator receiving the following immunizations according to current AAP guidelines: • DtaP • HepB • IPV • Hib • PCV

Neonates with a length of stay .60 days

Documented parent refusal and mortalities. The developer recommends that the measure be suspended when there are vaccine shortages rather than including vaccine unavailability as an exclusion

Retrospective review of both administrative and medical records data. Manual collection is required for parent refusal and crossreference to administrative data

*All specifications as of October 20, 2008. Refer to IP owner for most recent specifications. † IP, Intellectual property. ‡ Time-limited endorsement. § Candidate standard numbers assigned by NQF during the consensus process. ¶ Level III subspecialty NICUs have the personnel and equipment to care for infants ,1500 g. Hospitals that do not have level III NICUs should have low rates for this measure, indicating appropriate transfer of a mother at risk of preterm delivery to a facility capable of providing level III care for a very low birthweight infant. ¶ American Academy of Pediatrics Guidelines for Levels of Care: http://aappolicy.aappublications.org /cgi/reprint/pediatrics;114/5/1341.pdf **Previously endorsed measure; evaluated as part of NQF’s ongoing measure maintenance activities. AAP, American Academy of Pediatrics; DtaP, diphtheria, tetanus, and pertussis vaccine; DVT, deep venous thrombosis; HepB, hepatitis B vaccine; Hib, Haemophilus influenzae type b (Hib) conjugate vaccine; IM, intramuscularly; IPV, inactivated polio vaccine; NICU, neonatal intensive care unit; NOS, not otherwise specified; NQF, National Quality Forum; PCV, pneumococcal conjugate vaccine; TPN, total perenteral nutrition.

The hospital discharge abstract can be a useful data source for evaluating neonatal care.83 The U.S. Department of Health and Human Services mandates that a uniform hospital discharge data set, which includes 14 core data items, be submitted for each acute patient whose care is paid for by Medicare or Medicaid. The most widely used format for these submissions is the Uniform Bill, introduced in 1992 (UB-92).75,83 UB-92 contains data items for patient identification, insurance coverage, total charges, and entries for up to five diagnostic and three procedural codes. These codes are assigned based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM).59 UB-92 is required for hospitals submitting claims to Medicare, Medicaid, Blue Cross, and other commercial insurers. Although hospitals are not required to submit UB-92 for all neonates, most hospitals do complete this form. The Uniform Bill can provide useful information on procedures, diagnoses, and charges. The major advantage of this data source is its widespread use at a

large number of institutions and its ready availability as a computer dataset at many hospitals. Several weaknesses of this source of data must be considered, however. The Uniform Bill was designed for reimbursement, not for monitoring institutional performance or for clinical research. As a result, distortions in the data may result from attempts by hospitals to code diagnostic and procedural data with the goal of maximizing reimbursement. Significant errors in diagnostic coding may occur. Studies of the reliability of hospital discharge abstracts have found that the principal diagnosis identified in the discharge abstract agrees with the actual diagnosis based on chart review only 65% of the time.43 Although some perinatal conditions such as third- and fourth-degree lacerations were coded very completely,76 the validity of a variety of perinatal conditions and procedures was quite variable, ranging from quite good to very poor.102 The accuracy of coding for neonatal conditions has not been studied in detail. However, neonatal discharge

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

81

TABLE 5–2  Alternative Diagnosis-Related Group Classification Systems DRG System

Developer

Users

CMS DRG (formerly known as HCFA DRG)

Medicare

Medicare and some Medicaid programs

PM-DRG

NACHRI

No longer maintained

AP-DRG

New York State

New York and some other states

CHAMPUS DRG

Department of Defense

CHAMPUS

R-DRG

HCFA

APR-DRG

3M Health Information Systems and NACHRI

State health departments and health data commissions

AP, all patient; APR, all patient refined; CHAMPUS, Civilian Health and Medical Program of the Uniformed Services; CMS, Centers for Medicare and Medicaid Services; DRG, diagnosis-related group; HCFA, Health Care Financing Administration; NACHRI, National Association of Children’s Hospitals and Related Institutions; PM, pediatric modified; R, refined. From John Muldoon, National Association of Children’s Hospitals and Related Institutions, March 2004.

abstracts tend to contain less reliable data on race than do birth certificates. Another problem with hospital discharge abstract data is the absence of birthweight as a data item. It is not included on UB-92. Because of the powerful predictive relationship of birthweight to both resource use and neonatal outcomes, the value of discharge abstract data for the care of newborns would greatly increase. It is also possible to use linkage strategies to enhance the usefulness of secondary data sources for quality improvement. Highly successful links can be established without using personal data such as names, hospital numbers, or Social Security numbers; instead, links can employ virtual identifiers such as postal code of residence, clinical factors, and demographic factors. California’s Office of State Health Planning and Development sponsored a project to link the state-linked infant birth and death file with a modification of the UB-92 file.32 This database allows one to select outcomes from the ICD-9-CM and procedure codes available on the discharge billing file and adjust these outcomes based on the birthweight and clinical, demographic, and socioeconomic information from the birth certificate. Further links to this database have included the mother’s discharge file for the current pregnancy as well as all infant readmission discharge files during the first year of life. Examples of population-based studies using this linked database include the relationship between discharge timing after birth and infant readmission,15 shoulder dystocia, risk factors and neonatal outcomes,21 and neonatal outcomes in childbearing beyond the age of 40.22 In a project sponsored by the Pacific Business Group on Health with the technical oversight of the CPQCC, this database was used to perform a risk-adjusted analysis of primary cesarean section rates in California hospitals (www.healthscope.org). The technical details of the analysis are available at the CPQCC website (www.CPQCC.org). DRGs are another example of secondary data used to evaluate perinatal care.58 Defining case mix to be able to compare outcomes and resource use across institutions is of great importance to payers. DRG systems are classification schemes that use data that are routinely available in hospital discharge abstracts to group patients into relatively homogeneous categories. Outcomes and resource use are compared for similar DRGs across institutions. Because of the widespread use of DRGs, it is important for neonatologists to be familiar with these systems.

Ideally, each DRG category should contain patients who are clinically similar, whose care requires the same resource intensity, and who are at similar risk for adverse events, mortality, and morbidity. These systems, which were originally developed to guide prospective payment to hospitals, are used increasingly to classify patients for risk stratification in analyses of outcomes and costs. Several alternative DRG classification systems have been developed (see Table 5-2). They are updated periodically and exist in different versions. Individual DRGs are grouped into major diagnostic categories that contain related DRGs.

COSTS AND RESOURCES It is important for neonatologists to understand that they will be under increasing pressure to justify and reduce the costs of neonatal care. Neonatal intensive care is one of the most expensive types of hospital care,73,75,80 and it has come under increasing scrutiny by both public and private insurers seeking to contain health care costs. Insurers seek to compare treatment costs across institutions to determine whether costs at a given institution are excessively high. Meaningful comparisons of neonatal intensive care treatment costs across institutions are difficult to make. Most insurers have access to data from only a small set of hospitals, whose case mix may vary considerably. If case mix adjustments are made at all, they are typically based on the Medicare DRGs. However, the Medicare DRGs contain only seven categories for newborn infants and explain only 22% of variation in costs.65 Thus, the Medicare DRGs do not provide a good method for adjusting for case mix. This fact underscores the need for information systems that use alternative DRG systems to collect more detailed information on clinical aspects of care for infants treated in the NICU. Comparisons of treatment costs across institutions are not straightforward, even in the absence of case mix differences. Because of the wide variation in pricing policies across hospitals, comparisons of charged amounts may be misleading. To compare treatment costs across institutions, costs must be computed. Data from hospital billing systems or UB-92s are typically used to generate measures of treatment costs. These data contain information on charges for hospital services, which are converted to costs

82

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

using information on the internal pricing structure of hospital services. Standard methods exist for such conversions.75 Calculating conversions based on detailed hospital bills can be a daunting task because of the volume of services (tens of thousands) that must be converted. Conversions based on the Uniform Bill forms are more easily performed because of the aggregation of charges on them. Cost conversion methodologies must allocate both direct and indirect costs to each hospital service. Direct costs, such as the cost of a prescription medication, are relatively easy to identify. However, decisions about how to allocate indirect costs, which include facility costs and services such as administrative salaries, security services, laundry, housekeeping, and individual services, are more difficult to make. There are many ways to make these assignments. How indirect costs are allocated has a major effect on the ultimate calculated costs. If costs at different institutions are to be compared, it is crucial that these assignments be made in a uniform manner.75 Computerized hospital discharge abstracts can be used to create hospital-specific reports that address perinatal care and its costs.83 The nonprofit National Perinatal Information Center (Providence, RI) uses data from UB-92s supplemented with birthweight submitted by more than 50 participating hospitals with approximately 180,000 yearly births to create detailed reports that document and compare hospital performance and costs.83 Average length of stay is often used as a proxy for cost. In making average-length-of-stay comparisons, it is essential to adjust for differences in case mix. The Vermont Oxford Network has developed multivariate risk models for predicting length of stay that are used in routine reporting to members. As with mortality, there is substantial variation among the network hospitals even after adjusting for case mix, with adjusted total length of stay for surviving infants 501 to 1500 grams ranging from less than 40 days at some institutions to more than 75 days at others.35 Documenting comparative outcomes in the context of comparative costs is extremely important to managed care organizations with respect to selecting cost-effective hospitals and negotiating reimbursement rates. However, valid comparisons across institutions are difficult to make. The financial modules of generic hospital information systems are intended for application in all clinical areas of a hospital; therefore, these systems have broad appeal to hospital managers and yet may not have the granularity, clinical precision, or reporting flexibility to meet the needs of the perinatal unit. Neonatologists must become familiar with the management tools in use at their hospitals. They must understand their analytic shortcomings and their interpretation, because decisions regarding resource allocations within hospitals increasingly will be based on these tools.

Role of Networks Quality improvement activities for any given hospital are greatly facilitated by participating in a network. The networks facilitate valid comparisons by standardizing data definition and collection standards. Their reports allow

confidential risk-adjusted comparisons to peer institutions, and the networks provide important resources for organized data-driven improvement activities. Several NICU networks have been formed to evaluate the effectiveness and efficiency of neonatal intensive care and facilitate the use of data for quality improvement (see Box 5-1). Examples include the Australian and New Zealand Neonatal Network, the Canadian NICU Network (www. caneonet.org/nicu.html), the Indian Neonatal Forum Network, the International Neonatal Network,45 the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network,29,55 Pediatrix,8 and the Vermont Oxford Network.34,35 The NICHD Neonatal Research Network is a group of 16 academic NICUs whose government-funded activities include randomized trials and observational studies. Participants in the network are chosen based on competitive application to the NICHD. The NICHD Neonatal Research Network maintains a database for infants with birthweights of 401 to 1500 grams who were born at participating centers or admitted to them within 14 days of birth (starting in 2008, the database includes only inborn infants with birthweights of 1000 grams or less or gestational age of less than 29 weeks) who were treated at participating NICUs. Uniform definitions for data items and attention to maintenance of data quality make the NICHD Neonatal Research Network database a valuable resource for neonatologists. The published reports from the database can be used by other NICUs for comparison. The validity of making comparisons with these data depends on the definitions of data items used by an individual NICU and the similarity of their patient populations to those treated at NICUs in the NICHD Network. The Vermont Oxford Network is a collaborative network of neonatologists and other health care professionals, representing more than 800 institutions from North America and around the world (www.vtoxford.org). Membership is voluntary and open to all who are interested. The nonprofit network is supported by membership fees, research grants, and contracts. The primary philosophy of the Vermont Oxford Network is to improve the quality, safety, and efficiency of medical care for newborn infants and their families through a coordinated program of research, education, and quality improvement projects. In support of all three aspects of this program, the network maintains a database of infants with very low birthweight (401 to 1500 grams or 22 to 29 weeks’ gestation) who were born at participating centers or admitted to them within 28 days of birth. Members of the Vermont Oxford Network submit data using standardized definitions.95 Approximately 85% of members submit data electronically using either the network’s eNICQ software or other systems that employ standardized file formats (see instructions at www.vtoxford.org). Strict attention is paid to maintenance of data quality.37 The database provides core data for network clinical trials, is used for observational studies and outcomes research, and generates reports for members that compare their performance with that of other NICUs in the network. These reports, available in print, on CD, and in real time on the Nightingale Internet Reporting System, are intended for use in local quality management efforts. In 2007, the Vermont Oxford Network database enrolled more than 49,000 infants weighing 401 to 1500 grams from

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

682 NICUs; in the 20-year period from 1990 to 2009 approximately 500,000 infants were enrolled in the very low birthweight database. A major advantage of participating in a network database is that comparisons among NICUs based on uniform definitions are then possible. A limitation of the databases maintained by the NICHD and the Vermont Oxford Network is that historically they have been limited to infants with birthweights of 1500 grams or less. The Vermont Oxford Network has now expanded data collection and reporting to include infants weighing more than 1500 grams and to include a registry devoted to infants with neonatal encephalopathy. Another example is the California Perinatal Quality Care Collaborative, a statewide outgrowth of an initiative proposed by the California Association of Neonatologists and supported by a start-up grant from the David and Lucile Packard Foundation; the California Department of Health Services, Maternal and Child Health Branch; and California Children’s Services. The collaborative exists to improve the health of pregnant women, infants, and children by collecting high-quality information on perinatal outcomes and resource use and using these data for performance improvement and benchmarking processes in perinatal care and NICUs throughout California. The CPQCC forms a public and private alliance of stakeholders, including the Maternal and Child Health Branch, California Children’s Services, and the Office of Vital Records (all within the California Department of Health Services); the Office of Statewide Health Planning and Development; the Hospital Council; Regional Perinatal Programs of California; the American College of Obstetricians and Gynecologists; the California Association of Neonatologists; the Pacific Business Group on Health; the David and Lucile Packard Foundation; and the Vermont Oxford Network. The collaborative has three goals: (1) to provide a timely analysis of perinatal care, outcomes, and resource use based on a uniform statewide database; (2) to provide mechanisms for benchmarking and continuous quality improvement (CQI); and (3) to serve as a model for other states. The Vermont Oxford Network has provided major input to the development of the CPQCC database and CQI activities. The Vermont Oxford Network database for infants weighing less than 1500 grams has provided the foundation for the CPQCC database. In 2000, this was expanded to include a subset of the sickest infants weighing more than 1500 grams. In 2007, acute infant transport and in 2009 high-risk infant follow-up data to age three were added. In the 2007 CPQCC, members cared for 11,270 high-morbidity infants weighing more than 1500 grams and 6836 very low birthweight infants. Indications for transport, time intervals, initial clinical condition, and change in condition awaiting and during the course of transport were collected on the 7430 infants who were acutely transported in the neonatal period. The data are processed by the CPQCC data center using a confidential web-based data entry and real-time report system for inborn infants, outborn infants, and infants readmitted to an NICU during the first 28 days of life (www. cpqcc.org). Level of care specific metrics based on all California as well as each of California’s perinatal regional region are available. Beginning with 2005, the database was expanded to include individual state birth certificate and death certificate

83

data and hospital discharge summary data. The goal of this linkage is to allow both hospital level and community level analysis of perinatal risks and outcomes throughout California. Currently there are 127 member hospitals with NICUs admitting nearly 90% of California’s newborn infants who require critical care. Timely high-quality data fuel the project, whose major purpose is to improve the quality of perinatal care.

Work of Quality Improvement In the previous sections strategies have been presented for collecting and evaluating data using primary and secondary databases, the importance of risk adjustment in being able to compare outcomes and processes across institutions, and the advantages of being part of a network. In this section we discuss strategies for translating data into action, the work of quality improvement. To address quality of care requires the ability to specify outcomes. Outcomes may be objective or subjective. Typically, objective outcomes consist of mortality, morbidity, and long-term neurodevelopmental status. The most important morbidities to record are those that could be influenced by the quality of care, for example, nosocomial infection.36 In some cases, we presume that certain morbidities can be minimized by optimal care, although the specifics of what constitutes optimal care have not been defined. Conversely, it is clear that suboptimal practices can increase morbidity. For example, excessive ventilation can lead to pneumothorax or bronchopulmonary dysplasia, fluid overload can increase the risk for patent ductus arteriosus, and inadequate maintenance of thermal environment and suboptimal nutritional practices can result in prolonged hospitalization because of poor weight gain.8 In addition to the traditional clinical outcomes, subjective outcomes, such as parent satisfaction and quality of life, will become increasingly important measures of the quality of neonatal care.14,78,79 Specifying and measuring the practices and outcomes of care are only the starting point for improvement. The information must be analyzed, synthesized, and presented so that opportunities for improvement can be identified and the results of improvement efforts can be monitored and evaluated. However, information and performance feedback alone cannot cause the profound changes in care processes and in the behavior of caregivers that are necessary to improve the quality of medical care. These will occur only when all members of the care team have the knowledge, skills, motivation, and organizational support required to make continuous quality improvement an integral and ongoing component of their work. Multidisciplinary collaborative quality improvement has been applied successfully in a number of health care settings.66 However, the evidence underlying quality improvement collaboratives is positive but limited, the effects cannot be predicted with great certainty, and the factors associated with success are not well understood.81 Furthermore, not all quality improvement efforts in neonatology have been successful. In a cluster randomized trial of a quality improvement intervention designed to reduce chronic lung disease, Walsh and colleagues at the NICHD Neonatal Research Network did not find a benefit to the intervention.97 Future research into

84

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

quality improvement must explore the contextual factors that might explain why two hospitals in the same collaborative or two collaboratives using similar methods might experience different results.2,56,63 The management of quality in the field of health care has borrowed heavily from the techniques of quality management science in use in general industry. Berwick has pioneered the application of these techniques to medical care and applied them to a number of clinical problems in the Breakthrough Series of the Institute for Health Care Improvement.5,6,42 O’Connor and colleagues have shown that multidisciplinary collaborative improvement based on feedback of performance data, quality improvement training, and collaborative learning through site visiting can reduce mortality for cardiovascular surgery.60 Many health care organizations are now using these general methods to improve the quality of medical practice. The CPQCC and the Vermont Oxford Network provide two examples of how quality improvement is being applied to neonatal care. The CPQCC has established a permanent subcommittee, the Perinatal Quality Improvement Panel, made up of neonatologists, perinatologists, perinatal nurses, state and hospital representatives, and health outcomes researchers with experience in quality improvement and outcomes measurement. The panel reviews the statewide data and recommends quality improvement objectives, provides models for performance improvement, and assists providers in transforming data into information that can help to improve care.99 The panel’s first CQI on antenatal steroids enrolled 25 NICUs.100 The program was structured to improve the performance of all hospitals, and it endeavored to assist all participating hospitals in meeting established goals at the conclusion of the CQI cycle. The results achieved by the participating NICUs demonstrated substantial improvements and were publicly released on the Pacific Business Group on Health website (www.healthscope.org). The second topic selected for a network-wide CQI was nosocomial infection. Activity has greatly expanded, and in partnership with California Children’s Services, the 2008 catheter-associated blood stream infections (CABSI) initiatives, which were based on the Institute for Health Care Improvement model, have had the active participation of all 22 regional NICUs as well as 19 community NICUs. The goal is to move to population base quality improvement involving all 127 member NICUs. In addition to these formal public release cycles, CPQCC has conducted educational programs and workshops and has developed tool kits to facilitate CQI for several important areas. Ten quality improvement tool kits are available for download on the CPQCC website, www.cpqcc.org. These include Antenatal Corticosteroid Administration (underuse), Reducing Postnatal Steroid Administration (overuse), Nosocomial Infection Prevention (misuse), Nutritional Support of the Very Low Birthweight (VLBW) Infant, Perinatal HIV, and Care of the Late Preterm. Each tool kit has an established cycle for review and update. With multiple downloads in 2008, CPQCC is proving to be a popular source of quality improvement information for providers in California, the United States, and abroad. The Vermont Oxford Network has been working since the early 1990s to adapt collaborative quality improvement methods and apply them to neonatal intensive care.34,35,38,40,41 The

network’s initial collaborative improvement project, known as the NIC/Q (neonatal intensive care quality) Project, involved multidisciplinary teams from 10 institutions. Teams consisting of neonatologists, neonatal nurses, administrators, allied professionals, and quality improvement coaches from the institutions worked closely together to set common improvement goals, to identify potentially better practices for achieving those goals, and to implement the practices in their own NICUs. The clinical improvement goals included reductions in nosocomial infection and chronic lung disease for infants of very low birthweight; resource-related goals included reductions in length of stay and more appropriate use of blood gas testing and x-ray procedures. The teams received training in quality improvement from a professional quality improvement trainer and held discussions in a series of facilitated large group and small focus group meetings and conference calls. The potentially better practices were identified based on review and evaluation of the evidence in the literature, detailed analysis of the processes of care, and site visits to other participating institutions and benchmark units with superior performance outside the project. The Vermont Oxford Network database was used to provide performance feedback to participants and to monitor the results. The preliminary results of the project demonstrated significant reductions in nosocomial infections (six institutions) and chronic lung disease (four institutions) in the subgroups that focused on those goals, both when compared with themselves over time and when compared with a comparison group of 66 nonparticipating units that were members of the Vermont Oxford Network during the same period.38 In addition, the costs of care for infants with very low birthweight at the 10 NIC/Q sites decreased over the course of the project. This project demonstrated the potential for a multifaceted intervention of training in evidence-based collaborative quality improvement, feedback of performance data, and site visiting to improve the quality of neonatal intensive care. As a result of this initial experience, the Vermont Oxford Network organized an ongoing series of evidence-based NIC/Q Quality Improvement Collaboratives for Neonatology.35 These intensive collaboratives, composed of multidisciplinary teams from member institutions, have applied four key improvement habits to a broad range of clinical, operational, and organizational improvement goals.66 As in the original NIC/Q project, participants receive training in quality improvement; work together closely in facilitated large group meetings, focus groups, and conference calls; and use data from the Vermont Oxford Network database for feedback to monitor performance. During the course of the five completed and one ongoing NIC/Q collaboratives, teams have addressed improvements in a broad range of NICU domains related to quality and safety Recent NIC/Q collaboratives have organized around the six themes identified by the Institute of Medicine (family centered, safe, effective, efficient, timely, and equitable) plus an additional theme, social and environmental responsibility, added by the Vermont Oxford Network. These themes are shown in a graphic used by the collaboratives to stress that family centeredness and safety are at the center of improving our care (Fig. 5-4).

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

Efficient

Timely

Patient and family centered Equitable Safe

Socially and environmentally responsible

Effective

Figure 5–4.  Seven critical domains for improvement with

family always at the center of care.  (Adapted from the 6 domains in Crossing the Quality Chasm: A New Health System for the 21st Century13 and Battles JD: Quality and Safety by Design.4 Graphic from Vermont Oxford Network NICQ 2007 Collaborative. Courtesy of James Handyside.) The Vermont Oxford Network NIC/Q collaboratives have explored the use of on-site coaching, site visiting, and other approaches to help teams understand the unique contexts in which their care is provided. Research has pointed to the crucial importance of local context when evidence is applied to improve care. As pointed out by Batalden and Davidoff and by Pawson and Tilley, scientific evidence must be adapted to work in the unique setting of an individual institution.2,63 It is only by adapting the evidence to work in the local context that improvement in quality and safety can be achieved (Fig. 5-5). In recognition of this idea, Paul Plsek coined the term potentially better practice for the NIC/Q collaboratives, as opposed to the commonly used terms best or better practice, to indicate that a practice is not truly better or best until adapted, tested, and implemented to work in the local context.66 The development of regional quality initiatives represents an increasing trend in the United States and abroad. Examples in the United States include California (CPQCC, www.CPQCC. org), North Carolina (PQCNC, www.pqcnc.org), Tennessee (www.tipqc.org), Massachusetts (NeoQIC), Illinois, Michigan, Ohio (OPQC, www.opqc.org), and a rapidly growing list of others as well. An important requirement is that hospitals within the region be able to compare themselves with respect to the members of their regional group. Many of these networks have taken advantage of the group option provided by Vermont Oxford Network (VON’s) Nightingale data system. In addition to this option, California, one of the first of the regional networks, developed a statewide confidential Internet-based data system that interfaces with Nightingale and also produces

Generalizable scientific evidence

+

Particular context

Measured performance improvement

Figure 5–5.  A simple formula highlighting the importance of

adapting the scientific evidence for application in the specific local context of care. (From Batalden PB, Davidoff F: What is “quality improvement” and how can it transform healthcare? Qual Saf Health Care 16:2, 2007.)

85

California reports. Level of care–specific comparisons can be made with reference to all 127 California NICUs or with reference to only the specific hospitals within each of California’s 11 perinatal regions. The latter has facilitated the development of regional quality improvement activities. Although the VONbased neonatal data set provides a solid foundation, regional initiatives could find it advantageous to link their NICU data with other sources of perinatal data and outcomes to assess other opportunities to improve perinatal health and quality care. For example, CPQCC has partnered with California Maternal Child Health (MCH) to develop and incorporate a confidential Internet data set and quality improvement report for the more than 7000 yearly acute neonatal transports and has partnered with California Children’s Services (CCS) to develop and incorporate a confidential Internet-based statewide highrisk infant follow-up data set. Linkages of this sort expand the opportunities for regional initiatives to improve perinatal health and outcomes at both the hospital and community level.

Internet-Based Improvement Collaboratives In addition to the intensive face-to-face improvement collaboratives just described, the Vermont Oxford Network has also conducted a series of Internet-based improvement collaboratives called iNICQ. The iNICQ collaboratives have addressed topics in quality and safety. As a result of the NIC/Q and iNICQ collaboratives, more than 300 multidisciplinary teams from NICUs around the world have participated in formal collaborative efforts to improve the quality and safety of neonatal intensive care.

Four Key Habits for Improvement The four key habits for improvement were identified by Paul Plsek based on his work with the Vermont Oxford Network improvement collaboratives and now serve as foundation for the NIC/Q and iNICQ collaboratives (Fig. 5-6).66 The first is the habit for change. Change is difficult both for people and for organizations, yet without change, there cannot be improvement. Participants in NIC/Q and iNICQ collaboratives are taught to use a simple model for change developed by Langley and colleagues.50 The model is based on three questions: What are we trying to improve? (measurable improvement goal); How will we know that a change is an improvement? (measurement); and What changes can we make that will lead to improvement? (better practices). Multidisciplinary NICU teams answer these questions and test their proposed changes in a series of plan-do-study-act cycles within their institution. The focus is on rapid trial and learning cycles with measurable improvement goals. Improvements are applied to a broad range of clinical, operational, and organizational domains of performance. The second key habit is the habit for evidence-based practice. Participants learn to evaluate the strength and quality of the evidence for different practices and to apply the principles of evidence-based medicine in their daily practice.77,87 The third key habit is the habit for collaborative learning. This process involves collaboration among the disciplines and specialties within an institution and among multidisciplinary teams from different institutions.

86

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE Figure 5–6.  The four key habits for clini-

Change

P D A S

Evidencebased practice

P D A S

P D A S

FOUR KEY HABITS

Process and system thinking

cal improvement are applied by participants in the Vermont Oxford Network NIC/Q Evidence-Based Quality Improvement Collaborative for Neonatology to identify and implement better clinical, operational, and organizational practices for the care of newborn infants and their families.  (Courtesy of the Vermont Oxford Network, Burlington, Vt.)

Collaborative learning

The fourth key habit is the habit for process and systems thinking. This habit requires participants to see neonatal intensive care as a multifaceted process linking many people and organizational subsystems. NICUs are complex adaptive systems. By analyzing these systems and understanding the structures, patterns, and processes of care, it is possible to redesign them to be more effective and efficient. Furthermore, monitoring the implementation of and adherence to evidence-based processes of care within an institution provides a powerful method for quality improvement that in many instances is quicker, more practical, and more efficient than monitoring outcomes alone. The participants in the NIC/Q and iNICQ collaboratives contribute the results of their learning to a growing archive of improvement knowledge maintained by the Vermont Oxford Network. This archive is being organized into an Internet site (www.nicqpedia.org) that will provide the neonatal community with unique resources and tools for collaborative evidence-based quality improvement.

Patient Safety The report To Err is Human has focused national attention on the problem of medical errors and patient safety.48 There is widespread agreement that medical errors and the adverse events they cause represent a serious challenge to the health care system. We are just beginning to learn about the frequency and types of medical errors that occur in the NICU. In a study of medication errors in hospitalized children, Kaushal and colleagues found that in the NICU more than 90% of infants were subjected to a medication error.47 It has been suggested that the use of modern information technology and computerized physician order entry will reduce the frequency of medication errors and mitigate their effects.3 The integration of enhanced information technology in the NICU requires study. Nonmedication errors are also frequent in the NICU. Using data from an anonymous, voluntary, Internet-based

error-reporting system established by the Vermont Oxford Network, Suresh and colleagues documented a broad range of errors and near errors in NICU patients.89 Based on more than 1200 error reports from 54 NICUs over a 17-month period, the most frequent event categories were wrong medication, dose, schedule, or infusion rate (including nutritional agents and blood products, 47%); error in administration or method of using a treatment (14%); patient misidentification (11%); other system failure (9%); error or delay in diagnosis (7%); and error in the performance of an operation, procedure, or test (4%). The most frequent contributory factors were failure to follow policy or protocol (47%), inattention (27%), communication problem (22%), error in charting or documentation (13%), distraction (12%), inexperience (10%), labeling error (10%), and poor teamwork (9%). Serious patient harm was reported in 2% and minor harm in 25% of events. A selected list of events is shown in Box 5-2. A recent study at eight Dutch NICUs identified 5225 safety incident reports for 3859 admissions, of which 4846 were eligible for analysis. The most frequent events included medication errors (27%), followed by laboratory (10%) and enteral nutrition (8%).86 The Center for Patient Safety in Neonatal Intensive Care, funded by the Agency for Healthcare Research and Quality, has been established to conduct research in NICU patient safety. Preliminary studies by center investigators have categorized the wide range of NICU errors,89 reported the value of random safety audits built into daily NICU rounds,93 documented the risks of patient misidentification in the NICU,26 and explored the family perspective on NICU errors using the Internet tool for families, www.howsyourbaby. com.28 The random safety audit developed by Lee and coworkers is a simple and effective tool for addressing safety in the NICU.94 Future research will evaluate strategies to prevent and mitigate medical errors in the NICU. James Reason has pointed out that human error can be viewed in two ways.68 The “person approach” focuses on individuals, blaming them for their mistakes, whereas the “systems approach” focuses on the conditions in which

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

BOX 5–2 Examples of Medical Errors Reported to NICQ.ORG n Chest

tube inserted on wrong side milk infused intravenously n CT scan with IV contrast on wrong infant n Positive HIV test result recorded as negative n Misplaced catheter causing liver injury n Dislodged catheter causing blood loss n Penis burn from hot Mogen clamp n Full day’s IV fluids infused in 2 hours n Bili blanket wrong side up n Ultrasound examination on wrong twin n Phototherapy without eye shields n Human milk given to wrong infant n 100 3 dose of insulin; 10 3 dose of pancuronium bromide n Ligation of carotid artery rather than jugular vein n Infusion pump occlusion undetected for 24 hours n Human

Note: NICQ.org is an internet site maintained by the Vermont Oxford Network. Medical errors were reported voluntarily and anonymously by health professionals from 54 NICUs. These errors represent a selection from among nearly 2000 errors that have been reported as of June 2004. CT, computed tomography; HIV, human immunodeficiency virus; IV, intravenous; NICU, neonatal intensive care unit.

people work, concentrating on building safeguards into the system to prevent errors and mitigate their effects. Reason suggests that high-reliability organizations with low accident rates such as naval aircraft carriers, nuclear power plants, and air traffic control centers are successful because of a culture of safety that replaces individual blame with a system designed to be robust in the face of the unavoidable human and operational hazards. Health care organizations must now move from the person approach to the systems approach to achieve high reliability in medical care. The Joint Commission for Accreditation of Healthcare Organizations (JCAHO) is now focusing on patient safety in its hospital audits and has issued a series of National Patient Safety Goals (Box 5-3).46 It will be important for NICU teams to apply the “four key habits” to address these safety goals in the unique setting of the NICU and to develop methods for identifying and monitoring errors. Of course, safety is not really separate from the overall issue of quality of care. A culture of safety is an essential characteristic of the high-quality NICU. Although errors can never be totally eliminated, we must design the safest systems possible for the patients and families for whom we provide care.

CONCLUSION The recognition that there is widespread variation among physicians and hospitals in clinical practice and patient outcomes and the growing pressure to increase the quality, safety, and cost-effectiveness of medical care have resulted in unprecedented interest in assessing, evaluating, and improving medical practice. If health professionals are to function successfully in this environment, they must understand how to evaluate their own performance and how their performance will be

87

BOX 5–3 Joint Commission for Accreditation of Healthcare Organizations 2009 National Patient Safety Goals* n Improve

the accuracy of patient identification Improve the effectiveness of communication among caregivers n Improve the safety of using medications n Reduce the risk of health care–associated infections n Accurately and completely reconcile medications across the continuum of care n Reduce the risk of patient harm resulting from falls n Reduce the risk of surgical fires n Improve the effectiveness of clinical alarm systems n Reduce the risk of influenza and pneumococcal disease in older adults n Encourage patients’ active involvement in their own care as a patient safety strategy n Prevent health care–associated pressure ulcers n Identification by the organization of safety risks inherent in its patient population n

*http://www.firstassist.com/forms/JCAHO%202009%20National%20Patient %20Safety%20Goal.pdf

evaluated by others. These evaluations require accurate and reliable information. Most importantly, neonatologists and other health care professionals must learn how to use the information to improve the quality and safety of the medical care they provide. In this part of the chapter we reviewed some of the available data sources and discussed a number of issues related to using them for improvement. Although neonatologists should not be expected to become experts in databases and evaluation methods, they do need to develop a basic understanding that will allow them to work effectively with other professionals in the changing health care environment. In this “era of assessment and accountability,” we must all develop the knowledge, skills, and motivation necessary to assume leadership roles in multidisciplinary collaborative quality improvement within our institutions, in larger health care organizations, and across regions. Only then can the potential benefits of modern databases and information systems be translated into better medical care for newborn infants and their families.

ACKNOWLEDGMENTS We thank Jeannette Rogowski, PhD, of the University of Medicine and Dentistry of New Jersey for her assistance with the cost and resource section of this chapter.

REFERENCES 1. American College of Obstetricians and Gynecologists: Evaluation of cesarean delivery, Washington, D.C., 2000, American College of Obstetricians and Gynecologists. 2. Batalden PB, Davidoff F: What is “quality improvement” and how can it transform healthcare? Qual Saf Health Care 16:2, 2007.

88

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

3. Bates DW, Gawande AA: Improving safety with information technology, N Engl J Med 348:2526, 2003. 4. Battles JD: Quality and safety by design, Qual Saf Health Care 15 (Suppl 1):i1, 2006. 5. Berwick DM: A user’s manual for the IOM’s “Quality Chasm” report, Health Aff (Millwood) 21:80, 2002. 6. Berwick DM: Improvement, trust, and the healthcare workforce, Qual Saf Health Care 12:448, 2003. 7. Blondel B, et al: The impact of the increasing number of multiple births on the rates of preterm birth and low birthweight: an international study, Am J Public Health 92:1323, 2002. 8. Bloom BT, et al: Improving growth of very low birth weight infants in the first 28 days, Pediatrics 112:8, 2003. 9. Boyce JM, Pitt D: Guidelines for hand hygiene in health-care settings. Recommendations of the healthcare infection control practices advisory committee and the HICPAC/SHEA/APIC/ IDSA hand hygiene task force, MMWR Morb Mortal Weekly Rep 51:1, 2002. 10. Brodie SB, et al: Occurrence of nosocomial bloodstream infections in six neonatal intensive care units, Pediatr Infect Dis J 19:56, 2000. 11. California Perinatal Quality Care Collaborative (website). http://www.cpqcc.org. Accessed May 24, 2010. 12. Committee on Fetus and Newborn of the American Academy of Pediatrics: Levels of neonatal care, Pediatrics 114:1341, 2004. 13. Committee on Quality of Health Care in America: Crossing the quality chasm: a new health system for the 21st century, Washington, D.C., 2001, National Academy Press. 14. Conner J, Nelson EC: Neonatal intensive care: Satisfaction measured from a parent’s perspective, Pediatrics 103(1 Suppl E):336, 1999. 15. Danielsen B, et al: Newborn discharge timing and readmissions: California, 1992-1995, Pediatrics 106:31, 2000. 16. Darlow BA, et al, on behalf of the Australian New Zealand Neonatal Network: Variation in rates of severe retinopathy of prematurity among neonatal intensive care units in the Australian and New Zealand neonatal Network, Br J Ophthalmol 89:1592, 2005. 17. Dooley SL, et al: Quality assessment of perinatal regionalization by multivariate analysis: Illinois, 1991-1993, Obstet Gynecol 89:193, 1997. 18. Edwards W, et al: Parents’ perspectives on errors in neonatal intensive care, Pediatr Res 55:435A, 2004. 19. Fanaroff AA, et al: NICHD Neonatal Research Network: Trends in neonatal morbidity and mortality for very low birthweight infants, Am J Obstet Gynecol 196(2):147.e1, 2007. 20. Gibberd R, Pathmeswaran A, Burtenshaw K: Using clinical indicators toidentify areas for quality improvement, J Qual Clin Pract 20:136 2000. 21. Gilbert WM, et al: Associated factors in 1611 cases of brachial plexus injury, Obstet Gynecol 93:536, 1999. 22. Gilbert WM, et al: Childbearing beyond age 40: Pregnancy outcome in 24,032 cases, Obstet Gynecol 93:9, 1999. 23. Gould JB: Vital records for quality improvement, Pediatrics 103:278, 1999. 24. Gould JB: Quality improvement in perinatal medicine: assessing the quality of perinatal care, NeoReviews 5:e33, 2004. 25. Gould JB, et al: Very low birth weight births at non-NICU hospitals: the role of sociodemographic, perinatal, and geographic factors, J Perinatol 19:197, 1999.

26. Gould JB, et al: Cesarean rates and neonatal morbidity in a low-risk population, Obstet Gynecol 104:11, 2004. 27. Gray JE, et al: Neonatal therapeutic intervention scoring system: a therapy-based severity-of-illness index, Pediatrics 90:561, 1992. 28. Gray JE, et al: Patient mis-identification in the neonatal intensive care unit (NICU): qualification of risk, Pediatr Res 55:519A, 2004. 29. Hack M, et al: Very-low-birth-weight outcomes of the National Institute of Child Health and Human Development Neonatal Network, November 1989 to October 1990, Am J Obstet Gynecol 172:457, 1995. 30. Hack M, et al: Outcomes in young adulthood for very-low-birthweight infants, N Engl J Med 346:149, 2002. 31. Hadorn DC, et al: Assessing the Performance of Mortality Prediction Models, Santa Monica, Calif, 1993, RAND. 32. Herrchen B, et al: Vital statistics linked birth/infant death and hospital discharge record linkage for epidemiological studies, Comput Biomed Res 30:290, 1997. 33. Horbar JD: Birthweight-adjusted mortality rates for assessing the effectiveness of neonatal intensive care, Med Decis Making 12:259, 1992. 34. Horbar JD: The Vermont-Oxford Neonatal Network: integrating research and clinical practice to improve the quality of medical care, Semin Perinatol 19:124, 1995. 35. Horbar JD: The Vermont Oxford Network: evidence-based quality improvement for neonatology, Pediatrics 103(1 Suppl E):350, 1999. 36. Horbar JD, Carpenter J: Nosocomial infection in very low birth weight infants: We can do better! Pediatr Res 55:404A, 2004. 37. Horbar JD, Leahy KA: An assessment of data quality in the Vermont-Oxford Trials Network database, Control Clin Trials 16:51, 1995. 38. Horbar JD, et al: Collaborative quality improvement for neonatal intensive care, Pediatrics 107:14, 2001. 39. Horbar JD, et al: Trends in mortality and morbidity for very low birth weight infants, 1991-1999, Pediatrics 110:143, 2002. 40. Horbar JD, et al, eds: Evidence based quality improvement in neonatal and perinatal medicine: the NICQ 2000 experience, Pediatrics 111:e395, 2003. (Supplement, http://pediatrics. aappublications.org/cgi/content/full/111/4/SE1/e395) 41. Horbar JD, et al, eds: Evidence based quality improvement in neonatal and perinatal medicine: the NICQ 2002 experience, Pediatrics 118:S57, 2006. (Supplement, http://pediatrics. aappublications.org/content/vol118/Supplement_2/) 42. Institute for Healthcare Improvement (website). http://www. ihi.org/ ihi. Accessed May 24, 2010. 43. Institute of Medicine: Reliability of hospital discharge abstracts. Washington, D.C., 1977, National Academy of Sciences. 44. International Neonatal Network: The CRIB (Clinical Risk Index for Babies) score: a tool for assessing initial neonatal risk and comparing performance of neonatal intensive care units, Lancet 342:193, 1993. 45. International Neonatal Network, Scottish Neonatal Consultants, Nurses Collaborative Study Group: Risk adjusted and population based studies of the outcome for high risk infants in Scotland and Australia, Arch Dis Child Fetal Neonatal Ed 82:F118, 2000. 46. Joint Commission on Accreditation of Healthcare Organizations (website). http://www.jcaho.org. Accessed May 24, 2010.

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine 47. Kaushal R, et al: Medication errors and adverse drug events in pediatric inpatients, JAMA 285:2114, 2001. 48. Kohn LT, et al, and the Committee on Quality of Health Care in America: To err is human: building a safer health care system. Washington, D.C., 2000, National Academy Press. 49. Kulynych J, Korn D: The effect of the new federal medicalprivacy rule on research, N Engl J Med 346:201, 2002. 50. Langley GJ, et al: The improvement guide: a practice approach to enhancing organizational performance, San Francisco, 1996, Jossey-Bass. 51. Leapfrog Group (website). www.leapfroggroup.org. Accessed May 24, 2010. 52. Lee L, et al: Random safety audits in the neonatal unit, Arch Dis Child Fetal Neonatal Ed 94:F116, 2009. 53. Lee SK, et al: Transport risk index of physiologic stability: a practical system for assessing infant transport care, J Pediatr 139(2):220, 2001. 54. Lee SK, et al: Cost-effectiveness and choice of infant transport systems, Med Care 40(8):705, 2002. 55. Lemons JA, et al, and the NICHD Neonatal Research Network: Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996, Pediatrics 107:e1, 2001. 56. Lindenauer PK. Effects of Quality Improvement Collaboratives, BMJ 336:1448 2008. 57. Lyman JA, et al: Mapping from a clinical data warehouse to the HL7 reference information model, Proceedings of the AMIA Symposium, 2003:920. 58. Muldoon JH: Structure and performance of different DRG classification systems for neonatal medicine, Pediatrics (1 Suppl E)103:302, 1999. 59. National Center for Health Statistics: International Classification of Diseases: Clinical Modifications 4th edition. Los Angeles, 1995, Practice Management Information Corporation. 60. O’Connor GT, et al, and the Northern New England Cardiovascular Disease Study Group: A regional intervention to improve the hospital mortality associated with coronary artery bypass graft surgery, JAMA 275:841, 1996. 61. Park RE, et al: Explaining variations in hospital death rates. Randomness, severity of illness, quality of care, JAMA 264:484, 1990. 62. Parry G, et al: CRIB II: An update of the clinical risk index for babies score, Lancet 361:1789, 2003. 63. Pawson R, Tilley N. Realistic evaluation. London, England, 1997, Sage Publications, Ltd. 64. Perinatal Profiles Project: California Perinatal Profiles, Berkeley, 2004, University of California School of Public Health. 65. Phibbs CS, et al: Alternative to diagnosis-related groups for newborn intensive care, Pediatrics 78:829, 1986. 66. Plsek P: Quality improvement methods in neonatal and perinatal medicine, Pediatrics 103(1 Suppl E):203, 1999. 67. Pollack MM, et al: Accurate prediction of the outcome of pediatric intensive care. A new quantitative method, N Engl J Med 316:134, 1987. 68. Reason J: Human error: models and management, BMJ 320:768, 2000. 69. Relman AS: Assessment and accountability: the third revolution in medical care, N Engl J Med 319:1220, 1988. 70. Richardson DK, et al: Score for neonatal acute physiology: a physiologic severity index for neonatal intensive care, Pediatrics 91:617, 1993.

89

71. Richardson DK, et al: Measuring illness severity in neonatal intensive care, J Intensive Care Med 9:20, 1994. 72. Richardson DK, et al: Risk adjustment for quality improvement, Pediatrics 103(1 Suppl E):255, 1999. 73. Richardson DK, et al: A critical review of cost reduction in neonatal intensive care. I. The structure of costs, J Perinatol 21:107, 2001. 74. Richardson DK, et al: SNAP-II and SNAPPE-II: simplified newborn illness severity and mortality risk scores, J Pediatr 138:92, 2001. 75. Rogowski JA: Measuring the cost of neonatal and perinatal care, Pediatrics 103(1 Suppl E):329, 1999. 76. Romano PS, et al: Coding of perineal lacerations and other complications of obstetric care in hospital discharge data, Obstet Gynecol 106:717, 2005. 77. Sackett DL, et al: Evidence-based medicine. how to practice and teach EBM, 2nd ed. London, 2000, Churchill Livingstone. 78. Saigal S: Perception of health status and quality of life of extremely low birth weight survivors. The consumer, the provider, and the child, Clin Perinatol 27:403, 2000. 79. Saigal S, et al: Parental perspectives of the health status and health-related quality of life of teen-aged children who were extremely low birth weight and term controls, Pediatrics 105:569, 2000. 80. St John EB, et al: Cost of neonatal care according to gestational age at birth and survival status, Am J Obstet Gynecol 182:170, 2000. 81. Schouten LMT, et al: Evidence for the impact of quality improvement collaboratives: systematic review, BMJ 336:1491, 2008. 82. Schulman J, Spiegelhalter Dj, Parry G: How to interpret your dot: decoding the message of clinical performance indicators, J Perinatol 28:588, 2008. 83. Schwartz RM, et al: Administrative data for quality improvement, Pediatrics 103(1 Suppl E):291, 1999. 84. Simpson JM, et al, on behalf of the Australian and New Zealand Neonatal Network: Analysing differences in clinical outcomes between hospitals, Qual Saf Health Care 12:257, 2003. 85. Snell LM, et al: Reliability of birth certificate reporting of congenital anomalies, Am J Perinatol 9:219, 1992. 86. Snijders C, et al, on behalf of the NEOSAFE study group: Specialty-based, voluntary incident reporting in neonatal intensive care: description of 4846 incident reports, Arch Dis Child Fetal Neonatal Ed 94:F210, 2009. 87. Soll RF, Andruscavage L: The principles and practice of evidence based neonatology, Pediatrics 103(1 Suppl E): 215, 1999. 88. Spiegelhalter DJ. Handling over-dispersion of performance indicators, Qual Saf Health Care 14:347, 2005. 89. Suresh G, et al: Voluntary anonymous reporting of medical errors for neonatal intensive care. Pediatrics 113:1609, 2004. 90. Toolkit Group of the Perinatal Quality Improvement Panel: Nosocomial Neonatal Infection Toolkit, Palo Alto, 2004, California Perinatal Quality Care Collaborative. Available at: http://www.cpqcc.org/NIToolkit.html. 91. Tyson JE, et al: Intensive care for extreme prematurity: Moving beyond gestational age, N Engl J Med 358:1672, 2008. 92. US Department of Health and Human Services: Healthy People 2010, Washington, D.C., 2000, US Government Printing Office.

90

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

93. Ursprung R, et al: Random audits for patient safety in the NICU, Pediatr Res 55:435A, 2004. 94. Vermont Oxford Network 2007 Annual Summary for Very Low Birth Weight Infants. Horbar JD, et al, editors. Burlington, VT, 2008, Vermont Oxford Network. 95. Vermont Oxford Network Database Manual of Operations for Infants Born in 2009. Release 13.2, Burlington, VT, 2009, Vermont Oxford Network. Available at http://www.vtoxford. org/tools/2009ManualofOperationswithindex13_2.pdf. Accessed May 14, 2009 96. Vohr BR, et al: Center differences and outcomes of extremely low birth weight infants, Pediatrics 113:781, 2004. 97. Walsh MC, et al, for the NICHD Neonatal Research Network: A cluster randomized trial of benchmarking and multimodal quality improvement to improve survival free of bronchopulmonary dysplasia in infants ,1250 grams birthweight, Pediatrics 119:876, 2007. 98. Weed LL: Knowledge coupling: new premises and tools for medical care and education, New York, 1991, Springer-Verlag. 98a. Williams RL, Chen PM, Clingman EJ (eds): Maternal and Child Health Data Base Statistical Appendix, 1978 to 1982, University of California and Organization Research Institute, Santa Barbara, CA, 1984. 99. Wirtschafter DD, Powers RJ: Organizing regional perinatal quality improvement: global considerations and local implementation. NeoReviews 5:e50, 2004. 100. Wirtschafter DD, et al: Promoting antenatal steroid use for fetal maturation: results from the California Perinatal Quality Care Collaborative, J Pediatr v.148, 5:606, 2006. 101. Wright LL, et al: Evidence from multicenter networks on the current use and effectiveness of antenatal corticosteroids in low birth weight infants, Am J Obstet Gynecol 173:263, 1995. 102. Yasmeen S, et al: Accuracy of obstetric diagnoses and procedures in hospital discharge data, Am J Obstet Gynecol v. 194, 4: 992, 2006. 103. Zupancic JAF, et al: VON SNAP pilot project. Performance of the revised score for neonatal acute physiology in the Vermont Oxford Network, Pediatr Res 55:521A, 2004.

PART 2

Simulation in NeonatalPerinatal Medicine Louis P. Halamek Approximately 4 million babies are born in the United States every year; of these, around 10% require some degree of resuscitation, with 1% needing extensive resuscitative efforts such as chest compressions, intubation, and delivery of medication.44 Many different types of health care professionals are responsible for caring for newborns at the time of birth and in the days and weeks that follow. These professionals include, but are not limited to, neonatologists, pediatricians, family practitioners, obstetricians, midwives, neonatal nurse practitioners, nurses, and trainees at all levels in these disciplines. Given the large number of births, the frequency of resuscitation, and the diversity of professionals bearing

responsibility for caring for newborns, the need for effective means of acquisition and maintenance of the skill sets necessary to deliver safe and competent care is of tremendous importance. Traditionally, the apprenticeship model of assuming graduated responsibility for the care of real patients has been used to address this need. Unfortunately, the assumption underlying this model—that placing a trainee in a supervised clinical environment for a set period of time will allow him or her to experience a sufficient number and breadth of clinical cases to ensure the ability to practice independently and safely in the community—does not always prove to be true. Therefore, a new paradigm of learning is required.

TEACHING VERSUS LEARNING Whereas teaching is something that is (passively) done to trainees, learning is something that trainees must (actively) do themselves.30 Because not everything that is taught is necessarily learned, programs that best facilitate skill acquisition in trainees are those that focus on learning, rather than on teaching. Traditional didactic programs are passive by nature, and the settings in which they are held are typically isolated from realistic cues, distractors, and time pressure; thus, such programs are unable to prepare learners adequately for all of the challenges inherent when working in the real environment. Although learning in a real environment during the delivery of actual patient care may appear ideal at first glance, a deeper analysis reveals otherwise. The most obvious problem with using the real environment as the primary source of skill acquisition and practice is that any mistake could prove lethal to patients. The pace of actual clinical care conducted in the real environment with real patients is often too fast to allow trainees to take full advantage of the learning opportunities therein. Moreover, typically there is no way to ensure that all important learning opportunities will present themselves in the real environment during the time that the trainee is present. Finally, the real environment is also a very expensive environment and is populated with a number of professionals whose job description may not include providing learning opportunities for trainees.28 Learning is best facilitated when the learning opportunities are tailored to meet the needs of the learners. Training models that offer the same content in the same fashion to all learners (thus implying that competency can be attained simply by spending a particular, often arbitrary amount of time at a task) fail to recognize that adults have different strengths, weaknesses, and life experiences and acquire and maintain different skills at different rates. Some characteristics of effective learning strategies include the following: n A

focus on active rather than passive learning activities of skill sets while performing under realistic conditions n An emphasis on competency (the ability to perform successfully) rather than on compliance (adherence to rules, such as participation in an activity for a predetermined period of time) n Integration

It is much easier to design and implement training exercises that are teacher-centric and targeted at the needs of the average trainee rather than to develop programs that tailor the learning to meet the needs of individual learners; therefore, there are

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

few interventions that are truly effective at uniformly facilitating the acquisition of cognitive, technical, and behavioral skills in diverse groups of learners.

91

TABLE 5–4  S  kills Necessary for Successful Intubation Cognitive Skills

SKILLS TO BE LEARNED

Know the indications for intubation of the newborn

Any discussion of learning in health care must start with what it is that can be learned. There are three “skill sets” that may be acquired and refined by health care professionals:

Know how to recognize these indications when present

n What

Know the indications of a successful intubation

we know in our brains (cognitive skills or content knowledge) n What we do with our hands (technical skills) n How we employ the first two skill sets while caring for patients and working under realistic time pressure with our colleagues (behavioral skills) Content knowledge is the skill set most familiar to learners and is typically the major (or only) skill set that is formally evaluated, usually through written or online tests. Technical skills such as intubation are critical to neonatal care. Despite their importance, such skills are most commonly practiced at skills stations using models that poorly represent neonatal anatomy and physiology and are evaluated by a subjective assessment of performance that is isolated from the time pressure intrinsic to the real environment. Behavioral skills, such as leadership, teamwork, and effective communication, are critically important to successful patient outcomes (Table 5-3). Unfortunately, these important skills are rarely, if ever, addressed specifically in learning programs directed at health care professionals. Many patient care tasks actually incorporate elements of all three of these skills sets. Intubation of the newborn is one such example. Far from being simply a pure technical skill, effective and safe intubation requires coordination and integration of multiple cognitive skills (knowing the indications for intubation and the signs of successful and unsuccessful intubations), sequential discrete technical skills (assembling, testing, and inserting the laryngoscope), and a number of behavioral skills (effectively communicating observations and needs, evenly distributing the workload, and delegating responsibilities), all of which must also be accomplished in a time-efficient manner (Table 5-4).

TABLE 5–3  Key Behavioral Skills Know your environment Anticipate and plan Assume the leadership role Communicate effectively Delegate workload optimally Allocate attention wisely Use all available information Use all available resources Call for help when needed Maintain professionalism

Know what equipment (e.g., size of endotracheal tube, laryngoscope blade) to use to accomplish intubation

Technical Skills

Know how to assemble the laryngoscope Know how to hold the laryngoscope Know how to use the equipment to expose the airway to view Know how to insert the endotracheal tube into the airway Know how to assess for proper placement of the endotracheal tube in the airway Know how to secure the endotracheal tube in the airway Behavioral Skills

Communicate effectively with team members regarding the need for intubation, specific pieces of equipment, and the like Distribute the workload so that specific tasks are assigned to the team members most likely to carry them out successfully Delegate responsibility and supervise appropriately Call for help when necessary

SIMULATION-BASED LEARNING Simulation may be defined broadly as any exercise that allows an individual to experience a situation that, although not real, nevertheless generates authentic responses on his or her part.28 Simulation-based learning methodologies realistically recreate the key visual, auditory, and tactile cues of actual situations to provide learning experiences that closely mimic the conditions encountered when working in the real environment. Simulation in health care encompasses a wide spectrum of learning activities including those as diverse as the following: n Computer-based

interactive virtual environments populated not only by patients but also by unique computergenerated representations (avatars) of one’s human colleagues n Highly realistic physical environments in which real human health care professionals work as a team providing care for highly sophisticated patient simulators while using real working medical equipment Relatively simple types of simulation such as case studies, role playing, and task training (practice of one particular skill in isolation from other elements associated with that skill) have been used for many years in neonatal-perinatal

92

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

medicine. In general, however, the term simulation is typically reserved to describe learning opportunities that occur in highly realistic physical environments that include human actors and real working medical equipment and supplies. Learners in simulated environments suspend their knowledge that they are working in a contrived situation and display the same type of behavior as they do when working in the real environment. Examples of virtual and physical environments focused on neonatal patients may be found at http://storm.uni-mb.si/ (the Virtual Delivery Room) and http://www.cape.lpch.org/courses/ descriptions/index.html (the NeoSim training program in neonatal resuscitation). By providing key visual, auditory, and tactile cues, a high level of physical, biologic, and psychological fidelity to the real environment can be created.41 As a result of this degree of realism, learners respond as they would to a real-life situation; thus, their performance during a simulated situation can be hypothesized to resemble what it would be in the real world.28 Scenarios coupled with debriefings (in which discussion of what went well and what could be improved upon occur in a nonjudgmental fashion) provide rich learning experiences that equal or exceed those of other learning methodologies. Although undoubtedly some learning takes place during active participation in scenarios, trainees perceive that most of the learning occurs during the debriefing when time is allotted for self- or facilitated reflection on performance. This perception is supported by years of anecdotal experience utilizing simulation in high-risk domains as well as decades of research in the science of adult learning. Much has been made of the importance of the concept of fidelity in simulation-based learning, and debate continues in the health care simulation community as to how much fidelity is necessary.6 Although some may argue that the higher the fidelity of the scenario to real life the better the learning opportunity, in reality, as long as sufficient attention is paid to providing the key visual, auditory, and tactile cues for learners, allowing them to form a shared mental model of the nature of the situation that they are facing, they will be able to work effectively to resolve the clinical problems that become manifest during the scenario and therefore achieve the learning objectives. Three general levels of simulation fidelity should be considered: physical, biologic, and psychological. Physical fidelity refers to the realism of the physical space in which training occurs; this space is made to look real by including appropriate working medical equipment, fluids, pharmacologic agents, beds, and the other elements necessary for patient care. Biologic fidelity includes the patient simulators and standardized patients as well as the human beings acting as confederates during the simulation, playing roles designed to assist the evolution of the scenario. Finally, all of the previously mentioned elements interact with the mindset brought into the scenario by the learners to create a sense of realism or psychological fidelity. The overall goal of simulation-based learning is to provide training experiences that closely mimic the conditions encountered when working in the real environment; the major difference between the simulated environment and the real environment is the absence of real human patients. Simulation-based learning provides many obvious advantages over more traditional training methodologies. Because

patient simulators replace human beings, there is no risk to patients; invasive procedures can be practiced without the fear of patient harm or medical liability. Unlike what happens in the real environment, learning opportunities using simulation can be scheduled at convenient times and structured so that specific learning objectives are consistently achieved. Simulation-based learning is an ideal methodology for allowing learners to practice integration of multiple skill sets while working under highly realistic and often stressful conditions. Rather than being directed solely at the individual, simulation easily accommodates the learning needs of multidisciplinary teams. Simulation-based learning activities can easily be scaled in intensity to meet the needs of learners at all levels of experience, and they can be used to foster both the acquisition and maintenance of particular skills. It can also be hypothesized that learners who participate in simulation-based exercises likely will be better prepared and will need less supervision when entering or re-entering the real environment (Table 5-5). Simulation-based learning opportunities are the standard for skill acquisition in industries such as commercial aviation, aerospace, nuclear power, and the military, where the risk of death or severe injury to human beings is very real. The use of simulation in these domains is characterized by an emphasis on probing human beings and the systems they design (both technological and social) for weaknesses and then designing and testing solutions to address those weaknesses. This type of learning is the result of a culture that not only fosters but also demands a willingness to learn from mistakes made during simulated events, thus decreasing the

TABLE 5–5  Advantages of Simulation Presents no risk to human patients Permits training in environments usually inaccessible to less experienced trainees Can be tailored easily to the needs of individual trainees regardless of level of experience Allows practice without interruption or interference Fosters integration of cognitive, technical, and behavioral skills Facilitates multidisciplinary team training Creates training opportunities for rarely encountered but highly challenging or risky situations Provides structured learning opportunities with defined learning objectives Can be scheduled at times convenient to trainees and instructors Permits formal objective performance assessment Facilitates use of debriefings as a source of detailed constructive feedback Provides a very rich learning experience in a relatively short period of time Optimizes use of time, money, and other resources

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

chances of repeating such mistakes when working in the real environment. The National Aeronautics and Space Administration (NASA) was established in 1958 to conduct aeronautical research and administer the human and robotic exploration of space. Space travel is an inherently risky business; how else to describe a process that places human beings in a rocket filled with tons of liquid fuel and then ignites that fuel in the hope that it will propel those humans into the vacuum of space? The value of simulation was made clear during the Apollo 13 mission, launched on April 11, 1970, as the third mission to land humans on the moon. To prepare for this mission, the three members of the prime crew, Jim Lovell (captain), Tom Mattingly (command module pilot), and Fred Haise (lunar module pilot) trained in NASA’s flight simulators for months. One week before launch, Mattingly was exposed to the measles; because he was not immune to the disease, his backup, Jack Swigert, was given simulator time with Lovell and Haise in the week preceding launch to “ensure that Lovell, Swigert and Haise could function with unquestioned teamwork through even the most arduous and time-critical simulated emergency conditions.”3 A decision by the flight surgeon only 1 day before launch scrubbed Mattingly from the prime crew and placed Swigert in the left-hand seat as command module pilot for the mission. Fifty-six hours into the flight, as Apollo 13 was approximately 200,000 miles away from the earth en route to the moon, an explosion in the service module’s cryogenic oxygen system resulted in the uncontrolled venting of oxygen into space, creating a situation that threatened not only the success of the mission but also the lives of the crew (oxygen was the source of the crew’s breathing air, and substrate for the fuel cells that generate electrical power). As has been well documented, the crew did return safely to earth. What is less well known is that, well before Apollo 13 launched, the procedures that allowed the crew to survive and recover from this devastating event were devised by engineers charged with envisioning every possible failure and then designing procedures to address these failures.21 These procedures were tested under current mission parameters in the flight simulators almost continuously during the crisis and when deemed reliable were relayed to the crew.8 In the formal postmission debriefing, the crew referred to their experiences in flight simulation in excess of 40 times.4 Based largely on its successful use in other domains and its inherent face validity, simulation is being employed as a learning methodology with increasing frequency in health care. The rationale for employing simulation-based learning in neonatology is clear. Serious neonatal pathology is one example of the classic low-frequency, high-risk event that lends itself well to simulation-based learning.31 Many health care professionals who care for newborns have the opportunity to manage serious or rare disease processes on an infrequent basis. Even for those for whom a sufficient number of opportunities do exist, one must question whether it is acceptable to essentially practice on real living patients who are not capable of providing informed consent on their own. Although parents do act as surrogate decision makers for children below the age of consent, few want to contemplate that their child will be the first one on whom someone will

93

perform their first spinal tap, first intubation, or first thoracostomy tube placement. The subspecialty of neonatalperinatal medicine is unique in that one patient (the fetus) exists inside of another patient (the mother); in the case of multiple gestation, two or more neonatal patients await birth inside of the pregnant female patient. The possibility of a sick mother delivering (a) sick newborn(s) creates a situation in which optimal preparation occurs only when the neonatology and obstetric teams train together.29 The multitude of events that can complicate human birth and the neonatal period make this an especially appealing target for simulation-based learning.40 It may be argued that the ethical imperative for simulation is stronger in pediatrics in general and neonatal-perinatal medicine in particular than in any other field of health care.47

THE EVIDENCE BEHIND SIMULATION In 1999 the Institute of Medicine (IOM) published To Err is Human: Building a Safer Health System, a report on human error and patient safety in the United States.35 In this report, the authors estimated that between 44,000 and 98,000 Americans die each year as a result of medical errors. Although this figure has been highly debated, it is based on extrapolation of the data contained in studies out of Colorado, Utah, and New York published in peer-reviewed literature.7,36,43 The 1999 report was followed in 2001 by another from the IOM, Crossing the Quality Chasm: A New Health System for the 21st Century, in which the type of interventions, including training methodologies, necessary to improve patient safety were discussed.10 Subsequently in 2004 the Joint Commission (JC) published a Sentinel Event Alert describing ineffective communication as a major cause in almost 75% of the 47 cases of neonatal mortality or severe neonatal morbidity (lifelong serious neurologic compromise) reported to that agency; since that time, an additional 62 cases have been added.34 In response to these root cause analyses, the JC recommended that all health care organizations responsible for delivering newborns “conduct team training in perinatal areas to teach staff to work together and communicate more effectively” and “for high-risk events, such as shoulder dystocia, emergency cesarean delivery, maternal hemorrhage, and neonatal resuscitation, conduct clinical drills to help staff prepare for when such events actually occur, and conduct debriefings to evaluate team performance and identify areas for improvement.” Simulation-based learning is grounded in adult learning theory, supported by rational conjecture, and felt to be essential for achieving expert performance.22,23,39,46 Simulation-based learning in its broadest sense has been used for decades in other domains in which the risk to human life is high and is a core component of maintenance of certification programs in those domains. Although objective data exist in some of those domains to support the use of simulation, it is far from being definitive in nature or comprehensive across all domains.42 Despite the paucity of objective evidence, the use of simulationbased training remains standard operating procedure in commercial aviation, aerospace, the military, and other similar occupations, and no one working in those domains would consider conducting a prospective, randomized, controlled trial with subjects (and for those whose safety they bear responsibility) who are randomized to the “no simulation” group.

94

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

So what can be said about the effectiveness of simulationbased learning in health care in which the emphasis is clearly on evidence-based practice and historically the gold standard of evidence has consisted of the prospective, randomized, controlled, sufficiently powered clinical trial in which the outcomes focus on patients? In some ways this situation is similar to the debate that is currently taking place regarding the extent of evidence necessary to accept quality assurance/ improvement work. At one end of the spectrum are clinicians and investigators who insist that quality initiatives must be subject to the same rigorous testing that precedes the introduction of new pharmacologic therapies and medical instrumentation in order to prove that they actually improve quality and ensure patient safety. Alternatively, others note that requiring randomized controlled trials to assess the safety of innovations with high face validity may place humans at risk unduly and therefore prove impossible to conduct. In fact, some authors are of the opinion that not to use simulationbased training methodologies, relying instead solely on practice on real patients, is ethically indefensible.47 Thus, we are left with a situation in which the need for more definitive evidence, while desirable, is felt by at least some members of the health care education and training community not to be necessary or practical. The ever-expanding body of knowledge of the basic processes underlying normal and abnormal neonatal physiology has allowed those clinicians responsible for caring for newborns to generate evidence-based clinical practice guidelines under the auspices of the International Liaison Committee on Resuscitation (ILCOR) and its various neonatal delegations (Neonatal Resuscitation Program of the American Academy of Pediatrics, the American Heart Association, the Heart and Stroke Foundation of Canada, the Inter-American Heart Foundation, the European Resuscitation Council, the Australian and New Zealand Committee on Resuscitation, and the Resuscitation Councils of Southern Africa). ILCOR was founded in 1992 to facilitate international collaboration on issues involving neonatal, pediatric, and adult cardiopulmonary resuscitation and emergency cardiovascular care.9 As knowledge about the physiologic processes underlying neonatal cardiorespiratory decompensation and the list of therapeutic interventions grow, so, too, do the expectations for mastery of this knowledge and associated skill sets that are placed on those responsible for caring for the neonate in distress. In recognition of this fact, the 2010 ILCOR process includes a review of the evidence behind the use of simulation and debriefing in resuscitation training. Neonatal care occurs in environments that are extremely dynamic and complex, and the nature of the work performed in those environments requires that correct decisions be made and appropriate interventions be carried out, often while working as a member of a multidisciplinary team in the context of intense time pressure. Simulation is an ideal learning methodology to allow learners to practice working in such an environment. The first simulation-based learning program in neonatal-perinatal medicine (and one of the first in all of health care) is the NeoSim program developed at the Center for Advanced Pediatric and Perinatal Education (CAPE) located at Packard Children’s Hospital on the campus of Stanford University in Palo Alto, California. Launched in 1997, NeoSim has been a very successful innovation in

training in the cognitive, technical, and behavioral skills necessary for optimal care of the newborn in distress.32,40 The success of NeoSim is an example of the power of the simulationbased learning methodology; the NeoSim program serves as the basis for a series of changes taking place in the current national standard for training in neonatal resuscitation, the Neonatal Resuscitation Program (NRP) of the American Academy of Pediatrics (AAP).27 Simulation has been shown to be capable of producing short-term improvement in skills even in highly experienced professionals. In a study by Anderson and coworkers, nine nurses with expertise in extracorporeal membrane oxygenation (ECMO) participated in two sequential randomly assigned simulated ECMO emergencies using a sophisticated ECMO simulator.1 The simulated emergencies were captured on videotape and reviewed with the subjects during facilitated debriefings that occurred immediately following each scenario. All videotapes were scored for key technical and behavioral skills by reviewers masked to the sequence of the scenarios. Subjects performed key technical skills correctly more often in the second simulated ECMO emergency, and their response times for three out of five specific technical tasks improved from the first to the second simulated emergency by an average of 27 seconds. Subjects’ behavioral skills were uniformly rated more highly in the second simulated ECMO emergency, and the improvement in comprehensive behavioral scores from the first to the second scenario reached statistical significance in eight of the nine subjects. The researchers concluded that this simulation program creates a learning environment that readily supports the acquisition and refinement of the technical and behavioral skills that are important in solving clinically significant, potentially lifethreatening problems that can occur when patients are on ECMO.2 Similar short-term benefits ascribed to simulation have been documented in other health care disciplines.* Another example of simulation-based learning that is directly relevant to fetal and neonatal care is the Obstetrics Emergency Training Programme at Southmead Hospital, Bristol, United Kingdom. Draycott and colleagues used a variety of patient simulators in several simulated environments to show that simulation-based learning resulted in enhanced content knowledge and improved technical management of shoulder dystocia.11-14,19 In addition, the same group found in a retrospective multicenter cohort observational study of 19,460 infants that (1) the incidence of infants born with 5-minute Apgar scores of 6 or lower decreased from 86.6 to 44.6 per 10,000 births (P , .001), and (2) Hypoxic ischemic encephalopothy (HIE) decreased from 27.3 to 13.6 per 10,000 births (P 5 .032) over a 5-year period following the introduction of a training program consisting of a review of fetal heart rate tracings and hands-on drills in the management of shoulder dystocia, postpartum hemorrhage, eclampsia, twin delivery, breech presentation, and maternal and neonatal resuscitation. This is one of the first examples of improvement in clinically relevant outcomes in association with an educational intervention in healthcare.20 What is it about simulation-based learning programs in general and NeoSim and the Obstetrics Emergency Training

*References 5,15-18,24,25,33,37,38,45.

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine

Programme in particular that make them so effective in facilitating the acquisition and maintenance of cognitive, technical, and behavioral skills? Each element of these programs, including the scenarios, is built around specific learning objectives that are tailored to meet the needs of the learners. Every effort is made to ensure that the scenarios to which the learners are exposed are realistic in detail, challenging in scope, and relevant to their practice. Finally, expert feedback and facilitated debriefings assist the learners in their critical self-reflection on performance. Most important, the success of NeoSim and the Obstetrics Emergency Training Programme illustrates clearly that it is the methodology, not the technology, that determines the success or failure of any simulation-based learning program. Effective simulation is not dependent on the purchase and use of highly complex and expensive patient simulators costing tens or even hundreds of thousands of dollars; more important are carefully designed scenarios that align with the needs of the trainees, provision of important (not necessarily all) cues, and conduct of skillfully led debriefings. Although there are no prospective, randomized, sufficiently powered trials comparing the effect of simulation-based learning with traditional training methodologies on the acquisition by health care professionals of the content knowledge, technical skills, and behavioral skills required for effective and safe resuscitation, almost all studies to date yield evidence that indicates benefit of some kind. Therefore, it seems reasonable to recommend the use of simulation-based learning in view of the positive studies to date and its high level of face validity based on decades of adult learning theory and extensive practical experience in other high-risk domains while awaiting more definitive studies of its effects on health care professionals and the patients for whom they care.

CHALLENGES INHERENT IN USING SIMULATION Despite early successes in simulating care of the neonate, a number of issues continue to challenge those involved in this field. Identifying and replicating key visual, auditory, and tactile cues (and avoiding wasted efforts to simulate unimportant characteristics) allow learners to respond in a realistic fashion during a scenario and therefore be able to better appreciate their individual strengths and weaknesses. These cues come from the patient simulators, people, and equipment in the simulated environment. It is not possible to use standardized patients to simulate critically ill newborns; therefore, realistic neonatal patient simulators are required to achieve many of the learning objectives important in neonatal-perinatal medicine. For neonatal patient simulators to become more realistic and cost-effective, their development needs to be guided by health care professionals who are experienced in the care of newborns and who possess an understanding of simulation as a learning methodology. Appropriate metrics capable of assessing all three skill sets must be developed and validated if simulation is to be used for formal evaluation of any type, especially high-stakes evaluations such as state licensure or board certification. Although the objective assessment of content knowledge has long been achieved through written and oral responses to

95

multiple-choice and open-ended questions, the evaluation of technical and behavioral skills remains poorly defined. For example, although the technical skill of intubation can be broken down into a number of discrete steps and each step evaluated in great detail, in the end what is most clinically relevant is whether the patient is intubated successfully and safely, not the summary of the scores on each individual step of the procedure. The assessment of behavioral skills presents similar issues. How does one define and score skills such as leadership and communication? An evidence-based approach to skill assessment will require collaboration with professionals in fields such as statistics, human factors, psychology, and others. Ultimately calculation of the return on investment of simulation-based programs is necessary to document when simulation offers not just an optimal learning experience but also a tangible benefit in terms of improved patient outcomes, enhanced patient safety, and lowered costs. Professionals in quality improvement and risk management can serve as useful resources in determining priorities for training and in providing objective data to assess the outcomes of any intervention. Any improvements over historically accepted training models should be weighed against the costs averaged over time; this will require working with colleagues with expertise in health care finance. The current use of simulation in health care barely taps its full potential. Simulation-based learning has been employed primarily in preparing relative novices (students) to assume the responsibility of patient care; learning opportunities for experienced professionals functioning in the context of multidisciplinary teams are far less common. Although any training activity should concentrate primarily on successful achievement of learning objectives, too often in health care simulation the learning has been dictated by the specific technological features of the patient simulators in use. Another somewhat disturbing trend has been excessive emphasis on the emotional reaction of trainees to scenarios and debriefings (“How did you feel about that?”) rather than on the substance of their performance and its implications for patient care.

THE FUTURE OF SIMULATION IN NEONATAL-PERINATAL MEDICINE Since 1987, the NRP of the AAP has set a national standard and an international example for training in the resuscitation of the newborn and has enjoyed tremendous success by claiming more than 2,200,000 trainees and more than 27,000 instructors in the United States alone. The NRP’s Textbook of Neonatal Resuscitation has been translated into 25 languages, and NRP has been taught in 124 different countries around the world. Development of a career-long learning program in neonatal resuscitation that is relevant to professionals from multiple disciplines at all levels of experience and is embedded with robust learning opportunities and valid performance metrics is the ongoing focus of the NRP as it continually adapts to stay relevant and provide optimal learning experiences.27 A major shift in learning methodology is currently happening within the NRP. The NRP Steering Committee established

96

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

the Instructor Development Task Force in 2007 to develop appropriate training materials for instructors as their role shifts from that of a teacher responsible for imparting knowledge to trainees to that of a facilitator who fosters acquisition of skills by learners as they accept primary responsibility for their own education. The Steering Committee also developed a list of the characteristics desired in a cost-effective human neonatal patient simulator and published this online in 2005 as a request for proposals to industry. This marked the first time in the history of patient simulator development that a professional body rather than industry sought to drive development of a realistic patient simulator based on established learning objectives. With major program revisions enacted every 5 years, it is anticipated that the staged evolution of NRP into a program that fully integrates the simulation-based learning model pioneered by the Center for Advanced Pediatric and Prenatal Education (CAPE) NeoSim program will be complete by 2015. It makes sense that the neonatologists, obstetricians, and nurses from labor and delivery units and newborn nurseries who work closely together in the delivery room caring for patients should also conduct joint simulation-based learning exercises. This implies that the members of the obstetric team need tools that allow them to practice the cognitive, technical, and behavioral skills necessary for successful patient care in their domain. Human birth is characterized by nearly continuous changes in the physiology, anatomy, and spatial relationships among various physical structures in both mother and baby, and simulation of the process of labor and delivery is therefore a technically complex endeavor. In the near future, purely mechanical devices are unlikely to be able to simulate vaginal birth in a manner akin to real life in a costeffective manner nor to allow practice of highly invasive procedures such as cesarean section. However, it has been shown at CAPE and elsewhere that effective simulationbased learning can occur using tools possessing far less than 100% fidelity to actual biologic processes and structures. Development of hybrid technologies that combine materials similar to the plastics used for physical patient simulators with visual displays and haptic interfaces capable of generating the images and tactile sensations associated with patient care will create learning opportunities that are currently impossible to achieve in the absence of a real patient. Thus, combinations of physical whole body simulators with virtual reality interfaces designed to compensate for the physical simulators’ limitations will play a major role as pediatric and obstetric simulation evolves. In addition to physical and hybrid patient simulators, highly interactive web-based virtual environments will allow multiple professionals located in geographically distinct regions to participate in simulated clinical scenarios tailored to meet their specific learning needs. Close collaboration among physicians, computer scientists, biomedical engineers, medical artists, and others will allow the technical challenges of simulating birth to be overcome in a timely and cost-efficient manner.26-28,31

CONCLUSION Simulation-based learning in neonatal-perinatal medicine offers many advantages over more traditional and less interactive training methodologies. Although the field of neonatal simulation is still in its relative infancy and a number of

technical, financial, and cultural challenges must be met to realize the full potential of this powerful methodology, none of these challenges is insurmountable. The time to embrace simulation is now. Failure to do so will impede the practice of safe and effective health care and lead to harm of the smallest and most vulnerable patients.

REFERENCES 1. Anderson JM, et al: Simulating extracorporeal membrane oxygenation (ECMO) emergencies to improve human performance, Part I: methodologic and technologic innovations, Simul Healthc 1:220, 2006. 2. Anderson JM, et al: Simulating extracorporeal membrane oxygenation (ECMO) emergencies, Part II: qualitative and quantitative assessment and validation, Simul Healthc 1:228, 2006. 3. Apollo 13: The NASA mission reports. Post launch mission report, Page 81. Burlington, Ontario, Canada, 2000, Apogee Books. 4. Apollo 13 Technical Crew Debriefing: Mission Operations Branch. Flight Crew Support Division, Houston, Texas, 1970, NASA Manned Spacecraft Center, (Declassified, 1982). 5. Barsuk D, et al: Using advanced simulation for recognition and correction of gaps in airway and breathing management skills in prehospital trauma care, Anesth Analg 100:803, 2005. 6. Beaubien JM, Baker DP: The use of simulation for training teamwork skills in health care: how low can you go? Qual Saf Health Care 13(Suppl 1):i51, 2003. 7. Brennan TA, et al: Incidence of adverse events and negligence in hospitalized patients. Results of the Harvard Medical Practice Study I, N Engl J Med 324:370, 1991. 8. Chaikin A: A man on the moon. The crown of an astronaut’s career, Pages 285-336. New York, New York, 1994, Penguin Books. 9. Chamberlain D: The International Liaison Committee on Resuscitation (ILCOR)—Past and present: Compiled by the Founding Members of the International Liaison Committee on Resuscitation, Resuscitation 67(2-3):157, 2005. 10. Committee on Quality of Health Care in America: Crossing the quality chasm: a new health system for the 21st century, Washington, D.C., 2001, National Academy Press. 11. Crofts JF, et al: Change in knowledge of midwives and obstetricians following obstetric emergency training: a randomised controlled trial of local hospital, simulation centre and teamwork training, BJOG 114:1534, 2007. 12. Crofts JF, et al: Management of shoulder dystocia: skill retention 6 and 12 months after training, Obstet Gynecol 110:1069, 2007. 13. Crofts JF, et al: Training for shoulder dystocia: a trial of simulation using low-fidelity and high-fidelity mannequins, Obstet Gynecol 108:1477, 2006. 14. Crofts JF, et al: Shoulder dystocia training using a new birth training mannequin, BJOG 112:997, 2005. 15. Deering S, et al: Simulation training and resident performance of singleton vaginal breech delivery, Obstet Gynecol 107:86, 2006. 16. Deering S, et al: Improving resident competency in the management of shoulder dystocia with simulation training, Obstet Gynecol 103:1224, 2004. 17. DeVita MA, et al: Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator, Qual Saf Health Care 14:326, 2005.

Chapter 5  Quality and Safety of Neonatal Intensive Care Medicine 18. Dine CJ, et al: Improving cardiopulmonary resuscitation quality and resuscitation training by combining audiovisual feedback and debriefing, Crit Care Med 36:2817, 2008. 19. Draycott TJ, et al: Improving neonatal outcome through practical shoulder dystocia training, Obstet Gynecol 112:14, 2008. 20. Draycott TJ, et al: Does training in obstetric emergencies improve neonatal outcome? BJOG 113:177, 2006. 21. Elyea RL: Discussion of Task MSC/TRW-149.3, Transearth Injection Abort Processor Entry Range Functions. Personal communication to LPH. July 16, 2003. 22. Ericsson KA: Deliberate practice and the acquisition and maintenance of expert performance in medicine and related domains, Acad Med 79:S70, 2004. 23. Ericsson KA, et al: The role of deliberate practice in the acquisition of expert performance, Psychol Rev 100:363, 1993. 24. Falcone RA, et al: Multidisciplinary pediatric trauma team training using high-fidelity trauma simulation, J Pediatr Surg 43:1065, 2008. 25. Goffman D, et al: Improving shoulder dystocia management among resident and attending physicians using simulations, Am J Obstet Gynecol 199:294.e1, 2008. 26. Halamek LP: Improving performance, reducing error, and minimizing risk in the delivery room. In Stevenson DK, et al, editors: Fetal and neonatal brain injury: mechanisms, management, and the risks of practice, 4th ed. Cambridge, UK, 2009, Cambridge University Press. 27. Halamek LP: The genesis, adaptation, and evolution of the Neonatal Resuscitation Program, NeoReviews 9:e142, 2008. 28. Halamek LP: The simulated delivery room environment as the future modality for acquiring and maintaining skills in fetal and neonatal resuscitation, Semin Fetal Neonatal Med 13:448, 2008. 29. Halamek LP: A slippery slide into life. Agency for Healthcare Research and Quality Web M&M. Mortality and Morbidity Rounds on the Web. Spotlight Case. Obstetrics (website). http://webmm.ahrq.gov/case.aspx?caseID5112. Accessed December 1, 2008. 30. Halamek LP: Teaching versus learning and the role of simulationbased training in pediatrics, J Pediatr 151:329, 2007. 31. Halamek LP, et al: Simulation in Paediatrics. In Riley R, editor: Manual of simulation in healthcare. Cambridge, UK, 2008, Oxford University Press. 32. Halamek LP, et al: Time for a new paradigm in pediatric medical education: teaching neonatal resuscitation in a simulated delivery room environment, Pediatrics 106:e45, 2000. URL: http://www.pediatrics.org/cgi/content/full/106/4/e45. Accessed December 1, 2008.

97

33. Hall RE, et al: Human patient simulation is effective for teaching paramedic students endotracheal intubation, Acad Emerg Med 12:850, 2005. 34. The Joint Commission: Sentinel event alert: Preventing infant death and injury during delivery (website). http://www. jointcommission.org/SentinelEvents/SentinelEventAlert/ sea_30.htm. Accessed December 1, 2008. 35. Kohn LT, et al: To err is human: building a safer health system. Washington, D.C., 1999, National Academy Press. 36. Leape LL, et al: The nature of adverse events in hospitalized patients. Results of the Harvard Medical Practice Study II, N Engl J Med 324;377, 1999. 37. Mayo PH, et al: Achieving house staff competence in emergency airway management: results of a teaching program using a computerized patient simulator, Crit Care Med 32:2422, 2004. 38. Moorthy KY, et al: Surgical crisis management skills training and assessment: a simulation (corrected)-based approach to enhancing operating room performance, Ann Surg 244:139, 2006. 39. Moulert V, et al: The effects of deliberate practice in undergraduate medical education, Med Educ 38:1044, 2004. 40. Murphy AM, et al: Simulation-based training in neonatal resuscitation, NeoReviews 6:e489, 2005. 41. Murphy AM, et al: Validation of simulation-based training in critical care: use of heart rate variability as a marker for mental workload. In Patankar MS, editor: Proceedings of the first safety across high-consequence industries conference [book on CD-ROM]. St. Louis, MO, 2004, Saint Louis University Press (pp 157–160). 42. Salas E, et al: Team training in the skies: does Crew Resource Management (CRM) training work? Hum Factors 43:641, 2001. 43. Thomas EJ, et al: Incidence and types of adverse events and negligent care in Utah and Colorado, Med Care 38; 261, 2000. 44. Vento M, et al: Using intensive care technology in the delivery room: a new concept for the resuscitation of extremely preterm neonates, Pediatrics 122:1113, 2008. 45. Wayne DB, et al: Simulation-based training of Internal Medicine residents in Advanced Cardiac Life Support protocols: a randomized trial, Teach Learn Med 17:202, 2005. 46. Williams AM, et al: Perceptual-cognitive expertise in sport: some considerations when applying the expert performance approach, Hum Mov Sci 24:283, 2005. 47. Ziv A, et al: Simulation-based medical education: an ethical imperative, Simul Healthc 1:252, 2006.

CHAPTER

6

Practicing Evidence-Based Neonatal-Perinatal Medicine Suzanne M. Lopez, Kathleen A. Kennedy, and Jon E. Tyson

This chapter focuses on five key processes in practicing evidence-based neonatal-perinatal medicine: (1) asking a focused clinical question; (2) searching MEDLINE, the Cochrane Library, and other sources for high-quality evidence (primary reports and systematic reviews); (3) critically appraising the retrieved evidence for its validity; (4) extracting the data; and (5) applying the results to patient care. The role of the Cochrane Collaboration in the preparation, dissemination, and timely updating of systematic reviews of evidence from randomized clinical trials is highlighted. Strategies for promoting evidence-based clinical practice are presented. Evidence-based medicine has been described as “the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients.”31 The practice of evidence-based medicine requires efficient access to the best available evidence that is applicable to the clinical problem. It is essential, however, to make two disclaimers. First, not every clinical decision can be based on strong evidence because such evidence might not exist. Regarding a general internal medicine inpatient service in England, Ellis and coworkers11 estimated that principal treatments prescribed for patients’ primary diagnoses were based on strong evidence from randomized controlled trials (RCTs) in about 50% of cases; treatment was based on convincing non-RCT evidence in about 30% of cases; and there was no substantial evidence available in about 20% of cases. In a similar study in the neonatal intensive care unit at McMaster University Medical Centre, there was strong evidence from RCTs to support the choice of the principal prescribed treatment in 34% of cases.5 These findings were based on surveys in institutions emphasizing the practice of evidence-based medicine, and no attempt was made to identify the proportion of all

treatments prescribed that were based on strong evidence. The proportion would undoubtedly be substantially lower than 50% in these and other institutions. Many widely used therapies have not been well evaluated with respect to either effectiveness or safety.1 Second, evidence provides a necessary, but insufficient ground for clinical decisions. Clinical expertise is no less important under the evidence-based approach; an accurate history, physical examination, and clinical diagnosis are crucial to a properly directed search for evidence that is directly applicable to the patient’s problem. In addition, for some treatment decisions, it is essential to consider the values and preferences of parents with respect to the probable clinical outcomes of the treatments being considered for their infant.

ASKING A FOCUSED CLINICAL QUESTION A focused clinical question should contain the following elements: n Patients

of interest or exposure of interest n Nature of any comparisons to be made n Primary outcome of interest and other important outcomes n Treatment

The exact form of a focused clinical question depends on whether the question concerns treatment or prevention, etiology, diagnosis, or prognosis.15,41 For questions concerning treatment or prevention, a focused question has the following form: In [patient, problem, or risk factor] does [treatment A] compared with [control or treatment B] reduce [adverse outcome(s)]?

99

100

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Two examples follow: (1) In women carrying fetuses of 24 to 34 weeks’ gestation who are at risk of delivering, does corticosteroid (dexamethasone or betamethasone) compared with no treatment reduce the incidence of respiratory distress syndrome (RDS) in their infants? (2) In infants of 24 to 30 weeks’ gestational age, does prophylactic surfactant given immediately at birth in the delivery room compared with selective use of surfactant in infants who develop moderate or severe RDS reduce neonatal death or chronic lung disease? Armed with a focused clinical question based on an accurate delineation of the clinical problem, the treatment alternatives being considered, and the important clinical outcomes, a targeted search can be conducted for valid evidence that is applicable to the problem.

FINDING EVIDENCE Sources of Evidence Clinical evidence that is relevant to problems in neonatalperinatal medicine is appearing at an accelerating rate and can be found in numerous journals, books, conference proceedings, and other sources. Many published reports provide only weak evidence because strong research designs were not used. Evidence-based recommendations are constantly changing as new evidence becomes available. The challenge for a busy clinician is to be able to detect evidence that is valid, up-to-date, and applicable to the clinical problem using strategies that are comprehensive and yet efficient. These strategies may be directed to retrieving primary reports and reviews. Recent review articles might seem like an efficient source of best available evidence. Because most reviews do not use explicit review methods, however, there is a special need for the efficient retrieval of systematic reviews (discussed later). Although textbooks can provide valid evidence that is based on systematic methods of review, very few textbooks (except books that focus on evidence-based practice7,12,27,34) require contributors to use explicit and systematic methods when reviewing evidence and making treatment recommendations. There tends to be a long time gap between the appearance of new evidence and its impact on therapeutic recommendations found in textbooks.2 In neonatal-perinatal medicine and other fields in which new evidence is rapidly accumulating, it is especially important to be able to access systematic reviews that are frequently updated.

Efficient Strategies for Searching for Evidence PRIMARY REPORTS Primary reports that are relevant to neonatal-perinatal medicine are published in numerous journals. Most of these journals are indexed in MEDLINE, but additional reports may appear in journals indexed in other computerized databases, including CINAHL and EMBASE. With access to the Internet, one can now search MEDLINE for clinical evidence using PubMed; other databases maintained by the National Library of Medicine also can be accessed. PubMed can be accessed at www.ncbi.nlm.nih.gov/sites/entrez; a list of other databases supported by the National Library of Medicine can be found at www.ncbi/nlm.nih.gov/sites/ gquery. To define the topic of a search, one uses Medical Subject Headings (MeSH terms), text words, or a combination, combining them appropriately in a Boolean search with AND or OR (a medical librarian can quickly teach the logic of this). Help is also available online in the PubMed tutorial. Search terms for the patient population, the intervention, the comparison, the outcome of interest, or all of these may be included. Often the clinician may find that a MEDLINE search based only on topic descriptors yields a long list of reports that he or she does not have time to scan or read. Busy clinicians need to prune potentially cumbersome lists by incorporating into the search a strategy for limiting the retrieval to reports that are likely to be of high methodologic quality, and more likely to provide valid evidence. Such a strategy includes using methodologic filters that have been validated against hand-searching18,43 to detect articles that, depending on the type of focused question posed, have the methodologic quality attributes shown in Table 6-1. These methodologic filters are used together with topic descriptors (through the use of AND) so that only articles that are on topic in clinical terms and satisfy the methodologic criteria are retrieved. By choosing different methodologic filters, the clinician can maximize either the sensitivity (for comprehensiveness) or the specificity (for fewest methodologic falsepositive results) of his or her search. To do this, one uses PubMed’s Clinical Queries page (click on Clinical Queries on the PubMed page or access directly at www.ncbi. nlm.nih.gov/entrez/query/static/clinical.shtml), and selects Clinical Queries using Research Methodology Filters. One

TABLE 6–1  Searching MEDLINE for Sound Clinical Studies Using Methodologic Filters Type of Question

Criterion Standard for Methodologic Quality

Treatment

Random or quasi-random allocation of participants to treatment and control groups

Etiology

Formal control group using random or quasi-random allocation; nonrandomized concurrent controls; cohort analytic study with matching or statistical adjustment; or case-control study

Diagnosis

Provision of sufficient data to calculate sensitivity and specificity of the test, or likelihood ratios

Prognosis

Cohort of subjects who, at baseline, have the disease of interest, but not the outcome of interest

Table derived from Haynes RB et al: Developing optimal search strategies for detecting clinically sound studies in MEDLINE, J Am Med Inform Assoc 1:447, 1994.

Chapter 6  Practicing Evidence-Based Neonatal-Perinatal Medicine

is asked to click on the category type of the question one is asking (therapy, diagnosis, etiology, or prognosis), and on whether methodologic filters are wanted to emphasize sensitivity or specificity. If a clinician is reviewing a topic and wants to be comprehensive in retrieval of sound clinical studies, he or she would click on “sensitivity.” If the clinician has limited time and wants urgent access to perhaps only one or two reports that are likely to be methodologically sound, he or she would click on “specificity.”

REVIEWS Systematic reviews6,36 are distinguished from other types of reviews by the rigor of the review methods. The objectives and methods are explicitly planned a priori, and they are documented in the review. A review without a methods section is unlikely to be a systematic review. Systematic reviews of the results of RCTs attempt to identify all trials that have tested a defined therapy against an alternative in a defined population. Trials are included or excluded from the review on the basis of methodologic rigor (without consideration of the trial results). If the populations and the contrasting interventions are similar, the results may be summarized quantitatively by calculating a typical effect based on the results of all eligible trials. This latter step, called a meta-analysis, increases the precision of the estimates of treatment effect. A meta-analysis is not a necessary part of a systematic review, however; if there is clinical or statistical heterogeneity across trials, it may be inappropriate to calculate a typical effect. Systematic reviews can be found in MEDLINE by limiting the publication type to “MetaAnalysis” or by using PubMed’s Clinical Queries page (select Systematic Reviews). An example follows: The clinician wishes to find a systematic review, with meta-analysis, of studies of women at risk for preterm delivery that assesses the effect of antenatal corticosteroids on the incidence of RDS in their infants. Using PubMed, search terms are entered: corticosteroid AND respiratory distress syndrome. The search is limited by publication type to meta-analysis. Alternatively, topic descriptors could be entered into the Systematic Reviews section of the Clinical Queries page.

COCHRANE SYSTEMATIC REVIEWS The Cochrane Collaboration is an international organization that prepares, maintains, and disseminates up-to-date systematic reviews of health care interventions. The reviews are prepared by members of collaborative review groups, including the Pregnancy and Childbirth Review Group and the Neonatal Review Group. The reviews are published electronically in The Cochrane Library,8 which is published every 3 months and allows the reviews to be updated as new evidence appears. The reviews prepared by the Neonatal Review Group can also be found at a website maintained by the National Institute of Child Health and Human Development (www.nichd.nih.gov/cochraneneonatal/cochrane.cfm). Cochrane reviews are indexed in MEDLINE, so they can also be identified using PubMed searches (using the topic descriptor alone or limiting the search by using the topic descriptor AND Cochrane). A description of these reviews has also been published.37

101

CRITICALLY APPRAISING EVIDENCE FOR ITS VALIDITY The fundamental goal of clinical research is to obtain an unbiased answer to the question posed. Bias leads to an answer that is systematically different from the truth. Guides to assessing the validity of clinical research in the realms of therapy, etiology, diagnosis, prognosis, and reviews are available.16,17,19,21,22,29 Box 6-1 provides a simple distillation of the major methodologic issues to be considered. More comprehensive guides, with specific applicability to therapeutic studies in neonatal-perinatal medicine, have been published.4,30,42

BOX 6–1 Readers’ Guides for Appraising the Validity of Clinical Studies THERAPY Was the assignment of patients to treatments randomized? Was randomization concealed (so that the decision to enroll a patient could not be influenced by knowledge of planned group assignment)? Were all patients who entered the trial accounted for and attributed at its conclusion? Were outcomes assessed “blindly,” without knowledge of treatment group? When possible, were patients and caretakers blind to treatment? ETIOLOGY OR HARM Were there clearly defined comparison groups, similar with respect to important determinants of outcome, other than the one of interest? Were the outcomes and exposures measured in the same way in the groups being compared? Was follow-up sufficiently long and complete? Is the temporal relationship correct? DIAGNOSIS Was there an independent, blind comparison with a criterion standard? Did the patient sample include the kinds of patients to whom the diagnostic test would be applied in practice? Were the test results prevented from influencing the decision to perform the criterion standard (workup bias avoided)? Can the test be replicated on the basis of the method reported? PROGNOSIS Was there a representative, well-defined sample of patients at a uniform point in the course of the disease (inception cohort)? Was follow-up sufficiently long and complete? Were objective and unbiased outcome criteria used? Was there adjustment for important prognostic factors? Criteria from references 15, 16, 19, 21, and 22.

102

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Most studies on treatment or prevention use designs that can be classified into one of four categories, listed in order of increasing methodologic rigor: . Case series without controls 1 2. Nonrandomized studies using historical controls 3. Nonrandomized studies using concurrent controls 4. RCTs

The randomized trial is the strongest design for evaluating the effect of treatment. It offers maximum protection against selection bias that can invalidate comparisons between groups of patients. The allocation process should be truly random (not quasi-random, e.g., alternate) and blinded so that it is impervious to tampering or code breaking. In addition, follow-up should be complete, with all randomized patients being accounted for in the primary analysis, and outcome measurements should be made by observers who are blinded to the treatment allocation. When feasible, blinding of the caretakers, the patient, and the patient’s family to the treatment allocation should be accomplished. When reading reports of therapeutic studies, the clinician should scan the “Methods” section to assess validity using these criteria.

event rate. The complement of the relative risk (1 2 RR) is the relative risk reduction (RRR). Relative risk reduction is the percent reduction in risk in the treated group compared with the controls. RR of 0.75 represents 25% RRR. The risk difference (RD) or absolute risk reduction (ARR), [c/(c 1 d)] 2 [a/(a 1 b)], indicates the absolute magnitude of reduction in risk between the control group and treatment group. A risk difference of 0.05 represents an absolute 5-percentage point reduction of the event rate in the treated group. The reciprocal of the risk difference (1/RD) indicates the number of patients who must be treated to expect to prevent the event in one patient. In the example, 20 patients (1/0.05) need to be treated to prevent the event in 1 patient. The number needed to treat (NNT) is particularly relevant when deciding whether to use a treatment that is effective, but causes important clinical side effects or results in an important increase in economic costs. The patient’s expected event rate in the absence of treatment may be a crucial determinant of this decision. When outcome data are reported on a continuous scale (e.g., blood pressure measured in mm Hg), a different measure of effect, the mean difference, is computed.

APPLYING THE RESULTS TO PATIENT CARE EXTRACTING THE DATA AND EXPRESSING THE EFFECT OF TREATMENT Table 6-2 displays the structure of a typical study that assesses the effectiveness of a treatment. There are two exposure groups (labeled treated or control) and two possible outcome categories (labeled event or no event). An event is a categorical adverse outcome, such as occurrence of disease, adverse neurodevelopmental outcome, treatment side effect, or death. The effect of treatment is given by comparing the event rate in the treated and control groups, which can be accomplished using either relative or absolute treatment effect estimators. The relative risk (RR) is the ratio of risk in the treated group to the risk in the control group, [a/(a 1 b)] 4 [c/(c 1 d)]. This is a relative, but not absolute, measure of reduction in the

The results of randomized trials of therapy indicate the likely effects of the therapy—beneficial and adverse—on important clinical outcomes. These effects are average effects in the patients who are entered in the trials, however, and they may or may not accurately predict the net benefit to be expected in specific subgroups or in individual patients.10,13 This problem becomes especially important when a treatment has been shown to produce benefits and harm. Often, patients at high risk of the primary outcome are more likely than patients at low risk to benefit from an effective therapy. Patients at high risk and patients at low risk who receive a treatment are exposed to the adverse side effects of that treatment, however; the balance between likely benefits and harm can shift. This problem is compounded because individual patients may place different values on the relative importance of benefits and harm caused by treatment.

TABLE 6–2  Structure of a Study to Assess the Effect of a Treatment and Measures of Treatment Effect

OUTCOME Event

No Event

Treated

a

b

Control

c

d

Exposure

Treatment Effect Measures

Relative risk (RR)

a/(a 1 b) 4 c/(c 1 d)

Relative risk reduction (RRR)

1 2 RR

Odds ratio

ad/bc

Risk difference (RD)

c/(c 1 d) 2 a/(a 1 b)

Number needed to treat (NNT)

1/RD

103

Chapter 6  Practicing Evidence-Based Neonatal-Perinatal Medicine

PROMOTING EVIDENCE-BASED CLINICAL PRACTICE The evidence-based practice paradigm places responsibility on the physician to develop and maintain the skills needed to find relevant evidence efficiently, appraise it critically for its validity, and apply it to the clinical problem. The development of these skills should begin in undergraduate medical education, and

Gestational age (wk) 36 35 34 33 32 31

30

29

28

50

40 Number needed to treat

In deciding whether to use an effective therapy in an individual patient, particularly when the therapy results in important clinical side effects, one must consider the relative likelihood that the therapy would actually prevent the adverse target event, or cause adverse side effects, in that individual patient. One way of approaching this decision is to determine whether the report of the relevant trial or systematic review of trials described the size of risk reduction for the primary outcome, and risk increases for any side effects caused, according to patient subgroups defined by patient characteristics at entry. If so, it may be possible to derive a relative likelihood of being helped or harmed from the subgroup most similar to the individual patient. More often, however, either RRR across subgroups is not clearly different, or the clinician cannot judge this because data by subgroups are not presented, or the groups are too small for meaningful comparisons. Assuming that RRR is constant across the range of baseline risk, one can calculate a patient-specific absolute risk reduction (ARR) using the formula ARR 5 RRR 3 PEER, where PEER is the patient’s expected event rate in the absence of treatment. In neonatal-perinatal medicine, such information may often be available from estimates of risk based on gestational age, birthweight, and postnatal age. Using this approach, ARR (and its inverse, NNT) would be shown to vary with PEER. In patients at high risk, ARR would be high and NNT would be low, whereas in patients at low risk, the reverse would be true.14,17,35 An example of this form of analysis is shown in Figure 6-1. A systematic review of randomized trials of antenatal corticosteroid for the prevention of RDS in infants of mothers at risk of delivering prematurely showed that this therapy was effective in reducing the incidence of RDS, with RRR of 41%.9 RRR was fairly constant across subgroups based on gestational age. Because the expected risk for RDS is high at short gestation, but decre­ases markedly with increasing gestation, NNT to prevent one case of RDS is low when gestation is less than 30 weeks, but it increases sharply after 34 weeks. Although the trials did not show short-term toxic effects, few of them undertook the assessment of long-term effects. Given the uncertain balance at gestation periods beyond 34 weeks between the small likelihood of short-term benefits and the undocumented but not well-studied possibility of long-term risks, the National Institutes of Health (NIH) Consensus Conference on prenatal corticosteroids recommended that women carrying fetuses of 34 weeks’ gestation or less who threaten to deliver prematurely be considered candidates for steroid treatment (see Chapters 17 and 44).28 Other examples and a more detailed discussion of how to balance the risks and benefits in prescribing therapies for individual patients have been published.39

30

20

10

20

40

60

80

Baseline risk (%)

Figure 6-1.  Number of fetuses who must be treated (numbered needed to treat [NNT]) with antenatal corticosteroid to prevent one case of respiratory distress syndrome (RDS), as a function of baseline risk. The NNT is derived from the typical relative risk reduction of 41% calculated from the data of the trials included in the systematic review of Crowley.9 The shaded zone indicates the 95% confidence interval. As the ges­ tational age increases, baseline risk for RDS decreases, and NNT increases.

there is evidence that teaching critical appraisal skills can be incorporated successfully into clinical clerkships.3 A useful tool for acquiring these skills and enabling evidence-based care is the critically appraised topic,33 which comprises asking a focused question about a patient, finding and appraising relevant articles quickly, and synthesizing the evidence into a 1- or 2-page summary. Attainment of the required skills by individual practitioners poses a challenge. A survey of English general practitioners25 revealed that although they were favorably disposed to the concept of evidence-based clinical practice, they believed they lacked the necessary knowledge and skills to carry it forward. Only 16% had formal training in searching for evidence, and only about 33% believed they had sufficient understanding of key terms, such as relative risk and number needed to treat, that they could explain them to others. Most believed that the best way to promote evidence-based practice was the introduction of evidence-based practice guidelines or protocols.38 The preparation and dissemination of practice guidelines or consensus recommendations does not ensure their use in practice.26 Several strategies aimed at promoting behavior change among clinicians have been tested, and some have been found successful. Two of these strategies—introducing guidelines through opinion leaders and providing audit and

104

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

feedback—were included in an intervention package designed to encourage the use of antenatal corticosteroids in eligible women, in accordance with the NIH Consensus recommendations.28 In a randomized trial in which this package was compared with a control intervention consisting of the usual dissemination of recommendations, the evidence-based use of antenatal corticosteroids was found to be increased in the experimental group.23,24 Randomized trials are being employed not only to test the effectiveness of new therapies, but also to evaluate competing strategies for promoting the use of evidence in clinical practice.24,38 There is also a need for the further development of practical methods for measuring the importance or value that patients, caregivers, and the lay public attach to clinical outcomes.40 In neonatal-perinatal decision making, these values are usually sought from the parents, by using informal and unsystematic approaches. More formal and systematic methods for measuring preferences,20 including rating scales and the standard gamble, were used by teenage survivors of extremely low birthweight and their parents in a study of the quality of life of the adolescents.32 Such methods for rating health outcomes could be used increasingly in the future to assess the importance that parents attach to the probable outcomes of neonatal-perinatal treatment alternatives and to help guide evidence-based decision making for individual patients or patients in specific risk categories.39

REFERENCES 1. Ambalavanan N, Whyte RK: The mismatch between evidence and practice: common therapies in search of evidence, Clin Perinatol 30:305, 2003. 2. Antman EM et al: A comparison of results of meta-analyses of randomized control trials and recommendations of clinical experts, JAMA 268:240, 1992. 3. Bennett KJ et al: A controlled trial of teaching critical appraisal of the clinical literature to medical students, JAMA 257:2451, 1987. 4. Brok J et al: Agreement between Cochrane Neonatal reviews and clinical practice guidelines for newborns in Denmark: a cross sectional study, Arch Dis Fetal Neonatal Ed 93:F225, 2008. 5. Cairns PA et al: Is neonatal care evidence-based [abstract]? Pediatr Res 43:168A, 1998. 6. Chalmers I, Altman DG, editors: Systematic reviews, London, BMJ Publishing Group, 1995. 7. Chalmers I et al: Effective care in pregnancy and childbirth, Oxford, Oxford University Press, 1989. 8. Cochrane Library (website). http://www3.interscience.wiley.com/ cgi-bin/mrwhome/106568753/HOME. Accessed January 2010. 9. Crowley P: Antenatal corticosteroid therapy: a meta-analysis of the randomized trials, Am J Obstet Gynecol 173:322, 1995. 10. Dans AL et al: Users’ guides to the medical literature, XIV: how to decide on the applicability of clinical trial results to your patient, JAMA 279:545, 1998. 11. Ellis J et al: In-patient general medicine is evidence-based, Lancet 346:407, 1995. 12. Enkin M et al: A guide to effective care in pregnancy and childbirth, 3rd ed, Oxford, Oxford University Press, 2000. 13. Glasziou P et al: Applying the results of trials and systematic reviews to individual patients, ACP J Club 129:A15, 1998.

14. Glasziou P, Irwig LM: An evidence based approach to individualizing treatment, BMJ 311:1356, 1995. 15. Guyatt G, Rennie D: Users’ guides to the medical literature: a manual for evidence-based clinical practice, 2nd ed. Chicago, AMA Press, 2008. 16. Guyatt GH et al: Users’ guides to the medical literature, II: how to use an article about therapy or prevention. A. Are the results of the study valid? JAMA 270:2598, 1993. 17. Guyatt GH et al: Users’ guides to the medical literature, IX: a method for grading health care recommendations, JAMA 274:1800, 1995. 18. Haynes RB et al. Developing optimal search strategies for detecting clinically sound studies in MEDLINE, J Am Med Inform Assoc 1:447, 1994. 19. Jaeschke R et al: Users’ guides to the medical literature, III: how to use an article about a diagnostic test. A. Are the results of the study valid? JAMA 271:389, 1994. 20. Kaplan RM et al: Methods for assessing relative importance in preference based outcome measures, Qual Life Res 2:467, 1993. 21. Laupacis A et al: Users’ guides to the medical literature, V: how to use an article about prognosis, JAMA 272:234, 1994. 22. Levine M et al: Users’ guides to the medical literature, IV: how to use an article about harm, JAMA 271:1615, 1994. 23. Leviton LC et al: Methods to encourage the use of antenatal corticosteroid therapy for fetal maturation: a randomized controlled trial, JAMA 281:46, 1999. 24. Leviton LC, Orleans CT: Promoting the uptake of evidence in clinical practice: a prescription for action, Clin Perinatol 30:403, 2003. 25. McColl A et al: General practitioners’ perceptions of the route to evidence-based medicine: a questionnaire survey, BMJ 316:361, 1998. 26. McKinlay RJ et al: Systematic reviews and original articles differ in relevance, novelty and use in an evidence based service for physicians: PLUS project, J Clin Epidemiol 61:449, 2008. 27. Moyer VA, Elliott EJ: Evidence-based pediatrics and child health, London, BMJ Books, 2004. 28. NIH Consensus Development Panel: Effect of corticosteroids for fetal maturation on perinatal outcomes, JAMA 273:413, 1995. 29. Oxman AD et al: Users’ guides to the medical literature, VI: how to use an overview, JAMA 272:1367, 1994. 30. Reisch JS et al: Aid to the evaluation of therapeutic studies, Pediatrics 84:815, 1989. 31. Sackett DL et al: Evidence-based medicine: what it is and what it isn’t, BMJ 312:71, 1996. 32. Saigal S et al: Self-perceived health status and health-related quality of life of extremely low-birth-weight infants at adolescence, JAMA 276:453, 1996. 33. Sauve S et al: The critically appraised topic: a practical approach to learning critical appraisal, Ann R Coll Physicians Surg Can 28:396, 1995. 34. Sinclair JC, Bracken MB: Effective care of the newborn infant, Oxford, Oxford University Press, 1992. 35. Sinclair JC et al. When should an effective treatment be used? Derivation of the threshold number needed to treat and the minimum event rate for treatment, J Clin Epidemiol 54:253, 2001. 36. Sinclair JC et al: Introduction to neonatal systematic reviews, Pediatrics 100:892, 1997.

Chapter 6  Practicing Evidence-Based Neonatal-Perinatal Medicine 37. Sinclair JC et al: Cochrane neonatal systematic reviews: a survey of the evidence for neonatal therapies, Clin Perinatol 30:285, 2003. 38. Sinclair JC: Evidence-based therapy in neonatology: distilling the evidence and applying it in practice, Acta Paediatr 93:1146, 2004. 39. Sinclair JC: Weighing risks and benefits in treating the individual patient, Clin Perinatol 30:251, 2003.

105

40. Straus SE, McAlister FA: Evidence-based medicine: past, present, and future, Ann R Coll Physicians Surg Can 32:260, 1999. 41. Straus SE, et al: How to practice and teach EBM, 3rd ed, Philadelphia, Churchill Livingstone, 2005. 42. Tyson JE, et al: An evaluation of the quality of therapeutic studies in perinatal medicine, J Pediatr 102:10, 1983. 43. Wilcznyski NL, et al: Search strategies for identifying qualitative studies in CINAHL, Qual Health Res 17:705-710, 2007.

CHAPTER

7

Perinatal and Neonatal Care in Developing Countries Dharmapuri Vidyasagar and Anil Narang

Recognizing the importance of human health as a global issue, the World Bank initiated the Global Burden of Disease Study.24 This study indicates that perinatal conditions, including infant mortality, form a significant portion (39%) of the global burden of disease. Developing countries are the major contributors to global perinatal mortality. It is one of the top 10 leading causes of death in developing countries. These findings underscore the importance of improving perinatal and neonatal care in these countries, as declared in the Millennium Development Goals (MDG) in 2000.43 Over the subsequent decade, progress has been considerable, but in 2010, great challenges remain. This chapter provides an overview of global perinatal problems and discusses some of the strategies to address the issues of global health.

GLOBAL BURDEN OF MATERNAL DEATHS It is estimated that greater than 600,000 women die annually from causes related to pregnancy and childbirth.48 Asia and Africa have the highest numbers of maternal deaths; in developed countries, only about 4000 maternal deaths occur each year. The causes of maternal death include maternal hemorrhage (25%), sepsis (15%), abortion (13%), hypertensive disorders of pregnancy (12%), and obstructed labor (8%). About 20% of women die of diseases that are aggravated by pregnancy (i.e., malnutrition, iron deficiency anemia, heart disease, and tuberculosis). Numerous women also die of human immunodeficiency virus/acquired immunodeficiency syndrome (HIV/AIDS). Approximately 50 million more women are known to have other significant complications of pregnancy. Maternal death adds another dimension to the survival of the child. The likelihood of the death of a surviving infant after maternal death is five times that of an infant with a surviving mother. Greater than 50% of the deaths related to pregnancy and childbirth are estimated to be preventable using simple, well-accepted interventions.

GLOBAL BURDEN OF NEONATAL DEATHS Figure 7-1 shows the number of annual births and deaths in the world by region. Of 129 million annual births, 60% occur in Asia, 22% occur in Africa, 9% occur in Latin America, and 6% occur in Europe. Only 3% of births occur in North America.36 Despite improvements in the reduction of infant mortality around the world, neonatal mortality continues to be a major concern. Infant mortality rate (IMR) (death before 1 year of age) comprises neonatal mortality (death at ,28 days) and postneonatal mortality (death from 28 days to 1 year). Globally, of 129 million infants born annually, 4 million die during the neonatal period, and 8 million die before 1 year of age. Most of the global burden of IMR comes from the least developed and developing countries. Globally, 90% of births occur in developing countries (e.g., countries of Africa, Asia, and South America, and Mexico). Whereas only 10% of births occur in the developed world, 98% of all neonatal deaths occur in developing countries.48

ALMA-ATA DECLARATION AND OTHER GLOBAL HEALTH INITIATIVES Because of concerns of high IMR across the globe, various global interventions have been proposed and implemented by world organizations over the past three decades (Table 7-1). The Alma-Ata Declaration of 197812 by the World Health Organization (WHO) and the United Nations Children’s Fund (formerly the United Nations International Children’s Emergency Fund [UNICEF]) set a goal to achieve “Health for all by the year 2000.” The declaration included a goal for the reduction of maternal and infant mortality through primary care. In 1987, the International Conference on Safe Motherhood drew attention to high maternal mortality rates (MMR) in developing countries and encouraged policymakers to

107

108

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE GLOBAL ANNUAL BIRTHS North America 3% Latin America 9%

Oceania 0% Africa 22%

Europe 6%

Asia 60%

A

ANNUAL NEONATAL DEATHS North America Oceania 0% 0% Latin America/Caribbean 5% Europe 1%

B

Africa 30%

Asia 64%

Figure 7–1.  ​A, Global annual births; B, global burden of neo-

natal deaths. develop new strategies to reduce maternal mortality. This action led to the worldwide program of Safe Motherhood, placing an emphasis on antenatal care. A similar resolution, the Bamako Initiative, was instituted at the Annual Meeting of the African Ministries of Health in 1987. The objective was the achievement of universal maternal-child health coverage at the periphery by 2000. In 1988, the Task Force for Child Survival set objectives for a reduction in MMR and IMR.42 One of these objectives was a worldwide reduction in mortality by at least half, or by 50 to 70 per 1000 live births, whichever can be achieved first, for children 5 years old and younger. Another objective was the reduction of MMR worldwide by at least half. In 1994, WHO developed the Mother-Baby Package to help reduce MMR and the neonatal mortality rate (NMR) further. Recognizing that MMR remained high, in 1999 WHO reemphasized the need for an acceleration of national and international efforts to decrease MMR and perinatal mortality rate through the Safe Motherhood program. At the 48th World Assembly of WHO, the concept of integrated management of childhood illness was adopted to improve the well-being of the whole child younger than 5 years old. This concept included an emphasis on improved care of newborns. In 2000, the MDGs for health

were pledged by participating countries to ensure a two-thirds reduction of child mortality by 2015 (Box 7-1).43

INFANT MORTALITY RATE: A REGIONAL PERSPECTIVE There are wide variations in IMR among different regions of the world. Figure 7-2 shows global trends in IMR over the last five decades (1950-2000). Overall, there is a decreasing trend in IMR among all regions of the world.26 In 1950, IMR was lowest in industrialized Western countries (33 per 1000); this decreased to 5 per 1000 by 2000, a reduction of 85%. SubSaharan Africa had the highest IMR (157 per 1000) in 1950 and has experienced the least reduction (only 32%) among all the regions of the world during the last 50 years. South Asia, the Middle East/North Africa, and Latin America/the Caribbean have experienced 53%, 70%, and 74% decreases. IMR continues to remain significantly high, however: 70 per 1000 in South Asia, 45 per 1000 in the Middle East, and 27 per 1000 in Latin America. Selected countries with high IMR from each region are reviewed later. Another observation is that in 1950, IMR was high among all countries with great divergence among the different regions, with highest IMRs seen in sub-Saharan countries, followed by South Asia, and lowest IMRs being seen in developed countries. Over the next five decades, IMR steadily decreased in all regions showing a convergence toward a lower overall IMR in 2000. IMR continues to remain high in sub-Saharan countries and South Asia. The persistence of high IMRs in these two regions is a major global concern.

MILLENNIUM DEVELOPMENT GOALS As we fast approach the targeted date of 2015, the progress of different countries in meeting their MDG targets is being monitored closely. Table 7-2 shows the status of each region in regard to attainment of MDGs 4 and 5 as of 2006. Of the 68 priority countries analyzed, 16 were on track to meet MDG 4.8 Of these 16 countries, 7 had been on track in 2005 when the Countdown initiative was launched, 3 (including China) moved into the on-track category in 2008, and 6 were included in the Countdown process for the first time in 2008. Data for postnatal care either were unavailable or showed poor coverage in almost all 68 countries. The most rapid increases in coverage were seen for immunization, which also received significant investment during this period. There is a concern that many developing countries would not meet the target in MDGs 3, 4, and 5, the reduction of IMR and MMR, by 2015.

MAJOR CAUSES OF GLOBAL NEONATAL MORTALITY Factors that influence NMR and MMR include biologic factors, the commonly known maternal and neonatal medical problems; nonbiologic factors, such as socioeconomic conditions, maternal literacy, poverty levels, and gender equity, are equally important. Nonbiologic factors directly or indirectly affect the health access and health care of the mother and the infant. Some of these important factors are discussed.

Chapter 7  Perinatal and Neonatal Care in Developing Countries

109

TABLE 7–1  Global Initiatives to Decrease Neonatal, Child, and Maternal Mortality Initiative

Organization

Year and Occasion

Goals

Declaration of Alma-Ata

WHO, UNICEF

Sept 1978, USSR, International Conference on Primary Health Care

Health for all by 2000

Safe Motherhood Initiative

WHO, UNICEF, UNFPA, World Bank, and others

1987, Nairobi, International Safe Motherhood Conference

Reduce maternal mortality to half the present rate by 2000

Bamako Initiative

UNICEF, WHO

Sept 1987, Bamako, Mali, Annual Meeting of African Ministers of Health

Achieve universal maternal and child health coverage at the peripheral level by 2000 Revitalize peripheral public health systems Supply basic drugs Establish revolving funds Involve communities in health care

Task Force for Child Survival

WHO, UNICEF, World Bank, UNDP, Rockefeller Foundation

March 1998, Talloires, France, Protecting the World’s Children, an Agenda for the 1990s

Global eradication of polio Virtual elimination of neonatal tetanus, 90% reduction in cases of measles and 95% reduction in its fatalities, 25% reduction in fatalities owing to ARI Reduction of IMR and MMR by half or 50-70/1000, whichever is greater Reduction of MMR by at least half

Mother-Baby Package

WHO

Sept 1994, Cairo, International Conference on Population and Development

Reduce maternal mortality to half of 1990 levels by 2000

Reduce perinatal and neonatal mortality from 1990 levels by 30%-40% and improve newborn health Making Pregnancy Safer

WHO

1999, Safe Motherhood Initiative

Accelerate reduction of high maternal and perinatal mortality and morbidity by refocusing WHO strategies in national and international health sectors

Millennium Developmental Goals

WHO

2000-2015

Reduce child mortality Reduce by two thirds the mortality rate for children ,5 years old between 1990-2015

Newborn Care Training

NIH

2010

Essential newborn care training of community-based birth attendants has reduced the stillbirth rate.*

*From Carlo WA et al: Newborn-care training and perinatal mortality in developing countries, N Engl J Med 362:614, 2010. ARI, acute respiratory infection; IMR, infant mortality rate; MMR, maternal mortality rate; UNDP, United Nations Development Program; UNFPA, United Nations Fund for Population Activities; UNICEF, United Nations Children’s Fund (formerly United Nations International Children’s Emergency Fund; WHO, World Health Organization.

110

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

BOX 7–1 Millennium Development Goals n Goal

1: Eradicate extreme poverty and hunger 2: Achieve universal primary education n Goal 3: Promote gender equality and empower women n Goal 4: Reduce child mortality; reduce U5MR by two thirds between 1990 and 2015 n Goal 5: Improve maternal health; reduce maternal mortality rate by three fourths between 1990 and 2015 n Goal 6: Combat HIV/AIDS, malaria, and other diseases n Goal 7: Ensure environmental sustainability n Goal 8: Develop a Global Partnership for Development n Goal

Note: The Millennium Development Goals (MDGs) are eight goals adopted at the United Nations Millennium Summit held in September 2000. These goals contain 21 quantifiable targets that are measured by 60 indicators. These targets are to be achieved by 2015. The Millennium Declaration was adopted by 189 nations and signed by 147 heads of state and governments. MDG 4 presented in Table 7-2 relates to improvements in child health.43 U5MR, under-five mortality rate.

access to a toilet. NMR and IMR are higher among populations who do not have access to a flush or pit toilet (Fig. 7-4). Similarly, the lack of access to piped (i.e., running) water has an adverse effect on IMR.

PHENOMENON OF HIGH INFANT MORTALITY RATE IN RESOURCE-RICH COUNTRIES Resource-rich countries that depend on oil and minerals as a sizable part of their economy constitute a unique group because of their wide disparities of wealth and health.7 Countries in this group are mostly in the Middle East, Africa, and Latin America. Most of them, despite high per capita gross national products, do not do well in terms of health status. They have high IMRs, high mortality rates for children 5 years old and younger, and low life expectancy. Although the overall IMR has decreased from 200 per 1000 in the 1950s to 50 per 1000 in 2000, in some countries IMRs are far higher than the countries in Latin America and East Asia.34 This disparity is aptly referred to as “the oil curse.”7

INFLUENCE OF CULTURAL BELIEFS

Nonbiologic Factors Influencing Neonatal Mortality Biologic factors play a pivotal role in NMR and IMR of developing countries. The nonbiologic factors influencing IMR include social, economic, and environmental factors. The status of women, literacy in women, and women’s health policies also play an important role. In India, although overall literacy is low (53%), it is improving at accelerated rates in females more than in males in the age group of 15 to 49 years. IMR is higher in women who are illiterate (Fig. 7-3A). Severe gender bias practices beyond the neonatal period are reflected in the mortality rate, with the mortality rate for girls exceeding that of boys (see Fig. 7-3B). Gender discrimination is not unique to a particular region of the world. Concerns have been expressed regarding this problem in South American countries.28 Gender-biased discrimination begins to manifest early in life, leading to poorer nutrition and health in girls compared with boys. Other nonbiologic factors include housing, access to potable water, and

Care during pregnancy, childbirth, and the neonatal period worldwide is greatly influenced by regional, religious, ethnic, and cultural beliefs. Some of these practices may be helpful or harmless; however, some of them may be harmful. Educational programs should focus on the harmful effects of traditional practices. Because of the increasing migration of different ethnic groups to the Western world, all practicing neonatologists should develop awareness and cultural sensitivity to serve their patients better. Understanding cultural beliefs and their impact on the care of the mother and infant is crucial to the development of programs to reduce MMR and IMR. Practicing neonatologists should be familiar with the cultural practices of the people they serve.37

WAR CONFLICTS AND INFANT MORTALITY Although war inflicts heavy human casualties, little is known about the impact on perinatal mortality. Modern wars have inflicted major injury to civilians including children of all ages. War influences adversely perinatal and neonatal outcome directly and indirectly. A study from Bosnia30 shows startling

Figure 7–2.  ​Trends in six regions of the

Infant mortality/1000 live births

200

world in infant mortality rate (IMR), 1950-2005.26, 42a Note the steady decrease in IMR in all regions. Note also that in 1950 there was a wide variation in IMR among the different regions (divergence). In 2005, despite differences, there is a convergence of IMR among regions. See text for details.26

160 120 80 40 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 –55 –60 –65 –70 –75 –80 –85 –90 –95 –2000 –2005 Period World Africa Asia Europe

Latin America and the Caribbean North America Oceania

Chapter 7  Perinatal and Neonatal Care in Developing Countries

111

TABLE 7–2  Progress Report of MDG 4* in Different Regions of the World as of 2006 U5MR

AARR (%)

2006

Sub-Saharan Africa

187

160

1

Eastern/Southern Africa

165

131

1.4

West/Central Africa

208

186

0.7

79

46

3.4

6.2

Insufficient progress

Middle East/North Africa South Asia

Observed 1990-2006

Required 2007-2015

Progress Toward MDG Target

1990

10.5

Insufficient progress

9.6

Insufficient progress

11

No progress

123

83

2.5

7.8

Insufficient progress

East Asia/Pacific

55

29

4

5.1

On track

Latin America/Caribbean

55

27

4.4

4.3

On track

CEE/CIS

53

27

4.2

4.7

On track

Industrialized countries Developing countries World

10

6

3.2

6.6

On track

103

79

1.7

9.3

Insufficient progress

93

72

1.6

9.4

Insufficient progress

Note: The first two columns compare under-five mortality rates (U5MRs) among regions of the world in 1990 and 2006. The next two columns show the observed average annual rate of reduction (AARR in %) in U5MRs between 1990 and 2006, and the required improvement necessary to reach the targeted goals by 2015. The last column shows the progress made in each global region. Sub-Saharan Africa, South Asia, and India in particular have made insufficient progress compared with other regions of the world except sub-Saharan countries.8 *Goal 4: Reduce child mortality; reduce U5MR by two thirds between 1990 and 2015. CEE/CIS, Central Eastern Europe/Commonwealth of Independent States.

NMR AND IMR BASED ON MOTHER'S LITERACY

PERCENTAGE EXCESS FEMALE MORTALITY IN INDIA 60

80

50

Illiterate Literate

70

40

50

Percent

Deaths per 1000

60

40

30 20

40%

30

10

20

0

10

–10

19% 3% –14%

–20

0 NMR

A

IMR

Neonatal mortality

Postneonatal mortality

Infant mortality

Child mortality

B

Figure 7–3.  ​A, Nonbiologic factors influencing neonatal mortality rate (NMR) and infant mortality rate (IMR). B, Gender bias in female mortality. Child mortality rate refers to children younger than 5 years.

evidence of adverse effects on perinatal outcome. Similarly, Gulf War data showed similar effects on child mortality including a toll on neonates. War disturbs health services, including access to hospitals, and cuts off water, sanitation, and patient transport systems. Hamod and Sacy19 described the direct effect of war on neonates in a neonatal intensive care unit (NICU) in Lebanon. The major cause for increased perinatal mortality in the Bosnia study was an increase in

prematurity and increase in early neonatal mortality. In this study, the prematurity rate during war was twice as high as that before the war. Pediatricians and neonatologists serving in areas of war conflicts and those working in the armed forces should be well prepared to manage these situations. Similarly, natural disasters, such as hurricane Katrina and tsunamis, call for preparedness to care for pregnant women, parturient mothers, and neonates.27

112

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE 120

Deaths per 1000

100

Congenital malformations 10%

Access Other

Others 5% Infections 32%

80 60 40

Prematurity 24%

20 0 NMR

IMR Birth asphyxia and injury 29%

Figure 7–4.  ​Neonatal mortality rate (NMR) and infant mor-

tality rate (IMR), based on access to flush and pit toilets.

Figure 7–5.  ​Direct causes of neonatal mortality.

Biologic Causes of Neonatal Mortality In developing countries, birth asphyxia, sepsis, pneumonia, and prematurity continue to constitute a major portion of IMR (Fig. 7-5).36 WHO estimates that 40% to 60% of neonatal deaths from these causes are preventable. In a rural study, Baqui and colleagues5 studied the age of occurrence of each of the aforementioned clinical conditions and their impact on neonatal mortality (Fig. 7-6). Each of the conditions is discussed. In contrast, congenital malformations and prematurity are the dominant causes of neonatal mortality in the developed world.

BIRTH ASPHYXIA The impact of birth asphyxia–related neonatal mortality lasts during the first 2 weeks of life as shown in Figure 7-6. Birth asphyxia is noted to occur at the time of birth. In developing countries, an estimated 4 to 9 million infants per year experience birth asphyxia, and only about 1 to 2 million are resuscitated successfully.52 Annually, 1 million neonatal deaths are estimated to result from birth asphyxia. Birth asphyxia contributes

to 20% to 40% of all neonatal deaths. Many factors contribute to the high incidence of birth asphyxia in developing countries. They include the poor health of pregnant women, higher prevalence of pregnancy and labor complications, inadequate care during labor and delivery, and high rates of prematurity. The lack of proper resuscitation is another major factor contributing to asphyxia-related deaths. In addition, there is a severe shortage of skilled personnel and equipment. In recent years, great strides have been made to decrease the number of deaths from birth asphyxia. Many countries have initiated training programs to improve resuscitation skills at the grass roots level. In addition to the earlier efforts of WHO and UNICEF, the Neonatal Resuscitation Program (NRP), developed by the American Academy of Pediatrics and American Heart Association, has been adopted in its full or modified form by more than 72 countries worldwide, including India and China (see Chapter 26). In India, the NRP has become a standard skill-training module since 1990.45 A significant reduction in deaths related to birth asphyxia after

100

Cumulative mortality (%)

90 80 70 60 50 40 Tetanus Preterm birth Birth asphyxia, birth injury Sepsis or pneumonia

30 20 10 0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Age (days)

Figure 7–6.  ​Causes of neonatal deaths in rural India. Lines show cumulative

percent of deaths for each cause. See text for details.5

Chapter 7  Perinatal and Neonatal Care in Developing Countries

NRP training was reported in India.14,15 China has declared NRP to be a national priority.47 NRP is expected to have a continuing major global impact in reducing asphyxia-related deaths in years to come.

INFECTIONS Infections in the neonatal period can manifest at birth, but most often within the first few days of life. Sepsis-related deaths in the neonatal period occur at birth and peak at 2 weeks of life (see Fig. 7-6). Of the estimated 129 million children born annually worldwide, 20% die of neonatal infection. The incidence of sepsis is estimated to be 5 to 6 per 1000 live births among hospitalized patients (600,000 to 750,000). Infections in the neonatal period account for 30% to 40% of all neonatal deaths. About 1.5 to 2 million deaths related to neonatal sepsis occur each year.41 These numbers translate to 4000 to 5000 deaths every day. Meningitis accounts for 0.7 to 1 per 1000 live births (88,000 to 126,000 cases per year), and acute respiratory infections account for 800,000 deaths in neonates. Among these deaths are cases of pneumonia, bronchiolitis, and laryngotracheitis; cord infections occur in 2% to 54% of live births. Case-fatality rates are 15%. Neonatal tetanus is a preventable major acquired infection. The initial goal was to achieve universal coverage of at least 90% of pregnant women with at least two doses of tetanus toxoid by 2000; however, about 400,000 cases each year are still reported, with a fatality rate of 85% in untreated cases (370,000). The maternal transmission of HIV to the newborn is a major factor contributing to neonatal deaths in some countries, particularly sub-Saharan Africa. In 2000, 90% of the children infected with HIV were in Africa. Reported transmission rates range from 25% to 48% in developing countries.50 (See also Chapters 23 and 39.)

HYPOTHERMIA Neonatal hypothermia (see Chapter 30, Part 1) is defined as a core body temperature less than 36.5°C (,97.7°F). Hypothermia is more common in developing countries. In Ethiopia, an 8-year study showed that 67% of high-risk infants and infants with low birthweight (LBW) were hypothermic on admission. In China, the incidence of sclerema was 6.7 per 1000.47 In Nepal, more than 80% of infants became hypothermic at birth, and 50% remained hypothermic at 24 hours after birth.21 A hypothermic infant is at an increased risk of developing health problems and dying. The causes of hypothermia in developing countries are many. A lack of understanding of the importance of keeping the infant warm because of various cultural and regional practices is a major factor. Giving routine baths soon after birth without prior attention to the possibility of heat loss is one typical cause of hypothermia. Efforts are being made worldwide to increase awareness of the deleterious effects of hypothermia. Kangaroo care is becoming a universal practice. It has proven to be a very inexpensive but very effective method of providing thermal protection to the newborn in the immediate neonatal period. In addition, the immediate initiation of breast feeding provides skin-to-skin contact and early provision for calories. Newborns are also prone to heat loss during transport from home to hospital or hospital to hospital. In the absence of sophisticated and expensive incubators, simple techniques can be adopted.10 Plastic bags and wrapping with plastic

113

wrap or aluminum foil at birth are some of the recommended methods.

HEALTH CARE DELIVERY SYSTEMS: VARIOUS LEVELS OF PERINATAL HEALTH SERVICES Health is a governmental responsibility in developing countries. Maternal child health programs are under the ministries of health. Most countries have well-developed maternal child health services; generally, there are three levels of maternal child health services according to the location and level of service provided: the subcenter, primary health center, and tertiary center or district hospital (level III). Subcenters are located in rural areas and serve a population of 3000 to 5000 residents. They provide curative and preventive services; all of them have delivery rooms. Primary health centers are sometimes graded as level I or II. Level II primary health centers serve a population of about 1 million residents; they are located in rural areas. The staff typically includes a professional nurse or midwife, a health inspector, and a sanitary assistant. These centers are the “first-referral hospitals” in the rural community. Level II centers also have an operating room, basic laboratory, and delivery room. Neonatal facilities contain equipment for resuscitation. Workers in these centers are usually trained in midwifery, family planning, and prenatal care. Resuscitation at this level includes mouth to mask and bag and mask. The district hospital, a level III center, serves the local community and the region as the ultimate referral center. It provides secondary and tertiary curative services. The district hospital is staffed with physicians and a multidisciplinary team capable of providing a range of services, including those that are curative, to the mother and child. Because 65% to 75% of the deliveries in the developing world occur in rural areas and at home, the focus of discussion here is on the services provided at home and in the primary health centers.

Staffing At a district hospital, the personnel consists of different specialists, nursing staff, and other support staff, whereas at level I and II primary health centers, the physician is supported by an auxiliary nurse midwife, traditional birth attendants, and community health workers. At each health center, the physician is responsible for providing neonatal care, developing the protocols, and maintaining the quality of care. The physicians at primary health centers usually have a short course of formal training in neonatal care, with special reference to level I and II care, including the stabilization and transport of sick infants. The auxiliary nurse midwife stationed at the center is formally trained in the nursing aspects of various components of level I care. The auxiliary nurse midwife and the physician are available at the center for consultation and services. The primary health center staff is responsible for maintaining the skills and knowledge of health care workers. Educational programs are conducted by using different modalities, including pictographs and manikins. Protocols for patient management are prepared at the center.

114

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Most newborn care is provided at home by the mother and health care workers. The major role of the physician is at the district hospital. Health care workers assume an increasing role at the primary health centers, subcenter, and home. The cost of care is less at the health centers than at the hospital. There has been an increasing trend toward “institutional deliveries” (i.e., at primary health centers), with the objective of decreasing the currently high MMRs and NMRs noted with home deliveries. This trend is increasingly being adopted in China and India.

Available Equipment The facilities and services provided at the primary health centers depend on its functions for obstetric and neonatal care. Facilities for the care of some moderately sick infants may be available at some centers. The problems of maintaining the effective functioning of centers in rural areas include an uninterrupted supply of water and electricity. Each center should have an alternate back-up system in case of power failure or interruption of the supply of piped water. A refrigerator, suction machine, oxygen cylinder, flowmeter, microscope, hemoglobin meter, clock, disinfectants, soap, essential lifesaving drugs, and consumables such as syringes and needles are all components of the standard requirements for any primary health center. At level II and III centers, special equipment for the monitoring of vital signs such as heart rate, respiration, and oxygen saturation and special treatment devices such as Ambu bags, laryngoscopes, endotracheal tubes, and oxygen hoods for oxygen administration, with a supply of at least two oxygen cylinders with flowmeters, are needed. Additional necessary equipment includes intravenous pumps and phototherapy units. The investigative facilities at level II and III centers should include radiologic, hematologic, and bacteriologic services. Also important are simple investigative tools including a hemoglobin meter, glucometer, bilirubinometer for screening, microscope, and consumable supplies to carry out the tests. Most subcenters, primary health centers, and district hospitals in developing countries are far less well equipped.

Facilities for Delivery at Home Although institutional deliveries are highly recommended, the facilities for delivery of the infant at home should be improvised in regions where access to institutional deliveries is unavailable. Traditional birth attendants and community health workers usually work with families to prepare for the delivery well in advance. The family is asked to prepare a clean area in a well-lighted part of the house for delivery. Families are asked to prepare clean old sheets and clothes for use at the time of delivery. The health care professional responsible for home delivery and neonatal care carries a disposable delivery kit containing a mucus extractor, basic equipment for resuscitation (resuscitation bag with neonatal mask), a clinical thermometer, and a portable spring balance for weighing the infant (Fig. 7-7).

CARE OF THE MOTHER AND THE NEWBORN IN THE COMMUNITY The Child Survival and Safe Motherhood Program48 emp­ hasizes the primary care of the mother and newborn through the promotion of the concepts of essential obstetric care and essential care of the newborn. The components of essential obstetric care include the early detection of pregnancy, at least four antenatal checkups, the identification of high-risk pregnancies, immunization against tetanus (at least two doses), supplementation with iron and folic acid, and provision for adequate nutrition. Various home-based models of obstetric care have been tested in different countries. Although the focus of these studies was to improve maternal outcome, they also benefit neonatal outcome. The major issues are the delayed recognition of potential problems, a delay in decision making, a delay in the transport of the mother, and a delay in receiving quality care. The components of essential obstetric and newborn care are shown in Box 7-2 and Figures 7-8 to 7-13 and are discussed in detail.

Figure 7–7.  ​Infants are weighed using a color-coded spring balance. Color codes obviate

reading the exact weight by the traditional birth attendants. Weight indicator in red zone is very low birthweight; yellow zone, low birthweight; green, normal birthweight. Risk assessment is based on color code. See color insert.

Chapter 7  Perinatal and Neonatal Care in Developing Countries

BOX 7–2 Essential Components of Newborn Care . Care at birth—aseptic techniques at delivery time 1 2. Prevention and management of hypothermia— ensuring maintenance of warmth for the infant 3. Resuscitation of infant not crying—identification and referral of at-risk neonates 4. Physical examination of infant and identification of risk—identifying infants with low birthweight for home care; identifying infants with low birthweight who need referral 5. Ensuring early and successful breast feeding and breast milk feeding by spoon—ability to identify feeding problems; successfully initiating breast feeding soon after birth 6. Identifying signs of illness—providing essential newborn care 7. Grading severity of illness (see Table 7-5)

Care at Birth ASEPTIC CARE AT BIRTH Whether the delivery occurs at home or at the health center, the same principles of cleanliness must be observed. The principles of the “five cleans” (clean hands, clean blade, clean cloth, clean tie, and clean umbilical cord) are followed (Fig. 7-8). Improper cord care has been responsible for neonatal tetanus. Cord cutting with a clean blade and tying with clean thread are two important procedures of newborn care. When it is tied and cut, the cord must be left open without any dressing. It usually falls off by 5 to 10 days. The health care worker should examine the cord 2 to 4 hours after ligation for any bleeding and afterward for any discharge and redness of the skin around the base. Infants with these signs of infection should be treated with antibiotics.

115

The eyes should be cleaned at birth. If a discharge is noticed, it must be treated with antibiotic ointment. With regard to skin care, the infant is immediately cleaned and dried at birth. Bathing immediately after birth is not recommended (Fig. 7-9). The tradition of bathing should be postponed until a later age, when the infant is more stable. This process must be strictly observed in cases of premature infants and infants with LBW.

MANAGEMENT OF HYPOTHERMIA Hypothermia in the immediate newborn period is a major problem in infants delivered at home or at poorly equipped community centers. The principles of the “warm chain”51 are used to minimize heat loss soon after birth (Box 7-3). The mother is encouraged to maintain close physical contact with the infant, who is wrapped in clothing that includes a head cap and covering for hands and feet. At the health center, the auxiliary nurse midwife or nurse records axillary temperature. At home, the mother and traditional birth attendant or other health worker should be trained to assess the infant’s temperature without a thermometer by placing the dorsum of the hand on the abdomen and feet alternately and comparing the difference (Fig. 7-10). The difference in warmth between the two sites (abdomen and foot) may provide useful information. The kangaroo care method has been shown to have better outcomes in developing countries (Fig. 7-11). In a study in Nepal,21 the incidence of early hypothermia in the first 2 hours after delivery was reduced by 50% and the incidence of late hypothermia in the first 24 hours after birth was reduced by 30% by implementing one of three interventions after delivery: kangaroo care, traditional mustard oil massage under a radiant heater, or plastic swaddling. It was shown that 90% of infants in whom skin-to-skin contact with their mothers was maintained reached normal temperatures compared with only 60% of infants in incubators. Skin-to-skin contact has been shown to have other positive effects.17 It promotes mother-infant bonding, enhances lactation, leads

Figure 7–8.  ​Components of five cleans are shown: clean hands, clean blade, clean cloth, clean tie, clean umbilical cord.

116

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

A

Figure 7–9.  ​Bathing the newborn soon after birth is discour-

aged using the universal sign X as shown in this poster.

BOX 7–3 Warm Chain Steps

1. Keeping the place of birth (delivery room) warm 2. Immediately drying the infant 3. Keeping the infant warm during resuscitation 4. Maintaining skin-to-skin contact between the mother and infant 5. Initiating breast feeding 6. Postponing bathing and weighing 7. Providing appropriate clothing and bedding 8. Keeping mother and infant together 9. Keeping the infant warm during transport 10. Increasing the awareness of health care workers regarding the importance of maintaining neonatal temperature

to stimulation of the infant, and improves the psychological state of the mother. In a study on skin-to-skin contact in preterm infants, the infants not only remained clinically stable, but also showed more efficient gas exchange. At the health center, hypothermia is better managed by using the clinical skills of assessment (e.g., touch) in addition to the basic instruments (Table 7-3). Skin-to skin contact is easy to promote and implement and is the least expensive; it should be encouraged in developing countries.

Care of an Infant Not Crying at Birth The newborn infant is placed under a warming device, the mouth and nostrils of the infant are cleaned, and the infant is dried. If the infant still does not cry, gentle tactile stimulation is given on the side. If the infant does not respond, ventilation

B Figure 7–10.  ​A and B, Assessment of infant’s skin tempera-

ture using the dorsum of the hand (A) with simultaneous assessment of abdominal and foot temperatures (B). See text for details.

Figure 7–11.  ​Kangaroo care is encouraged soon after birth to

keep the newborn warm. The infant is wrapped with a blanket, and the mother holds the infant close to her body.

is initiated. Various forms of ventilation have been used: bag and mask, mouth to mask, and mouth to mouth. Self-inflating bags are commonly used. In their absence, mouth-to-mask ventilation may be carried out (Fig. 7-12). The mask must form a tight seal around the chin, and the attendant should deliver a steady rate of 30 to 40 times/min while watching the

Chapter 7  Perinatal and Neonatal Care in Developing Countries

117

TABLE 7–3  Management of Hypothermia* Category

Temperature Range (°C)

By Touch

Clinical Features

Management

36.5-37.5

Warm trunk/warm extremities

Normal infant

Cover adequately Keep next to mother

36-36.5

Warm trunk/cold extremities

Extremities bluish and cold Poor weight gain if chronic cold stress

Cover adequately Warm room or bed Skin-to-skin contact Provide warmth Consult physician

Moderate hypothermia

32-36

Cold trunk/cold extremities

Poor feeding, lethargy Fast breathing

Cover adequately Skin-to-skin contact Provide warmth Transfer to hospital Reassess every 15 min; if no improvement, provide additional heat

Severe hypothermia

,32

Extremely cold trunk/cold extremities

Lethargic, poor perfusion/ mottling Fast breathing, apnea Bleeding

Rapid rewarming until infant is 34°C, then slow rewarming Oxygen Intravenous fluids (dextrose [warm]) Consult physician and transfer to hospital

Normal Cold stress

*Range of temperatures that can be detected clinically by health workers or mother by touching trunk and feet simultaneously.

movements of the chest wall and observing the infant for an improvement in color and respiratory movements. Deficiencies in equipment (e.g., oxygen sources) and skilled personnel are major constraints that hinder adequate resuscitation. Self-inflating resuscitation bags are commercially available; they are relatively inexpensive, and physicians have been extensively trained to use them. Some paramedics have developed innovative ways of making masks from disposable plastic bottles. Oxygen is important in neonatal resuscitation, but it is expensive and a scarce

Figure 7–12.  ​Traditional birth attendants and village health

workers are trained in neonatal resuscitation using a mouth-tomask device and a manikin.

commodity in rural areas. Clinical trials have confirmed the efficacy of room air32,35 in neonatal resuscitation (see Chapter 26, Part 3). The long-term follow-up of these infants is also favorable. Although most studies were limited to term infants, resuscitation should be carried out using room air in the absence of available oxygen for the resuscitation of an asphyxiated newborn in rural areas.

Breast Feeding Exclusive breast feeding of all infants is promoted as an important part of essential newborn care. Infants delivered at an institution must be given only breast milk. To promote breast feeding worldwide, in 1991 UNICEF and WHO launched the campaign “Baby-Friendly Hospital Initiative.”49 The objective of this global initiative is to create a hospital environment that ensures breast feeding as the only means of nutrition. The principles of this program are listed in Box 7-4. Other policies that have influenced the global rates of breast feeding involve substitutes provided by the International Code of Marketing of Breast Milk. These efforts are having a major international impact. More than 14,500 hospitals around the world have been certified as “Baby Friendly.” Because some infants, particularly infants with LBW or very low birthweight (VLBW) may experience difficulty in latching on to the breast, they need special techniques to receive feeding. The use of spoons or traditional

118

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

BOX 7–4 Ten Steps to a Baby-Friendly Hospital 1. Have a written breast-feeding policy that is routinely communicated to all health care staff. 2. Train all health care staff in the skills necessary to implement this policy. 3. Inform all pregnant women about the benefits and management of breast feeding. 4. Help mothers initiate breast feeding within 1 hour of birth. 5. Show mothers how to breastfeed and maintain lactation even if they should be separated from their infants. 6. Give newborn infants no food or drink other than breast milk unless medically indicated. 7. Practice allowing mothers and infants to remain together 24 hours a day. 8. Encourage breast feeding on demand. 9. Give no artificial teats or pacifiers to breast-feeding infants. 10. Foster the establishment of breast-feeding support groups, and refer mothers to them at discharge.

feeding cups (paladai) (Fig. 7-13) in place of feeding tubes is an inexpensive way of providing preterm or sick infants with nourishment.

Primary Care by the Mother (“the Best Neonatologist”) It is often forgotten that the mother is the best natural and instinctive caregiver of the newborn. She is the best person to monitor the infant’s course. Emphasis should be placed on “health education” of the mother before and during pregnancy regarding self-care and newborn care. This approach is extremely important in rural and community settings, where skilled professionals are not readily available. Besides providing the mother with information regarding her own health, the importance of folic acid and iron intake, and vaccination against tetanus, she should be given structured education regarding normal infants and infants with LBW. The mother should be given skill training in reference to keeping the infant warm (kangaroo care), breast feeding, monitoring

growth, and immunization. In particular, mothers should be trained to recognize the danger signs of illness. They should be encouraged to seek help from health care workers in the community when the infant manifests any one or more of these signs: poor feeding, poor sucking, or crying; cold to the touch; difficulty breathing; change in color; or convulsions.

Home Visits by the Health Care Worker A suggested schedule for home visits is given in Table 7-4.33 All infants delivered at home must be seen by the health care worker within 48 hours of birth. The assessment includes all the items of essential newborn care. If the infant is doing poorly or the weight is less than 1500 g, the infant should be transferred to a health center. An infant whose birthweight is less than 2000 g must be given kangaroo care at home. In addition to routine care, exclusive breast feeding must be given. Infants with poor sucking and poor overall activity should be visited more frequently (once or twice a week). Infants whose birthweights are greater than 2000 g should be visited once every 4 weeks unless they present a problem at the time of the first visit or the mothers observe any sign of illness. After the initial stabilization, the health care worker should assess each infant carefully. The assessment is necessary to decide if the infant can be left in the care of the mother or needs transfer to a health center or from a level I center to a higher level of care. Health care workers must be trained to identify high-risk infants so that they can use simple clinical skills of observation of the infant (Table 7-5). The concept of the severity of illness and its consequences must be well understood. The grading of illness in newborns is based on activity, feeding behavior, color, temperature, breathing, and the presence of convulsions. After these observations are considered, one can determine the need for care at a health center or hospital. The following manifestations are indications for such a referral: the presence of moderate to severe illness with any birthweight, mild illness in infants weighing 1500 to 2000 g, and any weight of less than 1500 g. Any infant with a suspected major congenital malformation requiring surgery should be transferred to a district hospital. An infant who receives care at home or at a health center should be closely followed by the health care worker. The timing and frequency of visits depend on the birthweight and behavior.

Figure 7–13.  ​Infants with low

birthweight and premature infants are fed with a spoon or a paladai, a cup with a long beak.

Chapter 7  Perinatal and Neonatal Care in Developing Countries

119

TABLE 7–4  Home Visit Schedule Activities

Birthweight

Timing

Tasks

Initial visit

All birthweights

Within 48 h

Essential newborn care Transfer ,1500-g infant

Subsequent visits

.2500 g

Once every 2 wk

Essential newborn care Check for signs of illness

2000-2500 g

Every week for 4 wk

Essential newborn care Check for signs of illness

1500-2000 g

Every week for 4 wk

Kangaroo care Cup/spoon feeding Recognize signs of illness

Role of the Traditional Birth Attendant and Community Health Worker The lack of availability of a skilled attendant at the time of delivery is one of the major reasons for the increased number of maternal and infant deaths in developing countries. Community traditional birth attendants play an important role in attending deliveries at home.23,31 Focus is initially on only the mother. Later, workers in the field find the traditional birth attendant to be an important resource person to care for the newborn as well. Other workers find community health workers to be more appropriate in providing newborn care.3 There are differences between the two types of health care worker. Traditional birth attendants are illiterate, are older,

and learn by tradition; it is difficult for them to change their perceptions and practices. Conversely, community health workers are younger, literate, and adaptable to new principles and practices. Although most traditional birth attendants are women, in some parts of the world traditional birth attendants could be men. Traditional birth attendants are also known as village health workers in some regions. Their roles have been examined by numerous field studies, which validate the effectiveness of traditional birth attendants in reducing NMR and IMR. These studies have been conducted in several regions of the world.22 They show that neonatal mortality can be reduced if community-based health care workers are well trained, well equipped, and supervised. Some of the important studies are reviewed next.

TABLE 7–5  Grading Severity of Illness in a Newborn Mild

Moderate

Severe

Activity

Normal

Mild lethargy, relieved by feeding

Very lethargic, with poor response to warming

Feeding behavior

Poor feeding by spoon

By tube or IV line for #3 days

By IV line for .3 days

Color

Pink without O2

Pink with O2

Poor response to O2 or signs of shock

Jaundice

Onset after 24 h, with jaundice Onset within 24 h or jaundice over trunk and up to face up to arms and legs

Jaundice over palms and soles, or infant symptomatic

Hypothermia

Body temperature 35-36.5°C

Temperature 32-35°C

Temperature ,32°C

Breathing

Rate 60-80 breaths/min; no grunting or retractions

Rate .80 breaths/min or presence of grunting or retractions

Presence of apneic attacks or failure to respond to treatment

Seizures

Irritable only

Seizure relieved with treatment

Poor response to treatment

Low birthweight

2000-2500 g Weak sucking, color pale

1500-2000 g Weak sucking, blue, difficulty breathing, cold, or jaundiced

All ,1500 g With any of the additional symptoms, follow recommended home visit schedule in Table 7-4

IV, intravenous. Modified from Bhakoo ON et al: Facilities required for primary care (level 1) of newborn infants, J Neonatol 16:9, 2002.

120

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

HOME-BASED ANTENATAL CARE Home-based antenatal care can be provided by health care workers. The health care provider at the community level is expected to identify pregnant women in the community, ascertain the risks of pregnancy, ensure vaccination against tetanus, and administer folic acid and iron. After assessing the risk factors, health care workers also develop a plan for residential or institutional delivery. They seek assistance from the primary health center in these matters. The training of health care workers at the village level is crucial to the success of these programs. Because the educational background of village-based traditional birth attendants varies from illiterate to minimally literate, innovative techniques such as pictograms have been used to train them. Kumar and Walia23 showed that when a pictorial card is used, illiterate traditional birth attendants could be taught to keep useful maternal data records that help provide timely and important interventions. The cards illustrate the events of pregnancy, which the traditional birth attendant can record by crossing them out with a pen. In their study of homebased medical records, they are able to provide a minimum standard of care. These cards can be modified to suit local health care needs. Similar home-based records have been used in Nigeria and other countries.25 Extensive studies have been conducted to determine the effectiveness of home-based neonatal care provided by traditional birth attendants or community health workers in rural settings in India and Bangladesh. The essential components of home delivery include the preparation of a clean space, provision of clean linen, and preparation of a disposable delivery kit. The delivery kit includes soap, sterilized cord ties, a clean blade, a clean dry cloth sheet, and a gauze pad, all in one package. Health care workers are trained to observe the “five cleans” (see earlier discussion).

HOME-BASED NEONATAL CARE Raina and Kumar31 studied the impact of training traditional birth attendants in resuscitation techniques. There was a 19% reduction in perinatal mortality. Mortality was also reduced in cases where advanced techniques were used. Traditional birth attendants have a significant effect on reducing asphyxia-related neonatal mortality in the community. The pioneering work of Bang and coworkers3 in a rural setting showed that by training traditional birth attendants and community health workers to diagnose pneumonia with the use of a WHO algorithm and treat patients with a standard dose of antibiotic, the case-fatality rate for neonatal pneumonia was reduced to 0.8% compared with 23.6% among untreated patients. The IMR was 89 per 1000 compared with 121 per 1000 in the control areas, and the NMR was 67.8 per 1000 compared with 97.2 per 1000 in the control areas. The addition of traditional birth attendants to their study increased the coverage, and with supervision the traditional birth attendants made few errors. Bang and coworkers2 conducted another double-blind study on the treatment of sepsis, showing how traditional birth attendants and community health workers are trained to diagnose sepsis by using set clinical criteria. The traditional birth attendants and community health workers were instructed to treat the infant with a standard dose of antibiotics. Among

infants receiving antibiotics, there was a 58% reduction in NMR during the study, and the case-fatality rate from neonatal sepsis declined from 16.6% before the intervention to 2.8% after the intervention. Similar observations have been made in studies of neonatal resuscitation given by the traditional birth attendant or community health worker. Traditional birth attendants and community health workers are very important in health care delivery. Although community health workers are preferable in certain interventions, such as sepsis treatment, traditional birth attendants have shown that when trained they are just as useful and important for effective neonatal health care. The use of traditional birth attendants and community health workers is integral to neonatal health care in developing countries. Health care workers can use a simple algorithm based on clinical observations of a sick neonate at home or in the community to triage an infant. Figure 7-14 shows the elements of observation and action based on certain findings. The findings include common functions of the infant: feeding, crying, variations in skin temperature on touching, activity, breathing pattern, and condition of the cord. These findings can be graded from mild to severe. Any two milder forms of distress can be treated and managed at home by the health care worker and the mother. Infants showing more severe forms of distress must be treated and transferred to level II or III health centers.

STRATEGIES TO IMPROVE GLOBAL PERINATAL OUTCOME Strategies to reduce perinatal mortality in developing countries should include social and medical programs. At the social level, the emphasis should be on improving women’s health and avoiding gender bias against the female child, which can be achieved through an increase in female literacy. A related issue is female empowerment, which can be achieved through economic self-sufficiency. Female literacy is a proxy for one or more socioeconomic variables.

Social Engineering Figure 7-15 shows the interaction of education, economic power, and health. Gender bias plays a role at different phases of life. Figure 7-15 also shows how education and the economic autonomy of women engender political empowerment to overcome gender bias. Evidence from many countries shows that a woman who has more education has fewer children, better economic status, a lower fertility rate, a lower IMR, and improved quality of life. It has been shown that even in developed countries every additional year of education for women decreases mortality rates at all ages. Education increases earning potential, and income reduces mortality. Earning power provides women with economic, social, and political empowerment. In short, education is seen as a means of ensuring health. The microcredit program for women in Bangladesh has become an important model for other countries.20 Other social programs include improving household conditions, such as access to clean water and toilet facilities. Health strategies should include increasing awareness of maintaining health, improving health-seeking behavior, and improving access to health care. Finally, health

Chapter 7  Perinatal and Neonatal Care in Developing Countries

121

Neonate

Feeding well?

Stopped

No

Yes feeding

No

Sucking definitely weak or reduced

No

Warm trunk, cold extremities

No

Weak cry

Yes Temperature to touch Yes Cold trunk, cold extremities

No

Warm trunk, warm extremities Yes Baby’s cry? Yes

If any 2 criteria present then severe sepsis Treat and transfer immediately

No cry

No

Good cry Yes Mental state? Yes

Seizure

No

Alert and active

No

Unconscious/drowsy

If 2 or more criteria present then mild/moderate sepsis Treat

Yes Breathing pattern Yes Grunting or chest indrawing RR-60/min

No

Regular breathing. No chest indrawing Resp. rate 30-60/minute

No

RR-60/min after 2 counts

No

Moist and red

Yes State of umbilicus Yes Draining

No

Clean and dry Yes Normal baby

Figure 7–14.  ​Assessment and management of neonatal sepsis by traditional birth attendants and village health workers.

care professionals should focus on the preventive and curative aspects of care for the mother and child. The principles of essential obstetric and newborn care should be implemented so that 90% of the population in poor rural and urban areas can be served.

IMPROVEMENTS IN NEONATAL CARE PRACTICES We should take advantage of the highly developed scientific basis of clinical practice and use it at the grass roots level to prevent the most common causes of death. Several

evidence-based interventions and treatment modalities have reduced the NMR and IMR (Table 7-6). We should make these technologies universally available, considering them more a friend than foe in achieving these health care goals. Harnessing highly developed perinatal medicine and highly sophisticated medical and Internet technology to reach neonates has an impact on the global burden of IMR. One report13 showed that modern Internet technology was effectively used for consultations and patient management at rural primary health centers through e-mail communication with the level III health center at a medical school. It resulted in better management of mildly to moderately sick infants at primary health centers and reduced neonatal

122

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

Gender bias

Poor health status

Gender bias

Gender bias

Figure 7–15.  ​Interaction of education, health, and

wealth. Female child birth

No economic power

No social/political empowerment

Poor education Better Increased empowerment of women

transfers to the tertiary care hospitals. This approach was also found to be very cost-effective.

Relevance of Newer Technologies in Developing Countries Modern technology has been a major contributor to a reduction in neonatal mortality, especially in the developed world. Technology has been very helpful in improving the survival of infants with LBW and VLBW. Incubators were one of the earliest technological advances used in the care of premature infants. Newer and improved electronic monitoring and supportive technologies for neonatal use did not appear until the middle of the 20th century. Although the newer technologies greatly enhanced the care of critically ill neonates, they proved to be very expensive, which hindered NICUs in developing countries for a long time. Emphasis was placed on “low technology” and “appropriate technology” that was affordable and sustainable.

Many noninvasive devices developed in the 1980s to monitor vital signs are now readily available in developing countries (Table 7-7). Even though the initial cost is relatively high, the long-term benefits are many. Blood sampling and the need for a sophisticated laboratory can be avoided. The investment is cost-effective, and these devices are simple to use; health care workers can be easily trained to use them. There is clear evidence for the beneficial effect of diffusion of technology into developing countries on IMR.29 A direct relationship between technology import and reduction of IMR in developing countries has been shown (Fig. 7-16). Among monitoring devices, the oxygen saturation monitor is the most useful tool because of its simplicity in application and fast response. It provides important real-time information on heart rate and oxygenation without having to draw blood. In skilled hands, this unit can be very effective at any level for monitoring infants in distress. Pulse oximetry has been found to be an extremely useful noninvasive tool to assess status of oxygenation in developing countries. A study

TABLE 7–6  Strategies That Have Proved Effective in Prevention and Treatment of Neonatal Illnesses Strategy

Prevention

Strength of Evidence for Effectiveness

Breast feeding

Diarrhea Pneumonia Neonatal sepsis

Strong

Management of temperature

Hypothermia

Good

Clean delivery

Neonatal sepsis Neonatal tetanus

Strong

Tetanus toxoid vaccine

Neonatal tetanus

Strong

Maternal history with antenatal steroid therapy in prematurity

RDS HMD/RDS

Strong

Antibiotics for PROM

Preterm delivery and sepsis

Good

Water, sanitation, and hygiene

Diarrhea

Strong

HIV treatment

HIV

Strong

Antibiotics

Sepsis and pneumonia

Strong

Neonatal resuscitation

Prevention of birth asphyxia

Good

HIV, human immunodeficiency virus; HMD, hyaline membrane disease; PROM, premature rupture of (fetal) membranes; RDS, respiratory distress syndrome. Adapted from Gareth J et al: How many child deaths can we prevent this year? Lancet 362:65, 2003.

Chapter 7  Perinatal and Neonatal Care in Developing Countries

123

TABLE 7–7  Bedside Diagnostic and Treatment Devices Monitors

Item Description

Where Useful

O2 saturation monitor

Measures oxygenation

All levels

Noninvasive BP monitor

Measures blood pressure

All levels

Glucometer

Checks blood glucose

All levels

Bilirubinometer

Measures bilirubin levels

All levels

Bag and mask

Resuscitates patient

All levels

Nasal CPAP

Treats RDS

All levels

Fiberoptic transilluminator

Checks for pneumothorax

All levels

BP, blood pressure; CPAP, continuous positive airway pressure; RDS, respiratory distress syndrome.

MEDICAL IMPORTS AND INFANT MORTALITY 5.5 5 Infant mortality

4.5 4 3.5 3 2.5 2 1.5 –2

–1

0

1

2

3

4

5

Medical imports

Figure 7–16.  ​The relationship of technology transfer on in-

fant mortality rate (IMR) in developing countries. Scattergram shows per capita medical technology imports on the x-axis and IMR on the y-axis in 63 developing countries over a decade. There was good negative correlation (arrow) between medical imports and IMR (coefficient of correlation 20.79). (Modified from Papageorgiou C et al: International medical technology diffusion, J Internat Econom 72:409, 2007.)

from Papua New Guinea and Indonesia attests to this fact.46 The glucometer requires the use of blood from a heel stick and a chemical strip. The test can be performed at the bedside and requires consumable supplies. The transcutaneous bilirubinometer does not require any drawing of blood; no consumable supplies are required. The noninvasive blood pressure monitor requires a one-time investment. The transillumination device may be highly useful for detecting air leaks such as a pneumothorax without the need to obtain a chest x-ray. The bag and mask is the most useful device for resuscitation, providing short-term ventilation until the patient is placed on a mechanical ventilator. The use of bag-and-mask or mouth-to-mask devices in neonatal resuscitation is an example of low-cost technology that has a major impact on neonatal outcome. The Neonatal Resuscitation Program (NRP), which was introduced in India in 1989,45 has become the standard of

care in the delivery room for all newborns. There is evidence that asphyxia-related deaths have decreased since the implementation of NRP in several hospitals in India.15 An infant showing respiratory distress after birth or resuscitation needs closer observation. In the absence of laboratory facilities, the physician and nursing personnel must depend on their clinical skills to recognize and manage a critically ill newborn. Clinical scoring systems may be helpful in objectively assessing and monitoring the infant, especially at level I and II health centers. The respiratory distress syndrome (RDS) score, which was developed by Downes and associates16 (Table 7-8), correlates clinical features with acid-base and blood gas levels, with scores greater than 6 indicating the need for further evaluation and sometimes assisted ventilation. This scoring system is simple because it uses the five common clinical assessments and requires no biochemical determinations. The scores correlate well with physiologic parameters. The score provides therapeutic and prognostic guidance. This scoring system can be used in primary, secondary, and tertiary level health centers. In a more recent prospective study18 in Indonesia, the investigators evaluated the correlation between the RDS score and oxygen saturation using pulse oximetry. They found that a score of 4 or more accurately correlated with oxygen saturations (,92%), suggesting that the RDS score is a valuable clinical assessment tool to assess oxygenation in infants in respiratory distress. When an infant is found to have increasing respiratory distress, the infant is transferred to a health center of the appropriate level. Most level II and III health centers may have only oxygen and no ventilator. Simple devices such as nasal continuous positive airway pressure (CPAP) would be very valuable in managing such infants.

Use of Nasal Continuous Positive Airway Pressure The use of CPAP has proved to be a very simple intervention to support the infant in respiratory distress. CPAP technique is simple and cost-effective and can be used in developing countries. A study in Oman showed that in cases of mild to moderate disease in premature infants there was significant improvement of RDS in patients in whom nasal CPAP was used.4 At level II and III health centers in developing countries, which have limited resources, this method of treating moderate RDS is very useful.

124

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

TABLE 7–8  Clinical Scoring System* for Respiratory Distress Syndrome SCORE 0

1

2

Respiratory rate (breaths/min)

60

60-80

.80 or apneic episode

Cyanosis

None

In room air

In 40% oxygen

Retractions

None

Mild

Moderate to severe

Grunting

None

Audible without a stethoscope

Audible without a stethoscope

Air entry (crying)

Clear

Delayed or decreased

Barely audible

*The respiratory distress syndrome scoring system is useful at level II and III centers that manage infants with moderate to severe respiratory distress.

Surfactant Therapy Premature infants who do not respond to nasal CPAP require intubation and ventilatory support. Providing ventilatory support requires skilled nursing and support staff and laboratory facilities. The cost could be very high. Surfactant therapy is known to shorten the time needed for assisted ventilation. Despite the effectiveness of surfactant in improving the survival of premature infants with RDS, its use in developing countries is hindered by its high cost. The cost per dose exceeds the per capita annual income of many of the citizens in developing countries. In addition to its high cost, the use of surfactant therapy is impeded because of the lack of ventilatory support after surfactant administration. Nevertheless, many studies show that the use of surfactant could be cost-effective, even in developing countries. Verder and associates44 showed that surfactant therapy followed by nasal CPAP is an equally effective strategy for managing RDS in infants without the support of mechanical ventilation. Similarly, the use of surfactant in the treatment of preterm infants with respiratory distress in Latin America9 reduced mortality by 70% during the first few days of life, although mortality at 28 days was reduced by only 20%. These studies show that endotracheal surfactant, followed by nasal CPAP, in established cases of RDS can be used effectively. Infants who continue to require endotracheal intubation and ventilation and do not respond to nasal CPAP must be transferred to a level III tertiary care unit for alternative therapy. New technologies including surfactant therapy and mechanical ventilators are rapidly penetrating developing countries.

Establishment of Tertiary Care Units in Developing Countries The evolution of NICUs in developing countries has been understandably slow because of the high cost and lack of skilled personnel. In India, where there were only 4 or 5 NICUs in the 1980s, there were 27 accredited level II NICUs in the 1990s.45 The results of a national survey in 1987 and 1994-1995 showed that in 1995, an increasing number of centers became fully equipped. Newborn care facilities, particularly level III NICUs able to provide assisted ventilation, are steadily increasing in India and in China.45,47 The same trend is seen in other developing countries. By 2010 there has been a proliferation of level II and level III units in the private sector in India. Most recent data show about 72 level II NICUs and 68 accredited NICUs.

The evolving NICUs are having a major impact on survival of infants of institution-based deliveries, particularly infants with LBW and VLBW. Infants requiring assisted ventilation must be transported to NICU facilities where available. Generally, properly organized neonatal transport systems in developing countries are nonexistent, although well-organized transport systems are beginning to evolve in some countries (EMRI. WWW). Despite various limiting factors, Singh and colleagues39 in India and Bhutta6 in Pakistan have documented that intensive care, including ventilation, can be provided in developing countries, with good short-term and long-term outcomes. These emerging data indicate that as the overall survival of infants with LBW and VLBW increases, there will be an increased demand for NICU facilities with ventilatory support. The overall cost of neonatal intensive care at the tertiary level in developing countries is still high and affordable to only a few middle and upper income groups. Bhutta6 showed that infants with RDS can be successfully managed with ventilatory support. In that study, the overall mortality of infants with RDS treated with assisted ventilation was 39%, the average length of stay was 24.6 6 21.1 days, and the average cost was U.S. $1391 per survivor. In terms of a country’s own economy, this cost amounts to multiples of the country’s per capita gross national product (U.S. $420). Surfactant therapy was not used in this group of infants. Shanmugasundaram and colleagues38 in India produced a similar cost analysis. In India, although surfactant is an expensive drug (two to three times the country’s per capita gross national product), its use has reduced the overall hospital stay and the overall cost. Neonatologists must be aware of the economic implications for families and must inform them of the expected clinical outcome, particularly when treating extremely premature or critically ill infants. The real goal should be to consider preventive antenatal measures and newer treatment modalities that are less expensive, such as antenatal steroids in preterm labor, implementation of delivery room neonatal resuscitation, early use of CPAP, and surfactant therapy. With the increase in the survival of infants receiving neonatal intensive care in developing countries, the survival of premature infants and infants with VLBW is also increasing. Increased survival is also associated with increased morbidity (e.g., intraventricular hemorrhage, retinopathy of prematurity, chronic lung disease). Although some data regarding the prevalence of retinopathy of prematurity is emerging at an institutional level,1 there are no good data at a national level. There is an urgent need for a policy at the international level to monitor these emerging new diseases across the globe.

Chapter 7  Perinatal and Neonatal Care in Developing Countries

Ethical Dilemmas With increasing awareness of the importance of access to technology for improved neonatal outcome, there is an increasing interest to develop NICUs even in countries with economic constraints. A report from Ghana showed that introduction of new technology (ventilators) and training of existing health care personnel in a teaching hospital NICU resulted in improved survival of at-risk infants. These trends in developing countries raise concerns regarding economic, social, and ethical issues. These dilemmas are faced daily by practicing neonatologists in developing countries. As stated by Singh,40 undue emphasis may be placed on saving an individual infant at a greater cost to the family. He suggests that to ensure the principle of cost-effectiveness in resource-poor countries, the narrow principle of the “best interests” of the child should be replaced by the concept of a broader benefit to the family, society, and state. His point is valid and should be considered as a policy matter in resourcepoor countries. The welfare of an individual who can afford the cost of expensive treatment cannot be denied, however. In her analysis of health care policy, Deaton11 indicates that “a policy that harms no one, while making at least some people better off, is a good thing.” The criterion also says that innovations are beneficial and should be encouraged. For the proponents of public health, this concept is contradictory to the principle of equality. That is, that inequality is inherently bad, and innovations that increase it should be discouraged. Policies based on such positions result in the deaths of some people who could have lived, however, without neglecting others. Such policies also delay the diffusion of knowledge or technology that in most cases also would benefit poor people, although with some delay. These arguments are important for policymakers at the national level. In recent years, rapid demographic and economic transitions have been developing in different countries, particularly in India, China, and many Latin American countries. In countries such as these, while the health care policies of the national government focus on the care of the poor using public health principles, the private sector is rapidly expanding its role to cater to individuals who are economically more viable and can afford high-tech, high-cost, modern medicine. The same is true for neonatal care. Pediatricians and neonatologists in developing countries must be well trained in the use of modern technology and cultivate a sensitivity to the economic implications of NICU care for families. They must provide accurate information regarding complications and the quality of life for the infant at risk. It is also important to recognize that there are no supportive therapies after discharge for infants who subsequently manifest developmental disabilities. It is incumbent on institutions with NICUs to organize a developmental follow-up clinic and other support services. Families must be well informed.

SUMMARY There is evidence that NMR and IMR are steadily decreasing across the globe. These decreases are due to rapid diffusion of technologies and improvement in overall health care. There are wide regional differences, however. Many countries

125

are lagging behind in meeting their MDGs of reducing NMR and IMR by 2015. Most of the burden of perinatal disease is in rural communities and developing countries. Policymakers should strengthen rural primary health centers in accordance with the recommendations of WHO and UNICEF. These health centers should be provided with basic equipment needs. Health care personnel should focus on the preventive and primary care of the mother and infant, adhering to the principles of essential obstetric and newborn care. Communications among the community (i.e., its members or their representatives or both), subcenter, primary health center, and district hospital can be improved by using modern information technology. Early recognition of highrisk cases and early consultations by e-mail or telephone should be encouraged. At social and political levels, policymakers should direct their efforts toward eliminating gender bias and improving female literacy and economic empowerment. Public health plans should focus on providing running water, electricity, and access to pit or flush toilets. Preventive public health programs are very cost-effective in reducing NMR and IMR. Overall, during the past 50 years there has been a global trend toward a decrease in IMR. The wide divergence of high IMR in 1950 is showing a convergence toward lower IMR across the globe. This is a positive trend.

REFERENCES 1. Agarwal R et al: Changing profile of retinopathy of prematurity, J Trop Pediatr 48:239, 2002. 2. Bang AT et al: Pneumonia in neonate: can it be managed in the community? Arch Dis Child 68:550, 1993. 3. Bang AT et al: Effect of home-based neonatal care and management of sepsis on neonatal mortality: field trial in rural India, Lancet 355:1955, 1999. 4. Bassiouny MR et al: Nasal continuous positive airway pressure in the treatment of respiratory distress syndrome: an experience from a developing country, J Trop Pediatr 40:341, 1994. 5. Baqui AH et al: Rates, timing and causes of neonatal deaths in rural India: implications for neonatal health programmes, Bull World Health Organ 84:706, 2006. 6. Bhutta ZA: Is management of neonatal respiratory distress syndrome feasible in developing countries? Experience from Karachi (Pakistan), Pediatr Pulmonol 27:305, 1999. 7. Birdsall N, Subramanian A: Saving Iraq from its oil, Foreign Affairs 83:77, 2004. 8. Bryce J et al: Countdown to 2015 for maternal, newborn, and child survival: the 2008 report on tracking coverage of interventions, Lancet 371:1247, 2008. 9. CLAP Scientific Publication 1524.02 Montevideo-Uruguay. Nov 2003. ISBN 9974-622-31-X. www.clap.ops.oms.org. Accessed February 15, 2010. 10. Daga SR, Daga AS: Reduction in neonatal mortality with simple interventions, J Trop Pediatr 35:191, 1989. 11. Deaton A: Policy implications of the gradient of health and wealth, Health Affairs 21:13, 2002. 12. Declaration of Alma-Ata International Conference on Primary Health Care, Alma-Ata, USSR, September 6-12, 1978. 13. Deodhar J: Telemedicine by email: experience in neonatal care at a primary care facility in rural India, J Telemed Telecare 8(suppl):20, 2002.

126

SECTION I  THE FIELD OF NEONATAL-PERINATAL MEDICINE

14. Deorari AK et al; Medical College Network: Impact of education and training on neonatal resuscitation practices in 14 teaching hospitals in India, Ann Trop Pediatr 21:29, 2001. 15. Deorari AK et al: The national movement of neonatal resuscitation in India, J Trop Pediatr 46:315, 2000. 16. Downes JJ et al: Respiratory distress syndrome of the newborn: a clinical score with acid-base and blood gas correlations, Clin Pediatr 9:325, 1970. 17. Fohe K et al: Skin to skin contact improves gas exchange in premature infants, J Perinatol 20:311, 2000. 18. Haksari EL: Downes scores as clinical assessment in hypoxemia for neonates in respiratory distress; personal communication, April 7, 2008. 19. Hamod D, Sacy R: Neonatology in war, J Arab Neonatal Forum 3:34, 2006. 20. Hassan Z: Assessing the impact of micro-credit on poverty and vulnerability in Bangladesh. Policy Research Working Paper. Developmental Economics. The World Bank, Washington, DC, 1999. 21. Johanson RB et al: Effects of post-delivery care on neonatal body temperature, Acta Paediatr 81:859, 1992. 22. Kumar R: Reducing neonatal mortality through primary care, J Neonatol 16:15, 2002. 23. Kumar V, Walia I: Pictorial maternal and neonatal records for illiterate traditional birth attendants, Int J Gynaecol Obstet 19:281, 1981. 24. Lopez AD et al: Global and regional burden of disease and risk factors, 2001: systematic analysis of population health data, Lancet 367:1747, 2006. 25. Matthews MK et al: Training traditional birth attendants in Nigeria—the pictorial method, World Health Forum 16:409, 1995. 26. Moser K et al: World mortality 1950-2000: divergence replaces convergence from the late 1980s, Bull Health World Organ 83:202, 2005. 27. Orlando S et al: Neonatal nursing care issues following a natural disaster: lessons learned from the Katrina experience, J Perinat Neonatal Nurs 22:147, 2008. 28. PAHO program on women’s health and development address special needs (website). www.paho.org/English/DPI/100feature10.htm. Accessed February 18, 2010. 29. Papageorgiou C et al: International medical technology diffusion, J Internat Econom 72:409, 2007. 30. Radonicic F et al: Perinatal outcomes during 1986-2005 in Tuzia Canton, Bosnia and Herzegovina. J Matern Fetal Neonatal Med 21:567, 2008. 31. Raina N, Kumar V: Management of birth asphyxia by traditional birth attendants, World Health Forum 10:243, 1989. 32. Ramji S et al: Resuscitation of asphyxic newborn infants with room air or 100% oxygen, Pediatr Res 34:809, 1993. 33. Ramji S: Training paramedics in primary newborn care: curriculum and skills, J Neonatol 16:4, 2002.

34. Roudi F: Population: trends and challenges in the Middle East and North Africa. Washington, DC, Population Reference Bureau, 2001. 35. Saugstad OD: Resuscitation with room air or oxygen supplementation, Clin Perinatol 25:741, 1998. 36. Save the children: state of the world’s newborns. Washington, DC, September 2001. 37. Shah MA, editor: Transcultural aspects of perinatal health care: a resource guide, Elk Grove Village, IL, National Perinatal Association and American Academy of Pediatrics, 2004. 38. Shanmugasundaram R et al: Cost of neonatal intensive care, Indian J Pediatr 65:249, 1998. 39. Singh M et al: Assisted ventilation for hyaline membrane disease, Indian J Pediatr 32:1267, 1995. 40. Singh M: Ethical and social issues in the care of the newborn, Indian J Pediatr 70:417, 2003. 41. Stoll BJ: The global impact of infection, Clin Perinatol 24:1, 1997. 42. Task Force for Child Survival: Protecting the world’s children: an agenda for the 1990s. Paper presented at Tufts University European Center, March 10-12, 1988, Talloires, France. 42a. United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects: The 2008 Revision, New York, 2009. Accessed February 15, 2010. 43. UN General Assembly, 56th Session: Roadmap towards the implem entation of the United Nations Millennium Declaration: Report of the Secretary General (UN document #A/56/326). New York, United Nations, 2001. 44. Verder H et al: Nasal continuous positive airway pressure and early surfactant therapy for RDS in newborns of ,30 weeks gestation, Pediatrics 103:e24, 1999. 45. Vidyasagar D et al: Evolution of neonatal and pediatric critical care in India, Crit Care Clin 13:331, 1997. 46. Wandi F et al: Hypoxaemia among children in rural hospital in Papua New Guinea: epidemiology and resource availability—a study to support a national oxygen programme, Ann Trop Paediatr 26:277, 2006. 47. Wei KL: personal communication, 2004. 48. World Health Organization Maternal Health and Safe Motherhood Program: Perinatal mortality: a listing of available information. Geneva, World Health Organization, 1996. 49. World Health Organization: Evidence for the ten steps to successful breastfeeding. Geneva, World Health Organization, 1998. 50. World Health Organization: Prevention of mother to child transmission of HIV: selection and use of nevirapine. Technical notes. Geneva, World Health Organization, 2001. 51. World Health Organization: Thermal protection of the newborn: s practical guide. Geneva, World Health Organization, 1994. 52. World Health Organization: World Health Report 1998: life in the twenty-first century: a vision for all. Geneva, World Health Organization, 1998.

SECTION

II

The Fetus

127

CHAPTER

8

Genetic Aspects of Perinatal Disease and Prenatal Diagnosis Komal Bajaj and Susan Gross

Genetics is a fundamental part of every aspect of medicine. Constantly expanding knowledge of the human genome and the ability to perform testing in an efficient manner have made genetics a cornerstone of public health and clinical practice. This chapter highlights essential concepts regarding the genetic basis of disease and issues surrounding prenatal evaluation and diagnosis. Principles of inheritance, teratogens, genetic screening, and diagnostic modalities are discussed in detail.

PRINCIPLES OF INHERITANCE Chromosomal Disorders In humans, normal gametes are composed of 23 chromosomes each. A normal human somatic cell contains 46 chromosomes. In both genders, 22 pairs of chromosomes, also known as autosomes, are identical. Women have a homologous pair of sex chromosomes, known as the X chromosome. Men have a nonhomologous pair, an X and a Y chromosome. A chromosome is composed of a linear DNA molecule that is complexed with structural proteins known as histones to form chromatin. Each chromosome has a centromere, which divides the chromosome into a short arm (the p arm) and a long arm (the q arm). Where the centromere is located helps describe chromosomes as metacentric, submetacentric, and acrocentric. In metacentric chromosomes, the arm length is equal, whereas in submetacentric chromosomes, one arm is larger than the other. If the p arm contains such small amounts of genetic material that it is almost negligible, the chromosome is considered acrocentric. In humans, the acrocentric chromosomes are 13, 14, 15, 21, and 22. The ends of each chromosome are known as telomeres. During cell division, the chromosomes

condense more than 10,000-fold, resulting in compact structures that can segregate. To analyze chromosomes, a karyotype is produced (Fig. 8-1). The chromosomes are paired and organized according to size. The overall structure and banding pattern is evaluated and is reported according to the International System for Cytogenetic Nomenclature. According to this nomenclature, a karyotype designation includes the total chromosome number followed by the sex chromosome constitution. Females are 46,XX, and males are 46,XY. If there are any variants or abnormalities, this is reported after the sex chromosomes (Table 8-1).28 Chromosome disorders can be either structural or numerical. The consequence of the abnormality depends on the amount of genomic imbalance and the genes involved.

ADVANCED MATERNAL AGE Epidemiologic studies suggest that women are having fewer children, often later in life. The birthrate for women 40 to 44 years old increased 51% between 1990 and 2002.41 With the advent of assisted reproductive technology (ART), women in their 50s and 60s can achieve pregnancy. Although the effects of increasing age occur as a continuum, the term advanced maternal age typically refers to pregnant women who will be 35 or older on their expected date of confinement. Chromosomal analysis of samples from spontaneous abortions, prenatal diagnosis, and live births reveals that there is a steady increase in aneuploidy as a woman ages (Fig. 8-2). The basis for this increase is unknown, although it may be related to a decrease in the number of normal oocytes available or cumulative oxidative stress on the finite number of oocytes with which females are born. Along with chromosomal abnormalities, it has been observed that congenital anomalies increase with increased maternal age. The FASTER

129

130

SECTION II  THE FETUS Figure 8–1.  Karyotype of a normal male.

Notice the presence of an X and Y chromosome and 22 pairs of autosomes. 2

1

3

6

7

8

13

14

15

19

20

4

9

10

16

5

11

12

17

18

21

22

Sex chromosomes

TABLE 8–1  Abbreviations Used for Description of Chromosomes and Their Abnormalities Abbreviation

Meaning

Example

Condition

46,XX

Normal female

46,XY

Normal male

1

Gain of

47,XX,121

Female with trisomy 21

2

Loss of

45,XX,222

Female with monosomy 22

t

Translocation

46,XY,t(2;8)(q22;p21)

Male with balanced translocation between chromosome 2 and 8, with breaks in 2q22 and 8p21

/

Mosaicism

46,XX/47,XX,18

Female with two populations of cells, one with a normal karyotype and one with trisomy 8

From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 66.

trial reported rates of congenital anomalies for women younger than 35 years old as 1.7%; women 35 to 39 years old and 40 years old or older had rates of 2.8% and 2.9%.22

ABNORMALITIES OF CHROMOSOME NUMBER The mere presence of additional genetic material, albeit of normal makeup, can result in clinically significant phenotypes. Following is a discussion of the various types of numerical abnormalities.

Triploidy and Tetraploidy Triploid fetuses have three sets of chromosomes, for a total number of 69. Triploid fetuses are rarely born alive; when they are, survival is poor. Most triploidy is the result of fertilization

by two sperm. Tetraploids, fetuses with 96 chromosomes, are usually miscarried in the first trimester.

Aneuploidy In humans, the term aneuploid is used to describe any genotype in which the total chromosome number is not a multiple of 23. Most aneuploid patients have either a monosomy (only one representative of a particular chromosome) or a trisomy (three copies of a particular chromosome). As a rule, monosomies tend to be more deleterious than trisomies. Complete monosomies are generally not viable except for monosomy X (Turner syndrome). Trisomies for chromosomes 13, 18, 21, X, and Y are compatible with life, with trisomy 21 (Down syndrome) being the most common trisomy in live-born infants.

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

agents, such as ionizing radiation. Structural abnormalities can be divided into two categories—balanced and unbalanced. Balanced rearrangements have the normal complement of chromosomal material. Also, a balanced rearranged chromosome must have a functional centromere and two functional telomeres. Unbalanced rearrangements are either missing or have additional genetic information. Structural rearrangements include deletions, insertions, ring chromosomes, isochromosomes, and translocations (Fig. 8-4). One unique type of translocation is the robertsonian translocation, in which two acrocentric chromosomes lose their short arms and fuse near the centromeric region. Because the short arms of acrocentric chromosomes contain only genes for ribosomal RNAs, loss of the short arm is rarely deleterious. The result is a balanced karyotype with only 45 chromosomes, including the translocated chromosome, which comprises the long arms of two chromosomes. Carriers of robertsonian translocations are phenotypically normal, but have the risk of producing unbalanced gametes. The main clinical relevance of a robertsonian translocation is that one involving chromosome 21 could result in a child with Down syndrome. About 4% of cases of Down syndrome have 46 chromosomes, one of which is a robertsonian translocation between chromosome 21 and another chromosome.

1/25 1/28 1/33

1/50 1/87

Risk estimates

1/40

1/100 1/200

45 Maternal age

Figure 8–2.  Risk of fetal aneuploidy as a function of mater-

nal age. The most common mechanism for aneuploidy is meiotic nondisjunction, in which a pair of chromosomes fails to separate during either of the meiotic divisions (Fig. 8-3). Nondisjunction can rarely occur during a mitotic division after the formation of the zygote. If this happens early in cleavage, mosaicism may occur. In this situation, two or more different chromosome complements are present in one individual. The clinical significance of mosaicism is difficult to evaluate and depends on the developmental timing when the mosaicism occurred, the tissues affected, and the proportion of tissue affected.

Single-Gene Disorders Mendel studied the offspring characteristics of garden peas and observed that certain phenotypic characteristics occurred in fixed proportions. Single-gene traits for which mutations cause predictable disease are described as exhibiting mendelian inheritance because they follow the rules that he originally described. Currently, almost 4000 diseases are known to exhibit mendelian patterns of inheritance. Among hospitalized children, 6% to 8% are thought to have single-gene disorders. Variants of a gene are called alleles. For many genes, there is one prevailing allele, which is referred to as the wild-type

ABNORMALITIES OF CHROMOSOME STRUCTURE Chromosomal structural abnormalities are the result of chromosome breakage followed by anomalous reconstitution. Rearrangements result spontaneously or are due to inducing

MEIOSIS I

MEIOSIS II

Normal

Normal

Nondisjunction

Normal

Normal

131

Normal

Normal

Normal

Nondisjunction

Figure 8–3.  Consequences of nondisjunction at meiosis I (center) and meiosis II (right) compared with normal

disjunction (left). If the error occurs at meiosis I, the gametes either contain a representative of both members of the chromosome 21 pair or lack chromosome 21 altogether. If nondisjunction occurs at meiosis II, the abnormal gametes contain two copies of one parental chromosome 21 (and no copy of the other) or lack chromosome 21. (From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 68.)

132

SECTION II  THE FETUS

a b b

x

c

Terminal deletion

A

Interstitial deletion

Ring

C

E

Robertsonian translocation

a b b

a b

c

c

a b b b c

Unequal crossing over

B

D

Figure 8–4.  Structural rearrangements of chromosomes. A, Terminal and interstitial deletions, each generating an acentric fragment. B, Unequal crossing over between segments of homologous chromosomes or between sister chromatids (duplicated or deleted segments indicated by brackets). C, Ring chromosome with two acentric fragments. D, Generation of an isochromosome for the long arm of a chromosome. E, Robertsonian translocation between two acrocentric chromosomes. F, Insertion of a segment of one chromosome into a nonhomologous chromosome.  (From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 69.)

Isochromosome

F

allele. The other versions of the gene are mutations, not all of which may cause disease. Mutations can be inherited or de novo, meaning that neither parent possessed the mutation. Instead, the mutation occurred as a random error during gametogenesis.

AUTOSOMAL DOMINANT DISORDERS Approximately half of mendelian disorders are inherited in an autosomal dominant fashion. Inheritance usually exhibits a vertical pattern of transmission, meaning that the phenotype usually appears in every generation, with each affected person having an affected parent (Fig. 8-5). For each offspring of an affected parent, the risk of inheriting the mutated allele is 50%. An example of a disorder inherited in an autosomal dominant fashion is osteogenesis imperfecta. Biochemical defects in either the amount or the structure of collagen result in various clinical phenotypes depending on the mutation.

Insertion

Advanced Paternal Age The link between advanced maternal age and genetic abnormalities has been well established. The role of advanced paternal age, defined as 40 or older, is not as clear. It has been established that the rate of base substitution mutations during spermatogenesis increases as a man ages. The risk of de novo autosomal dominant disorders in offspring of fathers 40 years old or older is estimated at 0.3% or lower.21 Some evidence has suggested that advanced paternal age is associated with an increased risk for complex disorders, such as schizophrenia, autism, and congenital anomalies. The relative risk for these conditions is 2 or less. Although there may be slightly increased risk for a range of disorders associated with advanced paternal age, the overall risk remains low. No screening or diagnostic tests target conditions associated with advanced paternal age. Pregnancies that are fathered by men 40 years old or older should be treated according to standard guidelines established by the American College of

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis I 1

2

II 1

2

3

4

III 1

2

3

4

Affected (D/d) × Unaffected (d/d) 1/2 = D/d 1/2 affected 1/2 = d/d 1/2 unaffected

Figure 8–5.  Pedigree showing the typical inheritance of an autosomal dominant disorder. j 5 affected male; d 5 affected female; h 5 unaffected male; s 5 unaffected female. Affected (D/d) 3 Unaffected (d/d) 1 ⁄2 affected ⁄2 5 D/d 1 ⁄2 5 d/d ⁄2 unaffected Affected (D/d) 3 Affected (D/d) 1 3 ⁄4 5 DD ⁄4 affected 1 1 ⁄2 5 D/d ⁄4 unaffected 1 ⁄4 5 d/d 1 1

Medical Genetics (ACMG) and the American College of Obstetrics and Gynecology (ACOG).34

AUTOSOMAL RECESSIVE DISORDERS An autosomal recessive condition occurs when an individual possesses two mutant alleles that were inherited from heterozygous parents. For autosomal recessive diseases, an individual with one normal allele does not manifest the disease because the normal gene copy is able to compensate. Autosomal recessive disorders exhibit horizontal transmission, meaning that if the phenotype appears in more than one family member, it is typically in the siblings of the proband, not in parents, offspring, or other relatives (Fig. 8-6). If both

parents are carriers of a mutated allele, 25% of offspring have the autosomal recessive disease. Consanguineous unions, mating between individuals who are second cousins or closer, are at increased risk for an autosomal recessive disorder because there is a higher likelihood that both individuals carry the same recessive mutation. A common autosomal recessive disease is cystic fibrosis. Carrier screening and prenatal implications are discussed in a later section.

SEX-LINKED DISORDERS X chromosome inactivation is a normal process in females in which one X chromosome is randomly inactivated early in development. Females are normally mosaic with respect to X-linked gene expression. Disorders of genes located on the X chromosome have a characteristic pattern of inheritance that is affected by gender. Males with an X-linked mutant allele are described as being hemizygous for that allele. Males have a 50% chance of inheriting a mutant allele if the mother is a carrier. Females can be homozygous wild-type allele, homozygous mutant allele, or a heterozygote. An X-linked recessive mutation is phenotypically expressed in all males, but is expressed only in females who are homozygous for the mutation. As a result, X-linked recessive disorders are generally seen in males and rarely seen in females. An example of such a condition is hemophilia A. X-linked dominant disorders may manifest differently among heterozygous females in the same family because of different patterns of X chromosome inactivation. X-linked inheritance is classically characterized by the lack of male-to-male transmission because males transmit their Y chromosome to their sons, not their X chromosome (Fig. 8-7).

Nonmendelian Patterns of Inheritance MITOCHONDRIAL INHERITANCE Mitochondrial DNA (mtDNA) is organized as a 16.5-kb circular chromosome located in the mitochondrial organelles of a cell, not the cell nucleus. mtDNA contains 37 genes that

I

II

III

IV

Figure 8–6.  Typical pedigree showing autosomal recessive inheritance. Unaffected carrier (A/a) 3 Unaffected carrier (A/a). h 5 unaffected male; s 5 unaffected female; j 5 affected male; d 5 affected female. 1

unaffected ⁄4 5 AA ⁄2 5 Aa unaffected carrier 1 ⁄4 5 aa affected (From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 123.) 1

133

134

SECTION II  THE FETUS

heart, and kidneys. Poor growth, muscle weakness, loss of coordination, or developmental delay not explained by more common causes should alert a neonatologist or pediatrician to the possibility of a mitochondrial disease. When a mitochondrial disease is suspected, the child should be referred to a specialized medical center where comprehensive evaluation, including genetic studies, can be performed.

I

II

III

EPIGENETICS AND UNIPARENTAL DISOMY

IV

Figure 8–7.  Pedigree pattern showing X-linked dominant inheritance. h 5 unaffected male; s 5 unaffected female; j 5 affected male; d 5 affected female.  (From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 133.)

encode for important proteins, including proteins involved in oxidative phosphorylation. Mitochondrial inheritance has a few distinct features that differ from mendelian inheritance: maternal inheritance, replicative segregation, and heteroplasmy. Because sperm mitochondria are eliminated from the forming embryo, mtDNA is inherited entirely from the maternal side, with very rare exception. At cell division, the mitochondria sort randomly between two daughter cells, a process known as replicative segregation. A cell containing a mix of mutant and wild-type mtDNA can distribute variable proportions of mutant or wild-type DNA to daughter cells. By chance, a daughter cell may receive all wild-type or all mutant mtDNA, a state known as homoplasmy. Heteroplasmic daughter cells can result in variable penetrance and expression depending on the amount of mutant mtDNA present. More than 100 different mutations in mtDNA have been identified to cause disease in humans.42 Most of these involve the central nervous system or musculoskeletal system (Table 8-2). Mitochondrial disease typically manifests as dysfunction in high energy–consuming organs, such as the brain, muscle,

Epigenetics refers to modification of genes that determines whether a gene is expressed or not (see also Chapter 13). These modifications, an example of which is methylation, affect the expression of a gene, but not the primary DNA sequence itself. Imprinting refers to a phenomenon where genetic material is differentially expressed depending on whether it was inherited from the father or the mother. A different phenotype can result depending on the parent of origin because for certain genes, only the allele from one parent is transcriptionally active. Uniparental disomy is the inheritance of a pair of homologous chromosomes from one parent, rather than the normal scenario where one chromosome is inherited from each parent. This situation is thought to arise most commonly by a process called trisomy rescue, during which a trisomic cell is converted into a disomic cell. It is a matter of chance as to which chromosome drops out. When trisomy rescue occurs, both chromosomes are from one parent a third of the time. Classic examples of disorders related to genomic imprinting are Prader-Willi and Angelman syndromes. Both these syndromes involve the long arm of chromosome 15 (15q1115q13). At birth, Prader-Willi syndrome is characterized by hypotonia, low birthweight, and almond-shaped eyes. During childhood, other features, such as short stature, obesity, indiscriminate eating habits, small hands and feet, mental retardation, and hypogonadism develop. In most of these cases, there is paternally derived deletion, which means that all the genetic information in the region is maternal in origin. Angelman syndrome, characterized by mental retardation, short stature, abnormal facies, and seizures, is the opposite situation, in which the deletion is maternally derived, and the genetic information in the region is paternal

TABLE 8–2  Common Mitochondrial Diseases and Their Manifestations Name

Abbreviation

Disease Characteristics

Myoclonic epilepsy associated with ragged red fibers

MERRF

Progressive myoclonic epilepsy, short stature, clusters of diseased mitochondria accumulated in subsarcolemmal region of muscle fiber (appear as “ragged red fibers” when stained)

Mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke

MELAS

Muscle weakness, headaches, loss of appetite, seizures, lactic acidosis, stroke

Neuropathy, ataxia, and retinitis pigmentosa

NARP

Numbness and tingling in limbs, muscle weakness, ataxia, deterioration of light-sensing cells of retina

Leber hereditary optic neuropathy

LHON

Acute onset of visual loss and optic atrophy usually in early young adulthood

Myoneurogastrointestinal disorder and encephalopathy

MNGIE

Ptosis, external progressive external ophthalmoplegia, diffuse leukoencephalopathy, gastrointestinal motility dysfunction

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

only in origin. Approximately 30% of Prader-Willi cases and 5% of Angelman cases are the result of uniparental disomy. In this scenario, there is no cytogenetically detectable deletion. Because of imprinting, Prader-Willi syndrome results from uniparental disomy in which both chromosomes derive from the mother. The loss of the paternal contribution results in Prader-Willi syndrome. Paternal uniparental disomy in the same region results in Angelman syndrome because of the loss of maternal contribution of genes in the 15q11-q13 region.

TRINUCLEOTIDE REPEAT EXPANSION Most mutations, when they occur, remain unchanged as they get passed from one generation to the next. There is a subset of disorders, however, for which an expansion of an area of DNA containing repeating units results in disease. In this case, as the gene is passed on, the number of repeats (usually consisting of three nucleotides each) can increase to beyond polymorphic range and begin to affect gene function. Although the mechanism of how this expansion occurs is not completely elucidated, it is thought to be the result of a slipped mispairing mechanism in which an insertion occurs when a newly synthesized strand temporarily dissociates from the template strand. A dozen or so diseases, including congenital myotonic dystrophy, Huntington disease, Friedreich ataxia, and fragile X syndrome, are the result of unstable repeat expansions. Fragile X syndrome, the most common hereditary form of mental retardation, has an incidence of approximately 1 in 4000 births. The normal number of triplet repeats in the Xq27.3 region (FMR1 gene) is less than 45; a full mutation is considered to be greater than 200 repeats. Individuals with 45 to 54 repeats are referred to as intermediate carriers— these individuals are not at risk for any phenotypic abnormalities and are not at risk for expansion to a full mutation in their offspring. Individuals with 55 to 200 repeats are known as premutation carriers. Besides the risk of having an offspring with the full mutation, these premutation carriers are also at risk for adult-onset cerebellar dysfunction (known as fragile X–associated tremor/ataxia syndrome) and premature ovarian failure.20 Individuals who have any family or personal history of developmental delay, mental retardation, ovarian dysfunction, or tremor should be offered screening for fragile X syndrome. At this time, population carrier screening is not recommended.32

Multifactorial Inheritance There are disorders that affect certain families more than others, but do not follow mendelian patterns of inheritance or fit into the nonmendelian inheritance phenomenon. These disorders are thought to be the result of interplay between genetic and environmental factors and gene-gene interactions. Known as multifactorial or complex inheritance, these disorders have a greater incidence than disorders secondary to chromosomal or single-gene mutations. These disorders provide unique genetic counseling dilemmas regarding recurrence risks because although genotypes predisposing to disease may aggregate in families, the phenotypic expression is discordant owing to differences in nongenetic exposures.

135

An illustrative example of multifactorial inheritance is the occurrence of neural tube defects (NTDs). Spina bifida and anencephaly are NTDs that cluster in families and are a leading cause of fetal loss and handicap. Spina bifida is the result of incomplete fusion of vertebral arches and manifests in various degrees of severity. Anencephaly is a devastating condition in which the forebrain, overlying meninges, bone, and skin are absent. Most fetuses with anencephaly are stillborn. Although some NTDs can be explained by teratogens, amniotic bands, or chromosomal disorders, most are multifactorial. Decreased levels of maternal folic acid have been inversely correlated with the risk of NTDs. Folic acid levels are affected by two factors—dietary intake and enzymatic processing. Folic acid levels are detrimentally affected by a mutation in the enzyme 5,10-methylenetetrahydrofolate reductase (MTHFR). Fifteen percent of the population is homozygous for the mutation. It has been shown that the mothers of infants with NTDs were twice as likely to have MTHFR mutations than controls. Preconceptual supplementation of folic acid has been shown to decrease the risk of NTDs.39 All reproductive-age women should consume 0.4 mg of folic acid daily. Prenatal screening for NTDs is discussed later in this chapter.

TERATOGENS Environmental exposures—medications, maternal conditions, or infections—are the etiology of malformations in 10% of cases. The impact of the agent relates to the timing and amount of exposure (duration and dosage). The susceptibility of a fetus depends on its stage of development when an exposure takes place (Fig. 8-8).26 (see Chapter 12.) Exposure during the first 2 weeks after conception usually either is lethal to the embryo or has no adverse effect. This window is known as the “all-or-none” period. During organogenesis, teratogenic exposure may result in major morphologic abnormalities because of disruption of the forming organ systems. Most commonly prescribed medications can be used with relative safety during pregnancy. For the medications that are suspected or known teratogens, genetic counseling should emphasize relative risk. That is, risk increases should be presented in relation to a woman’s baseline risk of a birth defect, which is 2% to 3%. The possibility that certain medical conditions if left untreated pose greater threat to the fetus than the medications used to treat the condition should be addressed. The U.S. Food and Drug Administration has categorized the potential risk according to available safety evidence (Table 8-3). Use of these drugs in pregnancy ultimately requires expertise and access to up-to-date databases specializing in teratogenesis, available through a prenatal genetics service. Certain drugs may be a significant risk, but only during a particular trimester. Trimethoprim-sulfamethoxazole should be avoided in the third trimester to avoid kernicterus in the newborn, but may be used in the first and second trimesters. Likewise, a certain drug may be a genuine teratogen, but the risk of stopping the drug would be of even greater consequence. Lithium use during the first trimester of pregnancy has been associated with cardiac malformations, including Ebstein anomaly. Although efforts should be made to avoid lithium

Figure 8–8.  The developing embryo in stages. Black bars represent when the organ system is most

susceptible to teratogenic exposure. C.N.S., central nervous system. (From Moore K: The developing human: clinically oriented embryology, Philadelphia, 1982, Saunders.)

2

Embryonic disc

Amnion

Death of embryo and spontaneous abortion common

Not susceptible to teratogenesis

Blastocyst

Morula

Embryonic disc

Period of dividing zygote, implantation, and bilaminar embryo

1

3

5

Upper lip

Lower limb

Palate

9

TA—Truncus arteriosus; ASD—Atrial septal defect; VSD—Ventricular septal defect

Ears

16

CNS

External genitalia

Teeth

Eyes

32

Fetal Period (in weeks)

Functional defects and minor anomalies

Masculinization of female genitalia

Cleft palate

Enamel hypoplasia and staining

Major congenital anomalies

Highly sensitive period

Less sensitive period

Common site(s) of action of teratogens

8

Mental retardation Heart Upper limb

7

Low-set malformed ears and deafness

Cleft lip

6

Microphthalmia, cataracts, glaucoma

Amelia/Meromelia

Amelia/Meromelia

TA, ASD, and VSD

Neural tube defects (NTDs)

4

Main Embryonic Period (in weeks)

38

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

137

TABLE 8–3  FDA Guidelines Categorizing Pregnancy-Related Risk Category

Interpretation

A

Controlled studies show no risk. Adequate, well-controlled studies in pregnant women have failed to show risk to the fetus

B

No evidence of risk in humans. Either animal findings show risk (but human findings do not), or, if no adequate human studies have been done, animal findings are negative

C

Risk cannot be ruled out. Human studies are lacking, and animal studies are either positive for fetal risk or lacking as well. Potential benefits may justify potential risk

D

Positive evidence of risk. Investigational or postmarketing data show risk to fetus. Potential benefits may outweigh risk

X

Contraindicated in pregnancy. Studies in animals or humans, or investigational or postmarketing reports have shown fetal risk that clearly outweighs any possible benefit to the patient

FDA, U.S. Food and Drug Administration.

during the first trimester, if an alternative therapy is inappropriate, the lowest possible lithium dose should be used. Finally, certain drugs are known teratogens, but the underlying disorder may contribute to birth defects independently. Women with epilepsy are at increased risk of fetal malformations such as orofacial clefts, independent of whether or not they are taking anticonvulsant therapy.

Diagnostic Imaging Radiographic imaging modalities are widely used in inpatient and outpatient settings. Safety of ionizing radiation exposure in pregnancy is an important concern because of the association of ionizing radiation with prenatal death, malformation, and carcinogenesis. It is necessary to weigh the risk of exposure versus the risk to the mother and fetus of a delayed diagnosis. With proper test selection, shielding, and procedure modifications, fetal risk can be significantly reduced or eliminated. Common radiologic studies, such as computed tomography scan of the chest or dental x-rays, typically expose the fetus to much less than 5 rad of radiation (Table 8-4). At these levels, there has been no evidence of increased risk of mental retardation, fetal anomalies, or pregnancy loss. The risk of childhood leukemia with an exposure of 1 to 2 rad increases to 1 in 2000, a 1.5-fold increase above the baseline risk of 1 in 3000.1 Exposure greater than 10 rad has been associated with an increased risk of malformations. Diagnostic imaging rarely, if ever, exposes the fetus to this level of radiation, however. Iodinated contrast materials can cross the placenta and produce transient effects in the developing fetal thyroid. Although not contraindicated, iodinated contrast material should be used with great care. Magnetic resonance imaging (MRI), which uses electromagnetic radio waves to generate images, has not been associated with any harmful effects to the developing fetus. Gadolinium, the contrast agent most commonly used in MRI, crosses the placenta. There has been limited evaluation of its safety, and it is not recommended for use in pregnancy. Ultrasonography is an important tool in obstetrics. Although there has been some concern regarding elevation of tissue

temperature with prolonged use, there is no reliable evidence of harm to the human fetus as the result of standard, medically indicated ultrasonography. When used appropriately, ultrasonography can provide important information regarding gestational age and fetal anatomy. The casual use of ultrasonography for fetal pictures or sex determination is not justified.6

TABLE 8–4  E  stimated Average Fetal Exposure from Select Imaging Studies Procedure

Fetal Dose (mrad) for an Average Study

Chest x-ray (PA and lateral)

,1

Abdominal plain film

200-300

Intravenous pyelogram

400-900

Barium enema

700-1600

Cervical spine x-ray

,1

Dorsal spine x-ray

,1

Lumbar spine x-ray

400-600

Lumbosacral area

200-600

Upper GI series

50-400

Dental x-rays

0.01

Mammography

Negligible

CT scan of chest

30

CT scan of abdomen

250

Perfusion lung scan with Tc 99m

6-12

Ventilation lung scan

1-19

Pulmonary angiography via femoral route

221-374

Pulmonary angiography via brachial route

,50

CT, computed tomography; GI, gastrointestinal; PA, posteroanterior.

138

SECTION II  THE FETUS

Airline Flights The amount of cosmic radiation received during commercial flights is much less than the levels associated with fetal risk. A transcontinental flight across the United States would result in an exposure of approximately 6 mrad.9 Pregnant women should keep well hydrated and move their lower extremities when possible to avoid venous stasis and thromboembolic events.

CONGENITAL ANOMALIES AND ULTRASONOGRAPHY Ultrasonography has become an important tool in obstetrics and gynecology over the past two decades. Geneticists use ultrasonography as a way to characterize dysmorphology in utero. Congenital abnormalities occur in 3% to 4% of live births and have several causes—single-gene defects, chromosomal abnormalities, teratogens, multifactorial, or unknown reasons. Congenital anomalies can be divided into three categories depending on the mechanism underlying the defect: malformations, deformations, and disruptions. Malformations involve intrinsic abnormalities in the genetic programs controlling development. The syndactyly that is associated with Apert syndrome is the result of a mutation in a gene encoding for fibroblast growth factor receptors. Conversely, deformations are caused by extrinsic factors physically impinging on otherwise normal tissue. An example of this is arthrogryposis, or contractures of the extremities. This condition can be caused by a prolonged leakage of amniotic fluid, resulting in fetal crowding. Disruptions, the final category, are the consequence of fetal tissue destruction. They can be the result of vascular insufficiency or mechanical damage. Amniotic band syndrome results in the partial amputation of a limb. Assessment of an infant with birth defects requires a careful review of prenatal exposures and maternal illnesses. When evaluating birth defects, it is important also to consider whether a defect occurred in isolation or is part of a pattern, which may suggest a syndrome. Imaging studies or laboratory examinations may also help elucidate the underlying pathophysiology of an anomaly. ACOG states that “ultrasound examination is safe for the fetus and can be used for the diagnosis of many major and minor fetal anomalies.”6 Ultrasound examinations fall into three categories: limited, standard, or specialized. A limited examination is one that is performed to address a specific question, such as placental location, or to confirm fetal cardiac activity. It cannot replace a standard ultrasound examination, which is ideally performed between 18 and 20 weeks’ gestation. A specialized ultrasound examination, such as fetal echocardiography, fetal Doppler studies, or additional biometry, is performed when an anomaly is suspected based on family history, laboratory aberrations, or discovery during standard ultrasound examination. An abnormality seen on ultrasound examination is a common reason that patients opt to undergo invasive prenatal diagnosis. An example of the utility of ultrasonography is the case of fetal pyelectasis, or hydronephrosis. Defined as a renal pelvic

diameter of greater than or equal to 4 mm, this finding is present in 3% of euploid fetuses and is a relatively common finding. Postnatal evaluation shows that although most cases resolve, conditions such as ureteropelvic junction obstruction and vesicoureteral reflux may be identified in some infants.40 The fact that the potential issue can be identified prenatally and results in postnatal follow-up obviates the possibility of a child with chronic renal impairment owing to silent infection and inflammation. In a review of 36 published studies, the overall sensitivity for detecting fetal anomalies was 40.4% (range 13.3% to 82.4%), indicating that these detection rates greatly depend on the incidence of anomalies in the population studied and sonographer experience. It has been observed that 50% of fetuses with Down syndrome have no findings on prenatal ultrasound examination. When an examination is performed, the patient should be counseled regarding the benefits and limitations of ultrasonography.6 First-trimester ultrasonography has been traditionally used to confirm an intrauterine pregnancy, to estimate gestational age, and to evaluate pelvic anatomy. Since the introduction of first-trimester aneuploidy screening (see separate section), there has been interest in screening for structural anomalies in the first trimester. There is now extensive literature on the use of first-trimester testing in the detection of fetal anomalies, and in expert hands, a very thorough examination is possible. Prospective studies have shown a 44% to 50% detection rate of major anomalies in the first trimester.15 Further refinement in this technique as a screening tool is required before it can be considered part of routine practice. Three-dimensional ultrasonography has some unique advantages over two-dimensional ultrasonography. Its ability to acquire and manipulate an unlimited number of planes allows for the quantification of organ volume and the ascertainment of images inaccessible by two-dimensional technology. Despite these technical advantages, however, consistent utility has not been determined. Potential areas of clinical benefit include evaluation of NTDs and facial anomalies.

SCREENING MODALITIES Screening for Aneuploidies Current guidelines from ACMG and ACOG recommend that all pregnant women, regardless of age, have the option to undergo invasive diagnostic testing for fetal aneuploidy. After reviewing the benefits and limitations of screening and diagnostic modalities with their health care provider, patients have the option to undergo screening tests for fetal aneuploidy and NTDs (Table 8-5).

FIRST-TRIMESTER SCREENING First-trimester screening, which involves an ultrasound examination and ascertainment of serum markers, is performed between 10 4/7 and 13 4/7 weeks’ gestation. The ultrasound examination involves measurement of the nuchal translucency, a fluid-filled space behind the fetal neck. An increased nuchal translucency is significantly associated with aneuploidy, including Down syndrome. ACOG recommends that patients with a fetal nuchal translucency of 3.5 mm or greater be offered targeted ultrasonography

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

139

TABLE 8–5  E  levation and Depression of Parameters Used in First-Trimester and Second-Trimester Screening Tests FIRST-TRIMESTER SCREEN

SECOND-TRIMESTER SCREEN

Nuchal Translucency

PAPP-A

Free b-hCG

uE3

AFP

Free b-hCG

Inhibin A

Trisomy 21

Increased

Decreased

Increased

Decreased

Decreased

Increased

Increased

Trisomy 18

Increased

Decreased

Decreased

Decreased

Decreased

Decreased

Unchanged

Trisomy 13

Increased

Decreased

Decreased

Decreased

Decreased

Decreased

Unchanged

Neural tube defect

NA

NA

NA

Unchanged

Increased

Unchanged

Unchanged

AFP, a-fetoprotein; b-hCG, human chorionic gonadotropin b subunit; NA, not applicable; PAPP-A, pregnancy-associated plasma protein A; uE3, unconjugated estriol. From Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, 2007, Saunders, p 449.

with or without echocardiogram. Detection of Down syndrome with nuchal translucency alone is 70%. When combined with two serum markers, pregnancy-associated plasma protein A (PAPP-A) and human chorionic gonadotropin (hCG), the detection rate increased to 79% to 90%. Results are reported as an adjusted risk for Down syndrome. Patients should be informed of this number to make decisions.12

SECOND-TRIMESTER SCREENING The second-trimester maternal serum screen for aneuploidy, particularly trisomies 18 and 21, and NTDs is performed between 15 and 20 6/7 weeks of pregnancy. Maternal serum a-fetoprotein (AFP), hCG, and unconjugated estriol, known as the “triple screen,” has an aneuploidy detection rate of 65%. With the addition of a fourth marker, inhibin-A, to create the “quad screen,” the detection rate increases to approximately 80%. The detection of NTDs is discussed subsequently. Results are reported as age-adjusted risks and, similar to first-trimester screening, should be reported to patients to make decisions regarding their desire for invasive testing. Strategies to increase the detection rate to greater than 90% include the integrated screen, in which nuchal translucency measurement and PAPP-A results from the first trimester are combined with the quad screen in the second trimester. The downside for this strategy is that the result is not available until the second trimester. Sequential screening involves disclosure of the patient’s first-trimester screening results. If the patient opts not to undergo invasive testing, the quad screen is performed in the second trimester, and the second-trimester risk is calculated with consideration of the first-trimester screening results.13

Screening for Multifactorial Disorders SERUM a-FETOPROTEIN Maternal serum AFP is an important tool in screening for NTDs. AFP is a fetal-specific molecule that is synthesized by the fetal yolk sac, gastrointestinal tract, and liver. The maternal serum AFP concentration is usually significantly lower than fetal plasma or amniotic fluid. Drawn between 15 and 20 weeks’ gestation, maternal serum AFP screening should be offered to all pregnant

women. Although it is intended as a screening test for open NTDs, other abnormalities, such as ventral wall defects, may also cause an increase in maternal serum AFP. Results from the examination are reported as multiples of mean (MoM) based on gestational age. A value greater than 2.5 MoM is considered abnormal. Clinicians may opt to repeat the test after one moderately elevated result because one third are below the threshold on repeat analysis. If this occurs, a return to below threshold has not been associated with an increase in false-negative results.14 If the value remains persistently elevated, or the clinician opts not to repeat the test, the next step is to perform a detailed ultrasound examination and discuss the option of amniocentesis with the patient. Amniotic fluid AFP and amniotic fluid acetylcholinesterase (AChE) are the two main analytes evaluated for detection of NTDs. An elevation of AFP and AChE values suggests an open fetal NTD with 96% accuracy with a false-positive rate of 0.14%. Contamination of the amniotic fluid sample with blood accounts for half of false-positive results.37 Fetal karyotype should also be performed using the amniotic fluid obtained. An elevated maternal serum AFP in the second trimester that cannot be explained by fetal structural abnormality or underlying maternal conditions is associated with poor fetal outcome, including increased risk of intrauterine fetal demise, placental abruption, and preeclampsia.25

Screening for Mendelian Disorders Preconceptual and prenatal parental carrier screening is an effective way to prevent or prepare for neonatal diseases. Discussions of testing strategies for common mendelian disorders follow.

SCREENING FOR HEMOGLOBINOPATHIES As part of routine obstetric care, all women in the United States have a complete blood count early in pregnancy. This simple test helps identify women at risk for hemoglobinopathies and thalassemias. According to ACOG guidelines, women of African descent should have a hemoglobin electrophoresis along with a complete blood count. Additional factors to help distinguish couples at risk for a child affected by a hemoglobinopathy or thalassemia include microcytic anemia with

140

SECTION II  THE FETUS

normal iron status; family history of a hemoglobinopathy or thalassemia; or a high-risk ethnic background, such as Mediterranean or Southeast Asian descent. Women who have any of these risk factors should also undergo hemoglobin electrophoresis. If a hemoglobin variant or thalassemia is discovered, the woman’s partner should be evaluated. When the father is determined also to be a carrier or is unavailable for testing, the patient should be offered genetic risk assessment and prenatal testing.5

CARRIER SCREENING FOR CYSTIC FIBROSIS Cystic fibrosis, the most common autosomal recessive disease in live-born infants, is a multisystem disorder that is the result of mutations of a large gene on chromosome 7 that encodes the cystic fibrosis transmembrane conductance regulator (CFTR) protein. The condition is characterized by chronic pulmonary disease, pancreatic insufficiency, and liver disease. Although there have been more than 1200 mutations identified in the CFTR gene, certain mutations, such as DF508, have a much greater incidence in the population. In whites, the carrier rate is 1 in 25, and DF508 accounts for 75% of cystic fibrosis cases. The carrier rate differs based on ethnic origin; individuals of Asian descent have the lowest carrier rate, approximately 1 in 90 (Table 8-6).2 ACOG recommends that carrier screening be offered to all couples regardless of ethnicity. It is important for providers to discuss with their patients that a negative screen decreases the risk of being a carrier, but does not eliminate it completely because screening does not test for all possible mutations.

JEWISH CARRIER SCREENING The founder effect refers to the overrepresentation of specific alleles in an inbred population. This mechanism has been documented in the Jewish community, particularly individuals of Ashkenazi heritage (eastern European). Tay-Sachs disease is the most well known of these disorders. Because of communal efforts and professional guidelines, the incidence of this disease has declined remarkably, however, to the point that most children born with this disorder are not of Jewish heritage. Carrier screening, prenatal diagnosis, and culturally sensitive marital planning in regard to this autosomal recessive disease all were applied in this effort dating back to the 1970s. Since that time, the underlying genetic mutations are now known for several more similar diseases

known collectively as the Jewish genetic disorders. As a group, the carrier frequency is very significant—one in five Ashkenazi Jews carries a mutation for one of these autosomal recessive disorders. Culturally sensitive preconceptual and prenatal carrier screening is an important tool in preventing or preparing for these devastating diseases. The current recommendation from ACMG is to offer carrier screening to individuals who identify their background as Ashkenazi Jewish for the following diseases: cystic fibrosis, Canavan disease, familial dysautonomia, Tay-Sachs disease, Fanconi anemia (group C), Niemann-Pick (type A) disease, Bloom syndrome, mucolipidosis IV, and Gaucher disease (Table 8-7). Other disorders, such as glycogen storage disease type Ia and maple syrup urine disease, are likely to join this panel in the future. In couples with one partner of Ashkenazi Jewish background, that partner should be screened first, and if he or she is found to have a positive result, the other partner (regardless of background) should be screened for that disorder.16

CARRIER SCREENING FOR SPINAL MUSCULAR ATROPHY Spinal muscular atrophy (SMA) is a progressive neuromuscular disease resulting from degeneration of spinal a motor neurons. Childhood SMA is divided into three clinical groups, types I, II, and III SMA, depending on age of onset and clinical sequelae. With a live birth rate of 1 in 10,000, SMA is the second most common fatal autosomal recessive disorder after cystic fibrosis. The SMA-determining gene, SMA1, is located on 5q13. The carrier rate of an abnormal allele is 1 in 40 to 1 in 60, and is consistent throughout all ethnic groups. In 95% of cases, this abnormality is a deletion in exon 7 of SMA1. ACMG has recommended that voluntary SMA carrier testing be offered to all couples regardless of race or ethnicity. The sensitivity of the carrier screening test is approximately 94%. A negative screening test for both partners reduces the possibility of an affected offspring significantly, but does not eliminate it entirely.30

DIAGNOSTIC MODALITIES After genetic counseling, some couples may opt to undergo diagnostic testing as opposed to prenatal screening. Others choose diagnostic testing after an abnormal screening result or ultrasound finding.

TABLE 8–6  C  arrier Rates Based on Ethnicity and Likelihood of Being a Carrier for Cystic Fibrosis after a Negative Screening Result Carrier Rate Before Testing

Carrier Risk after Negative Test Result

Ashkenazi Jewish

1/24

,1/400

Non-Hispanic white

1/25

,1/208

Hispanic American

1/46

,1/164

African American

1/65

,1/186

Asian American

1/94

,1/184

Racial or Ethnic Group

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

141

TABLE 8–7  Jewish Genetic Disorders: Carrier Frequency and Disease Characteristics Disease Name

Carrier Frequency

Disease Characteristics

Gaucher disease type 1

1:18

Lysosomal storage disease caused by deficiency of the enzyme glucocerebrosidase. Enlarged spleen, liver malfunction, bone marrow dysfunction result from an accumulation of its substrate, glucocerebroside. Depending on disease severity, affected individuals can survive into adulthood

Tay-Sachs disease

1:31

Deficiency of enzyme hexosaminidase A, which results in accumulation of fatty acids known as gangliosides in the brain. Children with Tay-Sachs appear normal and then develop progressive mental and physical deterioration, including blindness, deafness, and dysphagia. Different variants are characterized by age of onset

Familial dysautonomia

1:31

Result of mutations in the IKBKAP gene, which encodes for the IKAP protein (IkB kinase complex associated protein). Presence of the abnormal protein affects the development of neurons in the sensory and autonomic nervous system resulting in pain insensitivity, poor growth, and labile blood pressure

Canavan disease

1:41

Defective ASPA gene results in accumulation of N-acetyl aspartate, which interferes with the neuronal myelin sheath. Symptoms include mental retardation, loss of motor function, and megalocephaly

Fanconi anemia group C

1:89

Mutations in 13 different genes involved in DNA repair complex are known to cause Fanconi anemia, characterized by skeletal anomalies, short stature, and increased risk of malignancy and aplastic anemia

Niemann-Pick disease type A

1:90

Lysosomal storage disease caused by a mutation in the SMPD1 gene, which results in accumulation of sphingolipids in neuronal, hepatic, and splenic tissue

Bloom syndrome

1:107

Mutations in the BLM gene, a member of the DNA helicase family, result in increased susceptibility to early cancers and phenotypic features that include high-pitched voice, distinct facies, telangiectasias, skin pigment changes, and hypogonadism

Mucolipidosis IV

1:127

Disruption of a nonselective cation channel, TRPML1, results in abnormal lysosomal storage and neurodegeneration

Chorionic Villus Sampling Chorionic villus sampling (CVS) involves procuring a small sample of the placenta for genetic diagnosis. Although not fetal tissue, the placenta is embryologically derived from the same trophoblastic cells as the fetus, and most often has the same karyotype as the fetus. Usually performed between 10 and 13 weeks’ gestation, CVS is an ambulatory procedure. Using ultrasound guidance, the placental villi can be obtained through a transcervical or transabdominal approach depending on placental location. CVS enables diagnosis of genetic disorders in the first trimester, giving patients more time to make decisions about the pregnancy, including the opportunity for first-trimester termination if they choose. When the villi have been obtained, the medium is placed onto a plastic tissue culture dish to evaluate the sample. If the

sample appears inadequate, the operator may opt to take a second pass to obtain more villi. Blood clots and maternal decidua can be separated from the villi, and these cleaned villi can be transferred to different media for further evaluation, including fluorescence in situ hybridization (FISH) studies and long-term culture. AFP testing cannot be performed on CVS tissue samples. It is impossible to assess risk of NTDs through CVS. The most serious complications associated with CVS are damage to the fetus and pregnancy loss. CVS has been associated with an increased risk of transverse limb and oromandibular defects. It is hypothesized that these defects are caused by disruption of the vascular system. The overall risk for transverse limb defects after CVS is approximately 1 in 3000, with most defects occurring if CVS is performed before 9 weeks’ gestation.11 Women contemplating CVS can be reassured that

142

SECTION II  THE FETUS

if performed after 9 weeks’ gestation, the risk of limb defects associated with CVS is low and approaches the baseline population risk.4 The pregnancy loss rate after CVS is complicated by the fact that the background risk of pregnancy loss in the first trimester is greater than in the second trimester. Operator experience, number of insertions, and gauge of catheter used all play a part in determining the procedurerelated risk of CVS. More recent data suggest that the pregnancy loss rate attributable to CVS approaches the rate of midtrimester amniocentesis, approximately 1 in 500 pregnancies.27 Two other important considerations with CVS are sampling success and the concept of confined placental mosaicism. On first attempt, transabdominal CVS results in an adequate sample 98% of the time versus 96% for transcervical CVS.33 Culture of mesenchymal cells of CVS samples reveals the existence of trisomies not present in the conceptus in approximately 1% of cases. In this scenario, an amniocentesis is warranted to confirm the fetal karyotype. Patients should be informed of the small risk of an inadequate sample or the need for additional testing secondary to placental mosaicism during their genetic counseling session, before opting for diagnostic testing.

soon after the procedure is over. Patients are encouraged to keep themselves well hydrated and to avoid strenuous physical activity, including intercourse, for 2 days. Fetal loss is the most devastating potential complication, and the risk attributable to amniocentesis is approximately 1 in 500, although more recent data suggest that it may be even lower.3 Increased maternal age, a history of bleeding associated with pregnancy, and abnormal serum screening results all increase the risk of fetal loss. Patients are cautioned regarding fetal injury, although this is an extraordinarily rare event with the advent of ultrasound guidance. Leakage of fluid can occur in approximately 1% of cases; however, in contrast to spontaneous rupture of membranes that occurs outside the setting of amniocentesis, 90% are self-limited and resolve spontaneously within 1 week.10 If a patient reports leakage, a conservative approach—one that monitors amniotic fluid volume, fetal growth, and signs of infection—should be adopted because the prognosis is excellent. Overall, long-term follow-up on offspring of women who had undergone amniocentesis did not show a significantly higher rate of major disabilities compared with matched controls.8

Amniocentesis

A third and much more infrequently used modality of diagnostic testing is cordocentesis, also known as percutaneous umbilical blood sampling. This procedure involves puncturing the umbilical vein under ultrasound guidance to obtain fetal blood cells for genetic analysis. Usually performed after 18 weeks, the procedure-related pregnancy loss rate is approximately 1 in 100 when the procedure is performed for genetic diagnosis.24 Given the high loss rate compared with CVS or amniocentesis, use of this test for genetic diagnostic purposes is limited to further evaluation of chromosomal mosaicism discovered on a CVS or amniocentesis result. Instead, most cordocenteses are performed to eva­ luate and treat fetal anemia, especially in the case of Rh sensitization.

Amniocentesis, a procedure to withdraw amniotic fluid from the uterine cavity, is most commonly performed either for prenatal genetic studies or for evaluation of fetal lung maturity (performed in the third trimester). It can also be used as a diagnostic tool for evaluation of intra-amniotic infection or as a therapeutic procedure to remove excess amniotic fluid. The rest of this section focuses on amniocentesis for the purposes of genetic diagnosis. Although technically possible after 11 weeks’ gestation, an amniocentesis is usually performed between 15 and 22 weeks’ gestation. Before 14 weeks’ gestation, the complication rate, particularly orthopedic malformations, is higher because of incomplete fusion of the amnion, chorion, and decidua parietalis. After 22 weeks, two potentially serious scenarios exist: (1) the procedure provokes labor or rupture of membranes resulting in a periviable delivery, or (2) abnormal results arrive after the fetus has entered into the 24th week of gestation, eliminating the option of termination in most states. Similar to a transabdominal CVS, the patient is placed in supine position. An ultrasound survey to assess viability, fetal position, placental location, and a gross anatomic survey is performed before the procedure. Amniotic fluid, 20 to 30 mL, is aspirated into sterile syringes. The amniocytes and desquamated fetal cells floating in the amniotic fluid provide a source of mitotically active cells for cytogenetic evaluation and culture. Levels of AFP and AChE in the amniotic fluid can help identify fetuses at high risk for NTDs. FISH results (see later) are usually available in 2 days, whereas results from cell culture take approximately 7 to 14 days. The fetal heart rate is assessed and documented after the procedure. It is normal for a patient to experience uterine cramping during the procedure, but cramping should resolve

Cordocentesis

Preimplantation Genetic Diagnosis Preimplantation genetic diagnosis (PGD) is performed by genetic analysis of a single cell biopsy specimen from an in vitro embryo at the eight-cell stage. Alternatively, polar body biopsy specimens may likewise be obtained before transfer. The nuclear material can be analyzed using molecular or cytogenetic techniques. There is now a wealth of clinical experience using this technology, and it is particularly useful in the assessment of known genetic disorders at the time of preconception. Although some families undergo PGD because of ethical concerns related to termination of pregnancy, some patients choose this approach after having undergone several affected pregnancies because they do not want to take the risk of another untoward and emotionally devastating outcome. Although this technology can also be used to determine whether or not a parent has transmitted a known chromosomal abnormality to his or her offspring, routine screening for aneuploidy remains controversial, and current evidence does not support its use for preimplantation screening of embryos.29

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

The advantage of PGD is that it allows couples to avoid intrauterine transfer of embryos affected by a detectable disorder. The disadvantage is that to perform PGD, a woman must undergo in vitro fertilization (IVF), even if fertility is not an issue, to conceive. PGD does not obviate the risk of fetal structural anomalies as the pregnancy progresses. Infants that are the result of pregnancies with PGD show no difference in the rate of fetal anomalies compared with infants that result from ART alone.36

Fetal DNA from Maternal Serum Cell-free fetal DNA is present in maternal circulation throughout pregnancy. Examination of this DNA has resulted in determination of fetal rhesus D (RhD) status, gender, and the presence or absence of aneuploidy. At this time, this technique is not reliable or efficient for wide clinical use, however. The idea of noninvasive fetal genetic screening is very appealing, and ongoing research is likely to result in more refined use of cell-free fetal DNA in the future.23

TECHNICAL ADVANCES IN MOLECULAR CYTOGENETICS Molecular and cytogenetic diagnostics are an important part of genetic counseling. Increasing understanding of the human genome has led to the development of efficient, accurate testing modalities. Following is an overview of the most commonly used diagnostic tools. Chromosomal analysis, or karyotyping, is best performed when the cell is in prometaphase or metaphase and the chromosomes are condensed. The most common staining technique, a Giemsa stain, results in a distinct pattern of alternating light and dark bands that is dependent on the composition of the underlying DNA sequence. Each chromosome has a unique banding pattern. Analysis of this banding pattern can show structural abnormalities. The disadvantage of this technique is that the sensitivity of banding is limited, which means that small structural abnormalities or mutations would go undiagnosed. The advantage of this technique is that the whole genome is visualized at one time. FISH is a diagnostic tool that has better resolution than traditional chromosomal banding. After denaturation, a DNA probe labeled with a fluorochrome is directly hybridized onto cells that have been fixed onto a glass slide. The fluorescent signal is immediately visible by fluorescence microscopy. A double signal corresponds to one signal on each chromosome pair. Cells with one nuclear signal are monosomic for the chromosomal region being evaluated. Conversely, trisomic cells have three nuclear signals. The advantage of FISH is that it can be applied to nondividing and dividing cells. It can identify several different types of mutations—deletions, duplications, and aneuploidy. Its major disadvantage is that structural abnormalities are missed because the DNA probe detects only the presence of a genetic sequence, not its location. Although FISH is a rapid test, its use is currently limited to a few aneuploidy and deletion syndromes. FISH is an important tool for patient care in labor and delivery and the neonatal intensive care unit when quick decision making is required. Because of its limitations,

143

however, a full karyotype is still performed before formulating a definitive report. Polymerase chain reaction (PCR) is a process to amplify a single DNA molecule into thousands of copies in a matter of hours. The double-stranded DNA of interest is heated to separate into two single strands. Single-stranded primers, usually 20 to 30 nucleotides in length, are mixed in to allow them to anneal to the opposite DNA strands of the desired target. Next, a thermostable DNA polymerase and single nucleotides are introduced, and the complementary strand is synthesized using the primers as sites of initiation. The advantage of PCR is that it is fast and extremely sensitive, meaning that PCR can amplify DNA from a single cell. Because of this sensitivity, however, contamination with even small amounts of other DNA can produce an erroneous result. To design appropriate primers, the nucleotide sequence information of the target region must be known. Quantitative PCR is a newer application of standard PCR in which the accumulation of PCR products over time is measured directly with the use of fluorescent probes that hybridize to the target sequence. This probe begins to fluoresce when the DNA polymerase cleaves it. The fluorescent signal increases proportionally to the amount of PCR product, and this signal is quantified by comparing the cycle number at which the sample reaches a predetermined level of fluorescence with a standardized curve of a control. In some countries, this technique is preferred over FISH analysis.19 Duplications or deletions that are too small to see by chromosomal banding can still have important clinical sequelae. These small changes can be detected by comparative genomic hybridization. This technique measures the differences in copy number, or dosage, of DNA segments being evaluated. Labeled control and sample DNA should fluoresce in a 1:1 ratio. If the sample DNA has a deletion, this ratio of sample to control changes to 0.5:1. In addition to prenatal diagnosis, comparative genomic hybridization has applications in pediatric diagnosis. Although this technology holds great promise in reaching the goal of rapid, high-resolution cytogenetic evaluation, an area of concern is how to interpret copy number variants. In humans, the amount of variability can cause difficulty in interpretation of results, particularly when using oligoarray technology. Determining the significance of copy number variants would be an important development in the use of this promising technology.35

ASSISTED REPRODUCTIVE TECHNOLOGIES More than 40,000 infants born in the United States each year are conceived through ART. The outcomes of techniques such as IVF and intracytoplasmic sperm injection have been well examined. The risks of ART can be divided into two categories—obstetric risks and fetal risks. Thirty percent of ART pregnancies are twins or higher order multiples, which inherently carries an increased risk of premature delivery. Singleton IVF pregnancies are also at an increased risk of preterm delivery, a twofold risk, compared with spontaneously conceived singleton pregnancies. IVF singleton pregnancies also have increased risk for other morbidities, including abnormal placentation, preeclampsia, and cesarean delivery.31

144

SECTION II  THE FETUS

Fetal risks when comparing IVF singletons versus spontaneous conception singletons include low birthweight, intrauterine fetal demise, and cerebral palsy.31 All these risks are approximately twice baseline risk. Large population-based studies revealed that the adjusted odds ratio for major congenital malformations for IVF fetuses was 1.3. Chromosomal abnormalities when adjusting for maternal age are similar between IVF and spontaneous conceptions.18 The risk of de novo chromosomal abnormalities seems to triple, however, if intracytoplasmic sperm injection is performed. It has been observed that pregnancies that are the result of ART are at an increased risk for rare imprinting disorders, such as Beckwith-Wiedemann syndrome, suggesting that epigenetic changes may occur as a result of ART.17,38 Although there may be increased morbidity related to ART, there is still no clear evidence as to whether these issues are related to the procedure or to the underlying cause of infertility, as there can never be a true control group of children who were born to infertile couples who did not undergo ART. In other words, there are two important considerations in this regard—the impact of gamete manipulation and the inherent characteristics of the infertile population itself on pregnancy outcome. Handling of oocytes or embryos during critical periods of development could affect important processes. The average maternal age at time of conception with ART is higher than spontaneous conceptions. Male patients with infertility who require intracytoplasmic sperm injection have an increased incidence of genetic abnormalities themselves. Future studies to clarify the role of each of these considerations will be important in clearly defining the risks to the developing fetus.

GENETIC EVALUATION AND COUNSELING Genetic counseling is an ever-evolving field that has become an important part of clinical practice. Full of ethical and practical challenges, genetic counseling focuses on assessing a patient’s genetic risk to provide individualized information and options. Some level of genetic counseling is provided by health care professionals in all fields of medicine. Obstetricians spend time with their pregnant patients discussing prenatal screening and diagnosis, whereas oncologists review hereditary cancer susceptibility. Patients can also be referred to genetic counselors, formally trained professionals who have received a Master’s level degree and are certified by the American Board of Genetic Counseling. Genetic counselors are trained not only in pedigree construction and analysis, but also in communication education and counseling skills, and are a distinct discipline from medical geneticists, who are certified by the American Board of Medical Genetics after completing a residency program. Although such physicians also provide counseling, their focus is more on diagnosis, treatment, and management of genetic disease. In many centers, physicians and counselors work as a team to provide comprehensive genetic services. Indications for referral are varied, but include (1) family history of earlyonset cancer, (2) personal or family history of known or suspected hereditary disease, (3) ethnic background associated with an increased risk of a heritable disorder, (4) teratogen exposure during pregnancy, (5) abnormal prenatal ultrasound

or abnormal first-trimester or second-trimester screening results, and (6) recurrent pregnancy loss. Any pregnancy at risk of birth defects warrants such a referral. A genetic counseling visit entails obtaining a detailed medical and family history, including the age and health status of first-degree, second-degree, and third-degree relatives. For the prenatal patient, additional information such as genetic screening results, ultrasound findings, and possible teratogenic exposures is discussed. This information allows for a targeted discussion regarding the likelihood of developing disease, testing options for the condition, the impact that an illness could have on the patient and family, and the possible interventions available to modify the disease. Ideally, genetic counseling is provided in a nondirective manner; emphasis is placed on educating the patient on his or her options and the consequences of those options. Before initiating any testing, a provider should ensure that proper consent is obtained. It is up to the provider to ensure that the patient understands the nature of the test, its limitations, and its potential sequelae. This becomes more complex when testing children or adolescents is considered. In this situation, the benefits of timely genetic testing should be the primary justification. The American Society of Human Genetics (ASHG) suggests that “counseling and communication with the child and family about genetic testing should include advocacy on behalf of the interests of the child.”7 Generally, if there is no immediate benefit to the child, testing is usually deferred until adulthood, when the individual can make his or her own choices.7 Another ethical dilemma of genetic testing that clinicians face is the balance between maintaining patient confidentiality versus the duty to protect other family members who might be affected. An individual’s genetic testing results have implications for an entire family. The ASHG encourages voluntary disclosure by the person tested (proband) whenever possible. If this does not happen, the ASHG recommends that the degree of disclosure to other family members depends on the magnitude and immediacy of risk faced, stating that disclosure is acceptable if “harm is likely to occur, and is serious, immediate, and foreseeable.”3 A concern that many patients have is the impact on their genetic testing results on future employment and ability to obtain health insurance. In 1995, the Equal Employment Opportunity Commission issued guidelines that individuals discriminated against based on their genetic testing had the right to sue. A year later, the Health Insurance Portability and Accountability Act (HIPAA) was enacted to prevent insurance companies from denying coverage based on genetic testing. On a federal level, the Genetic Information Nondiscrimination Act (GINA) was passed in early 2008.43 It prohibits U.S. insurance companies and employers from discriminating on the basis of information derived from genetic tests. Insurers and employers are not allowed under law to request or demand a genetic test. GINA prevents insurance companies from discriminating through reduced coverage, and prohibits employers from making employment decisions based on an individual’s genetic code. State laws that have an impact on the provision of genetic services also exist, and there may still be issues with respect to procurement of life insurance despite the aforementioned laws.

Chapter 8  Genetic Aspects of Perinatal Disease and Prenatal Diagnosis

CONCLUSION An understanding of the genetic basis of disease and the tools available for prenatal genetic evaluation is vital in the management of patients and their offspring in the peripartum and newborn period. Reproductive genetic counseling and diagnostic techniques are a vital means of defining a potential clinical dilemma, allowing for seamless coordination between obstetricians and neonatologists. Medical genetics is a rapidly evolving field with new advances and discoveries reported frequently. The underlying genetic core principles and the clinician’s respect for patient autonomy to make good choices based on complete information have remained unchanged, however.

REFERENCES 1. ACOG Committee Opinion #299: Guidelines for diagnostic imaging during pregnancy, Obstet Gynecol 104:647, 2004. 2. ACOG Committee Opinion #325: Update on carrier screening for cystic fibrosis. 2005. 3. ACOG Committee Opinion #410: Ethical Issues in Genetic Testing. 2008. 4. ACOG Practice Bulletin #77: Screening for fetal chromosomal abnormalities, Obstet Gynecol 109:217, 2007. 5. ACOG Practice Bulletin #78: Hemoglobinopathies in pregnancy, Obstet Gynecol 109:229, 2007. 6. ACOG Practice Bulletin #98: Ultrasonography in pregnancy, Obstet Gynecol 112:951, 2008. 7. American Academy of Pediatrics. Ethical issues with genetic testing in pediatrics, Pediatrics 107:1451, 2001 8. Baird PA et al: Population-based study of long-term outcomes after amniocentesis, Lancet 344:1134, 1994. 9. Barish RJ: In-flight radiation exposure during pregnancy, Obstet Gynecol 103:1326, 2004. 10. Borgida AF et al: Outcome of pregnancies complicated by ruptured membranes after genetic amniocentesis, Am J Obstet Gynecol 183:937, 2000. 11. Botto LD et al: Chorionic villus sampling and transverse digital deficiencies: evidence for anatomic and gestational-age specificity of the digital deficiencies in two studies, Am J Med Genet 62:173, 1996. 12. Driscoll DA, Gross SJ: First trimester diagnosis and screening for fetal aneuploidy, Genet Med 10:73, 2008. 13. Driscoll DA: Second trimester maternal serum screening for fetal open neural tube defects and aneuploidy, ACMG Policy Statement 2004. 14. Evans MI et al: Impact of folic acid fortification in the United States: markedly diminished high maternal serum alphafetoprotein values, Obstet Gynecol 103:474, 2004. 15. Flood K, Malone FD: Screening for fetal abnormalities with ultrasound, Curr Opin Obstet Gynecol 20:139, 2008. 16. Gross SJ et al: Carrier screening in individuals of Ashkenazi Jewish descent, Genet Med 10:54, 2008. 17. Halliday J et al: Beckwith-Wiedemann syndrome and IVF: a case-control study, Am J Hum Genet 75:526, 2004. 18. Hansen M et al: Assisted reproductive technologies and the risk of birth defects—a systematic review, Hum Reprod 20:328, 2005. 19. Jackson L: Cytogenetics and molecular cytogenetics, Clin Obstet Gynecol 45:622, 2002.

145

20. Kronquist KE, Sherman SL: Clinical significance of trinucleotide repeats in fragile X testing: a clarification of American College of Medical Genetics guidelines, Genet Med 10:845, 2008. 21. Lazarou S, Morgentaler A: The effect of aging on spermatogenesis and pregnancy outcomes, Urol Clin North Am 35:331, 2008. 22. Malone FD et al: First-trimester or second-trimester screening, or both, for Down’s syndrome, N Engl J Med 353:2001, 2005. 23. Maron JL, Bianchi DW: Prenatal diagnosis using cell-free nucleic acids in maternal body fluids: a decade of progress, Am J Med Genet 145:5, 2007. 24. Maxwell DJ et al: Fetal blood sampling and pregnancy loss in relation to indication, Br J Obstet Gynaecol 98:892, 1991. 25. Milunsky A et al: Predictive values, relative risks, and overall benefits of high and low maternal serum alpha-fetoprotein screening in singleton pregnancies: new epidemiologic data, Am J Obstet Gynecol 161:291, 1989. 26. Moore K: The developing human: clinically oriented embryology, Saunders, Philadelphia, 1982. 27. Mujezinovic F, Alfirevic Z. Procedure-related complications of amniocentesis and chorionic villous sampling: a systematic review, Obstet Gynecol 110:687, 2007. 28. Nussbaum RL et al: Thompson and Thompson genetics in medicine, 7th ed, Philadelphia, Saunders, 2007. 29. Practice Committee of the Society for Assisted Reproductive Technology; Practice Committee of the American Society for Reproductive Medicine. Preimplantation genetic testing: a Practice Committee opinion, Fertil Steril 88:1497, 2007. 30. Prior TW: Carrier screening for spinal muscular atrophy, Genet Med 10:1, 2008. 31. Reddy U et al: Infertility, assisted reproductive technology, and adverse pregnancy outcomes, Obstet Gynecol 109:967, 2007. 32. Sherman S et al: Fragile X syndrome: diagnostic and carrier testing, Genet Med 7:584, 2005. 33. Smidt-Jensen S et al: Randomised comparison of amniocentesis and transabdominal and transcervical chorionic villus sampling, Lancet 340:1237, 1992. 34. Toriello HV, Mech JM: Statement on guidance for genetic counseling in advanced paternal age, Genet Med 10:457, 2008. 35. Veltman JA: Genomic microarrays in clinical diagnosis, Curr Opin Pediatr 18:598, 2006. 36. Verlinsky Y, Kuliev A: Current status of preimplantation diagnosis for single gene disorders, Reprod Biomed Online 7:145, 2003. 37. Wald N et al: Amniotic fluid acetylcholinesterase measurement in the prenatal diagnosis of open neural tube defects. Second report of the Collaborative Acetylcholinesterase Study, Prenat Diagn 9:813, 1989. 38. Wilkins-Haug L: Assisted reproductive technology, congenital malformations, and epigenetic disease, Clin Obstet Gynecol 51:96, 2008. 39. Wilson RD et al: The use of folic acid for the prevention of neural tube defects and other congenital anomalies, J Obstet Gynaecol Can 25:959, 2003. 40. Woodward M, Frank D: Postnatal management of antenatal hydronephrosis, BJU Int 89:149, 2002. 41. www.cdc.gov/nchs/data/nvsr/nvsr52/nvsr52_19acc.pdf\. Accessed October 30, 2008. 42. www.ncbi.nlm.nih.gov/omim/. Accessed November 1, 2008. 43. www.ornl.gov/sci/techresources/Human_Genome/elsi/legislat. Accessed November 10, 2008.

CHAPTER

9

Perinatal Imaging Nancy E. Judge, Stuart C. Morrison, and Noam Lazebnik

Ultrasound has permanently changed the image of perinatology, increasing expectations and improving outcomes in neonatal care. In the past decade, sonographic genetic screening, Doppler vascular measurements, and threedimensional examinations have moved from investigational techniques to widely accepted diagnostic tools. Gradually, the field of obstetric ultrasound (US) has matured, establishing its relevance and its limitations in antenatal diagnosis. Most recently, magnetic resonance imaging (MRI) has joined sonography for the prenatal diagnosis of an ever broader spectrum of disorders.

FETAL IMAGING TECHNIQUES Real-time ultrasound, in which image brightness varies with the intensity of returning signals (B-mode), is the standard method of fetal imaging (Fig. 9-1). Typically, brief signal bursts are followed by relatively long receptive intervals; lower signal frequencies encounter less interference but reflect from fewer informative interfaces. Image quality varies with distance to the target, structure size, movement relative to the signal, and tissue transmission characteristics. Ideally, the transducer is in proximity to its target; transvaginal scans facilitate gynecologic examinations and early gestation and cervical studies. Suboptimal images are common with obesity, limited fluid interfaces, intervening structures, and poor positioning with respect to the beam. Diagnostic difficulty also increases when different materials have similar echo characteristics, such as blood, urine, and ascites. M-mode ultrasound is a direct representation of beam reflection by moving edges (e.g., in cardiovascular imaging). Interpretation requires hard-to-achieve, standardized stable views; fetal M-mode imaging has been deemphasized because of improved B-mode resolution. M-mode is useful in assessing arrhythmia, contractility, and pericardial effusion. M-mode “snapshots” also document cardiac activity and rate (Fig. 9-2). Doppler ultrasound uses the frequency shift that occurs when the beam is reflected off moving objects to demonstrate the presence, velocity, and direction of blood flow.

Pulsed waves are used to determine flow velocity from individual vessels (Fig. 9-3). Direct volume calculations from narrow, tortuous fetal and uterine vessels are inaccurate. To compensate, prenatal Doppler findings, ideally obtained at beam angles less than 60 degrees, are generally expressed as ratios. Color Doppler ultrasound semiquantitatively assigns direction to flow; by convention, warm colors denote movement toward the transducer, and saturation is keyed to velocity. Color Doppler illuminates cardiac, arterial, and venous structures (Figs. 9-4 and 9-5); any moving structure is potentially amenable to color Doppler detection. Color Doppler energy (power Doppler) is based on signal intensity; amplitude corresponds to blood cell motion. Color Doppler is effectively independent of angulation and is sensitive to very low flow, thus it is helpful for mapping vascular beds and for quickly spotting umbilical, pulmonary, middle cerebral, pericallosal, and other vessels (Figs. 9-6 and 9-7). Three-dimensional ultrasound (US) acquires and analyzes returning echoes along a third axis. Images are then manipulated electronically to show both surfaces and volumes from multiple perspectives, both as static and real-time (fourdimensional) views. Surface rendering of subtle facial details and of fetal small parts, multiplanar views, and stratified slices enhance detection of anomalies and partly overcome positional limitations of standard scans (Fig. 9-8). Threedimensional studies have facilitated volume calculations, analysis of complex spatial relationships, and have better explicated abnormal findings.9 Spatiotemporal correlation of heart movement with color and power Doppler augments standard cardiac imaging (Fig. 9-9). Skilled practitioners report significant gains in efficiency by off-line interpretation of acquired volumes in lieu of real-time scanning.8 Threedimensional data sets are likely to become a standard format for research, remote interpretation, and consultative purposes. Post hoc analysis of stored volumes during litigation might also be expected. Three-dimensional studies can be significantly hampered by artifacts and erratic resolution, memory-intensive archival requirements, and inherent dependence on two-dimensional signal quality.

147

148

SECTION II  THE FETUS

Figure 9–1.  Transabdominal B-mode, two-dimensional scan.

Profile of 20-week fetus.

Figure 9–4.  Color flow Doppler demonstration of a right

pelvic kidney (long arrow), descending aorta (thick arrow), and inferior vena cava (short arrow) in a 28-week fetus. Flow toward the transducer is the color of the upper bar; flow away corresponds to the lower bar colors. See color insert.

UC UA UB UA FHR

Figure 9–2.  Transvaginal M-mode demonstration of embry-

Figure 9–5.  Transverse view of fetal pelvis, umbilical cord

onic cardiac activity at 5 5/7 weeks’ gestation. Upper frame: Embryo with cursor across thorax. Lower frame: M-mode display of wall movement during two cardiac cycles, 114 beats per minute (between vertical lines).

(UC), and urinary bladder (UB) with power Doppler showing bifurcation of the umbilical arteries (UA).

Figure 9–3.  Upper frame: Color Doppler highlights a loop of

umbilical cord (trapezoid). The gate (transverse parallel lines) identifies the sampling site within the umbilical artery. Lower frame: Pulse Doppler waveform recorded from the umbilical artery (see Fig. 9-4). See color insert.

Figure 9–6.  Color Doppler demonstration of the internal carotid, middle cerebral, and pericallosal arteries (thick, thin, and curved arrows). Lower velocities are shown as more saturated hues. Turbulent flow and flow more rapid than the scale parameters demonstrate aliasing (mixture of colors). Compare with Figure 9-7, power Doppler of same structures. See color insert.

Chapter 9  Perinatal Imaging

Figure 9–7.  Power Doppler parasagittal cranial image of

internal carotid (short thick arrow), middle cerebral (long arrow), and pericallosal arteries (curved arrow). See color insert.

149

Figure 9–9.  Spatiotemporal color flow imaging of normal four-chamber heart at 18 weeks. Transverse chest with fourchamber view. Upper left: Cardiac apex facing right. Upper right: Axial view. Lower left: Coronal view. Lower right: Cardiac apex facing left (three-dimensional). See color insert.

sequences detect recent hemorrhage and meconium-filled bowel. Diffusion imaging is useful for identifying ischemic injury to the brain. The use of gadolinium is relatively contraindicated but may be justified for assessment of placental accretion or serious maternal disease. Claustrophobia can be difficult for some patients.

*

BIOEFFECTS AND SAFETY

*

Figure 9–8.  Two-dimensional profile (compare with normal profile in Fig. 9-1) and three-dimensional coronal rendering of a 26-week fetus with a large, right-sided cleft lip and palate (asterisk). The fetus is in a frank breech position with the foot (arrow) visible on the head. See color insert.

MRI can now be performed quickly, usually eliminating the need for maternal or fetal sedation. Maternal oblique positioning prevents inferior vena caval compression. Each image is obtained with an ultrafast sequence in less than 1 second. Individually tailored to specific problems or anatomic areas, MRI is not used for comprehensive screening, unlike ultrasound. The large field of view, excellent soft tissue contrast, and multiple planes of construction make MRI an appealing imaging modality. MRI can supplement ultrasounds limited by maternal obesity or oligohydramnios. MRI is usually used in the second and third trimesters to elucidate problems found on earlier ultrasound examinations. Fast T2-weighted sequences with single-shot fast spin-echo techniques are the most commonly performed sequences. Images are acquired in the axial, coronal, and sagittal planes to the fetus or orthogonal to the maternal pelvis. T1-weighted

Most pregnant women with access to modern obstetric care undergo sonographic examinations during the first and later trimesters. There has been no convincing evidence of harm in humans to date,2 although the endemic nature of early ultrasound exposure in modern obstetrics may require ingenuity in future study designs. By definition, ultrasound is inaudible; however, sound energy can produce both heat and pressure effects in tissue. Although not oncogenic, under nonclinical conditions, ultrasound may cause cell lysis, intracellular shearing, streaming effects, altered membrane permeability, and abnormal chromosome function. Small mammals experience smaller litters, impaired growth, and more anomalies after in utero exposures that do not always exceed comparable human levels. Current energy outputs greatly exceed those used in most of the original safety studies, with particular concern aimed at energy output during embryogenesis. Heat exposure triples with each change in modality: from B-mode and three-dimensional, to M-mode, to color flow, before peaking during pulse Doppler. Harmful levels should not be reached during routine studies, but might occur during focal cranial Doppler interrogation or in a febrile patient, without unusually long exposures.1 Mechanical disruption from cavitating gas bubbles is improbable in the fetus. Both temperature and disruption risks are displayed on equipment (Fig. 9-10); thermal index (TI) is a ratio between transducer output and the energy needed to warm tissue 1°C, with a desired value less than 2. Thermal index is further categorized by tissue type: soft tissues, cranial structures, and bone. Mechanical index (MI) refers to pulse amplitude tissue effects of compression and decompression, ideally maintained below

150

SECTION II  THE FETUS

d1 d2

d3

Figure 9–10.  Transvaginal longitudinal and transverse axis

ultrasound images of a 4 4/7-week intrauterine pregnancy, showing a small yolk sac (thin arrows) within the gestational sac (calipers) and echogenic surrounding dicidual reaction (thick arrows). The embryo is not yet visible. Note MI and TI notations at top center. 0.4 in fetal studies.1 These levels may be exceeded by practitioners trying to achieve interpretable images during early gestation or in obese and other difficult to scan patients. For safety, only medically essential examinations should be performed; settings and duration should be the minimum required to achieve adequate views.5 MRI uses strong magnetic fields and radiofrequency waves with no known harmful biologic effects, but large longitudinal studies are lacking. As with ultrasound, heat delivery to the fetus is a recognized hazard; in MRI, however, the maternal surface receives the greatest thermal exposure. Noise from the coils (up to 120 dB) is capable of causing acute hearing damage; tissue attenuation decreases fetal intensities by 25%, to relatively safe levels. Direct magnetic bioeffects remain unproven; the FDA45 states that safety to the fetus “has not been established.”

ETHICAL CONSIDERATIONS The education of physician sonologists encompasses visual recognition and interpretative tasks common to diagnostic imaging, but also demands mastery of specific mechanical skills (or at least an ability to assess the latter in sonographers). Ultrasound is an abbreviated module in general radiology and obstetric training with proportionate underrepresentation on certification examinations. Graduates may be obliged to acquire experience in the field to an undesirable degree, given the realities of medical practice and the serious consequences of error. Both US and MRI are rapidly evolving; clinically relevant frontiers are often explored by collaborative fields that benefit from pooled data, prerelease technology, and funding for staff and statistical analysis. Laudable advice from these investigators to confirm applicability or to undergo formal training before adopting practices may be deemed unrealistic or self-serving by readers with fewer resources. Nuchal translucency (NT), a deceptively simple method for aneuploidy

screening, sounds a cautionary note (Fig. 9-11).44 Despite orderly dissemination, certification courses, ongoing audits, professional society support, and laboratory coordination, in contrast to the near-viral dissemination of many advances, reliability remains a concern, complicating accuracy and counseling. Practitioners should be candid when informing patients of their ability to provide a requested service and assiduous in improving their skills. Professional judgment remains paramount in deciding how and when to incorporate new developments into clinical practice. Prenatal identification of fetal sex for the purpose of selective termination is available for serious X-linked disorders, but has been more widely applied to abort normal females because of a lower perceived value (Fig. 9-12). No fully satisfactory response, whether acquiescence, nondisclosure, or noninspection, has been found for this abhorrent societal bias.25 Multiple gestations may result in a number of dilemmas. Advances and regulation in assisted reproductive technology (ART) have decreased the incidence of multifetal pregnancies, but fetal reduction remains a painful choice for parents

Figure 9–11.  Nuchal lucency measurement (calipers) in a

13-week fetus. The nasal bone is visible (thin arrow) and the amnion (thick arrow) is clearly distinguished from the skin fold. Note TI and MI settings are displayed on this image.

Figure 9–12.  Vulvar skin folds (arrow) of normal female

fetus in the midtrimester.

Chapter 9  Perinatal Imaging

facing the prospect of extreme prematurity in higher order multiples. The management of twin-twin transfusion syndrome, anomalous co-twins, and discordant growth or distress far from term necessitate choosing among unsatisfactory options. Nondiagnostic studies for a suspected diagnosis, varying prognoses for a given diagnosis, and the inherent limitations of US and MRI lead to ethical issues in management and counseling. Physicians and patients both share unrealistic expectations for the predictive accuracy of targeted diagnoses.27 Anomalies and variants linked to Down and other syndromes (“markers”) discovered during routine studies present patients and caregivers with unanticipated, unwelcome choices, particularly if patients explicitly refused serum screening or direct genetic testing.23 Ideally, an informed consent discussion that addresses risks, benefits, consequences, and limitations of US or MRI should be provided by the referring clinician to all patients before they choose to undergo such imaging.34 Given the irreversible nature of birth and abortion, ultimately prospective parents must judge for themselves their tolerance for uncertainty in diagnosis and for imperfection in their offspring.

151

Figure 9–13.  Fetus with trisomy 21 and large atrioventric-

ular canal defect, transverse thorax. Without the presence of a normal crux, color flow shows flow from a common atrium (thin arrow) into a common ventricle (thick arrow). See color insert.

APPLICATIONS OF ULTRASOUND Genetic Screening Genetic screening combines ultrasound and biochemical testing to enhance the detection of chromosomal abnormalities, resulting in greater scrutiny of younger patients and less frequent invasive testing for women older than 35 years (see Chapter 8).4 Different algorithms yield noticeable variations in sensitivity, specificity, and predictive values; cost or convenience may factor in choosing a strategy.21 At least one third of fetuses with Down syndrome will have anomalies, notably endocardial cushion defects (Fig. 9-13), duodenal atresia (Fig. 9-14), and less detectably, small atrioseptal and ventriculoseptal cardiac defects (Fig. 9-15). Two thirds may have associated markers, not limited to increased nuchal lucency and hypoplastic nasal bones in the first trimester; thickened nuchal fold (Fig. 9-16) and nasal hypoplasia; ventriculomegaly; choroid cysts (Fig. 9-17); hypoplasia of the fifth digit, decreased long bone to biparietal diameter ratios, enhanced echogenicity of papillary muscles and bowel, and renal pyelectasis (Fig. 9-18) reported in the second trimester. Screen components may be affected by ethnicity, habitus, build, diet, and fetal sex, as well as by operator-dependent detection rates. The permutations stymie counselors attempting to provide (and patients trying to grasp) basic descriptions of risks and benefits. Doppler documentation of tricuspid regurgitancy and increased ductus venosus resistance in the first trimester seems a promising addition to early screening protocols, but raises concern about the effects of exposures required to obtain optimal views.

Assisted Reproduction Ultrasound is essential for timing and guiding egg retrieval and helpful in embryo transfer; its role in judging endometrial receptivity is less clear. ART results in more twins and

D

S

Figure 9–14.  “Double bubble” sign of duodenal atresia in a

35-week pregnancy, with the dilated stomach (S) and the obstructed proximal duodenum (D) seen on transverse abdominal view, spine to right of image. The finding is uncommon before the late second trimester.

higher-order multiples; early US of embryos, amnionicity, and chorionicity is essential to subsequent management. ART patients also risk ectopic and heterotopic pregnancies at more than double the usual rates. US rapidly identifies abnormal implantations, often allowing for medical or minimally invasive therapy, and guides care of hyperstimulation. Assignment of color to three-dimensional follicle scans promises to speed enumeration of follicles (Fig. 9-19). Saline US studies routinely complement or replace hysterosalpingograms in assessment of uteri and adnexa. MRI may play an expanded role in structural and functional evaluation in the future.

152

SECTION II  THE FETUS

Kidneys

rt

lt 1

2

GB

Figure 9–15.  Ventriculoseptal defect (pointing finger) demon-

strated by color flow Doppler in 28-week fetus. See color insert.

Figure 9–18.  Transverse abdominal view, backup. Gallblad-

der, (arrow GB) at right, with asymmetric renal pelves (calipers) illustrating the normal left renal appearance (short thin arrow) and right pyelectasis/caliectasis (thick arrow). Between 1% and 5% of normal fetuses have pyelectasis, limiting its utility as an isolated marker.

First-Trimester Studies NF

CSP

C

Figure 9–16.  Thickened nuchal fold (NF) calipers at a level

demarcated by the cavum septum pellucidum (CSP) and the cerebellum (C), here noted at 18 weeks, is considered a marker for trisomy 21 (Down syndrome) and cardiac and other anomalies. The finding is infrequent, even in affected infants, limiting its utility.

Traditionally, first-trimester transvaginal scans exclude ectopic implantations and confirm viability by demonstrating an intrauterine “double sac,” yolk sac, and embryo with cardiac activity; identifiable embryos in the salpinx are uncommon. Heterotopic pregnancy (coexisting intrauterine and extrauterine implantation) occurs in less than 0.1% of patients, leading to the pragmatic approach that finding an intrauterine pregnancy usually excludes an ectopic.7 A “double sac” is usually seen transvaginally at levels of 1000 to 1500 international units of human chorionic gonadotropin before 5 1/2 menstrual weeks (see Fig. 9-10). Finding the yolk sac confirms the intrauterine site. The gestational sac diameter grows 1 mm daily in early pregnancy, and the embryonic disc should be visible once sac diameters exceed 15 mm. The maximum embryonic length also increases by 1 mm daily; transvaginal documentation of cardiac activity is expected by the end of the sixth week (4 mm embryonic length).32 Embryonic heart rates increase to more than 160 beats per minute by the ninth week and then decline slightly through the 13th week. Persistent rates less than 100 beats per minute are often linked with abortion, aneuploidy, and anomalies.22 Embryonic surveys are limited and necessarily provisional; early fetal period scans can diagnose a number of entities accurately (Fig. 9-20). Midtrimester confirmation continues to be prudent for the majority of first trimester findings. Later studies retain advantages with respect to the natural history of many anomalies and for visualization of heart, spine, and other problematic structures.

Multiple Gestation

Figure 9–17.  Choroid plexus cysts. Coronal view of the brain shows bilateral choroid plexus cysts (arrows).

(See also Chapter 19.) With an increase in the number of fetuses, scans become more complex, time consuming, and error prone; additionally, determination of zygosity is an essential element of the study. The management of anomalous, discordant, or moribund

Chapter 9  Perinatal Imaging

153

Figure 9–19.  Three-dimensional

color-enhanced volume rendering of seven follicles in the right ovary. The echo-free follicles are outlined by the operator in orthogonal planes. The generated volumes are then displayed in the three-dimensional rendering (lower right) with color correlation. See color insert.

Rt ovary

NT

Yolk sac

Figure 9–20.  Eleven-week embryo with a thickened nuchal

lucency and an omphalocele (thick arrow) containing fetal liver. The fetal liver does not undergo physiologic herniation. A study at 18 weeks identified a lumbosacral spina bifida in addition to confirming the omphalocele. co-twins differs significantly based on chorionicity (Fig. 9-21). US categorization is most accurate for different-sexed twins, but approaches this accuracy in gender concordant pairs by assigning chorionicity based on sac appearance in early gestation. Successful attempts may be made throughout gestation by examining the dividing membranes at their placental origin. Dichorionic diamniotic twins are usually dizygotic, with independent risks for anomalies and placental malfunction. Monochorionic pairs are predictably monozygotic; attrition rates exceed 30% from early abortion, anomalies, and prematurity. Matched and isolated anomalies are both common in

monochorionic gestations; loss of a co-twin may kill its sibling outright or produce neurologic damage in up to one third of survivors. Monochorionic twin-twin transfusion syndrome (TTTS), more common in females, is characterized by unbalanced, shared perfusion that restricts growth and amniotic fluid production in the donor and causes volume overload, cardiac dysfunction, and polyhydramnios in the recipient. Ultrasound staging has been used to time and to guide a variety of ablation strategies. Serial amniocentesis may be helpful in milder cases. Management by these approaches has been modestly effective in decreasing stillbirths and prematurity in TTTS. Monoamniotic twins rarely experience TTTS but have cord entanglements, almost half of which are lethal. Carriage to term and labor are usually avoided in this type of twinning. For all multiple gestations, serial US monitoring to assess growth, well-being, placental performance, and cervical length is common practice.

PREGNANCY EVALUATION More than 50 fetal structures have now been measured by US, with results expressed as a function of gestational age. Commercial equipment comes with a variety of preinstalled nomograms and biometry-based dating and weight estimation algorithms. These should be reviewed for study limitations and population characteristics before using in place of userspecific curves. Biometry can evaluate the size and development of structures given sure dating, but one of the most powerful applications of prenatal US is using biometrics to establish or confirm gestational age.14 Through 22 weeks of gestation, most genetically normal individuals cluster closely on nomographic curves.18 If accelerated or restricted growth supervenes, ultrasound dating is compromised accordingly. One of the first dating parameters developed was the biparietal diameter (BPD), obtainable after parietal bone calcification

154

SECTION II  THE FETUS

Twin amioncity/chorionicity

Di/Di

Di/Mono

Mono/Mono

Figure 9–21.  Diagram (left) of variations in appearance of the dividing membrane for dichorionic, monochorionic, and monoamniotic twins. Upper image shows typical thick first trimester appearance of the chorion (arrow) in dichorionic diamniotic twins. Lower image shows the peaked dividing membrane (arrow) of dichorionic twins in the latter half of pregnancy.

in week 12. Biparietal diameter is measured from the outer to the inner table of the skull, perpendicular to the parietal bones and central falx cerebri. The proper plane contains the cavum septum pellucidum, thalami, third ventricle, and tentorial hiatus, in conjunction with the bony table head circumference (Fig. 9-22). Biparietal diameter and head circumference are

relatively spared in nutritional and perfusional disorders of growth; cranial measurements may be distorted by compression. The abdominal circumference (AC), measured along the outer margin of the abdominal skin line at the level of the gastric bubble and the intrahepatic portion of the umbilical vein at the bifurcation of the portal veins, is the best single predictor of growth aberrations but is less helpful for dating (Fig. 9-23).

HC

BPD

AC

Figure 9–22.  Measurement of the biparietal diameter (verti-

cal calipers) and head circumference. The required landmarks include the midline falx cerebri (long arrow), the cavum septum pellucidum (thin arrow), and the thalami (short arrow). The BPD calipers are placed on the upper outer table and the inner margin to compensate for signal scatter.

Figure 9–23.  The abdominal circumference indirectly reflects hepatic glycogen stores. The correct level should include the gastric bubble (thin arrow) and the junction of the portal veins (thick arrow); the circumference should be fitted to the outer diameter of the skin line.

Chapter 9  Perinatal Imaging

155

Femoral length is the fourth measurement commonly included in biometric formulae. Extraocular and transcerebellar diameters, and humeral and pedal lengths are common additional dating parameters. Special curves are available for nonCaucasian ethnicity or multiple gestation.

Fetal Growth Serial measurements over time are the best way to judge fetal growth. The sac and embryo grow perceptibly each day; by the second trimester, intervals of 2 or 3 weeks are more reliable. Problems arise when attempting simultaneously to assign age and weight percentile without reliable dating. The abdominal circumference is a better indicator of decreased perfusion or increased glycogen storage than the cranial measures; ratios of abdominal circumference to the head and femur amplify the differences but have poor sensitivity and specificity. Strategies to identify small-for-date fetuses have included risk panels, Doppler ratios, amniotic fluid volume, placental scores, and biometry-based weights; none have had completely satisfactory results, although outcomes appear to have improved.39 Even accurate estimation of fetal weight is an unrealized basic goal. Critical obstetric and neonatal decisions regarding viability or delivery method rely on ultrasound-derived weights, which vary from true values by more than 10%, and more at the extremes. Clinical algorithms and physician and maternal estimates compare well with ultrasound in estimating birth weight.

Placental Location Placental location can be confidently established by transvaginal US. Low-lying placentation persists as placenta previa in only 1% to 2% of patients at term (Fig. 9-24). Using color Doppler, fetal vessels can be seen near the cervix, facilitating the diagnosis of funic presentation and vasa previa (Fig. 9-25). US is also helpful in finding placental accretion, the absence of a normal cleavage plane between decidua and placental vessels. Accretion, deeper penetration of the vessels into the myometrium (incretion), and serosal penetration (percretion) were previously diagnosed when failed attempts to deliver the placenta were followed by massive hemorrhage. Placental

Figure 9–25.  Fetal vessels (thick arrows) within the mem-

branes overlying the internal os (arrow) are termed vasa previa. Rupture of the vessels can rapidly exsanguinate the fetus; artificial membrane rupture and labor are contraindicated. See color insert.

accretion is more likely in women with placenta previa and a history of cesarean section,24 after myomectomy or curettage, and with high parity. The frequency of accretion has increased more than 10-fold in the past 20 years to 1 in 2500, echoing rising cesarean rates.35 Diagnostic US findings of indistinct placental margins, attenuated myometrium, and large turbulent placental vessels have sensitivity and specificity in the range of 85%.19 Antepartum diagnosis permits planning for anesthesia, transfusion, balloon tamponade, arterial embolization, or a scheduled cesarean-hysterectomy (Fig. 9-26). Magnetic resonance imaging (MRI) can be helpful when there is a posterior placenta, after myomectomy, and when ultrasound findings are ambiguous.6 The diagnosis of placental abruption remains a clinical one. US diagnosis is reported to have only 24% sensitivity in the third trimester, although specificity may reach 88%.26 The role of imaging in this disorder is to exclude placenta previa, an equally common source of severe third-trimester bleeding. Subchorionic hematomas are often noted on transvaginal

P

Figure 9–24.  Placenta previa: the internal os (arrow) is

completely covered by the placental tissue (P) in this transabdominal view.

Figure 9–26.  Hysterectomy specimen with placenta accreta. There is no visible distinction between the uterine muscle and the placental tissue, with the exception of a small area in the lower right (arrow). See color insert.

156

SECTION II  THE FETUS

scans early in gestation; symptomaticity, size, and persistence have been linked to poorer outcomes.37 Later abruptions are more difficult to visualize; acute bleeding is isoechoic with placenta and can be mistaken for placentomegaly. Hypoechoic fluid collections and hyperechoic infarcted areas appear in more chronic presentations. Grading placental appearance to detect disturbed growth or maturation is of limited benefit. Persistent immaturity is linked to hydrops fetalis, although not as often as increased echogenicity and thickening. Precociously mature placentas may presage growth restriction (Fig. 9-27).

Amniotic Fluid Volume Amniotic fluid volume is initially determined by secretions from the amnion; however, by the 16th week of pregnancy the fetal renal production accounts for the majority (see

Chapter 22). Malformations of the esophagus and upper gastrointestinal tract, inhibited fetal swallowing, aneuploidy, intermittent renal obstruction, maternal diabetes, twin-twin transfusion, some cases of dwarfism, and impending fetal hydrops are associated with severe polyhydramnios. Severe growth restriction with polyhydramnios carries a poor prognosis. Polyhydramnios is idiopathic in almost half of cases; the mechanical consequences include prematurity, malposition, abruption, preeclampsia, and puerperal hemorrhage. Abnormal volumes are subjectively apparent to experienced examiners but quantification remains a research issue. Maximum single pocket sizes or summed pool depths across quadrants conveniently serve as proxies for volume in clinical settings (Fig. 9-28). Oligohydramnios may occur after amniorrhexis; fluid may also decrease after fetal renal compensation for placental hypoperfusion, from functional or obstructive urogenital anomalies, and with severe maternal dehydration. The measurement of amniotic fluid may be combined with nonstress testing and assessment of movements, tone, and breathing to provide reassurance of fetal well-being. The mechanisms of amniotic fluid dynamics, the role of fluid assessment in clinical care, and the appropriate therapies for abnormalities remain poorly understood.42

Cervical Length and Pelvic Structures

Figure 9–27.  Grade III placenta, characterized by echogenic

outlines of the cotyledons with hypoechoic centers.

Cervical length (CL) is often measured as part of antenatal US, using the vaginal approach in borderline or at-risk patients (Fig. 9-29). The length of the closed portion of the endocervical canal seems continuously correlated with duration of gestation; moreover, once values fall below 25 mm, preterm deliveries increase.20 In patients with known prematurity risks, shorter cervical length is strongly predictive of delivery before 36 weeks; combined with assays for fetal

Figure 9–28.  Four quadrant

Q1

Q2

Q3

Q4

vertical pocket assessment of amniotic fluid; other techniques include measurement of the single greatest vertical pocket or identification of a 2 3 2-cm cord-free area.

Chapter 9  Perinatal Imaging

CervixL

Figure 9–29.  Transvaginal study showing cervix (dotted line)

with MacDonald cerclage (arrows) in situ.

fibronectin and other biomarkers, US can be used to find those at highest risk for imminent delivery.29 Attempts to reduce prematurity using tocolysis have been ineffective; however, delaying birth long enough for steroid enhancement of lung maturity has proven more feasible. First trimester screening by US to detect the need for cervical cerclage is not reliable. Specific candidates may benefit, however, from either preventive and “rescue” procedures on the basis of midtrimester cervical length. Progesterone supplementation by injection or vaginal suppository has increased in recent years; the latter has some supporting evidence for effectiveness when a short cervix is noted by US.10 The gravid uterus is conveniently studied by US. Müllerian duplications and septations occur in about 0.5% of the population (Fig. 9-30). Patients with bicornuate uteri may experience irregular bleeding in early pregnancy, alterations in cervical competency, and rarely torsions or cornual ruptures. Poorly vascularized septations are etiologically associated with abruption. Myomas complicate 1% to 2% of pregnancies,

157

with more frequent cesareans and prematurity, abruption, degeneration, and fetal malposition; less common complications include fetal deformation, dystocia, puerperal hemorrhage, and hysterectomy.30,41 Normal adnexal structures are not palpable after the first trimester; thus ultrasound is currently the first choice for ovarian evaluation during pregnancy, often augmented by MRI and computed tomography (CT). Ovarian torsion is an acute surgical emergency; although rare, it is more likely with ovarian cysts and adnexal masses, favoring first trimester or puerperium onset. Absent Doppler flow, free fluid, and demonstration of twisted vessels or of an edematous, rapidly enlarging ovary aid in expedient recognition and treatment.17 Corpora lutea, hydrosalpinges, theca lutein, and dermoid cysts are frequent findings, expectantly managed based on reassuring ultrasound, Doppler, and MRI characteristics.13 Ultrasound is an initial step in the workup of appendicitis during pregnancy; after nondiagnostic studies, spiral CT, MRI, or exploratory surgery may be needed. In the course of second trimester studies of fetal size, number, placentation, and amniotic fluid, a number of structures are routinely imaged. Basic examinations usually include documentation of fetal cranial integrity; midline brain structures, cavum and thalami, lateral and third ventricles, choroid plexus, cerebellum, and posterior fossa. Increased nuchal thickness is associated with a range of adverse outcomes but is rarely found, yielding a low positive predictive value.43 Views confirming facial symmetry, intact orbits, lenses, nares, lips, and normal profile are routine in comprehensive studies. Fetal swallowing and respiratory movements may be noted. The spinal column is normally imaged in both the long-axis and coronal views. Some distinct advantages in the visualization of the facial features, small parts, and spine are offered by three-dimensional imaging (Fig. 9-31). Views of the thorax, particularly those of the axis, site, and relative proportions of the cardiac and mediastinal structures with respect to the fetal lung and pulmonary vessels, provide indirect support for the integrity of the diaphragm. The cardiac examination has gradually expanded from documenting axis, laterality, and rate to a requisite symmetric four-chamber view, with routine efforts to ascertain normal outflows, and ductal and aortic arches (Fig. 9-32). Prenatal cardiac studies may include M-mode rhythm, Doppler flow studies, and structural studies capable of approaching in detail and diagnostic accuracy postnatal echocardiology, although subject to some predictable limitations. Abdominal views confirm the closure of the ventral wall, normal situs of the liver and gastric bubble, unremarkable umbilical cord appearance, normal renal and adrenal contours, bowel dimensions and echogenicity, bladder filling, and the presence of the spleen and gallbladder. The accuracy of the assignment of fetal sex increases to greater than 98% halfway through the second trimester (see Figs. 9-12 and 9-70). Small parts, including digits, benefit from threedimensional and improved conventional resolution.

DOPPLER ULTRASOUND Figure 9–30.  Müllerian anomaly: Bicornuate uterus in trans-

verse view with two decidualized cavities (arrows) and indented external contours.

Initially greeted as a key imaging method, Doppler ultrasound has been hampered by costs, time constraints, safety concerns, and a lack of clear benefit for most patients. Current applications, still more frequent in research and referral

158

SECTION II  THE FETUS Figure9–31.  Three-dimensional rendering of the fetal spine permits localization of the neural tube defect (indicated by arrows). Compare with MRI in Figure 9-35. See color insert.

RA RV

LA LV

Figure 9–32.  Transverse view of the thorax showing a nor-

mal four-chamber heart, apex down and to the left of the picture. The right and left atria (RA, LA) and corresponding ventricles (RV, LV) lie in the transverse plane in the fetus. centers, include measurements of uterine artery bloodflow, umbilical cord, the ductus venosus, and middle cerebral artery, among others. Uterine bloodflow patterns reflect maternal vascular resistance and placental site.31 Abnormal patterns precede growth restriction and maternal hypertensive complications. Umbilical cord arteries are readily identified by color Doppler as they bifurcate around the bladder (see Fig. 9-5). A two-vessel cord may have associated cardiac and renal anomalies, aneuploidies, or growth restriction. Abnormally high umbilical artery resistance is characteristic of uteroplacental constraints on fetal growth or, less frequently, fetal anomalies. Doppler findings may hint at the etiology of a size/date discrepancy; weight estimation more reliably identifies small-for-gestational age (SGA) fetuses.

Persistent absence or reversal of end-diastolic flow, although uncommon, is associated with severe growth restriction and increased perinatal morbidity and loss. Middle cerebral artery resistance and venous patterns are altered later in the course of compromised perfusion; paradoxical ductus venosus or umbilical venous patterns may be premorbid findings. Schemes that combine clinical risk factors, biometry, Doppler findings, and biophysical parameters appear to have improved outcomes for SGA infants.3 Doppler measurement of peak systolic velocity in the middle cerebral artery has become incorporated in the management of fetal anemia and isoimmunization (Fig. 9-33). Previously, abnormal maternal antibody levels signaled the need for invasive cord blood sampling or amniocentesis. The former is associated with technical obstacles, fetal losses, and prematurity; the latter was informative only for hemolytic sources of anemia. The peak velocities of the middle cerebral artery are negatively correlated with hemoglobin values, are noninvasive, and are relatively easy to obtain. At a cutoff of 1.5 multiples of the mean (MOM), the sensitivity for moderate to severe anemia approaches 100%, with a false-positive rate from 0% to 28%; reliability drops near term.33 Color flow Doppler can distinguish between cystic and vascular lesions, confirm the presence or site of organs, and evaluate central nervous system, cardiac, and pulmonary vascular anatomy. When combined with three-dimensional scanning, color flow can provide a virtual vascular cast, although direct clinical applications are currently limited.

PROCEDURES Real-time ultrasound is well suited by virtue of safety, economy, and convenience for guiding placement of needles, cannulae, catheters, and other devices used in fetal diagnosis and therapy. The range of procedures includes egg retrieval and

Chapter 9  Perinatal Imaging

159

BOX 9–1 Indications for Antenatal Ultrasound

Figure 9–33.  Middle cerebral artery (in trapezoid) Doppler flow velocity in a fetus with Rhesus sensitization. The peak systolic velocity (PSV shown in lower portion) increases as fetal anemia progresses. Doppler measurements in the fetus are often expressed as ratios to compensate for inaccuracies introduced by narrow, tortuous vessels. See color insert.

embryo transfers, embryonic and fetal reduction, chorionic sampling, placental and skin biopsies, amniocentesis and amnioreduction, cord blood sampling, intrauterine transfusion, aspiration of fluid from various sites, and in conjunction with fetoscopy and vascular ablation. The effects of interventions can be continuously monitored during and after the procedures. Determination of fetal and umbilical cord positions during delivery, versions, and immediately before transfusions, surgical interventions, and the EXIT (ex utero intrapartum treatment) procedure are also important applications.

FETAL ANOMALIES (See also Chapter 29.) Neonatal and parental outcomes may be optimized by advance warning and timely preparations for a range of abnormalities. Unfortunately, more than one third of anomalies detectable after delivery are not recognized prenatally.27 False-positive rates are usually less than 10%, but discordancy with autopsy findings emphasizes limits inherent in US. Nonetheless, ultrasound is the accepted initial approach for finding fetal malformations (Box 9-1).

Central Nervous System FETAL VENTRICULOMEGALY (See also Chapter 40.) Strictly speaking, hydrocephalus and ventriculomegaly are not synonymous. Hydrocephalus connotes raised intracranial pressure; this functional observation cannot be made by ultra­ sound. The lateral ventricles may be enlarged (Fig. 9-34), without altered pressure, because of a developmental abnormality. Enlargement of the atrium and posterior horns of the lateral ventricles (colpocephaly) often occurs in association with agenesis of the corpus callosum or type II Chiari malformation. The lateral ventricles may also be wider as a result of

Pregnancy dating Uncertain dates Scheduled delivery Suspected fetal death Late registrant Pregnancy termination Fetal size and growth Size/dates discrepancy Suspected polyhydramnios Suspected oligohydramnios Suspected multiple gestation Maternal risk factors for abnormal growth Vaginal bleeding Suspected abruption Placenta previa Follow-up for previously identified previa Cervical assessment Incompetent cervix Adjunct to cerclage Prematurity risk Pelvic pain or pelvic mass Suspected hydatidiform mole Suspected uterine abnormality Fetal localization Suspected ectopic Fetal position/weight in labor or after amniorrhexis Adjunct to breech version Adjunct to amniocentesis and other procedures Fetal anatomic survey Abnormal serum screening/maternal age Prior history of anomaly Follow-up of detected anomaly Teratogen exposure Known maternal risks for anomalies Isoimmunization/fetal anemia risks

brain destruction from a variety of causes, including ischemia and in utero infection by cytomegalovirus. A single measurement at the level of the atrium of the lateral ventricle is used to define ventriculomegaly; measurements exceeding 10 mm are considered abnormal.15 The width of the atrium varies little during pregnancy. The normal choroid plexus position depends on gravity; on standard axial views the choroid, attached at the level of the foramen of Monro, rests on the dependent wall of the lateral ventricle. It marks the limit of the lateral ventricle even when the wall itself cannot be seen; a “dangling” choroid plexus establishes the extent and severity of ventricular enlargement (Fig. 9-35). The clinical outcome with markedly dilated ventricles has been discouraging. Survivors are often mentally and physically impaired. Unfortunately, it has proved impossible to define a group of fetuses with ventriculomegaly who might benefit from in utero shunting. Mild ventriculomegaly, more common in males (atrial size between 10 and 15 mm) has an uncertain prognosis; the majority of infants are normal. Agenesis of the corpus callosum can be an isolated finding or more commonly associated with other brain anomalies.

160

SECTION II  THE FETUS

B

Figure 9–36.  MRI at 25 weeks’ gestation with agenesis Figure 9–34.  Fetal MRI of hydrocephalus at 23 weeks’ gesta-

tion. Note massive enlargement of both lateral ventricles with thin rim of remaining brain parenchyma.

of corpus callosum. Enlarged posterior lateral ventriclescolpocephaly (thin arrow). Rectangular “wide open” Sylvian fissure (thick arrow) is normal for this gestational age. Maternal urinary bladder (B).

Figure 9–35.  “Dangling choroid” (thick arrow) with bilater-

ally dilated lateral ventricles secondary to an open neural tube defect. The latter is signaled by scalloping of the frontal bones, “lemon sign” (thin arrow). Ultrasound diagnosis can be challenging before 20 weeks, and MRI examination can be helpful (Fig. 9-36).

MENINGOMYELOCELE AND (TYPE II) CHIARI MALFORMATION A meningomyelocele (Fig. 9-37) can occur in association with downward displacement of the hindbrain and with other brain anomalies. It may or may not include ventriculomegaly. Often hydrocephalus does not develop in these children until after birth. In cases of an open neural tube defect (Fig. 9-38), the a-fetoprotein level is elevated in both maternal serum and amniotic fluid (see Chapter 8). A meningomyelocele is demonstrated by ultrasound as splaying or divergence of the posterior ossification centers,

Figure 9–37.  MRI of sacral meningomyelocele (thick arrow) at 21 weeks’ gestation; normal posterior fossa including cerebellum and fourth ventricle (thin arrow).

which is best appreciated on axial views of the spine. A fluidfilled sac may be seen, and the integrity of the overlying skin can be assessed. The level of the meningomyelocele can also be ascertained, with three-dimensional ultrasound and MRI (see Figs. 9-31 and 9-37) being the most accurate. This information can be helpful for predicting the outcome in affected children.

VENTRICULOMEGALY The “lemon sign” refers to the altered appearance of the calvaria, similar in shape to a lemon, on an axial scan (see Fig. 9-35).36 The biconcave frontal bones produce the calvarial distortion. This sign is dependent on gestational age, which is demonstrated between 18 and 24 weeks and is not

Chapter 9  Perinatal Imaging

161

Figure 9–40.  Two-dimensional image of anencephaly at Figure 9–38.  Fetal MRI of Chiari II malformation at 21 weeks’ gestation. Sacral meningomyelocele (thick arrow) and dilated lateral ventricles (thin arrow).

totally specific; it is sometimes found in otherwise normal children. The “banana sign” refers to an abnormal position of the cerebellum, which is curled in a crescent (bananalike) around the brainstem (Fig. 9-39). Obliteration of the cisterna magna and cerebellar distortion are secondary to downward displacement of the hindbrain, which is associated with type II Chiari malformation.

ANENCEPHALY Anencephaly, which is the absence of the normal brain and calvaria superior to the orbits, can be detected as early as 10 weeks but is recognizable throughout gestation (Fig. 9-40). Associated polyhydramnios (50%) appears late in the second trimester. Maternal serum a-fetoprotein is elevated in most cases. Approximately 50% have associated anomalies, such as meningomyelocele, cleft palate, and clubfoot. Occasionally the typical appearance is altered by the presence of echogenic material

POST

17 weeks’ gestation. The cranial vault is absent above the orbits (arrows). superior to the orbits, identified pathologically as angiomatous stroma (“area cerebrovasculosa”). Anencephaly is considered incompatible with meaningful postnatal survival.

ENCEPHALOCELE (See also Chapter 40.) The protrusion of brain tissue within a meningeal sac is usually a straightforward ultrasound diagnosis. Most encephaloceles in the Western world are occipital and midline, and a distinction should be made between them and soft tissue edema or cystic hygroma of the neck. The identification of the bony calvarial defect allows a specific diagnosis (Fig. 9-41). Encephaloceles may occur in isolation or be associated with amniotic band syndrome (when off midline) and genetic syndromes such as Meckel-Gruber.

DANDY-WALKER CYST Dandy-Walker malformation is a fluid-filled cyst of the posterior fossa with or without enlargement of the lateral ventricles (Fig. 9-42). Enlargement of the posterior fossa is always present

FOSSA

Figure 9–39.  Open neural tube defects may be associated

with an obliterated posterior fossa “banana sign” (arrow) in 60% of cases.

Figure 9–41.  Large occipital encephalocele with herniation of neural tissue through the defect (arrows).

162

SECTION II  THE FETUS

CHOROID PLEXUS CYST Cysts in the choroid plexus are identified in 1% to 2% of normal pregnancies in the second trimester and usually resolve by the early third trimester (see Fig. 9-17). Choroid cysts have been associated with aneuploidies, including trisomies 18 and 21, although rarely as isolated findings. Number, size, and persistence of cysts lend no additional clinical significance to the finding. F

Spine

P

Figure 9–42.  Fetal MRI of Dandy-Walker cyst (arrow) at 20

weeks’ gestation. Note the posterior placenta (P) and anterior fibroid (F).

with uplifting of the tentorium. The cerebellar hemispheres are rudimentary, separated by the posterior fossa cyst. Agenesis should not be diagnosed before 18 weeks because the development of the cerebellar vermis may not be complete until that time. The Dandy-Walker variant, associated with an increased incidence of chromosomal abnormalities, consists of direct communication between the fourth ventricle and cisterna magna without posterior fossa enlargement. The inferior cerebellar vermis is only mildly hypoplastic. These ultrasound findings can be subtle and confirmation from MRI helpful (Fig. 9-43).

Figure 9–43.  Fetal MRI of Dandy-Walker variant at 21 weeks’

gestation. Remaining cerebellum is uplifted with wide inferior communication of fourth ventricle (arrow).

Spinal anomalies demonstrable by ultrasound include congenital vertebral anomalies (Fig. 9-44). Sacrococcygeal teratoma is a large mass arising posteriorly from the rump of the fetus. The posterior elements of the lumbosacral spine are intact. This tumor is often associated with polyhydramnios and prematurity and more rarely with hydrops. The mass may be cystic or solid, extend into the fetal pelvis (Fig. 9-45) and abdomen, and rarely undergo malignant degeneration.

Head and Neck After 14 weeks, the visualization of the nose, orbits, forehead, lips, and ears is feasible. Orbits are clearly seen axially; the ocular diameters, interocular distance (defining hypertelorism and hypotelorism), and binocular distance are measured. The profile reveals the forehead, nose, and jaw sagittally (contrast Fig. 9-1 with Fig. 9-46). The coronal view, the best for facial structures, includes the orbits (and lenses), eyelids, nose, and lips. The coronal view demonstrates facial clefting abnormalities, including the cleft lip and cleft palate, which may be central or lateral. Sagittal views should include the nasal bone; nasal bone hypoplasia or absence is used by some as a marker for Down syndrome. The complex anatomy of the fetal face is shown exquisitely by three-dimensional sonography (see Fig. 9-8; Fig. 9-47).28 Cystic hygroma (lymphatic malformation) is a septate, cystic mass arising in the neck and occiput; it may extend to the remainder of the trunk (Fig. 9-48). Posterior septation of the nuchal ligament distinguishes this lesion from neural

Figure 9–44.  Coronal view through the fetal spine showing a hemivertebra. The spinal ossification center (upper left arrow) has no mate.

Chapter 9  Perinatal Imaging

163

Figure 9–47.  Unremarkable fetal face shown three dimenFigure 9–45.  Fetal MRI of sacrococcygeal teratoma at 27 weeks’

gestation. Tumor (thick arrow) extends into fetal pelvis between spine (thin arrow) and urinary bladder (curved arrow).

sionally at 34 weeks’ gestation. Views are optimized by adequate fluid interface, lack of fetal movement, limited maternal adipose and scarring tissues, and posterior placentation. See color insert.

Figure 9–46.  Fetal retrognathia (recessed chin) is a feature of

a variety of syndromes and aneuploidies, as noted in this fetus with trisomy 18. tube defects. Associated aneuploidies, including monosomy X, are common. Less frequent neck masses include anterior goiter, teratoma, and hemangioma. Increased nuchal translucency (subcutaneous sonolucent area) (see Fig. 9-11) and nuchal thickness (see Fig. 9-16) are associated with aneuploidies, including trisomies 21 and 18, cardiac malformations, a wide range of other anomalies, maternal diabetes, and fetal loss. With a normal karyotype and reassuring midtrimester cardiac and anatomic surveys, the outcome is generally good.11 The cause of the transient fluid accumulation in the neck is unknown; the appearance at birth is usually normal.

Heart (See also Chapter 45.) Most fetal echocardiograms are obtained at 18 to 20 weeks of gestation; however, the fetal heart may be evaluated as early as 12 weeks with transvaginal ultrasound. Optimal

Figure 9–48.  Transverse image through the head showing a posteriorly located cystic hygroma with multiple septate cysts (C).

timing is a compromise between offering a diagnosis as early as possible and adequately visualizing complex anatomy, recognizing that some lesions develop late in pregnancy. A four-chamber view of the fetal heart should be part of all obstetric ultrasound examinations (see Fig. 9-32). Abnormal four-chamber views will detect approximately 50% of all cases of congenital heart disease in a nonselected population. Fetal arrhythmias and abnormal situs may also be diagnosed (Fig. 9-49). An attempt should be made on each study to obtain aortic arch and outflow views, linking the aorta and neck vessels with the left ventricle and the bifurcating pulmonary artery with the right ventricle; these views allow identification of transposition of the great vessels, tetralogy of Fallot, truncus arteriosus, and some aortic abnormalities.16 Sensitivity for complex cardiac lesions exceeds that for isolated septal

164

SECTION II  THE FETUS

A Figure 9–50.  A 10-week fetus within its gestational sac.

Physiologic midgut herniation occurs from week 8 to week 11.

Cl

B Figure 9–49.  Fetal arrhythmia demonstrated by M-mode (A), consistent with paroxysmal atrial tachycardia at 218 beats per minute. Sustained rates greater than 200 beats per minute may be associated with fetal hydrops (B). The fetus has ascites (short arrow), pleural effusion (long arrow), and skin edema (thin arrow).

Figure 9–51.  Cross section of the abdomen at the level of the umbilical vessels, demonstrating heterogeneous-appearing, nondistended, fluid-filled loops of normal small bowel. The abdominal walls are clearly intact adjacent to the cord insertion (CI).

defects; however, the correct diagnosis is highly dependent on operator experience.

Gastrointestinal Tract (See also Chapter 47.)

NORMAL BOWEL APPEARANCE Physiologic bowel migration into the proximal umbilical cord occurs from 7 to 10 weeks of gestation and may be visible on first-trimester scans (Fig. 9-50) up to 11.5 weeks. The normal liver remains intra-abdominal; migration thus should never be mistaken for ventral abdominal wall defects.12 The normal fetal stomach is seen transvaginally by 9 to 10 weeks of gestation as an echo-free structure in the upper left abdomen that changes in size as it empties and fills. Fetal small bowel usually is not distinguished early in gestation but appears as fluid-filled loops in the central abdomen by the late second trimester (Fig. 9-51).

The large bowel is clearly identified by the third trimester as a hypoechoic tubular structure at the abdominal periphery. Meconium in the large bowel is a normal third trimester finding. Meconium in the second trimester may have increased echogenicity that resolves over time. Hyperechogenic bowel, as bright as adjacent osseous structures, is a nonspecifice abnormal finding linked to poor fetal outcomes. Echogenic bowel has been associated with trisomy 21, cystic fibrosis, bowel atresia, congenital infections (e.g., cytomegalovirus), intrauterine growth restriction, chronic abruption, and poor outcome. Obstruction of the gastrointestinal tract can be diagnosed by ultrasound. Blockage of the proximal alimentary canal interferes with amniotic fluid turnover, which involves swallowing and absorption by the fetus. In proximal bowel obstruction, polyhydramnios is an invariable finding (see Chapter 22).

Chapter 9  Perinatal Imaging

165

ESOPHAGEAL ATRESIA AND TRACHEOESOPHAGEAL FISTULA A nonvisualized fetal stomach combined with polyhydramnios should alert the examiner to the possibility of esophageal atresia. Unfortunately, these two signs are identified in only 40% of the cases. Nonvisualization of the stomach is not specific for esophageal atresia; a proximal esophageal pouch is almost never visualized. Polyhydramnios, which is seen in two thirds of the cases, may not develop until the third trimester. The VACTERL complex consists of vertebral, anal, cardiovascular, tracheal, esophageal, renal, and limb malformations; these systems should be scrutinized in the fetus when a tracheoesophageal fistula from poor gastric filling is suspected. The stomach may not fill when there is little to swallow, as in oligohydramnios, or when there are abnormalities of the neural or musculoskeletal system (e.g., PenaShokeir syndrome) that interfere with swallowing.

SMALL BOWEL OBSTRUCTION In duodenal atresia, a distended stomach and proximal duodenum produce a “double-bubble” sign akin to the neonatal x-ray appearance (see Fig. 9-14). The strong (25%) association with Down syndrome may warrant genetic amniocentesis when duodenal atresia is suspected. Other etiologies include annular pancreas, duodenal web, malrotation, and severe duodenal stenosis. Ileal and jejunal atresia usually cause multiple distended bowel loops, often exhibiting peristalsis (Fig. 9-52). More distal bowel obstruction, seen in meconium ileus, Hirschsprung disease, and anal atresia, may not produce noticeably dilated loops of bowel until the third trimester. Meconium peritonitis, a sterile but morbid chemical response to intrauterine bowel perforation, is characterized by echogenic peritoneal calcifications and free intraperitoneal fluid (Fig. 9-53).

ANTERIOR ABDOMINAL WALL DEFECTS With an approximate incidence of omphalocele of 1 in 5000 and of gastroschisis of 1 in 10,000 live births, abdominal wall defects are among the more common neonatal abnormalities. A midline defect involving the base of the umbilical cord is

Stomach

Figure 9–53.  Dilation of small bowel loops associated with ileal obstruction and perforation. The echogenic material (arrow) is a meconium pseudocyst.

characteristic of omphalocele (Fig. 9-54). A sac surrounds the herniated viscera and may also include the liver (Figs. 9-55 and 9-56). Omphaloceles are often associated with other anomalies (Beckwith-Wiedemann syndrome, pentalogy of Cantrell) and aneuploidies (trisomies 13 and 18). In gastroschisis, herniated bowel loops float in the amniotic cavity without a covering membrane (Fig. 9-57). The defect is usually to the right of the umbilical cord insertion; associated fetal abnormalities are rare.

DIAPHRAGMATIC HERNIA AND THORACIC LESIONS The diagnosis of congenital diaphragmatic hernia (Fig. 9-58) is made when the abdominal organs (usually the stomach and bowel) are seen in the thorax. Fluid-filled loops of bowel may show peristalsis in the thoracic cavity. Left cardiac hypoplasia, polyhydramnios, paradoxical breathing, and other anomalies are often associated (see Chapters 11, 44, and 47). Diaphragmatic hernia has a mass effect, with a mediastinal shift and lung compression (Fig. 9-59). Unfortunately, the extent of pulmonary hypoplasia, measured at the level of the four-chamber heart, cannot by itself predict postnatal outcome. Among the best current ultrasound predictors are measurements of the lung-to-head ratio and calculated threedimensional lung volumes. MRI can give precise volumetric measurement of the lungs that can be a helpful adjunct. The presence of liver herniation into the chest (Fig. 9-60) is slightly less ominous for right-sided defects than for those on the left. Cystic adenomatoid malformation (Fig. 9-61) may have a similar imaging appearance but lacks peristalsis and displacement of abdominal structures. Solid forms may be difficult to distinguish from sequestrations; the latter usually display feeding vessels (Fig. 9-62). Spontaneous regression during pregnancy may occur in both sequestrations and cystic adenomatoid lesions.

Gallbladder and Bile Ducts Figure 9–52.  Massive dilation of bowel loops secondary to early

in utero volvulus with obstruction. The stomach and duodenal bulb are marked with arrows. Polyhydramnios is present.

(See also Chapter 48.) A choledochal cyst is a localized dilation of the biliary system, often involving the common bile duct. It is seen as a fluid-filled cyst in the upper right abdominal quadrant, separate from the

166

SECTION II  THE FETUS Figure 9–54.  Three-dimensional views of an omphalocele (star) at 18 weeks’ gestation. Upper left: sagittal two-dimensional image; upper right: transverse trunk just above the defect; lower left: coronal image; lower right: three-dimensional view. Contents are best identified on the twodimensional proportionate size on the rendered image. The small dot is sited within the gastric bubble. See color insert.

Omphalocele

Liver

Femur

Umbilical cord loop

Figure 9–55.  Power Doppler transverse glass body view of a

massive omphalocele (outline) that includes the liver and hepatic vessels (color). The extruded portion and the sac are almost equal in diameter to the torso (to the right). See color insert. gallbladder. Echogenic material (suspected gallstones) is occasionally noted in the gallbladder lumen on third-trimester scans (Fig. 9-63). Follow-up studies in neonates generally confirm resolution; children are usually asymptomatic.

Genitourinary Tract (See also Chapter 47.) The kidneys are readily visible by 12 to 14 weeks’ gestation; they are located lateral to the fetal spine (Fig. 9-64). Normal kidney size can be plotted against gestational age; by the end

Figure 9–56.  Fetal MRI of omphalocele at 24 weeks’ gesta-

tion. Note membrane (thin arrow) covering the omphalocele that includes the liver. Fetal stomach and bladder are shown by thick arrows. Note normal smooth brain surface at 24 weeks without gyri or sulci.

Chapter 9  Perinatal Imaging

ACI

Left

167

Bowel loops Liver

Heart Lung Right

Figure 9–57.  Gastroschisis. Transverse abdominal image of

the spine (left, short arrow) demonstrates free bowel loops (thick arrow) to the right of the umbilical cord insertion (long arrow). This lesion is more common in younger gravidas and is usually isolated.

Figure 9–58.  MRI of posterior Bochdalek diaphragmatic hernia at 28 weeks’ gestation (thick arrow). Also note the umbilical vein (thin arrow) entering the fetal liver above the urinary bladder to become the ductus venosus.

of the third trimester, fat deposition creates an echogenic border that enhances demonstration of renal structures. The bladder is visible as early as 11 weeks of gestation, outlined laterally by the umbilical arteries (Fig. 9-65). Normal fetal ureters are never visible by ultrasound. Fetal adrenal glands are easily identified later in pregnancy because of the relatively large cortex. Lethal renal malfunctions such as bilateral multicystic kidneys or bilateral renal agenesis result in the absence of fetal urine production; oligohydramnios may not become apparent until 18 to 20 weeks of gestation. Amniotic fluid after 16 weeks is almost entirely fetal urine. Infants born after prolonged oligohydramnios have a characteristic facial appearance, limb deformities, and pulmonary hypoplasia,

Figure 9–59.  Ultrasound of a left-sided diaphragmatic hernia

at 22 weeks’ gestation. The heart is markedly displaced to the right, although it maintains a leftward axis. The left chest is filled by bowel loops and by the more homogeneous liver. There is a small amount of residual lung visible.

Figure 9–60.  Fetal MRI of right-sided diaphragmatic hernia at 32 weeks’ gestation. Liver (thick arrow) has herniated into right thorax with compressed right lung. Thin arrow shows incidental nuchal cord.

termed Potter syndrome (or sequence). Finding oligohydramnios on ultrasound should direct attention to the fetal urinary tract. If the bladder is not seen at first, the patient should be rescanned at 30-minute intervals. Unfortunately, a lack of surrounding amniotic fluid obscures fetal anatomy. Anhydramnios, a persistently empty bladder, and an inability to identify fetal kidneys or renal arteries is strongly suggestive of bilateral renal agenesis. Fetal adrenal glands lose their normal angulation in renal agenesis and may be mistaken for the kidneys. Fetal MRI can be ideal in establishing this lethal diagnosis. Care must be exercised when suspecting unilateral renal agenesis to exclude renal ectopy (pelvic kidney); color

168

SECTION II  THE FETUS

gb

Figure 9–63.  Fetal gallbladder (gb) (arrow) with echogenic

Figure 9–61.  MRI of left cystic adenomatoid malformation

(arrow) at 22 weeks’ gestation. Note the normal right lung and intact left diaphragm.

sludge and small calculi in the third trimester. These are usually transient, resolving once maternal hormonal levels decrease. Compare with normal echo-free appearance in Figure 9-18.

Renals

H

C

Figure 9–62.  Pulmonary sequestration shown as an echo-

Figure 9–64.  Transverse view of the fetal abdomen, spine up, demonstrating the normal renal appearance (arrows).

genic region on oblique view of the lower thorax. Power Doppler highlights the cardiac apex (C) and hepatic vessels (H) to the right. The feeding vessel is indicated by the arrow. See color insert.

Doppler may reveal the aberrant course of the renal artery (see Fig. 9-4). Mild dilation of the fetal renal pelvis is common; it is a normal feature that should not be misinterpreted as hydronephrosis. Pyelectasis has been reported to be a “weak” marker for Down syndrome. Identifying a single measurement to predict postnatal obstructive uropathy proved elusive. Pathologic dilation of the upper urinary tract was previously diagnosed when the renal pelvis measured 10 mm or more in the anteroposterior dimension. This measurement failed to identify hydronephrosis before the third trimester; cutoffs of 4 to 7 mm are now used in second and third trimesters, respectively. Obstruction of the urinary tract produces a variable

B

Figure 9–65.  Power Doppler demonstration of two umbilical

cord arteries (arrows) outlining the fetal bladder (B). See color insert.

Chapter 9  Perinatal Imaging

169

Figure 9–66.  Fetal MRI at 27 weeks’ gestation. A, Dilated right upper pole renal duplication (arrow). B, Dilated distal right ureter (arrow) of duplication.

A

B

sonographic appearance, depending mainly on timing of onset. Renal dysplasia, sometimes with cyst formation, is the consequence of early obstruction; obstruction occurring later in gestation is more likely to result in dilation of the collecting system (Fig. 9-66). Hydronephrosis may be unilateral, resulting from obstruction of the ureteropelvic junction, or bilateral, resulting from lower urinary tract obstruction by posterior urethral valves or urethral agenesis. A distended, thick-walled bladder may be seen in males with posterior valves. Hydroureters are usually present; urinary ascites may occur. “Keyhole” dilation of the posterior urethra helps distinguish this lesion from prune-belly syndrome (Fig. 9-67). Multicystic renal dysplasia appears as several noncommunicating cysts varying in size, lacking organization, and changing over time. Severe hydronephrosis, in contrast, is an orderly arrangement of the enlarged renal pelvis surrounded by smaller calyces. Bilateral renal anomalies (including bilateral multicystic renal dysplasia) are common (40%) in the fetus (Fig. 9-68).

Figure 9–67.  Posterior urethral valve. Coronal scan of the

fetal pelvis shows an enlarged bladder and dilated posterior urethra. Arrows are proximal to the posterior urethral valves. Note the oligohydramnios.

Dilated ureters may be seen in ureterovesical junction obstruction (primary megaureter). Dilated ureters have also been attributed to reflux. Autosomal recessive polycystic kidney disease causes bilaterally enlarged, echogenic kidneys, which retain their reniform shape (Fig. 9-69). Many other renal anomalies have been identified prenatally, including duplicated collecting systems and bladder exstrophy; in the latter condition, persistent nonvisualization of the bladder coincides with normal amniotic fluid and unremarkable kidneys. Distinguishing a male from a female fetus is reliably accomplished after 16 weeks’ gestation (Fig. 9-70) by external genital examination. Ovarian cysts, fluid-filled or septate with fluid levels, are seen in utero; these cysts are often abdominal, resembling gastrointestinal and other abnormal cysts. Hydrometrocolpos is another fluid-filled pelvic structure in females.

Musculoskeletal System Examination of all four extremities and femur measurements are routine parts of prenatal studies. Standardized tables exist for all the long bones, the feet, and other skeletal components. Shortening of the extremities beyond 3 standard deviations for age is very suspicious for skeletal dysplasia. Because measurements in normal populations are distributed below the 5th percentile, it is reassuring that most skeletal dysplasias show dramatic reductions; an abnormal femur-to-foot ratio also provides diagnostic support. The biparietal diameter continues to reflect gestational age unless the skull is involved in the dysplastic syndrome. Careful measurement of all bones in the peripheral skeleton is required in suspected cases, first to define the category of dysplasia and then, if possible, to make a diagnosis (see Chapters 29 and 54). Overall limb reduction is termed micromelia; rhizomelia refers to proximal (femurs and humeri), mesomelia to more distal (forearms and lower legs), and acromelia to the most distal (feet and hands) dysplasias. Fractures and curvatures, altered bone density, the appearance of the spine, skull, and ribs, and extraskeletal anomalies may be keys to a specific diagnosis. Polyhydramnios may also be present. Pulmonary hypoplasia is a frequent cause of death in fetuses with severe skeletal dysplasia. Lung volume can be

170

SECTION II  THE FETUS

Figure 9–68.  Multiple views of bilateral multicystic dysplastic kidneys. This diagnosis is considered incompat-

ible with postnatal survival and is always associated with severe oligohydramnios in the second half of gestation. Note lack of fluid interface.

P

Figure 9–69.  Congenital polycystic kidney disease. Trans-

verse abdominal scan shows echogenic, enlarged kidneys filling the fetal abdomen (arrows). The renal pelvis (P) is seen in the superior kidney. Oligohydramnios is present.

assessed by MRI or by three-dimensional imaging, but femur/abdominal circumference of less than 0.16 appears to function as a convenient proxy for more tedious estimators for lethality.40 Thanatophoric dysplasia is a uniformly lethal skeletal dysplasia that can often be recognized in utero. Features are severe micromelia, curved femurs, and pulmonary hypoplasia (Fig. 9-71). One form of thanatophoric dysplasia includes a cloverleaf skull deformity. Achondrogenesis is also lethal, signaled by severe micromelia and a lack of vertebral ossification. Lethal type II osteogenesis imperfecta produces short, fractured bones, often bent or curved (Fig. 9-72). Bone density is decreased, best appreciated in the calvaria. The lethal form of hypophosphatasia may have a similar ultrasound appearance of severe demineralization. Heterozygous achondroplasia is the most common nonlethal skeletal dysplasia. Shortening of the femur characterizes this autosomal dominant, rhizomelic disorder, but may not be noticeable until the third trimester, limiting prenatal diagnosis to known carriers. Other abnormalities of the musculoskeletal system include the malformation or absence (dysostosis) of various parts of the skeleton, for example, limb reduction anomalies

Chapter 9  Perinatal Imaging

A

B Figure 9–70.  The male fetus (A) can reliably be distinguished from the female fetus (B) by ultrasound after the first trimester. See color insert.

171

Figure 9–72.  Osteogenesis imperfecta. Humerus with marked

shortening and abnormal shape secondary to multiple fractures, indicated by cursors.

as in radial clubhand or hemimelia (Figs. 9-73 and 9-74), valgus deformities of the feet (Fig. 9-75), and amniotic band amputation. Pitfalls in the diagnosis of skeletal dysplasia are appreciable. A phenocopy gives a fetus the attributes of a specific syndrome, but the cause may be another factor, such as infection, another genetic abnormality, or teratogen exposure. An example of a phenocopy is the stippling of epiphyses caused by maternal warfarin use, mimicking stippled epiphyseal skeletal dysplasia.

Two-Vessel Umbilical Cord A single umbilical artery may be identified on transverse section and confirmed by color Doppler imaging (Fig. 9-76). Single umbilical artery may be associated with other congenital anomalies, prompting a careful anatomic survey; however, the reported anomalies do not display a consistent pattern. Lack of normal coiling is also associated with fetal abnormalities, particularly those affecting movements. It is unclear whether isolated single umbilical artery is associated with aneuploidy, nor does laterality of the absent vessel have significance. Small-for-gestational age is more common in neonates with single umbilical artery. Counseling, selective use of amniocentesis, and serial third-trimester examinations may be helpful in management.38

SUMMARY

Figure 9–71.  Thanatophoric dysplasia. Sagittal view of the fetus reveals severe deformation of the rib cage with protrusion of the abdomen (arrows).

As sonography nears maturity, it has shown a gratifying degree of success in fulfilling its early potential. Initial limitations have yielded to advances in technologies and techniques or have been remedied by genetic and biochemical breakthroughs. MRI is now well established as a versatile partner to US in prenatal diagnosis. The future of imaging continues to hold immense promise for continued advances in understanding and improving neonatal outcomes.

172

SECTION II  THE FETUS

Figure 9–73.  Fetus with short-ribbed polydactyly syndrome. Multiple images show a small chest with

hypoechoic ribs, large amounts of amniotic fluid, and postaxial polydactyly.

Figure 9–75.  Valgus deformity of the fetal foot may be iso-

lated, familial, positional, or secondary to central nervous system and spinal cord abnormalities. See color insert.

Figure 9–74.  Truncated fetal arm (arrow) associated with

trisomy 21 fetus. See color insert.

Chapter 9  Perinatal Imaging

Figure 9–76.  Two-vessel umbilical cord. A transverse scan of

the umbilical cord shows the vein and a single artery. On color flow, one of the branches of the normal “Y” shape will be absent (see Fig. 9-5)

REFERENCES 1. Abbott JG: Rationale and derivation of MI and TI: a review, Ultrasound Med Biol 25:431, 1999. 2. Abramowicz JS: Ultrasound in obstetrics and gynecology: is this hot technology too hot? J Ultrasound Med 21:1327, 2002. 3. American College of Obstetricians and Gynecologists: ACOG Practice Bulletin No 98: ultrasonography in pregnancy, Obstet Gynecol 112(4):951, 2008. 4. American College of Obstetricians and Gynecologists: ACOG Committee Opinion No. 296: first-trimester screening for fetal aneuploidy, Obstet Gynecol 104:215, 2004. 5. American Institute of Ultrasound in Medicine: AIUM Practice Guideline for the Performance of an Antepartum Obstetric Ultrasound Examination, J Ultrasound Med 22:1116, 2003. 6. Baughman WC et al: Placenta accreta: spectrum of US and MR imaging findings, Radiographics 28, 1905, 2008. 7. Bello G et al: Combined pregnancy: the Mount Sinai experience, Obstet Gynecol Surv 41:603, 1986. 8. Benacerraf BR et al: Three-dimensional US of the fetus: volume imaging, Radiology 238(3):988, 2006. 9. Benacerraf B: Three-dimensional fetal sonography, J Ultrasound Med 21:1063, 2002. 10. Berghella V: Novel developments on cervical length screening and progesterone for preventing preterm birth, Br J Obstet Gynecol 116(2):182, 2008. 11. Bilardo CM et al: Increased nuchal translucency thickness and normal karyotype: time for parental reassurance, Ultrasound Obstet Gynecol 30(1):11, 2007. 12. Bowerman R: Sonography of fetal midgut herniation: normal size criteria and correlation with crown-rump length, J Ultrasound Med 5:251, 1993. 13. Bromley B et al: Adnexal masses during pregnancy: accuracy of sonographic diagnosis and outcome, J Ultrasound Med 16:447, 1997. 14. Campbell S et al: Routine ultrasound screening for the prediction of gestational age, Obstet Gynecol 65:613, 1985. 15. Cardoza JD et al: Exclusion of fetal ventriculomegaly with a single measurement: the width of the lateral ventricular atrium, Radiology 169:711, 1988.

173

16. Carvalho JS et al: Improving the effectiveness of routine prenatal screening for major congenital heart defects, Heart 88:387, 2002. 17. Chang HC et al: Pearls and pitfalls in diagnosis of ovarian torsion, Radiographics 28(5):1355, 2008. 18. Chervenak FA et al: How accurate is fetal biometry in the assessment of fetal age? Am J Obstet Gynecol 178:678, 1998. 19. Comstock CH et al: Sonographic detection of placenta accreta in the second and third trimesters of pregnancy, Am J Obstet Gynecol 190:1135, 2004. 20. Crane JM et al: Transvaginal sonographic measurement of cervical length to predict preterm birth in asymptomatic women at increased risk: a systematic review, Ultrasound Obstet Gynecol 31(5):579, 2008. 21. Cuckle HS et al: Contingent screening for Down syndrome— results from the FaSTER trial, Prenat Diagn 28:89, 2008. 22. Doubilet PM et al: Long-term prognosis of pregnancies complicated by slow embryonic heart rates in the early first trimester, J Ultrasound Med 18:537, 1999. 23. Filly RA et al: Routine obstetrical sonography, J Ultrasound Med 21:713, 2002. 24. Finberg HJ et al: Placenta accreta: prospective sonographic diagnosis in patients with placenta previa and prior cesarean section, J Ultrasound Med 11:333, 1992. 25. Fletcher JC et al: Ethics in reproductive genetics, Clin Obstet Gynecol 35:763, 1992. 26. Glantz C et al: Clinical utility of sonography in the diagnosis and treatment of placental abruption, J Ultrasound Med 21:837, 2002. 27. Goldberg JD: Routine screening for fetal anomalies: expectations, Obstet Gynecol Clin North Am 31(1):35, 2004. 28. Hata T et al: Three dimensional sonographic visualization of the fetal face, Am J Roentgenol 170:481, 1998. 29. Iams JD et al: The length of the cervix and the risk of spontaneous delivery, N Engl J Med 334:567, 1996. 30. Katz VL et al: Complications of uterine leiomyomas in pregnancy, Obstet Gynecol 73:593, 1989. 31. Kofinas AD et al: Uteroplacental Doppler flow velocity waveform indices in normal pregnancy: a statistical exercise and the development of appropriate reference values, Am J Perinatol 9:94, 1992. 32. Laing FC et al: Ultrasound evaluation during the first trimester of pregnancy. In Callen PW, editor. Ultrasonography in Obstetrics and Gynecology, 4th ed. Philadelphia, WB Saunders Co, 2000. 33. Mari G et al: Noninvasive diagnosis by Doppler ultrasonography of fetal anemia due to maternal red-cell alloimmunization: Collaborative Group for Doppler Assessment of the Blood Velocity in Anemic Fetuses, N Engl J Med 342:9, 2000. 34. Marteau TM: Towards informed decisions about prenatal testing: a review, Prenat Diagn 15:1215, 1995. 35. Miller DA et al: Clinical risk factors for placenta previaplacenta accreta, Am J Obstet Gynecol 177:210, 1997. 36. Nicolaides KH et al: Ultrasound screening for spina bifida: cranial and cerebellar signs. Lancet 2:72, 1986. 37. Pearlstone M et al: Subchorionic hematoma: a review, Obstet Gynecol Surv 48:65, 1993. 38. Persutte WH et al: Single umbilical artery: a clinical enigma in modern prenatal diagnosis, Ultrasound Obstet Gynecol 6:216, 1995. 39. Resnick R: Intrauterine growth restriction, Obstet Gynecol 99:490, 2002.

174

SECTION II  THE FETUS

40. Rahemtullah A et al: Suspected skeletal dysplasias: femur length to abdominal circumference ratio can be used in ultrasonographic prediction of fetal outcome, Am J Obstet Gynecol 177:864, 1997. 41. Rice JP et al: The clinical significance of uterine leiomyomas in pregnancy, Am J Obstet Gynecol 160:1212, 1989. 42. Ross MG et al: National Institute of Child Health and Development Workshop Participants, J Matern Fetal Med 10:2, 2001. 43. Smith-Bindman R et al: Second-trimester ultrasound to detect fetuses with Down syndrome: a meta-analysis, JAMA 285:1044, 2001.

44. Snijders RC et al: First trimester trisomy screening, nuchal translucency measurement training and quality assurance to correct and unify technique, Ultrasound Obstet Gynecol 19:953, 2002. 45. United States Food and Drug Administration: Guidance for content and review of a magnetic resonance diagnostic device 510 (k) application: Safety parameter action levels. Center for Devices and Radiologic Health Report. Rockville, MD: US FDA; 1988.

CHAPTER

10

Estimation of Fetal Well-Being PART 1

Antepartum Fetal Surveillance

a fetus in jeopardy, the physician would have an opportunity to intervene before progressive fetal hypoxemia and acidosis lead to fetal death or to the delivery of a severely compromised neonate.

Susan E. Gerber

INDICATIONS FOR SURVEILLANCE The primary goal of antenatal fetal surveillance is avoidance of intrauterine fetal death. In the United States, the incidence of intrauterine fetal death of 20 weeks’ gestation or greater was 6.22 deaths per 1000 births in 2005. This rate has declined steadily over time, but has been relatively stable since 2003.20 In most pregnancies, routine prenatal care provides for fetal surveillance via clinical assessment of fetal growth and activity. In certain populations with an increased risk of fetal demise, a greater degree of fetal surveillance may be warranted, however. The first part of this chapter addresses the methods used to perform such surveillance and the evidence for their use.

Pregnancies at increased risk for intrauterine fetal demise fall into two categories: pregnancies at risk related to maternal conditions and pregnancies at risk related to pregnancyassociated conditions. Table 10-1 lists some conditions in which antenatal surveillance should be considered. There are numerous situations in which a population is known to have an increased risk of fetal death, but the etiology is unclear. One such population is the 1% of all pregnant women who are found to have an unexplained elevated maternal serum a-fetoprotein. Although numerous studies have confirmed the elevated risk of fetal death in this population, there is no consensus on whether antenatal surveillance reduces this risk.43 Antenatal testing in such a population is commonly performed, but it remains controversial.

RATIONALE FOR SURVEILLANCE Intrauterine fetal death may result from various etiologies, such as congenital malformations, fetomaternal hemorrhage, congenital infection, and uteroplacental failure. In many circumstances, deaths are precipitated by sudden catastrophic events, such as an abruptio placentae or cord prolapse. Such events are often unpredictable and not preventable by any form of antepartum surveillance. Women at risk for such events may not benefit from increased surveillance. Maternal cocaine use may increase the risk of abruptio placentae and intrauterine fetal death, but without underlying uteroplacental insufficiency or growth restriction, such an event would be unpredictable. The methods commonly used for antenatal fetal surveillance rely on fetal biophysical parameters that are sensitive to hypoxemia and acidemia, such as heart rate and movement. Blood flow in the fetal-placental circulation is responsive to these conditions as well. The surveillance tools are useful in a fetus at risk for hypoxemia because of chronic uteroplacental insufficiency. It is hoped that if fetal surveillance identifies

PHYSIOLOGIC BASIS FOR ANTENATAL SURVEILLANCE In experiments involving animal and human fetuses, hypoxemia and acidosis have been shown consistently to alter fetal biophysical parameters such as heart rate, movement, breathing, and tone.7,26,35 The fetal heart rate (FHR) is normally controlled by the fetal central nervous system (CNS) and mediated by sympathetic or parasympathetic nerve impulses originating in the fetal brainstem. The presence of intermittent FHR accelerations associated with fetal movement is believed to be an indicator of an intact fetal autonomic nervous system. In a study of fetal blood sampling of pregnancies resulting in healthy neonates, Weiner and colleagues42 established a range of normal fetal venous pH measurements. In this population, the lower 2.5 percentile of fetal venous pH was 7.37. Manning and associates28 showed that fetuses without heart rate accelerations had a mean umbilical vein pH of 7.28 (6 0.11), and fetuses with abnormal movement had a

175

176

SECTION II  THE FETUS

TABLE 10–1  Indications for Antenatal Surveillance Maternal Conditions

Pregnancy-related Conditions

Antiphospholipid syndrome

Pregnancy-induced hypertension

Hyperthyroidism (poorly controlled)

Decreased fetal movement

Diabetes mellitus

Oligohydramnios

Cyanotic heart disease

Polyhydramnios

Systemic lupus erythematosus

Intrauterine growth restriction

Hypertensive disorders

Multiple gestation

Chronic renal disease

Post-term pregnancy

Hemoglobinopathies (hemoglobin SS, SC, or S-thalassemia)

Previous fetal demise (unexplained or recurrent risk) Isoimmunization (moderate to severe) Preterm premature rupture of membranes Unexplained thirdtrimester bleeding

From American College of Obstetricians and Gynecologists: Antepartum fetal surveillance, ACOG Practice Bulletin 9, Washington, DC, 1999, ACOG.

[CST], in which the patient has regular uterine contractions, either spontaneously or induced). Sometimes a woman undergoing NST is found to have spontaneous contractions, adding the reassurance of a negative CST. With the patient in a recumbent, tilted position, FHR is monitored with an external transducer for up to 40 minutes. The FHR tracing is observed for the presence of accelerations above the baseline. A reactive test is one in which there are at least two accelerations that peak 15 beats/min above the baseline and last (not at the peak) for at least 15 seconds before returning to baseline (Fig. 10-1). Most NSTs are reactive within the first 20 minutes. For tests that are not, possibly because of a fetal sleep cycle, an additional 20 minutes of monitoring may be needed. A nonreactive NST is one in which two such accelerations do not occur within 40 minutes, or in which the acceleration peaks are less than 15 beats/min. Although the NST is noninvasive and easy to perform, it is limited by a high rate of false-positive results. Normal fetuses often have periods of nonreactivity because of benign variations such as sleep cycles. Vibroacoustic stimulation may be used safely in the setting of a nonreactive NST to elicit FHR accelerations without compromising the sensitivity of the NST.44 In this situation, the operator places an artificial larynx on the maternal abdomen and activates the device for 1 to 3 seconds. This technique is often useful in situations in which the FHR has normal beat-to-beat variability and no decelerations, but does not show any accelerations on the NST. If the test remains nonreactive, further evaluation with a biophysical profile or CST is warranted.

CONTRACTION STRESS TEST mean pH of 7.16 (6 0.08). These and similar observations were the basis for the development of antenatal fetal testing modalities that are currently in use.

NONSTRESS TEST In most institutions, the first-line assessment tool for fetal surveillance is the nonstress test (NST). Lee and coworkers19 first described the association between FHR accelerations and fetal movements in 1975. Monitoring for the presence or absence of both elements was proposed as a method of evaluation of fetal well-being. The NST is performed in a nonlaboring patient (as opposed to the contraction stress test

CST is designed to evaluate FHR response to maternal uterine contractions. The principles that are applied to the evaluation of intrapartum FHR monitoring (see Part 2) are used here. In response to the stress of the contraction, a hypoxemic fetus shows FHR patterns of concern such as late decelerations, indicating worsening hypoxemia or fetal compromise. Similar to NST, for CST the patient is placed in a recumbent tilted position, and FHR is monitored with an external fetal monitor. FHR pattern is evaluated while the patient experiences at least three contractions lasting 40 seconds within a 10-minute period. If the patient is not contracting spontaneously, contractions may be induced with nipple stimulation or intravenous oxytocin. If no late or significant

Figure 10–1.  Reactive nonstress test, showing accelerations occurring with fetal movement. Note the arrows on the contraction

channel. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins.)

Chapter 10  Estimation of Fetal Well-Being

variable decelerations are noted on FHR tracing, CST is considered to be negative. If there are late decelerations after at least 50% of the contractions, CST is positive. If late decelerations are present less than 50% of the time, or if significant variable decelerations are present, the test is considered to be equivocal. Contraindications to the performance of CST include clinical situations in which labor would be undesirable (e.g., placenta previa or previous classic cesarean section).

BIOPHYSICAL PROFILE The biophysical profile was developed by Manning and associates27 as an alternative tool to other methods of antenatal surveillance to evaluate fetal well-being. As originally described, it combines NST with four components evaluated by ultrasonography. In a 30-minute period, the following observations are sought: 1. Fetal breathing movements (one or more episodes lasting

at least 30 seconds)

2. Fetal movement (three or more discrete body or limb

movements)

3. Fetal tone (one or more episodes of active extension with

return to flexion of a limb or trunk, or the opening and closing of a fetal hand) 4. Amniotic fluid volume (originally described as a single vertical pocket of $1 cm, subsequently modified to $2 cm) (Fig. 10-2) 5. Reactive NST Each component is assigned a score of 2 if present and 0 if absent. A combined score of 8 or 10 indicates fetal well-being. A score of 6 is considered to be equivocal, and it usually merits delivery if the pregnancy is at term, or repeat testing in 24 hours if the pregnancy is preterm. A score of 4 or less is considered to be abnormal, and delivery is warranted except for extenuating circumstances. The biophysical profile also has been analyzed with the four ultrasound parameters alone, and when all are present, it has been shown to have a false-negative rate similar to the full biophysical profile.25

177

AMNIOTIC FLUID VOLUME ASSESSMENT Amniotic fluid volume is commonly estimated on ultrasound scan via one of two methods. The amniotic fluid index is calculated by measuring the maximal vertical pockets of fluid (without loops of umbilical cord) in each of the four quadrants of the maternal abdomen (see Fig. 10-2). Alternatively, the single deepest vertical pocket of fluid is measured. Decreased amniotic fluid volume, or oligohydramnios, is typically defined as either an amniotic fluid index of 5 cm or less or no measurable vertical pocket of fluid greater than 2 cm. Although both measurement techniques are equally effective in estimating amniotic fluid volume, both have relatively poor sensitivity for the detection of oligohydramnios.22 In most circumstances, oligohydramnios is thought to reflect fetal compromise. A decrease in placental perfusion results in decreased blood flow and consequently decreased oxygen delivery to the fetus. There is also decreased renal perfusion by the preferential shunting of blood to the fetal brain. Decreased renal perfusion results in decreased fetal urine output, which leads to a decreased amniotic fluid volume. Oligohydramnios is commonly associated with post-term pregnancy, fetal growth restriction, maternal hypertension, and preeclampsia. In various clinical scenarios, oligohydramnios has been found to be associated with an increased risk of preterm delivery, low or very low birthweight, low Apgar scores, intrauterine fetal death, meconium-stained amniotic fluid, admissions to a neonatal intensive care unit, and cesarean delivery for nonreassuring fetal status.3,18,21 In a term pregnancy, oligohydramnios is considered an indication for delivery. In a preterm pregnancy, immediate delivery may be undesirable, and in such cases increased surveillance is warranted. In a preterm fetus with oligohydramnios, delivery is indicated for nonreassuring or abnormal fetal surveillance, or when there is no interval growth on ultrasound.

MODIFIED BIOPHYSICAL PROFILE Although the NST reflects the present fetal neurologic status and oxygenation, amniotic fluid volume is a better measure of chronic placental function. Some authors have favored the use of the modified biophysical profile, which consists of the combination of NST and amniotic fluid index. In a study of 15,482 women undergoing antenatal testing, Miller and colleagues32 found modified biophysical profile to have a lower false-negative rate than NST. An abnormal result may be followed by a full biophysical profile or a CST.

DOPPLER FLOW VELOCIMETRY

Figure 10–2.  Ultrasound image of the maximal vertical pocket measured as part of either the biophysical profile or the amniotic fluid index. Note the absence of umbilical cord in the measured pocket.

Ultrasonography of fetal and maternal blood flow is also used to evaluate the health of a pregnancy. The most commonly studied vessel is the fetal umbilical artery. Doppler measurements of the pulsatile blood flow in the umbilical arteries directly reflect the status of the fetomaternal circulation (Fig. 10-3). A progressive decrease in placental function or blood flow is thought to manifest with an increased resistance to flow as evidenced by a diminution in the diastolic flow and eventual absence or reversal of flow during diastole in the fetal vessels. In clinical practice, commonly measured

178

SECTION II  THE FETUS

A

B

C Figure 10–3.  Composite images of three studies of fetal umbilical arterial velocimetry, ranging from normal to markedly abnormal. A, Normal velocimetry pattern. B, Absent diastolic flow, indicating increased placental resistance. C, Reversal of diastolic flow, indicating worsening placental function. D, diastolic velocity; S, systolic velocity.

indexes include the systolic/diastolic (S/D), ratio resistance index [(S-D)/S], and pulsatility index [(S-D)/A]. Doppler flow velocimetry of the fetal umbilical artery has been studied in numerous at-risk populations. Pregnancies with suspected intrauterine growth restriction have been extensively studied, and there is evidence that in this population, abnormal blood flow in the umbilical artery is associated with increased risk of perinatal morbidity and mortality.39 A Cochrane review of 11 randomized trials showed a trend toward decreased perinatal mortality with the use of Doppler assessment of the umbilical artery in high-risk pregnancies.36 The use of umbilical artery Doppler flow velocimetry as a primary testing method results in fewer antenatal tests and less intervention with similar neonatal outcome compared with pregnancies monitored by NST.17 There is also evidence that abnormal umbilical artery blood flow velocimetry precedes FHR abnormalities by a median of 7 days.2 Doppler flow velocimetry has also been studied in other vessels including the fetal ductus venosus, fetal middle cerebral artery, and maternal uterine artery. Evaluation of the fetal ductus venosus remains controversial. In the setting of severe preterm fetal growth restriction, abnormal flow in the ductus venosus has been shown to be independently associated with perinatal morbidity and mortality.5,6,12 The utility of this assessment has been questioned, however, given an apparent

brief latency between abnormality and fetal compromise and poor predictive values, particularly at early gestational ages.15 In a high-risk cohort, the presence of abnormal flow velocity or “notching” in the maternal uterine artery at 22 to 24 weeks’ gestation has been associated with an increased risk of subsequent preeclampsia and fetal growth restriction.11 More recent studies have also shown a possible association with fetal growth restriction for this finding in the first trimester.31 Controversy remains regarding the optimal timing of the evaluation, and the specific abnormalities (“notch” versus abnormal indexes) that are most predictive of adverse outcome.30 Doppler flow velocimetry of the fetal umbilical artery has not been found to convey any benefit in a low-risk population.1 Selective use of Doppler velocimetry for traditional indications within a general population has not been shown to result in decreased maternal hospitalization or improved neonatal outcome.38 Uterine artery Doppler flow velocimetry has also been shown to have limited predictive value in a low-risk population.10 Fetal cerebral arteries show altered blood flow in the setting of hypoxemia. In certain pregnancies, such as those complicated by isoimmunization, the fetus is at risk for the development of severe anemia. Traditionally, standard practice has been to evaluate such pregnancies with serial amniocentesis or cordocentesis or both to determine if a fetus is at risk for intrauterine death owing to anemia. Mari29 described the use of Doppler velocimetry of the fetal middle cerebral artery to evaluate fetuses at risk for anemia because of maternal isoimmunization. The peak systolic velocity of the middle cerebral artery is inversely correlated with fetal hemoglobin, and the measurement provides a noninvasive means of detecting an increased risk of fetal anemia (Fig. 10-4). These results have been shown to have equal or greater sensitivity in the detection of severe fetal anemia and have replaced the use of amniocentesis for this diagnosis in many centers.37

Figure 10–4.  Ultrasound image of the fetal cerebral vascula-

ture, with Doppler visualization of the middle cerebral artery and measurement of the peak systolic velocity. (From Dukler D et al: Noninvasive tests to predict fetal anemia: a study comparing Doppler and ultrasound parameters, Am J Obstet Gynecol 188:1310, 2003.)

Chapter 10  Estimation of Fetal Well-Being

INTERPRETATION OF TEST RESULTS The realistic goal of antepartum testing is to decrease the risk of intrauterine fetal demise or perinatal mortality in the tested population so that it approaches the rate for a low-risk population without an excessive or unacceptable false-positive rate that may result in unnecessary intervention. When corrected for congenital anomalies and unpredictable causes of intrauterine death, the rate of stillbirth in the tested population (after antepartum testing with normal results) has been reported to be approximately 1.9 per 1000 for NST, 0.3 per 1000 for CST, 0.8 per 1000 for biophysical profile, and 0.8 per 1000 for modified biophysical profile.1 These rates are comparable to the rates for the risk of fetal death in a low-risk population. The false-positive rate is more difficult to ascertain because a positive test usually results in obstetric intervention, significantly decreasing the likelihood of intrauterine death. One study showed, however, that 90% of nonreactive NSTs are followed by a negative CST result, consistent with a high falsepositive rate for NST.13 A study of CSTs in which physicians were blinded to the results found that in 61% of patients with positive tests there were no fetal late decelerations in labor, no low Apgar scores, and no significant neonatal morbidity.41 Manning and associates24 reported on a cohort of 913 infants delivered after biophysical profile score of 6 or less. Nearly 40% of infants with scores of 6 showed no markers of fetal compromise at delivery, as defined by fetal distress in labor, admission to the neonatal intensive care unit, a 5-minute Apgar score of 7 or less, or an umbilical cord pH less than or equal to 7.20. There was a significant inverse linear association, however, between biophysical profile score and these markers, and all fetuses with scores of 0 had at least one of these markers at delivery. In a clinically stable situation, reassuring tests (reactive NST, negative CST, and biophysical profile of 8 or 10) are considered reliable for 1 week, and so testing is usually performed on a weekly basis. Labile conditions may merit more frequent testing; the frequency is left to the discretion of the physician. If the indication for testing is not a persistent one (e.g., maternal perception of decreased fetal movement), there is no evidence to support the continuation of antenatal testing. In certain high-risk populations, the false-negative rate of NST may be unacceptably high. The stillbirth rate within 1 week of a reactive NST is markedly higher for patients with diabetes mellitus (14 per 1000) and fetal growth restriction (20 per 1000).4 Similarly, elevated results have been reported for patients with prolonged gestations.33 Boehm and coworkers8 found that the stillbirth rate decreased from 6.1 per 1000 to 1.9 per 1000 in their high-risk population when the frequency of testing was changed from once weekly to twice weekly. For this reason, testing twice weekly may be appropriate in certain populations, such as those described.

FETAL GESTATIONAL AGE AND ANTENATAL SURVEILLANCE FHR variability and reactivity vary with gestational age. Before 28 weeks of gestation, 50% of all NSTs may not be reactive. From 28 to 32 weeks of gestation, approximately 15% of normal fetuses have nonreactive NSTs.1 Although fetal breathing

179

movements and body movements are noted to decrease before the onset of spontaneous labor,9 the biophysical parameters that make up the biophysical profile score are present at early gestational ages, and are useful in the evaluation of a very premature fetus. The optimal gestational age at which to begin antenatal surveillance depends on the clinical condition. In making this decision, the physician must weigh the risk of intervention at a premature gestational age against the risk of intrauterine fetal death. The American College of Obstetricians and Gynecologists recommends initiating testing at 32 to 34 weeks’ gestation for most at-risk patients, with the acknowledgment that some situations may warrant testing earlier at 26 to 28 weeks of gestation.1

CLINICAL CONSIDERATIONS Many clinical situations not related to fetal well-being temporarily affect the interpretation of antenatal surveillance techniques. Maternal cigarette smoking and alcohol ingestion decrease fetal movement, breathing, and heart rate reactivity.14,16,23 Commonly used maternal medications such as opioids and corticosteroids significantly decrease various fetal biophysical parameters, including movement, breathing, and heart rate reactivity, without compromising neonatal outcome.34,40 Maternal medical conditions, such as an acute asthma exacerbation or diabetic ketoacidosis, may result in nonreassuring fetal surveillance, including worrisome biophysical profile scores. Delivery of a mother in such an unstable condition is dangerous and undesirable. The improvement or elimination of such conditions usually results in an improvement in fetal status. The improved fetal status is accompanied by an improvement in antenatal testing results, and it negates the need for a premature delivery.

SUMMARY In high-risk populations at increased risk of perinatal mortality, antenatal fetal surveillance plays a large role in prenatal care. Pregnancies at risk for progressive deterioration of placental function leading to fetal hypoxemia and acidosis are most likely to benefit from the methods currently in use. The various modalities, including NST, CST, biophysical profile, and Doppler velocimetry, rely on fetal biophysical parameters that are significantly associated with the presence or absence of fetal hypoxemia. Because all tests are associated with a false-positive rate, each test result should be interpreted within the clinical context presented by the patient.

REFERENCES 1. American College of Obstetricians and Gynecologists: Antepartum fetal surveillance, ACOG Practice Bulletin 9. Washington, DC: ACOG, 1999. 2. Arduini D et al: The development of abnormal heart rate patterns after absent end-diastolic velocity in umbilical artery: analysis of risk factors, Am J Obstet Gynecol 168:43, 1993. 3. Baron C et al: The impact of amniotic fluid volume assessed intrapartum on perinatal outcome, Am J Obstet Gynecol 173:167, 1995.

180

SECTION II  THE FETUS

4. Barrett J et al: The nonstress test: an evaluation of 1,000 patients, Am J Obstet Gynecol 141:153, 1981. 5. Baschat AA et al: Predictors of neonatal outcome in early-onset placental dysfunction, Obstet Gynecol 109:253, 2007. 6. Bilardo CM et al: Relationship between monitoring parameters and perinatal outcome in severe, early intrauterine growth restriction, Ultrasound Obstet Gynecol 23:119, 2004. 7. Boddy K et al: Foetal respiratory movements, electrocortical and cardiovascular responses to hypoxaemia and hypercapnia in sheep, J Physiol 243:599, 1974. 8. Boehm FH et al: Improved outcome of twice weekly nonstress testing, Obstet Gynecol 67:566, 1986. 9. Carmichael L et al: Fetal breathing, gross fetal body movements and maternal and fetal heart rates before spontaneous labour at term, Am J Obstet Gynecol 148:675, 1984. 10. Chien PFW et al: How useful is uterine artery Doppler flow velocimetry in the prediction of pre-eclampsia, intrauterine growth retardation and perinatal death? An overview, Br J Obstet Gynaecol 107:196, 2000. 11. Coleman MAG et al: Mid-trimester uterine artery Doppler screening as a predictor of adverse pregnancy outcome in high-risk women, Ultrasound Obstet Gynecol 15:7, 2000. 12. Cosmi E et al: Doppler, cardiotocography, and biophysical profile changes in growth-restricted fetuses, Obstet Gynecol 106:1240, 2005. 13. Evertson LR et al: Antepartum fetal heart rate testing, I: evolution of the non-stress test, Am J Obstet Gynecol 133:29, 1979. 14. Fox HE et al: Maternal ethanol ingestion and the occurrence of fetal breathing movements, Am J Obstet Gynecol 132:34, 1978. 15. Ghidini A: Doppler of the ductus venosus in severe preterm fetal growth restriction: a test in search of a purpose? Obstet Gynecol 109:250, 2007. 16. Graca LM et al: Acute effects of maternal cigarette smoking on fetal heart rate and fetal body movements felt by the mother, J Perinat Med 19:385, 1991. 17. Haley J et al: Randomised controlled trial of cardiotocography versus umbilical artery Doppler in the management of small for gestational age fetuses, Br J Obstet Gynaecol 104:431, 1997. 18. Hsieh TT et al: Perinatal outcome of oligohydramnios without associated premature rupture of membranes and fetal anomalies, Gynecol Obstet Invest 45:232, 1998. 19. Lee CY et al: A study of fetal heart rate acceleration patterns, Obstet Gynecol 45:142, 1975. 20. MacDorman MF, Kirmeyer S: Fetal and perinatal mortality, United States, 2005, Natl Vital Stat Rep 57:1, 2009. 21. Magann EF et al: Comparability of the amniotic fluid index and single deepest pocket measurements in clinical practice, Aust N Z J Obstet Gynaecol 43:75, 2003. 22. Magann EF et al: How well do the amniotic fluid index and single deepest pocket indices (below the 3rd and 5th and above the 95th and 97th percentiles) predict oligohydramnios and hydramnios? Am J Obstet Gynecol 190:164, 2004. 23. Manning FA, Feyerbend C: Cigarette smoking and fetal breathing movements, Br J Obstet Gynaecol 83:262, 1976. 24. Manning FA et al: Fetal assessment based on fetal biophysical profile scoring, IV: an analysis of perinatal morbidity and mortality, Am J Obstet Gynecol 162:703, 1990.

25. Manning FA et al: Fetal biophysical profile scoring: selective use of the non-stress test, Am J Obstet Gynecol 156:709, 1987. 26. Manning FA, Platt LD: Maternal hypoxemia and fetal breathing movements, Obstet Gynecol 53:758, 1979. 27. Manning FA et al: Antepartum fetal evaluation: development of a fetal biophysical profile score, Am J Obstet Gynecol 136:787, 1980. 28. Manning FA et al: Fetal biophysical profile score, VI: correlation with antepartum umbilical venous fetal pH, Am J Obstet Gynecol 169:755, 1993. 29. Mari G: Noninvasive diagnosis by Doppler ultrasonography of fetal anemia due to maternal red-cell alloimmunization, N Engl J Med 342:9, 2000. 30. Mari G: Doppler ultrasonography in obstetrics: from the diagnosis of fetal anemia to the treatment of intrauterine growth-restricted fetuses, Am J Obstet Gynecol 200:613.e1, 2009. 31. Melchiorre K et al: First-trimester uterine artery Doppler indices in the prediction of small-for-gestational age pregnancy and intrauterine growth restriction, Ultrasound Obstet Gynecol 33:524, 2009. 32. Miller DA et al: The modified biophysical profile: antepartum testing in the 1990s, Am J Obstet Gynecol 174:812, 1996. 33. Miyazaki F, Miyazaki B: False reactive nonstress tests in postterm pregnancies, Am J Obstet Gynecol 140:269, 1981. 34. Mulder EJ et al: Antenatal corticosteroid therapy and fetal behaviour: a randomised study of the effects of betamethasone and dexamethasone, Br J Obstet Gynaecol 104:1239, 1997. 35. Murata Y et al: Fetal heart rate accelerations and late decelerations during the course of intrauterine death in chronically catheterized rhesus monkeys, Am J Obstet Gynecol 144:218, 1982. 36. Neilson JP, Alfirevic Z: Doppler ultrasound for fetal assessment in high risk pregnancies, Cochrane Database Syst Rev (2): CD000073, 2000. 37. Oepkes D et al: Doppler ultrasonography versus amniocentesis to predict fetal anemia, N Engl J Med 355:156, 2006. 38. Omtzigt AM et al: A randomized controlled trial on the clinical value of umbilical Doppler velocimetry in antenatal care, Am J Obstet Gynecol 170:624, 1994. 39. Signore C et al: Antenatal testing—a reevaluation: executive summary of a Eunice Kennedy Shriver National Institute of Child Health and Human Development workshop, Obstet Gynecol 113:687, 2009. 40. Smith CV et al: Influence of intravenous fentanyl on fetal biophysical parameters during labor, J Matern Fetal Med 5:89, 1996. 41. Staisch KJ et al: Blind oxytocin challenge test and perinatal outcome, Am J Obstet Gynecol 138:399, 1980. 42. Weiner CP et al: The effect of fetal age upon normal fetal laboratory values and venous pressure, Obstet Gynecol 79:713, 1992. 43. Wilkins-Haug L: Unexplained elevated maternal serum alpha-fetoprotein: what is the appropriate follow-up? Curr Opin Obstet Gynecol 10:469, 1998. 44. Zimmer EZ, Divon MY: Fetal vibroacoustic stimulation, Obstet Gynecol 81:451, 1993.

Chapter 10  Estimation of Fetal Well-Being

PART 2

Evaluation of the Intrapartum Fetus Susan E. Gerber and Barrett K. Robinson

It has long been recognized that the process of labor and delivery is metabolically stressful for the fetus. Most healthy fetuses are able to tolerate this challenge with no adverse effect. The purpose of intrapartum fetal monitoring is to identify fetuses for whom the stresses of labor induce such significant acidemia and hypoxemia that permanent harm would be risked if delivery were not imminent. Identification of this risk should allow physicians to intervene and prevent intrapartum death or neonatal damage. This part addresses the methods currently in use to accomplish this goal and discusses their limitations.

FETAL HEART RATE MONITORING In the United States, the primary method used for intrapartum surveillance is the evaluation of FHR. Continuous electronic FHR monitoring, which was introduced in the 1970s, typically is used. Although electronic FHR monitoring has never been shown to result in better neonatal outcomes compared with intermittent auscultation of FHR,1 85% of all deliveries were managed with electronic FHR monitoring in 2003.12 During labor, uterine contractions produce increased intrauterine pressure, decreased maternal uterine artery blood flow, and intermittent reduction of placental gas exchange. Over the course of a normal labor, all fetuses show a progressive decrease in oxygenation and pH.13 A fetus is not at risk for permanent neurologic damage, however, until the levels of hypoxia and acidosis become extreme. The underlying rationale for intrapartum FHR monitoring is the same as the rationale behind antepartum surveillance with NST or CST. FHR is normally controlled by the fetal CNS, and the fetal CNS is sensitive to hypoxia. Experimentally induced hypoxia is associated with consistent and predictable changes in FHR. Most studies have been performed in fetal sheep, with intermittent cord occlusion used to produce hypoxia.20 These changes, when present in the FHR tracing, are thought to be indicators of fetal hypoxia.

181

Electronic FHR monitoring may be performed externally or internally. Telemetry units are available, so that a patient need not be confined to a bed to be monitored. Most external monitors use a Doppler device with computerized logic to interpret and count the Doppler signals. Internal FHR monitoring may be performed in circumstances in which external monitoring is unreliable or not feasible (e.g., maternal obesity). Internal monitoring uses an electrode that is attached to the fetal scalp. The FHR tracing is transmitted to a monitor and interpreted visually by the medical team.

INTERPRETATION OF FETAL HEART RATE PATTERNS Baseline Fetal Heart Rate Baseline FHR is the average FHR rounded to increments of 5 beats/min during a 10-minute segment, excluding periodic or episodic changes, periods of marked variability, or baseline segments that differ by more than 25 beats/min. In any given 10-minute window, the minimal baseline duration must be at least 2 minutes, or the baseline is considered indeterminate. In cases in which the baseline is indeterminate, the previous 10-minute window should be reviewed and used to determine the baseline. A normal FHR baseline rate ranges from 110 to 160 beats/min. If the baseline FHR is less than 110 beats/min, it is termed bradycardia. If the baseline FHR is greater than 160 beats/min, it is termed tachycardia. The initial response of FHR to intermittent hypoxia is deceleration, but baseline tachycardia may develop if the hypoxia is prolonged and severe. Tachycardia may also be associated with conditions other than hypoxia, such as maternal fever, intra-amniotic infection, thyroid disease, the presence of medication, and cardiac arrhythmia (Fig. 10-5).

Fetal Heart Rate Variability The presence of baseline FHR variability is a useful indicator of fetal CNS integrity. Variability is defined as fluctuations in FHR baseline of two cycles per minute or greater, with irregular amplitude and inconstant frequency. Moderate FHR variability is strongly associated (98%) with an umbilical pH greater than 7.15.16 In most cases, normal FHR variability provides reassurance about fetal status. In the absence of maternal sedation, magnesium sulfate administration, long-term beta-blocker treatment, or extreme prematurity, decreased variability, or

Figure 10–5.  Fetal tachycardia, with a baseline fetal heart rate of 165 beats/min. This was in the setting of a maternal fever.

(From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 70.)

182

SECTION II  THE FETUS

flattening of FHR baseline, may serve as a barometer of the fetal response to hypoxia. Because decreased variability or flattening is presumed to be a CNS response, in most situations decelerations of FHR precede the loss of variability, indicating the cause (Fig. 10-6). FHR variability is visually assessed as the amplitude of the peak to trough in beats per minute as follows: Amplitude range Undetectable 1 to 5 beats/min 6 to 25 beats/min .25 beats/min

Classification Absent Minimal Moderate Marked

Accelerations Accelerations in FHR are periodic elevations above the baseline, and they are usually associated with fetal movement. The presence of FHR accelerations during labor is always reassuring, as it is in a nonlaboring patient. The absence of accelerations during labor is not reflective of fetal hypoxemia, however.

Decelerations Decelerations in FHR are episodic decreases below the baseline. They are mediated by the fetal chemoreflex in response to hypoxia, presumably to reduce the myocardial work and oxygen requirements. Decelerations are quantified by the depth of the nadir in beats per minute and the duration from the beginning to the end of the deceleration. Decelerations are defined further as “recurrent” if they occur with at least 50% of the contractions. Decelerations are categorized by virtue of their morphology and timing, and this classification is thought to identify the underlying mechanism responsible for the decelerations.

Three types of decelerations were initially described by Hon and colleagues in 1967.9 Early decelerations are shallow and symmetric, gradual in onset and recovery, and associated with a contraction such that the nadir of the deceleration occurs at the same time as the peak of the contraction. They are thought to be caused by the compression of the fetal head in active labor and are associated with a favorable fetal outcome (Fig. 10-7). Variable decelerations are the most common form of decelerations in labor. They are typically associated with an abrupt onset and abrupt return to baseline. They vary in shape, depth, and duration, although they usually coincide with a contraction. They are also frequently preceded and followed by small accelerations in FHR (Figs. 10-8 and 10-9). Variable decelerations are often associated with compression of the umbilical cord and with favorable outcome, particularly if they are brief, shallow, and not recurrent. Repetitive prolonged and deep variable decelerations may result in progressive hypoxemia and acidemia. Late decelerations, by contrast, have a more gradual onset and return to baseline—typically 30 seconds or more from onset to nadir. The onset, nadir, and recovery of the deceleration occur after the onset, peak, and end of the contraction (Fig. 10-10). Although an occasional late deceleration is not indicative of fetal compromise, recurrent late decelerations are considered to be nonreassuring and cause for further evaluation, remediation, or delivery. A sinusoidal heart rate pattern consists of a regular oscillation of the baseline variability, in a smooth undulating pattern. This pattern typically lasts at least 10 minutes, has a relatively fixed period of three to five cycles per minute, and has an amplitude of 5 to 15 beats/min above and below the baseline (Fig. 10-11). This pattern is quite rare, but can be associated with severe chronic anemia or severe hypoxia and acidosis.

Figure 10–6.  Abnormal (absent) fetal heart rate (FHR) variability. Because there are no decelerations present, this would

qualify as a category II FHR tracing.  (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 138.)

Chapter 10  Estimation of Fetal Well-Being

183

Figure 10–7.  Early decelerations. Note the way in which the decelerations appear to “mirror” the uterine contractions. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 74.)

Figure 10–8.  Variable decelerations. Note the normal fetal heart rate baseline and variability between the decelerations. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 78.)

Figure 10–9.  Variable decelerations. In contrast to Figure 10–8, note the tachycardic fetal heart rate (FHR) baseline and absent

variability between the decelerations. Also note the prolonged FHR “overshoot” in the middle portion of the panel. This FHR tracing is more concerning for the presence of fetal hypoxia and acidosis, and would be considered category III.

Figure 10–10.  Late decelerations. Note the timing of the onset, nadir, and recovery of the deceleration, which occur after the

onset, peak, and end of the contraction. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 201.)

184

SECTION II  THE FETUS

Figure 10–11.  Sinusoidal fetal heart rate pattern. This fetus was severely anemic and hydropic because of maternal Rh isoim-

munization. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 166.)

INTERPRETATIVE SYSTEMS FOR CLASSIFICATION OF FETAL HEART RATE PATTERNS

Three-Tier Fetal Heart Rate Interpretation System17

Since its introduction, the utility of electronic fetal monitoring has been limited by the subjective nature of its interpretation. Although many interpretative systems exist for FHR tracings, the 2008 National Institutes of Child Health and Human Development (NICHHD) workshop dealing with Standardization of Nomenclature for Intrapartum Electronic Fetal Monitoring, jointly sponsored by NICHHD, the American College of Obstetricians and Gynecologists, and the Society for Maternal Fetal Medicine, had certain criteria. The system selected by the workshop is one that is evidence-based, simple, and applicable to clinical practice.11 The FHR response is a dynamic process that requires frequent reassessment. Categorization of a tracing is limited to the time period being assessed, and over time FHR tracings commonly migrate from one category to another. FHR tracing patterns provide information on the current acid-base status of the fetus and cannot predict the development of cerebral palsy. The two FHR findings that reliably predict the absence of acidemia are accelerations and moderate variability. Accelerations are defined by a peak of 15 beats/min or greater above the baseline, with the acceleration lasting 15 seconds or more, but less than 2 minutes from the onset to the return to the previously determined baseline. In pregnancies of less than 32 weeks’ gestational age, accelerations are defined as having a peak 10 beats/min or greater above the baseline and a duration of 10 seconds or more. Although either fetal accelerations or moderate FHR variability reliably predicts the absence of acidemia, the absence of accelerations, the presence of minimal vari­ ability, or the presence of absent variability do not reli­ ably predict the presence of fetal hypoxemia or metabolic acidemia. The significance of marked variability (formerly described as saltatory) is unclear. Although the entire asso­ ciated clinical circumstances must always be taken into account, the 2008 NICHHD workshop simplified categorization and interpretation of FHR tracings into a three-tier system described next.11

Normal tracings, which are strongly predictive of normal fetal acid-base status at the time of observation and can be followed in a routine manner without any specific action required, include all of the following:

CATEGORY I

n Baseline

rate of 110 to 160 beats/min variability n Absence of any late or variable decelerations n Early decelerations may or may not be present n Accelerations may or may not be present n Moderate

CATEGORY II Indeterminate tracings, although not predictive of abnormal fetal acid-base status, cannot be classified as category I or III, and require evaluation and continued surveillance and reevaluation. These tracings are frequently encountered in clinical care and include any of the following: n Baseline

rate

n Tachycardia n Bradycardia

not accompanied by absent baseline variability FHR variability n Minimal baseline variability n Absent baseline variability not accompanied by recurrent decelerations n Marked baseline variability n Absence of induced accelerations after fetal stimulation (e.g., scalp stimulation, vibroacoustic stimulation, direct fetal scalp sampling, transabdominal halogen light) n Periodic or episodic decelerations n Recurrent variable decelerations accompanied by minimal or moderate baseline variability n Prolonged deceleration 2 minutes or more, but less than 10 minutes n Recurrent late decelerations with moderate baseline variability n Variable decelerations with other characteristics, such as slow return to baseline, “overshoots,” or “shoulders” n Baseline

Chapter 10  Estimation of Fetal Well-Being

185

CATEGORY III

focusing on the relationship between such tracings and clinical outcomes.11

Abnormal tracings, which are predictive of abnormal fetal acid-base status at the time of observation, require prompt evaluation and initiation of expeditious attempts to resolve the abnormal FHR pattern, such as provision of maternal oxygen, change in maternal position, discontinuation of labor stimulation, treatment of maternal hypotension, or additional efforts. These tracings include either:

EVALUATION AND MANAGEMENT OF NONREASSURING FETAL HEART RATE PATTERNS

n Absent

baseline FHR variability along with any of the following n Recurrent late decelerations n Recurrent variable decelerations n Bradycardia n Sinusoidal pattern Standardization of terminology and subsequent categorization into one of three tiers should aid providers when attempting to decide whether the patterns suggest a lack of fetal acidemia or alternatively require intervention. Despite numerous studies having shown that interobserver and intraobserver variability is high when FHR tracings are reviewed,7,15 there is a common consensus that normal tracings classified as category I indicate an absence of fetal acidemia and are reassuring.9-11 Acidemia may be present in one of four fetuses with abnormal, nonreassuring category III FHR tracings.16 Although expeditious action is indicated to resolve the concerning aspects of the abnormal tracing or to consider moving toward delivery, because of the low prevalence of intrapartum fetal asphyxia, abnormal tracings have a wellrecognized false-positive rate greater than 90%.9 Patients with indeterminate tracings—classified as category II—may ultimately be the most difficult to manage in clinical practice. A mainstay of the recommended strategy is close, continuous evaluation and assessment. These tracings may ultimately fit criteria for normal, reassuring category I tracings as time passes or after subsequent evaluative strategies, at which point confidence in the nonacidemic status of the fetus may be gained. Alternatively, indeterminate tracings may ultimately meet the criteria for abnormal category III tracings, in which case the imperative to resolve concerning aspects or move expeditiously toward delivery becomes clear.17 Because of the potential uncertainty inherent in these nonpredictive tracings, a call has been issued for investigational research

The evaluation of a nonreassuring FHR pattern begins with a search for an etiology that would require immediate delivery. A prolonged deceleration might be the result of an umbilical cord prolapse, or an acute placental abruption or uterine rupture (Fig. 10-12). In the absence of such events, one must search for a remediable cause of the nonreassuring FHR tracing. Repetitive late decelerations or fetal bradycardia may result from contractions that are too frequent in timing or from uterine hyperstimulation (Fig. 10-13). Without adequate recovery between contractions, the fetus is unable to tolerate the episodes of hypoxemia and becomes progressively acidotic. When the hyperstimulation is stopped, either by discontinuing or decreasing an oxytocin infusion or by administering a betamimetic agent, the fetal status typically improves, however, and this pattern resolves. Nonreassuring FHR patterns may also be seen in the setting of maternal conditions such as severe anemia, hypotension, hypoxemia (e.g., in the setting of an acute respiratory event such as an asthma attack), or acidosis (e.g., an episode of diabetic ketoacidosis). Improvement of the maternal status typically results in the return of a reassuring FHR pattern. Not only is delivery not required in such a setting, but it is often undesirable because of the fragile maternal medical condition. Maternal position in labor can affect the FHR tracing as the supine position decreases uterine blood flow and placental perfusion. Repositioning a patient to the lateral recumbent position can often reverse a nonreassuring FHR pattern with no other intervention. Supplemental oxygen therapy for the mother is often administered and thought to increase fetal oxygen levels, although there are no data on the efficacy or safety of this therapy.9 When recurrent variable decelerations are present, amnioinfusion, in which fluid is infused into the uterine cavity, has been shown to decrease the rate of variable decelerations and cesarean delivery for nonreassuring fetal status.8

Figure 10–12.  Prolonged deceleration associated with acute prolapse of the umbilical cord. Note the reassuring fetal heart

rate tracing leading up to the deceleration, indicating an acute event. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 83.)

186

SECTION II  THE FETUS

Figure 10–13.  Prolonged deceleration associated with uterine hyperstimulation. Note the recovery of the deceleration after

discontinuation of the oxytocin infusion. (From Freeman R, Garite T: Fetal heart rate monitoring, Baltimore, 1981, Williams & Wilkins, p 85.) Interpretation of FHR baseline must take into account any maternal medication administration. Numerous commonly used medications, such as magnesium sulfate, corticosteroids, are known to result in decreased FHR variability and narcotics. Diminished variability in this setting may not reflect fetal hypoxia, but other criteria must be used for evaluation of the fetal status. The clinical context of the fetus must also be considered when evaluating the FHR tracing because not all fetuses have the same ability to tolerate a normal labor. A preterm fetus or fetus with growth restriction typically has less placental reserve and is more susceptible to a rapid progression in hypoxia and metabolic acidosis. Similarly, pregnancies complicated by conditions such as abruptio placentae, chorioamnionitis, and preeclampsia may also present lowered fetal reserve because of impaired placental gas exchange. When faced with a nonreassuring FHR tracing and no identifiable cause, the clinician may choose to seek reassurance from various other tests. Additional testing is often done because of the high false-positive rate for electronic fetal monitoring in an attempt to avoid the morbidity of an unnecessary cesarean section or operative vaginal delivery. An acceleration in FHR after vibroacoustic stimulation or fetal scalp stimulation with a digital examination provides reliable reassurance of a normal fetal pH and allows labor to continue.4,18 Blood sampling taken from the fetal scalp has also been used to assess the fetal pH directly, although this technique is used less frequently today.

LIMITATIONS OF FETAL HEART RATE MONITORING The introduction of electronic FHR monitoring in the 1970s was expected to reduce significantly the number of cases of intrapartum asphyxia leading to death or permanent neurologic damage and, in its most severe form, cerebral palsy. The rates of intrapartum fetal death have significantly declined since the 1970s.19 The rates of cerebral palsy have not decreased in developed countries in the past 30 years, however.5 There are multiple reasons for these findings. Although cerebral palsy was previously thought to result primarily from intrapartum asphyxia, it is now recognized that intrapartum events are responsible for only a small fraction of this condition, with most cases caused by antenatal events that occur

before the onset of labor. Despite continual efforts at standardization, interpretation of FHR tracing is subjective and prone to significant interobserver and intraobserver variation.7,15 This technique is fraught with false-positive results and poor positive predictive value. Using cerebral palsy as an endpoint, one study found the positive predictive value of FHR monitoring to be 0.14%.14 The rate of cesarean delivery in the United States has increased dramatically during this time frame, to nearly 32% in 2007.6 Many experts attribute this increase to the introduction of FHR monitoring and the increasingly contentious medicolegal environment. Despite the lack of evidence to support long-term benefit from its use, electronic FHR monitoring is likely to remain the most common form of intrapartum fetal assessment in the near future.

NEW TECHNOLOGIES Given the limitations of the current system, additional technologies have been developed in an attempt to identify the at-risk fetus better without unnecessarily intervening and introducing iatrogenic morbidity and mortality. In 2000, the U.S. Food and Drug Administration granted conditional approval of the OxiFirst Fetal Oxygen Saturation Monitoring System for use as an adjunct to traditional electronic FHR monitoring. This device was designed to provide continuous monitoring of the fetal oxygen saturation with a sensor placed against the fetal cheek. A multicenter randomized trial including 5341 women showed no significant difference in cesarean delivery rates or neonatal outcome when this technique was compared with traditional monitoring.3 More recently, attention has turned to the use of a fetal electrocardiogram, with specific focus on the ST segment. European studies have evaluated the use of an automatic ST waveform analyzer in conjunction with traditional electronic fetal monitoring, and indicate potential benefits, but further research is warranted.2

SUMMARY Despite the inherent stress of labor, most fetuses are able to tolerate the transient episodes of hypoxemia without harm. Rarely, the process of labor and delivery places a fetus in jeopardy of long-term neurologic damage or death as a result of profound hypoxemia and metabolic acidosis. Since the

Chapter 10  Estimation of Fetal Well-Being

1970s, electronic FHR monitoring has emerged as the most common technology to monitor fetuses during labor, in the hopes of identifying at-risk fetuses to effect delivery before permanent harm is done. This technology has been successful in largely eliminating intrapartum fetal death. It has likely resulted in a dramatic increase in unnecessary cesarean deliveries, however, without reducing the rate of cerebral palsy. It is hoped that future research and technology development will refine and improve this technology and result in benefit to mothers and infants.

REFERENCES 1. American College of Obstetricians and Gynecologists: Intrapartum fetal heart rate monitoring, ACOG practice bulletin no. 70. Washington, DC, 2005, ACOG. 2. Amer-Wahlin I et al: Cardiotocography only versus cardiotocography plus ST analysis of fetal electrocardiogram for intrapartum fetal monitoring: a Swedish randomized controlled trial, Lancet 358:534, 2001. 3. Bloom SL et al: Fetal pulse oximetry and cesarean delivery, N Engl J Med 355:2195, 2006. 4. Clark SL et al: The scalp stimulation test: a clinical alternative to fetal scalp blood sampling, Am J Obstet Gynecol 148:274, 1984. 5. Clark SL et al: Temporal and demographic trends in cerebral palsy—fact and fiction, Am J Obstet Gynecol 188:628, 2003. 6. Hamilton BE et al: Births: preliminary data for 2007, National Vital Statistics Report, Web release, vol 57, no. 12, Hyattsville, MD, 2009 National Center for Health Statistics. 7. Helfand M et al: Factors involved in the interpretation of fetal heart monitor tracings, Am J Obstet Gynecol 151:737, 1985. 8. Hofmeyr GJ. Amnioinfusion for umbilical cord compression in labour, Cochrane Database Syst Rev (1):CD000013, 1998. 9. Hon E et al: The classification of fetal heart rate, II: a revised working classification, Conn Med 31:779, 1967.

187

10. Krebs HB et al: Intrapartum fetal heart rate monitoring, I: classification and prognosis of fetal heart rate patterns, Am J Obstet Gynecol 133:762, 1979. 11. Macones GA et al: The 2008 National Institute of Child Health and Human Development workshop report on electronic fetal monitoring: update on definitions, interpretation, and research guidelines, Obstet Gynecol 112:661, 2008. 12. Martin JA et al: Births: final data for 2003, National Vital Statistics Reports, vol 54, no. 8, Hyattsville, MD, 2005, National Center for Health Statistics. 13. Modanlou H et al: Fetal and neonatal acid-base balance in normal and high-risk pregnancies: during labor and the first hour of life, Obstet Gynecol 43:347, 1974. 14. Nelson KB et al: Uncertain value of electronic fetal monitoring in predicting cerebral palsy, N Engl J Med 334:613, 1996. 15. Nielsen PV et al: Intra- and inter-observer variability in the assessment of intrapartum cardiotocograms, Acta Obstet Gynecol Scand 66:421, 1987. 16. Parer JT et al: Fetal acidemia and electronic fetal heart rate patterns: is there evidence of an association? J Matern Fetal Neonatal Med 19:289, 2006. 17. Robinson BK et al: A review of the proceedings from the 2008 NICHD Workshop on Standardized Nomenclature for Cardiotocography: updates on definitions, interpretative systems with management strategies, and research guidelines in relation to intrapartum EFM, Rev Obstet Gynecol 1:186, 2008. 18. Smith CV et al: Intrapartum assessment of fetal well-being: a comparison of fetal acoustic stimulation with acid base determinations, Am J Obstet Gynecol 155:726, 1986. 19. Walsh CA et al: Trends in intrapartum fetal death, 1979-2003, Am J Obstet Gynecol 198:47.e1, 2008. 20. Westgate JA et al: The intrapartum deceleration in center stage: a physiologic approach to the interpretation of fetal heart rate changes in labor, Am J Obstet Gynecol 197:236.e1, 2007.

CHAPTER

11

Surgical Treatment of the Fetus Timothy M. Crombleholme and Foong-Yen Lim

Prenatal diagnosis has become increasingly sophisticated, and technological advances have enhanced not only our range of diagnostic capabilities but our understanding of prenatal natural history as well. Invasive therapies have developed as a natural consequence of our expanded understanding of the natural history and pathophysiology of structural anomalies.7,70 In the 1960s and 1970s, despite rapid progress in prenatal diagnosis, few invasive therapies were considered, much less employed.1 Once a diagnosis was made prenatally, parents had only two alternatives: pregnancy termination, if the diagnosis was made before 24 weeks of gestation, or continuing to term.96 An additional option is altering the site of delivery so that the appropriate pediatric specialists would be available immediately to treat the newborn with a congenital anomaly. As the natural history of many prenatally diagnosed anomalies became better understood, early delivery was recognized as an option to avoid the continuing damage caused by the anomaly in utero.5,7 Today there are more alternatives. This chapter describes the treatment options available, how they were developed (Table 11-1), and new therapies that may be available in the future.

FETAL SHUNTING PROCEDURES A new era in invasive fetal therapy began in the early 1980s, when several independent groups introduced shunting procedures for hydrocephalus and hydronephrosis.30,66,78 These first few cases represented an extension of invasive fetal therapy from simple intrauterine blood transfusion for a medical illness to the first attempts at in utero treatment of structural anomalies (Table 11-2). During this period, hydronephrosis and hydrocephalus were being recognized more frequently with ultrasound examination. The prenatal natural history of these lesions was established by serial sonographic observation of untreated cases.* Fetuses with high-grade obstructive uropathy followed to term were often born with advanced hydronephrosis, type

IV cystic dysplasia, and pulmonary hypoplasia that were incompatible with life.23,45,167 In the case of obstructive hydrocephalus, it was known that shunting during the newborn period improved neurologic outcome, and it was reasoned that decompression in utero might avert progressive brain damage.131,137,208 However, the poor outcomes observed with shunting for hydrocephalus resulted in a moratorium on the use of shunts in the treatment of obstructive hydrocephalus.29 In other conditions, work by numerous investigators using appropriate animal models helped define the pathophysiology of these lesions and establish the theoretical basis for intervention.6,74-76,86,87

Hydronephrosis The first case of a fetus with obstructive uropathy treated in utero by vesicoamniotic shunting was reported by Golbus and colleagues in 1982.78 Advances soon followed in diagnosis, technique, shunt design, and patient selection.† The enthusiasm for treating fetal obstructive uropathy has continued unabated during the past two decades. The procedure became widely implemented before stringent selection criteria for treatment were developed and the therapeutic efficacy of the procedure was established. The widespread use of vesicoamniotic shunts also had the effects of shifting cases away from the centers where these questions were being studied and limiting attempts to better define the role of vesicoamniotic shunting in the management of fetal obstructive uropathy. The lack of a prospective, randomized trial makes the efficacy of prenatal decompression the most difficult question to address in the treatment of fetal obstructive uropathy. One of the few series that attempted to address this question, albeit in a retrospective analysis, was reported by Crombleholme

*References 23, 25, 31, 45, 74, 76, 157. † References 36, 38, 44, 100, 118, 175.

189

190

SECTION II  THE FETUS

TABLE 11–1  Treatable Fetal Malformations Fetal Malformation

Fetal Presentation

Fetal/Neonatal Consequences

Fetal Treatment Options

Acardiac/acephalic twin (TRAP sequence)

Polyhydramnios

IUFD, hydrops, cardiac failure Multifocal leukoencephalo­ malacia

Fetoscopic cord ligation Fetoscopic cord coagulation

Amniotic band syndrome

Edematous limb from constricting band Umbilical cord constriction

Limb amputation

Fetoscopic laser lysis of bands

Fetal death

—

Rapidly progressive isolated hydrocephalus

Neurologic damage

Ventriculoamniotic shunt Ventriculoperitoneal shunt

Complete heart block Hydrops, slow heart rate

IUFD, neonatal death

Fetal epicardial pacemaker

Congenital diaphrag­ Herniated viscera in chest matic hernia (CDH)

IUFD, pulmonary hypoplasia and respiratory insufficiency, neonatal death

Open fetal tracheal clip Fetoscopic tracheal clip Fetoscopic tracheal balloon

Cystic adenomatoid malformation of the lung (CCAM)

Chest mass, mediastinal shift Hydrops

Pulmonary hypoplasia, IUFD, neonatal death

Thoracoamniotic shunt (if there is dominant cyst) Fetal resection of CCAM (if it is solid tumor)

Fetal hydrothorax

Mediastinal shift, hydrops, polyhydramnios

Pulmonary hypoplasia

Thoracoamniotic shunt

Myelomeningocele (MMC)

Neural tube defect, “lemon” sign, “banana” sign

Paralysis, hindbrain herniation Hydrocephalus

Open fetal MMC repair

Neck masses (cervical Polyhydramnios, neck mass, teratoma, absent stomach bubble lymphangioma)

Inability to ventilate due to lack of airway Anoxic brain injury, neonatal death

EXIT procedure

Ovarian cyst

Wandering cystic abdominal mass

Ovarian torsion, polyhydramnios

Cyst aspiration

Posterior urethral valves

Hydronephrosis and oligohydramnios

Renal dysplasia and renal insufficiency Pulmonary hypoplasia and respiratory insufficiency

Vesicoamniotic shunt Cystoscopic laser ablation of valves Open vesicostomy

Sacrococcygeal teratoma

High output failure Hydrops

IUFD, prematurity, tumor rupture, hydrops, hemorrhage

Open resection Radiofrequency coagulation

IUFD, heart failure Multifocal leukoencephalo­ malacia

Serial amnioreduction Microseptostomy Cord coagulation Selective laser photocoagulation

Aqueductal stenosis

Twin-twin transfusion Oligohydramnios and syndrome polyhydramnios, growth discordance

EXIT, ex utero intrapartum treatment; IUFD, intrauterine fetal death; TRAP, twin reversed arterial perfusion

and coworkers.38 In fetuses predicted to have either good or poor prognoses by fetal urine electrolyte and ultrasound criteria, survival was greater among those who underwent decompression in utero, as opposed to those who did not undergo decompression. In the group of fetuses predicted to have a poor prognosis by selection criteria, 10 were treated. Three of these pregnancies were electively terminated, four neonates died of pulmonary hypoplasia or renal dysplasia, and three survived. All three survivors had restoration of normal amniotic fluid (AF) levels and no pulmonary complications, but in two of the three, renal failure subsequently developed and the patients underwent renal transplantation. Among the 14 patients with no intervention, there were no survivors (11 terminations and

three neonatal deaths from pulmonary hypoplasia). In the entire series, uncorrected oligohydramnios was associated with a 100% neonatal mortality rate. Normal or restored amniotic fluid volume was associated with a 94% survival rate.38 Although in utero decompression seems to prevent neonatal death from pulmonary hypoplasia, the effect on renal function is less clear. The severity of renal dysplasia at birth depends on the timing and severity of obstruction before birth.* Experimental work suggests that relief of obstruction

*References 38, 45, 91, 93, 95-97, 99, 143.

Chapter 11  Surgical Treatment of the Fetus

191

TABLE 11–2  Fetal Malformations Treatable by Needle Aspiration and Shunting Procedures Fetal Malformation

Fetal Presentation

Fetal/Neonatal Consequences

Posterior urethral valves

Hydronephrosis Oligohydramnios

Renal dysplasia and renal insufficiency Pulmonary hypoplasia and respiratory insufficiency

Cystic adenomatoid malformation of the lung

Mediastinal shift Hydrops

Pulmonary hypoplasia

Aqueductal stenosis

Hydrocephalus

Neurologic damage

Fetal hydrothorax

Mediastinal shift, hydrops, polyhydramnios

Pulmonary hypoplasia

Ovarian cyst

Wandering cystic abdominal mass, polyhydramnios

Ovarian torsion

during the most active phase of nephrogenesis (20 to 30 weeks of gestation) may obviate further damage and allow normal nephrogenesis to proceed.* The development of postnatal renal failure in two infants who were not treated because amniotic fluid volume remained normal raises the question of whether to treat fetuses with obstruction before oligohydramnios develops. Because renal development or maldevelopment is complete at birth, relief of obstruction in infancy or childhood may not prevent the progression to end-stage renal failure.199 Müller and associates reported on a group of fetuses with obstructive uropathy, a favorable prognostic profile, and normal amniotic fluid, in whom renal insufficiency developed by 1 year of age.155 The only feature that distinguished this group of fetuses was a urinary level of b2-microglobulin greater than 2 mg/L. Elevated fetal urinary levels of b2-microglobulin may identify fetuses at increased risk for ongoing renal damage from obstruction, even when amniotic fluid is normal. Other investigators have not found b2-microglobulin levels to be as useful, and levels greater than 10 mg/L can be a predictor of poor outcome in fetuses older than 20 weeks’ gestation.65 It remains an open question whether in utero decompression could prevent long-term renal insufficiency in these patients. The maternal morbidity associated with vesicoamniotic shunting has been reported to be minimal, but chorioamnionitis related to the procedure has been reported.38,77 In addition, there have been reports of shunt-induced abdominal wall defects with herniation of bowel through trochar stab wounds and maternal ascites from leakage of amniotic fluid through the uterine wall into the maternal peritoneal cavity.143,176,178 The utility of vesicoamniotic shunts is limited by the brief duration of decompression, the risk of infection, the risk of catheter obstruction or dislodgment, fetal injury during placement, and potentially inadequate decompression of the fetal urinary tract.36,38,61,77 These factors make vesicoamniotic shunts less effective in long-term decompression of the urinary tract early in gestation and are the impetus for development of open fetal surgical and fetoscopic techniques to treat obstructive uropathy in utero.36,40,61

*References 6, 11, 74, 76, 80, 91, 93, 163, 183.

Fetal Hydrothorax Thoracentesis is a diagnostic maneuver done to obtain pleural fluid for differential cell count and culture and to establish whether the effusion is chylous (Fig. 11-1). But even repeated thoracenteses provide inadequate decompression of the fetal chest. There have been several reports of thoracentesis for fetal hydrothorax that have resulted in either complete reso­ lution or a good outcome despite reaccumulation.12,127,164 However, other results of repeated thoracentesis for fetal hydrothorax have been disappointing because of rapid reaccumulation of the effusion and neonatal death resulting from respiratory insufficiency.135,159 Spontaneous resolution of fetal

FHT

Lung

Heart

Figure 11–1.  Fetal sonogram demonstrating larger tension fetal hydrothorax (FHT) compressing the adjacent lung and causing shift of the heart into the contralateral hemithorax. (From Shaaban AF et al: The role of ultrasonography in fetal surgery and invasive fetal procedures, Semin Roentgen 34:62, 1999.)

192

SECTION II  THE FETUS

hydrothorax may occur in as many as 10% of cases, and resolution following thoracentesis may or may not be related to the procedure. Thoracentesis alone cannot adequately decompress the fetal chest to allow pulmonary expansion and prevent pulmonary hypoplasia.129 Thoracoamniotic shunting for fetal hydrothorax, first reported by Rodeck and colleagues in 1988, provides continuous decompression of the fetal chest, allowing lung expansion.175 If instituted early enough, this allows compensatory lung growth and prevents neonatal death resulting from pulmonary hypoplasia. Nicolaides and Colleagues reported on 48 cases of thoracoamniotic shunting, but there was no attempt to distinguish isolated primary fetal hydrothorax from secondary fetal hydrothorax.159 Despite intervention, mortality was high. Four deaths were due to termination of pregnancy when a chromosomal abnormality was diagnosed. In addition, there were 12 neonatal deaths despite thoracoamniotic shunt placement, but these fetuses seemed to have severe hydrops and secondary fetal hydrothorax. Two fetuses that died in utero also seemed to have had secondary fetal hydrothorax and severe hydrops. If the cases that appear to be secondary fetal hydrothorax are eliminated, the survival of isolated primary fetal hydrothorax treated with thoracoamniotic shunting is 38 of 41 (92%) cases.153,159 A similar survival rate with thoracoamniotic shunting was found by

Hagay and coworkers in their review of fetal pleural effusions.83 This is a striking improvement when compared with a survival of only 50% in untreated fetal hydrothorax. The indications for thoracoamniotic shunting are not well defined. Most authors consider the presence of fetal hydrothorax-induced hydrops or polyhydramnios as indications for shunting (Fig. 11-2).135,159,175 We recommend thoracoamniotic shunting for primary fetal hydrothorax with evidence of effusion under tension even in the absence of hydrops.153 Because spontaneous resolution has been observed even in severe cases of fetal hydrothorax, we reserve thoracoamniotic shunting for cases in which tension hydrothorax recurs after two thoracenteses. The two currently available catheters are the Harrison double-pigtail catheter (VPI Company, Spenser, IN) and the KCH catheter (Rocket USA Inc., Branford, CN) (Fig. 11-3). The limited experience with thoracoamniotic shunting suggests that it is extremely effective in decompressing effusion and improving survival. The risks to mother and fetus of thoracentesis and shunt placement have been minimal and are far outweighed by the potential benefits. Few complications have been reported for either fetal thoracentesis or thoracoamniotic shunts. Procedure-related fetal death is rare, due to hemorrhage from an intercostal artery laceration or torsion of the umbilical cord.135 Migration of the shunt under

Fetus with Pleural Effusions

Secondary FHT

Diagnostic evaluation •Detailed sonography •Thoracentesis with cellular and genetic analysis •Fetal echocardiogram

Primary FHT

Repeat thoracentesis Reaccumulation

Large •Mediastinal shift •Hydrops

Small without progression

Resolution without recurrence

Thoracoamniotic shunt

Counsel

Follow with serial ultrasound

Other anomalies

3 - 4 Times/Interval)

1

Nasal Flaring

2

Respiratory Rate - 60/min

1

Respiratory Rate - 60/min with Retractions

2

Excessive Sucking

1

Poor Feeding

2

Regurgitation

2

Projectile Vomiting

3

Loose Stools

2

Watery Stools

3

Comments PM

TOTAL SCORE Initials of Scorer

Figure 38–13.  Neonatal abstinence score used for the assessment of infants undergoing neonatal abstinence. Evaluator should

check sign or symptom observed at various time intervals. Add scores for total at each evaluation. (Adapted from Finnegan LP, Kaltenbach K: The assessment and management of neonatal abstinence syndrome. In Hoekelman RA, et al, editors: Primary pediatric care, 3rd ed, St. Louis, 1992, Mosby, p 1367.)

monitored cases, breast feeding by the methadone-maintained mother may be cautiously permitted to enhance maternalinfant bonding.

obstetric attention appear to improve fetal growth, resulting in higher birthweights compared with heroin-exposed infants.

Fetal Growth Effects

Central Nervous System Effects

There is general consensus that neonates born to heroinaddicted mothers have lower mean birthweight, length, and head circumference than unexposed neonates. This detrimental effect on fetal growth has also been shown in experimental animal models of prenatal heroin exposure. Women in methadone maintenance also have babies with lower birthweights than matched unexposed controls, but their babies have significantly higher birthweights than babies whose mothers use heroin and were not maintained on methadone.88 Methadone treatment and improved

Gestational exposure to opioids decreases nucleic acid synthesis and protein production in fetal brain and impairs growth of neurons.106 Rat pups exposed to morphine in utero have changes in the packing density and morphology of neurons and neuronal processes that are significantly smaller than seen with controls. Prenatal opioid exposure has been shown to decrease exploratory behavior and to increase response latency to noxious stimuli in young animals. Furthermore, gestational exposure to morphine also increases self-administration of both heroin and cocaine in

748

SECTION V  PROVISIONS FOR NEONATAL CARE

adult animals. In preclinical studies in mice, Slotkin and colleagues observed that prenatal heroin exposure evokes neurochemical and behavioral deficits associated with specific interference with hippocampal cholinergic function.143 In addition, heroin alters elements of cell signaling cascades shared by cholinergic inputs and other neurotransmitter systems. Furthermore, significant deficits in norepinephrine and dopamine turnover that were not seen in the immediate postnatal period emerged later in young adulthood, when hippocampus and cerebral cortex showed almost complete inactivation of noradrenergic and dopaminergic tonic activity. Because methadone, in contrast to heroin, does not produce enduring alterations in norepinephrine or dopamine turnover, the authors speculate that this may have favorable implications for infants born to mothers taking methadone as an opioid substitution for heroin.

Neurodevelopmental Effects Clinical studies of infant neurodevelopmental outcome following prenatal heroin and methadone exposure are difficult to interpret because of the exposure to both agents and the difficulty with ascertaining adequate control groups and covariate control of other factors that may contribute to outcome, such as genetic factors, exposure to other teratogens, pregnancy and parturition complications, low socioeconomic status, and issues affecting the family and caregiving environment (e.g., drug and alcohol abuse by the mother or other adult household members, domestic violence, maternal HIV/AIDS). Mothers who enter methadone maintenance programs do so because they are addicted to heroin and affected by its multiple accompanying risk factors. Furthermore, many mothers continue to supplement methadone with street heroin as well as alcohol, cigarettes, cocaine, and other drugs. Also complicating the picture is the diversity among practitioners regarding methadone dosing practices. Because heroin abuse in pregnancy is a rather low-incidence disease, the available single-site studies of heroin addiction and methadone treatment in pregnancy have tended to be relatively small, limiting statistical power and generalizability. Nevertheless, follow-up reports shed light on the potential impact of opioid exposure on human infants and children. In a detailed review of infant and toddler outcomes, several controlled studies have shown various neurobehavioral deficits in infants exposed to methadone in utero, compared with various control groups, using the BNBAS and neurologic assessments.58 Findings included depressed interactive behaviors; decreased visual and auditory orientation and motor maturity; more jerkiness, tremulousness, hypertonicity, increased arousal, and irritability, and poorer quality of cuddling; and significant mean duration of crying. In a study of videotaped feeding and play interactions, poorer interaction ratings were observed in the methadone-exposed mothers and infants.84 In unstructured free play at 18 months, methadone-exposed children were noted to have less age-appropriate play—specifically, less time was spent in symbolic pretend play.114 Follow-up studies of methadone-exposed infants using the Bayley Scales of Infant Development (BSID) and neurologic examinations have shown motor tone difficulties and transiently lower mean Mental Developmental Index (MDI)

and Psychomotor Development Index (PDI) scores. Prenatal opioid exposure was also reported to affect the special sensory organs, leading to an observed increase in otitis media, strabismus, and nystagmus. In a large matched cohort study, there were no significant associations, after covariate control, between in utero opiate exposure and mental, psychomotor, or behavioral functioning.113 According to Bernstein and Hans, cumulative risk factors predict poorer outcome better than the prenatal exposure to methadone alone.25 In an investigation comparing children who were born to heroin-dependent mothers and adopted at a very young age, with those born to heroin-dependent mothers (or fathers) but raised at home, the effect of home environment was determined to be more important than in utero exposure.121

COCAINE Prevalence Few perinatal substances of abuse have engendered as much controversy as cocaine and crack cocaine. Well established in adults is cocaine’s propensity for causing severe CNS and cardiovascular toxicity, including convulsions, myocardial infarction, and death. Cocaine (C17H21NO4, benzoylmethyl­ ecgonine), an alkaloid derived from Erythroxylon coca shrub leaves, is classified as a stimulant. Cocaine hydrochloride, typically snorted as a powder, can also be ingested orally or administered intravenously. Smoking of crack cocaine, an inexpensive alkaloid free-base form of nearly pure cocaine, became widespread in the 1980s, reaching epidemic proportions in many of the nation’s inner cities and affecting large numbers of women of childbearing age. Hospitals providing obstetric and newborn services inevitably felt the impact of rising numbers of cocaine-exposed infants. Initially confined to the high-intensity, drug-trafficking areas, this epidemic subsequently spread to remote rural areas. Media attention became focused dramatically on the fate of so-called crack babies. The scientific community, recognizing the challenges of evaluating the effects of in utero cocaine exposure on the developing infant and child, responded with ongoing well-designed prospective longitudinal investigations, which have yielded numerous reports of outcomes within multiple domains of development from birth through early adolescence. The 1992 NIDA National Pregnancy and Health Survey of drug use by pregnant women showed that 1.1% (about 45,100 U.S. births) had been exposed to cocaine during pregnancy. The 2007 NSDUH showed that 0.4% of pregnant women reported some form of past-month cocaine use, in contrast to 1% of nonpregnant women. Regional differences exist, and cocaine use during pregnancy within some hospitals is substantially higher, as shown in investigations using systematic prenatal or postpartum interviews and bioassays.

Pharmacologic Effects of Cocaine The main effects of cocaine involve the norepinephrine, dopamine, and serotonin neurotransmitter systems. Inhibition of norepinephrine reuptake leads to accumulation on the synaptic cleft and stimulation of postsynaptic norepinephrine receptors, with manifestations of tachycardia, arrhythmias, hypertension, vasoconstriction, diaphoresis,

Chapter 38  Pharmacology

and mild tremors. Blocking of dopamine reuptake results in dopamine accumulation in the synaptic cleft and stimulation of dopamine neurotransmission, causing neurochemical magnification of the pleasure response, increased alertness, enhanced sense of well-being and self-esteem, and heightened energy and sexual excitement. Compulsive abusers experience anxiety, depression, and exhaustion, and chronic users may exhibit mood disorders, paranoid psychosis, sexual dysfunction, and addiction. Acute tolerance, rebound depression, crash, and craving are explained by regulatory changes in presynaptic and postsynaptic dopaminergic receptors after chronic cocaine use. Cocaine diminishes the need for sleep by decreasing serotonin biosynthesis as a result of decreased uptake of tryptophan. Cocaine readily traverses the placenta and is also excreted in breast milk. Metabolism of cocaine is primarily through plasma and hepatic cholinesterases to inactive compounds, ecgonine methyl ester and benzoylecgonine, excreted renally. Demethylation of cocaine yields norcocaine, an active metabolite with a greater capacity to inhibit norepinephrine uptake. In pregnant women, fetuses, and infants, plasma cholinesterase activity is diminished, and excretion may be prolonged. Individual susceptibility to cocaine’s effects may be related to genetic polymorphism for the cholinesterase enzyme, and the term placenta may also provide a degree of protection for individual fetuses by converting cocaine into less active metabolites, presumably by cholinesterase activity. Cocaine use during pregnancy alters fetal oxygenation by reducing uterine and placental blood flow and impairing fetal oxygen transfer. Fetal tachycardia and hypertension reflect fetal hypoxemia, or increased fetal levels of cocaine or fetal catecholamines, or a combination. In animal studies, significant increases in maternal and fetal arterial pressure and decreases in uterine blood flow have been reported. In humans, maternal and fetal deaths have been linked to cocaine use, and pregnant cocaine-abusing women are at increased risk for spontaneous abortions, abruptio placentae, and premature rupture of the membranes. Transient neonatal ventricular tachycardia has been associated with maternal cocaine abuse shortly before delivery. Higher arterial blood pressure and diminished cardiac output and stroke volume, presumably affected by increased norepinephrine levels, have also been reported in full-term cocaine-exposed infants compared with drug-free infants on the first postnatal day. There is no definitive dysmorphology syndrome (such as FAS) associated with prenatal cocaine exposure. In animal studies of gestational cocaine exposure, a wide variety of congenital anomalies, including exencephaly, anophthalmia, malformed or missing lenses, cryptorchidism, hydronephrosis, grossly distended bladder, limb reduction defects, ileal atresia, and cardiovascular anomalies, have been described. It has been hypothesized that cocaine-induced blockage of norepinephrine uptake is responsible for placental vasoconstriction and fetal hypoxia, which produce the defects. Cocaine has also been determined to be teratogenic in a dosedependent manner during late organogenesis or during the postorganogenic period in the Sprague-Dawley rat. The observed pattern of defects, principally reduction deformities of the limbs and tail and genital tubercle defects, occurred as a result of hemorrhagic necrosis, disruption, and amputation of existing and developing structures.161 An important

749

concept is that vulnerability to cocaine-induced structural defects may not be limited to the period of organogenesis but instead may be related to aberrations in fetal, placental, and uterine blood flow leading to fetal vascular disruption—that is, a structural defect may result from destruction of a previously formed, normal part. The structural defects described in the animal studies mirror those from human case reports and small series, especially observations from referral services such as neurology, cardiology, and nephrology-urology. However, most infants exposed prenatally to cocaine are born without defects, and systematic investigations have failed to show a discernible pattern or an increased prevalence of congenital malformations associated with gestational cocaine exposure in humans, with the possible exception of genitourinary tract defects.20 Cocaine-induced ischemia may play a role in the development of intestinal atresia, infarction, and necrotizing enterocolitis, which has been observed in term and preterm infants prenatally exposed to cocaine.

HIV/AIDS and Sexually Transmitted Diseases Cocaine-using women are more prone to a variety of sexually transmitted diseases both before and during pregnancy. Intravenous substance abuse by the woman or her sexual partner is a recognized HIV risk factor, but sexual promiscuity and trading of sex for cocaine or crack also increases the risk for heterosexual transmission of HIV infection and other sexually transmitted diseases in cocaine-abusing women of childbearing age. These infections occurring in pregnancy pose additional risks to the fetus.

Prematurity Although an increased rate of premature delivery or a shorter gestation has been reported in cocaine-using women, this finding is not consistent across studies. Prenatal cocaine exposure has been associated with shorter gestational age even in women receiving prenatal care.16,128 In larger samples, observed differences in gestational age between cocaine-exposed and unexposed infants, although perhaps statistically significant, have been modest (i.e., about 1 week or less).12,16

Intrauterine and Postnatal Growth The most frequently reported adverse consequence of prenatal cocaine exposure is intrauterine growth restriction. Studies in Long-Evans rats have shown that prenatal cocaine exposure results in dose-dependent decreases in birthweight and postnatal growth and alterations in fetal body composition, with lower levels of body fat, protein, and calcium.35 In humans, restricted intrauterine growth may be in part the result of malnutrition and poor weight gain in the mother because of cocaine’s suppression of appetite, but in utero cocaine exposure appears to result in depressed neonatal fat stores and diminished lean body mass, even after controlling for maternal weight gain.67 Structural equation modeling revealed a direct impact of prenatal cocaine exposure in term infants on fetal growth involving birthweight, length, and head circumference in the large single-site Miami Prenatal

750

SECTION V  PROVISIONS FOR NEONATAL CARE

Cocaine Study cohort of cocaine-exposed compared with unexposed term infants, after control of potential confounders.12 From the Maternal Lifestyles Study, which includes cohorts from a multi-site study of term and preterm infants, there is evidence that the impact of prenatal cocaine exposure on fetal growth may be more pronounced later in gestation, with a cocaine-associated deceleration in growth occurring after 32 weeks’ gestation.9 Other prospective clinical investigations of in utero cocaine exposure have reported somewhat inconsistent results, with some confirming cocaineassociated decreases in one or more anthropometric measures and others reporting no differences, especially after adjustment for selected covariates. Furthermore, higher cocaine levels in maternal hair have been correlated with lower head circumference, suggesting a significant dose-response effect.15 In a unique rural longitudinal prospective cohort on prenatally cocaine-exposed and unexposed infants, Behnke and colleagues highlighted the potential significance of this finding in their report of indirect effects of prenatal cocaineassociated birth head circumference decrement on preschool development.21 Data on long-term growth of prenatally cocaine-exposed children are sparse, and cross-sectional and longitudinal findings also vary from cohort to cohort and by specifically affected growth parameters. One investigation reported that cocaine-exposed children, particularly those born to older mothers, were shorter and more likely to have clinically significant height deficits at age 7 years.43 In a longitudinal growth curve analysis of growth from 1 to 10 years, prenatal cocaine exposure during the first trimester was related to a slower rate of growth in both weight and head circumference over time.126 In contrast, another investigation showed that the cocaine-exposed children experienced catch-up growth, mainly in the first 6 to 12 months, that resulted in no apparent deficits in height, weight, or head circumference by age 8 years.104 As has been suggested with prenatal tobacco exposure, early infancy catch-up growth in response to prenatal cocaine exposure may prove to be a risk factor for obesity in later childhood and adolescence.

Sudden Infant Death Syndrome Like tobacco and opioids, cocaine has been previously linked to SIDS. However, a meta-analysis concluded that an increased risk for SIDS could not be attributed to intrauterine cocaine exposure alone, after controlling for concurrent use of other drugs.60

Central Nervous System Effects Cocaine adversely affects CNS development by binding with monoaminergic receptors in the fetal brain. Animal studies have demonstrated altered neocortical development, including inappropriate placement of neurons, deficient lamination, and decreased volume, density, and total number of neocortical neurons, after prenatal administration of cocaine, and some of these abnormalities persist beyond the newborn period. Additionally, prenatal cocaine exposure has been shown to produce functional changes in the dopaminergic and serotonergic systems in offspring. Moreover, observed structural and functional alterations have been associated

with neurobehavioral abnormalities in animals (e.g., attention, learning) that closely parallel findings in clinical studies.101,109 Investigations in neonatal rats exposed prenatally to cocaine have shown significant neurobehavioral deficits, including hyperactivity, abnormal sexual behavior (demasculinization) in adulthood, and significantly increased metabolic activity of the limbic, motor, and sensory system of female, but not male, rat brains, as reviewed by Hutchings and Dow-Edwards.81 Cocaine-induced alterations in cerebral blood flow have been implicated in CNS hemorrhagic and ischemic lesions. Concern, raised initially by a single case report of a cocaineexposed term infant with neonatal cerebral infarction, was heightened by a larger series of full-term infants exposed to cocaine, amphetamines, or both, who were reported to have echoencephalographic evidence of white-matter cavities, acute infarction, intraventricular hemorrhage, subarachnoid hemorrhage, and ventricular enlargement.51 However, a subsequent well-controlled prospective study of mainly full-term infants did not substantiate an increased risk for CNS structural lesions associated with prenatal cocaine exposure.19 In another well-designed study of full-term and near-term infants, no group differences were noted between cocaine and comparison groups, but subependymal hemorrhages were found more frequently among more heavily cocaineexposed compared with unexposed infants.68 With regard to prematurity, known to be associated with intraventricular hemorrhage and periventricular leukomalacia, a relatively large study of premature infants with very low birthweight showed no group differences in the rate of these complications between cocaine-exposed and unexposed infants.53 In human infants with in utero cocaine exposure, neonatal electroencephalographic abnormalities, clinical seizures that may extend beyond the perinatal period, and aberrations in auditory brainstem response have also been noted.

Infant and Toddler Neurobehavior and Mental and Psychomotor Development Infant neurobehavioral functioning, measured by the Brazelton Neonatal Behavioral Assessment Scale (BNBAS), the NICU Network Neurobehavioral Scale, and other measures, has been a primary focus of nearly all of the prospective longitudinal cohort studies of in utero cocaine exposure. On the BNBAS, poorer state regulation, impaired autonomic control, irritability, poorer orientation, decreased habituation, diminished alertness, and abnormal reflexes, tone, and motor maturity have been reported among cocaine-exposed infants, although there has been little consistency across studies in the specific domains affected. Meta-analysis indicated reliable but modest cocaine-associated effect sizes for the motor and abnormal reflex clusters of the BNBAS.78 In a large cohort of full-term infants, prenatal cocaine exposure was modestly related to all BNBAS domains except abnormal reflexes, even with statistical control for other drug exposures and key infant and maternal characteristics.116 Several controlled studies with large sample sizes detected no cocaine-associated deficits on the BNBAS at birth, but two of these studies reported abnormalities in areas such as state regulation, autonomic regulation, and reflexes at 2 to 4 weeks postpartum.38,158 Prenatal cocaine exposure was related to lower

Chapter 38  Pharmacology

arousal, lower regulation, and high excitability on the NICU Network Neurobehavioral Scale at 1 month of age after adjustment for pertinent covariates.98 Prenatal cocaine exposure also has been associated with an increased number of cry utterances and short cries, suggestive of an overaroused (excitable) pattern of neurobehavioral functioning.42,98 Other neurologic and physiologic measures have also revealed neurobehavioral abnormalities such as increased jitteriness, irritability, and poorer auditory response and habituation among cocaine-exposed infants. Moreover, cocaine-associated alterations in physiologic regulation during infancy using measures of heart rate and respiratory sinus arrhythmia have been reported.134 A dose-response effect has been documented, with poorer neurobehavioral functioning found in infants exposed to higher or more frequent doses of cocaine prenatally.59,134,158 Although these early neurobehavioral abnormalities do not appear to be related to cocaine intoxication or withdrawal symptoms, it is not yet known to what degree they will persist beyond infancy. Recent studies suggest that these early neurobehavioral impairments may endure because cocaineexposed children have been shown to display greater reactivity and poorer regulation during stressful conditions later in infancy and in the preschool years.50,54 Numerous published studies have examined the effects of prenatal cocaine exposure on early cognitive and motor development. Prenatal cocaine exposure has been associated with lower MDI scores on the BSID during the infant and toddler years in some studies. Compared with unexposed children, children with gestational cocaine exposure were more likely to score in the delayed range and to show a more significant decline in MDI from 6 to 24 months of age.139 In another report, prenatal cocaine exposure was associated with lower overall MDI assessed at several time points from 3 to 36 months of age. The observed relationship was mediated by birthweight and gestational age, and the cocaineexposed children showed a decline in cognitive performance through age 24 months but did not differ from unexposed controls in developmental trajectories.110 Other studies have not found a relationship between prenatal cocaine exposure and early cognitive development after covariate adjustment.30,113,127 Most studies report no specific cocaine effect on BSID motor performance.113,139 In contrast, in longitudinal analyses of motor development, cocaine-exposed compared with unexposed infants demonstrated early deficits in motor development with recovery to normal functioning by 18 months.115 Several reports highlight the importance of the postnatal caregiving environment in influencing developmental outcome in cocaine-exposed children.30,113

Child Cognition, Neurodevelopment, and Behavior Most published reports on IQ in prenatally cocaine-exposed preschool- and school-aged children document no significant global IQ difference in cocaine-exposed compared with unexposed groups from similar backgrounds. In studies demonstrating cocaine-associated deficits in cognitive functioning, effects are generally subtle, are often mediated or moderated by other variables, and tend to be specific to more specialized domains of function such as perceptual reasoning, visual-motor skills, and memory.6,22,140 Large, well-controlled

751

investigations have documented a cocaine-specific effect on language-processing abilities through early school years after consideration of other prenatal exposures and other covariates and when considering a dose-response relationship. Studies specifically addressing longitudinal analyses of language found significant prenatal cocaine-associated deficits throughout the period of study during early childhood but no group differences in trajectory of language development over time.14,100 A recent report from a large cohort study, however, found no main prenatal cocaine exposure effects on standardized language measures at 6 and 9.5 years, although interaction effects included lower receptive scores at age 6 in prenatally cocaineexposed children, and moderating effects for birthweight and gender.18 Extant data, although limited, suggest that prenatally cocaine-exposed children do not differ from unexposed peers on standardized tests of achievement or indicators of academic performance, at least through the early school years.80,140 However, prenatal cocaine exposure has been linked to increased risk for learning disability at age 7 years,117 and receipt of special education services.99 Preclinical studies have associated the effects of cocaine exposure with structural and functional alterations in the monoamine system, involving norepinephrine and serotonin pathways throughout the brain and more targeted dopaminergic circuits.109 Dopamine depletion has been linked to long-term cocaine exposure in studies of laboratory animals and adult humans. An extensive review of the animal literature notes a consistent pattern of cocaine exposure’s direct effects on dopaminergic pathways central to arousal regulation and attentional reactivity in prenatal and preweaning rats.109 Emerging imaging data with school-aged children have also begun to reveal possible cocaine-associated structural and functional variations in dopamine-rich areas of the brain. Prenatal cocaine exposure has been linked to less mature development of frontal white-matter pathways,160 reduced caudate volume,8 and increased frontostriatal activation during a response inhibition task.138 Clinically, cocaine-exposed children have demonstrated subtle deficits on tests of executive functioning, a largely frontal lobe–mediated process that involves the ability to plan and accomplish future goals or actions and the ability to inhibit or control motor responses that interfere with goal attainment.108,130,160 Additionally, preschool- and school-aged cocaine-exposed children, when compared with unexposed controls, exhibit impaired attentional processing as measured by omission errors on computerized continuous performance tasks,13 and evidence greater restlessness, distractibility, and noncompliance as rated by observers in a clinic setting. Some studies indicate increased levels of behavior problems as rated by parents and teachers. As cocaine-exposed children enter later childhood and early adolescence, of particular interest is whether in utero cocaine exposure will influence the child’s own drug involvement. Although clinical data are not yet available, increased selfadministration of cocaine has been demonstrated in adult rats after in utero cocaine exposure.129 In addition to neurochemical and vascular mechanisms of prenatal cocaine exposure pathophysiology on fetal development, Lester and Padbury propose a third mechanism in which cocaine may also act as an intrauterine stressor, that is, a challenge that alters the internal milieu of the fetus and disrupts genetic programming of fetal-placental development, consistent with the concept of

752

SECTION V  PROVISIONS FOR NEONATAL CARE

fetal origin of adult disease.97 They hypothesize that alterations in hypothalamic-pituitary axis stress responses may underlie long-term behavioral and emotional problems, including compulsive disorders such as drug addiction in prenatally cocaineexposed offspring.97 Early fears of devastating effects of cocaine in vast numbers of affected infants have not been substantiated by systematic prospective studies or meta-analyses. However, reports in individual cases that show the potential for cocaine to be associated with severe anomalies or vascular disruptions may indicate low-incidence cocaine-associated lesions expressed in certain hosts, depending on dosage or pattern of exposure or individual susceptibility factors, such as genetics, metabolism, maternal nutrition, and maternal health conditions. Large prospective studies have now begun to show converging evidence for significant, albeit relatively subtle, cocaine-associated deficits in a number of domains of neurobehavioral and neuropsychological functioning, such as sustained attention, language functioning, and behavior. In many of the studies, the observed deficits appear to be statistically robust indicators of cocaine exposure because they persist even after controlling for numerous environmental confounding variables. Furthermore, in some studies, the effects have been shown to be dose dependent, which lends credence to a teratologic effect. The importance of cocaineassociated subtle deficits should not be overlooked because they may be costly in terms of increased numbers of children qualifying for special services in the school system.96 Significant deficits in executive function may not become apparent until the child reaches late adolescence or young adulthood as complex frontal lobe neurons complete their developmental trajectory of maturation into adulthood. Thus, “sleeper effects” of prenatal cocaine exposure may emerge later to affect real-life functionality as these children face increasing demands related to academics, social situations, and acceptance of responsibility for the well-being of themselves and others. More attention needs to be paid to gender modification of cocaine effects by addressing the potential for sex-related effects on specific outcomes, even when there are no apparent group differences in the overall sample. Although malefemale differences have been reported,22,49 sex and gender variations in the impact of prenatal cocaine exposure are not well understood. In clinical populations, the mothers seldom use cocaine as a single agent, and it has become increasingly evident that in studies of the effects of perinatal substance abuse, analyses should include assessment of the effects of drug-drug interactions on the outcomes of interest. Consideration should also be given to the myriad caregiving environmental factors that may influence these results. Addressing these analytic challenges are important in understanding the risk and resiliency factors and cumulative risks that predict individual and group outcomes. A growing number of studies show the contribution of environmental risk (e.g., poverty, caregiver instability, domestic and community violence exposure) in influencing a variety of outcomes in cocaine-exposed children. Finally, it should be noted that cocaine exposure in infants is not limited to the in utero environment. Cocaine exposure through breast milk has been reported to cause intoxication

in the newborn. Postnatal ingestion and passive inhalation of cocaine and crack cocaine have also been reported to cause toxicity in young infants and toddlers, and, of course, active and passive exposure could affect older children and adolescents in high-risk environments.

AMPHETAMINES In the western part of the hemisphere, abuse of amphetamines and methamphetamines is especially predominant, and many users of amphetamines and methamphetamines are women of childbearing age. Although published data are limited, the obstetric and neonatal complications associated with prenatal exposure to these substances appear to be similar to those observed with cocaine. Amphetamine (methylphenylethylamine) is a stimulant of norepinephrine, dopamine, and serotonin release, and the clinical effects have a longer duration of action than cocaine. Methamphetamine (also known as meth, speed, ice, and crystal), is the N-methyl homologue of amphetamine. With better blood-brain barrier penetration than amphetamine, methamphetamine has significantly higher CNS stimulant effect and less peripheral nervous system and cardiovascular stimulation. Gestational methamphetamine exposure has been reported to increase the incidence of miscarriage, prematurity, intrauterine growth restriction, and placental hemorrhage. Infant feeding and sleep patterns are disturbed, and hypertonia and tremors have been observed. Little and colleagues, in a retrospective medical record review, showed that infants exposed prenatally to methamphetamine had significantly lower birthweight, length, and head circumference than unexposed control infants.103 No difference in congenital anomalies was noted. In the large, prospective, longitudinal, multi-site study of the effects of prenatal methamphetamine exposure (the Infant Development, Environment and Lifestyle, or IDEAL, study), methamphetamine exposure was associated with lower birthweight and increased incidence of small for gestational age at delivery.144 This study also found a doseresponse relationship between prenatal methamphetamine exposure and infant neurobehavioral outcomes, with exposed infants demonstrating increased physiologic and CNS stress, lower arousal, and increased lethargy relative to unexposed controls.145 In one study using magnetic resonance imaging, children (ages 3 to 16 years) with a history of prenatal methamphetamine exposure were found to have smaller subcortical volumes that correlated with poorer performance on cognitive tests of visual-motor integration, sustained attention, and verbal and long-term spatial memory.34 Although limited by small sample size, this investigation provides preliminary clinical evidence that in utero methamphetamine exposure may have neurotoxic effects on brain development, with possible long-term implications for cognitive and behavioral functioning. A growing concern related to parental methamphetamine abuse is environmental exposure to methamphetamine and other chemicals in infants and children due to in-home manufacturing of the drug. “Meth labs” have become increasingly popular in certain areas of the United States, and case reports of accidental ingestion in infancy and childhood of methamphetamine and chemicals used in its production have been published.61,76 Large, systematic studies of children with

Chapter 38  Pharmacology

environmental exposure to methamphetamine are not available, but one initial report suggests greater behavioral problems in children with a history of residing in methamphetamineproducing homes.7 Ecstasy, or MDMA (3,4-methylenedioxymethamphetamine), has become increasingly popular as a recreational drug since the late 1980s. MDMA is an indirect monoaminergic agonist that stimulates release and inhibits reuptake of serotonin and other neurotransmitters, thus evoking elation and pleasure, accompanied by undesirable effects of hyperactivity and hyperthermia. Subsequent monoaminergic depletion leads to depression and lethargy. Cardiac arrhythmia, acute renal failure, rhabdomyolysis, disseminated intravascular coagulation, and death have been reported in young people using MDMA and engaging in “rave dancing” at crowded clubs. Corro­borated by animal studies, chronic MDMA use by adult humans has been shown to cause long-term neuropsychopharmacologic damage in terms of learning/memory, cognitive function, sleep, appetite, and loss of sexual pleasure. These effects persist even during abstinence, suggesting permanent axonal loss. Skelton and colleagues recently summarized evidence from a number of preclinical studies that indicates developmental exposure to MDMA is related to persistent learning and memory difficulties in offspring.141 One study of neonatal rats, postnatal days 11 to 20 (correlating with late third trimester in human brain development), shows that MDMA exposure causes dose-related sequential learning and spatial learning and memory deficits, but these effects are not seen with MDMA exposure at postnatal days 1 to 10 (correlating with first-trimester human brain development). Thus, timing of exposure appears to play a critical role in expression of effects.29 A study of ecstasy in humans described the characteristics of pregnant women in Toronto, Canada, who contacted the Hospital for Sick Children’s Motherisk Alcohol and Substance Use Hotline and reported ecstasy use during pregnancy.79 Compared with nonusers, MDMA users were younger and more likely to be white and single. MDMA users were also more likely to smoke cigarettes, drink alcohol, engage in binge drinking, use other illicit drugs, have coexisting psychiatric or emotional problems, and have unplanned pregnancies. Thus, the typical pregnant MDMA user appears to have multiple pregnancy risk factors. The United Kingdom National Teratology Information Service collected follow-up data on 136 ecstasy-exposed pregnancies and found a significantly increased risk for congenital defects, predominantly cardiovascular and musculoskeletal anomalies, in the offspring.111 The authors acknowledge that although the findings are clearly of interest, the study sample was quite small, with limited power to draw definitive conclusions regarding causality of the observed defects. Published reports of other neonatal and infant outcomes of gestational ecstasy exposure are not yet available. Many of the same caveats cited previously with regard to the interpretation of findings related to prenatal cocaine exposure are applicable to the study of these stimulants as well.

SEROTONIN REUPTAKE INHIBITORS Although not categorized as abuse substances, there is evidence for increasingly widespread maternal use of selective serotonin reuptake inhibitors (SSRIs) during pregnancy.163

753

These drugs are being commonly prescribed as the treatment of choice for major depression and anxiety. Although SSRIs vary in potency, pharmacokinetic effects, molecular structure, and half-life, the SSRIs act similarly by inhibiting serotonin reuptake at the presynaptic junction, leading to increased concentrations at the synaptic cleft and potentiating serotonergic neurotransmission. Thus, the fetus is exposed to increased serotonin levels during development. An early meta-analysis of several smaller studies of SSRI use in the first trimester was reassuring with regard to no demonstrable increased risk for congenital malformations,93 but subsequently concerns have been raised regarding an observed increase in congenital malformations associated with maternal therapy during pregnancy with SSRIs, especially paroxetine (Paxil).23 Gestational SSRI exposure has also been implicated in persistent pulmonary hypertension of the newborn, but this observation also requires further substantiation. There is an emerging literature that prenatal SSRI exposure and possibly withdrawal may be associated with neurobehavioral effects that may include changes in behavioral state and altered autonomic reactivity. Whether this actually constitutes a neonatal abstinence syndrome and whether there are long-term developmental consequences of in utero exposure to SSRIs remains unclear. Awaiting more definitive long-term studies of safety, clinicians should carefully consider the maternal diagnosis and indications as well as the potential risks and benefits before prescribing treatment with SSRIs during pregnancy.

REFERENCES 1. Abel EL: An update on incidence of FAS: FAS is not an equalopportunity birth defect, Neurotoxicol Teratol 17:437, 1995. 2. American Academy of Pediatrics: Fetal alcohol syndrome and alcohol-related neurodevelopmental disorders, Pediatrics 106:358, 2000. 3. American Academy of Pediatrics Committee on Drugs: Neonatal drug withdrawal, Pediatrics 101:1079, 1998. 4. American College of Obstetricians and Gynecologists: Smoking Cessation During Pregnancy, ed No 316, Washington, D.C., 2005, ACOG. 5. Anderson HR, Cook DG: Passive smoking and sudden infant death syndrome: review of the epidemiological evidence, Thorax 52:1003, 1997. 6. Arendt RE, et al: Children prenatally exposed to cocaine: developmental outcomes and environmental risks at seven years of age, J Dev Behav Pediatr 25:83, 2004. 7. Asanbe CB, et al: The methamphetamine home: psychological impact on preschoolers in rural Tennessee, J Rural Health 24:229, 2008. 8. Avants BB, et al: Effects of heavy in utero cocaine exposure on adolescent caudate morphology, Pediatr Neurol 37:275, 2007. 9. Bada HS, et al: Gestational cocaine exposure and intrauterine growth: maternal lifestyle study, Obstet Gynecol 100:916, 2002. 10. Bada HS, et al: Low birth weight and preterm births: etiologic fraction attributable to prenatal drug exposure, J Perinatol 25:631, 2005.

754

SECTION V  PROVISIONS FOR NEONATAL CARE

11. Baer JS, et al: A 21-year longitudinal analysis of the effects of prenatal alcohol exposure on young adult drinking, Arch Gen Psychiatry 60:377, 2003. 12. Bandstra ES, et al: Intrauterine growth of full-term infants: impact of prenatal cocaine exposure, Pediatrics 108:1309, 2001. 13. Bandstra ES, et al: Longitudinal investigation of task persistence and sustained attention in children with prenatal cocaine exposure, Neurotoxicol Teratol 23:545, 2001. 14. Bandstra ES, et al: Severity of prenatal cocaine exposure and child language functioning through age seven years: a longitudinal latent growth curve analysis, Substance Use Misuse 39:25, 2004. 15. Bateman DA, Chiriboga CA: Dose-response effect of cocaine on newborn head circumference, Pediatrics 106:e33, 2000. 16. Bauer CR, et al: Acute neonatal effects of cocaine exposure during pregnancy, Arch Pediatr Adolesc Med 159:824, 2005. 17. Bearer CF, et al: Ethyl linoleate in meconium: a biomarker for prenatal ethanol exposure, Alcohol Clin Exp Res 23:487, 1999. 18. Beeghly M, et al: Prenatal cocaine exposure and children’s language functioning at 6 and 9.5 years: moderating effects of child age, birthweight, and gender, J Pediatr Psychol 31:98, 2006. 19. Behnke M, et al: Incidence and description of structural brain abnormalities in newborns exposed to cocaine, J Pediatr 132:291, 1998. 20. Behnke M, et al: The search for congenital malformations in newborns with fetal cocaine exposure, Pediatrics 107:E74, 2001. 21. Behnke M, et al: Outcome from a prospective, longitudinal study of prenatal cocaine use: preschool development at 3 years of age, J Pediatr Psychol 31:41, 2006. 22. Bennett DS, et al: Children’s intellectual and emotionalbehavioral adjustment at 4 years as a function of cocaine exposure, maternal characteristics, and environmental risk, Dev Psychol 38:648, 2002. 23. Bérard A, et al: First trimester exposure to paroxetine and risk of cardiac malformations in infants: the importance of dosage, Birth Defects Res B Dev Reprod Toxicol 80:18, 2007. 24. Berghella V, et al: Maternal methadone dose and neonatal withdrawal, Am J Obstet Gynecol 189:312, 2003. 25. Bernstein VJ, Hans SL: Predicting the developmental outcome of 2-year-old children born exposed to methadone: impact of social-environmental risk-factors, J Clin Child Psychol 23:349, 1994. 26. Bookstein FL, et al: Corpus callosum shape and neuropsychological deficits in adult males with heavy fetal alcohol exposure, Neuroimage 15:233, 2002. 27. Bradley KA, et al: Alcohol screening questionnaires in women: a critical review, JAMA 280:166, 1998. 28. Breslau N, et al: Maternal smoking during pregnancy and offspring IQ, Int J Epidemiol 34:1047, 2005. 29. Broening HW, et al: 3,4-Methylenedioxymethamphetamine (Ecstasy)-induced learning and memory impairments depend on the age of exposure during early development, J Neurosci 21:3228, 2001. 30. Brown JV, et al: Prenatal cocaine exposure: a comparison of 2-year-old children in parental and nonparental care, Child Dev 75:1282, 2004. 31. Burd L, et al: Fetal alcohol syndrome: neuropsychiatric phenomics, Neurotoxicol Teratol 25:697, 2003. 32. Button TM, et al: The relationship of maternal smoking to psychological problems in the offspring, Early Hum Dev 83:727, 2007.

33. Center for Substance Abuse Treatment, Consensus Panel: Pregnant, Substance-Using Women. (DHHS Publication No. (SMA) 02-3677). Rockville, MD, 1993, Substance Abuse and Mental Health Services Administration. Treatment Improvement Protocol (TIP) Series #2. 34. Chang L, et al: Smaller subcortical volumes and cognitive deficits in children with prenatal methamphetamine exposure, Psychiatry Res 132:95, 2004. 35. Church MW, et al: Effects of prenatal cocaine on hearing, vision, growth, and behavior, Ann N Y Acad Sci 846:12, 1998. 36. Clark CM, et al: Structural and functional brain integrity of fetal alcohol syndrome in nonretarded cases, Pediatrics 105:1096, 2000. 37. Cnattingius S: The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes, Nicotine Tob Res 6:S125, 2004. 38. Coles CD, et al: Effects of cocaine and alcohol use in pregnancy on neonatal growth and neurobehavioral status, Neurotoxicol Teratol 14:23, 1992. 39. Coles CD, et al: Neonatal ethanol withdrawal: characteristics in clinically normal, nondysmorphic neonates, J Pediatr 105:445, 1984. 40. Committee to Study Fetal Alcohol Syndrome: Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention, and Treatment, Washington, 1996, National Academy Press. 41. Cornelius MD, et al: Prenatal tobacco effects on neuropsychological outcomes among preadolescents, J Dev Behav Pediatr 22:217, 2001. 42. Corwin MJ, et al: Effects of in utero cocaine exposure on newborn acoustical cry characteristics, Pediatrics 89:1199, 1992. 43. Covington CY, et al: Birth to age 7 growth of children prenatally exposed to drugs: a prospective cohort study, Neurotoxicol Teratol 24:489, 2002. 44. Coyle MG, et al: Diluted tincture of opium (DTO) and phenobarbital versus DTO alone for neonatal opiate withdrawal in term infants, J Pediatr 140:561, 2002. 45. Dahl RE, et al: A longitudinal study of prenatal marijuana use: effects on sleep and arousal at age 3 years, Arch Pediatr Adolesc Med 149:145, 1995. 46. Dashe JS, et al: Relationship between maternal methadone dosage and neonatal withdrawal, Obstet Gynecol 100:1244, 2002. 47. Day NL, et al: Prenatal marijuana exposure contributes to the prediction of marijuana use at age 14, Addiction 101:1313, 2006. 48. Day NL, Richardson GA: An analysis of the effects of prenatal alcohol exposure on growth: a teratologic model, Am J Med Genet 127C:28, 2004. 49. Delaney-Black V: Prenatal cocaine: quantity of exposure and gender moderation, J Dev Behav Pediatr 25:254, 2004. 50. Dennis T, et al: Reactivity and regulation in children prenatally exposed to cocaine, Dev Psychol 42:688, 2006. 51. Dixon SD, Bejar R: Echoencephalographic findings in neonates associated with maternal cocaine and methamphetamine use: incidence and clinical correlates, J Pediatr 115:770, 1989. 52. Donatelle RJ, et al: Incentives in smoking cessation: status of the field and implications for research and practice with pregnant smokers, Nicotine Tob Res 6:S163, 2004. 53. Dusick AM, et al: Risk of intracranial hemorrhage and other adverse outcomes after cocaine exposure in a cohort of 323 very low birth weight infants, J Pediatr 122:438, 1993. 54. Eiden RD, et al: Effects of prenatal cocaine exposure on infant reactivity and regulation, Neurotoxicol Teratol 31:60, 2009.

Chapter 38  Pharmacology 55. England L, Zhang J: Smoking and risk of preeclampsia: a systematic review, Front Biosci 12:2471, 2007. 56. England LJ, et al: Effects of smoking reduction during pregnancy on the birth weight of term infants, Am J Epidemiol 154:694, 2001. 57. English DR, et al: Maternal cannabis use and birth weight: A meta-analysis, Addiction 92:1553, 1997. 58. Eyler FD, Behnke M: Early development of infants exposed to drugs prenatally, Clin Perinatol 26:107, 1999. 59. Eyler FD, et al: Birth outcome from a prospective, matched study of prenatal crack/cocaine use: II. Interactive and dose effects on neurobehavioral assessment, Pediatrics 101:237, 1998. 60. Fares I, et al: Intrauterine cocaine exposure and the risk for sudden infant death syndrome: A meta-analysis, J Perinatol 17:179, 1997. 61. Farst K, et al: Methamphetamine exposure presenting as caustic ingestions in children, Ann Emerg Med 49:341, 2007. 62. Fergusson DM, et al: Maternal use of cannabis and pregnancy outcome, Br J Obstet Gynaecol 109:21, 2002. 63. Fernandez-Ruiz JJ, et al: Role of endocannabinoids in brain development, Life Sci 65:725, 1999. 64. Finnegan LP, Kaltenbach K: Assessment and Management of Neonatal Abstinence Syndrome. In Hoekelman, editors: Primary Pediatric Care, ed 3rd, St. Louis, 1992, Mosby pp 1367-1378. 65. Finnegan LP, et al: Opioid Dependence: Scientific Foundations of Clinical Practice. Pregnancy and Substance Abuse: Perspective and Directions. New York, 1990, Proceedings of the New York Academy of Medicine. 66. Fiore M, et al: Treating Tobacco Use and Dependence: 2008 Update, Rockville, MD, 2008, U.S. Department of Health and Human Services, Public Health Service. 67. Frank DA, et al: Neonatal body proportionality and body composition after in utero exposure to cocaine and marijuana, J Pediatr 117:622, 1990. 68. Frank DA, et al: Level of in utero cocaine exposure and neonatal ultrasound findings, Pediatrics 104:1101, 1999. 69. Fried PA, Makin J: Neonatal behavioral correlates of prenatal exposure to marihuana, cigarettes and alcohol in a low risk population, Neurotoxicol Teratol 9:1, 1987. 70. Fried PA, Smith AM: A literature review of the consequences of prenatal marihuana exposure: An emerging theme of a deficiency in aspects of executive function, Neurotoxicol Teratol 23:1, 2001. 71. Fried PA, Watkinson B: 12- and 24-month neurobehavioral follow-up of children prenatally exposed to marihuana, cigarettes and alcohol, Neurotoxicol Teratol 10:305, 1988. 72. Fried PA, et al: Growth from birth to early adolescence in offspring prenatally exposed to cigarettes and marijuana, Neurotoxicol Teratol 21:513, 1999. 73. Fried PA, et al: Differential effects on cognitive functioning in 13-to 16-year-olds prenatally exposed to cigarettes and marihuana, Neurotoxicol Teratol 25:427, 2003. 74. Gilliland FD, et al: Maternal smoking during pregnancy, environmental tobacco smoke exposure and childhood lung function, Thorax 55:271, 2000. 75. Goldschmidt L, et al: Prenatal marijuana exposure and intelligence test performance at age 6, J Am Acad Child Adolesc Psychiatry 47:254, 2008. 76. Gospe SM Jr: Transient cortical blindness in an infant exposed to methamphetamine, Ann Emerg Med 26:380, 1995.

755

77. Gray KA, et al: Prenatal marijuana exposure: effect on child depressive symptoms at ten years of age, Neurotoxicol Teratol 27:439, 2005. 78. Held JR, et al: The effect of prenatal cocaine exposure on neurobehavioral outcome: A meta-analysis, Neurotoxicol Teratol 21:619, 1999. 79. Ho E, et al: Characteristics of pregnant women who use ecstasy (3, 4-methylenedioxymethamphetamine), Neurotoxicol Teratol 23:561, 2001. 80. Hurt H, et al: School performance of children with gestational cocaine exposure, Neurotoxicol Teratol 27:203, 2005. 81. Hutchings DE, Dow-Edwards D: Animal models of opiate, cocaine, and cannabis use, Clin Perinatol 18:1, 1991. 82. Ioffe S, Chernick V: Development of the EEG between 30 and 40 weeks gestation in normal and alcohol-exposed infants, Dev Med Child Neurol 30:797, 1988. 83. Jacobson JL, Jacobson SW: Effects of prenatal alcohol exposure on child development, Alcohol Res Health 26:282, 2002. 84. Jeremy RJ, Bernstein VJ: Dyads at risk: Methadone-maintained women and their four-month-old infants, Child Dev 55:1141, 1984. 85. Johnson RE, et al: Use of buprenorphine in pregnancy: Patient management and effects on the neonate, Drug Alcohol Depend 70:S87, 2003. 86. Jones HE, et al: Treatment of opioid-dependent pregnant women: clinical and research issues, J Subst Abuse Treat 35:245, 2008. 87. Kakko J, et al: Buprenorphine and methadone treatment of opiate dependence during pregnancy: comparison of fetal growth and neonatal outcomes in two consecutive case series, Drug Alcohol Depend 96:69, 2008. 88. Kandall SR, et al: Differential effects of maternal heroin and methadone use on birth weight, Pediatrics 58:681, 1976. 89. Kandall SR, et al: Relationship of maternal substance abuse to subsequent sudden infant death syndrome in offspring, J Pediatr 123:120, 1993. 90. Kattwinkel J, et al: The changing concept of sudden infant death syndrome: Diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk, Pediatrics 116:1245, 2005. 91. Kelly SJ, et al: Effects of prenatal alcohol exposure on social behavior in humans and other species, Neurotoxicol Teratol 22:143, 2000. 92. Kodituwakku PW, et al: Specific impairments in self-regulation in children exposed to alcohol prenatally, Alcohol Clin Exp Res 19:1558, 1995. 93. Lattimore KA, et al: Selective serotonin reuptake inhibitor (SSRI) use during pregnancy and effects on the fetus and newborn: a meta-analysis, J Perinatol 25:595, 2005. 94. Law KL, et al: Smoking during pregnancy and newborn neurobehavior, Pediatrics 111:1318, 2003. 95. Leonardi-Bee J, et al: Environmental tobacco smoke and fetal health: systematic review and meta-analysis, Arch Dis Child Fetal Neonatal Ed 93:F351, 2008. 96. Lester BM, et al: Cocaine exposure and children: The meaning of subtle effects, Science 282:633, 1998. 97. Lester BM, Padbury JF: Third pathophysiology of prenatal cocaine exposure, Dev Neurosci 31:23-35, 2009. 98. Lester BM, et al: The maternal lifestyle study: Effects of substance exposure during pregnancy on neurodevelopmental outcome in 1-month-old infants, Pediatrics 110:1182, 2002.

756

SECTION V  PROVISIONS FOR NEONATAL CARE

99. Levine TP, et al: Effects of prenatal cocaine exposure on special education in school-aged children, Pediatrics 122:e83, 2008. 100. Lewis BA, et al: Prenatal cocaine and tobacco effects on children’s language trajectories, Pediatrics 120:E78, 2007. 101. Lidow MS: Consequences of prenatal cocaine exposure in nonhuman primates, Dev Brain Res 147:23, 2003. 102. Lipsitz PJ: A proposed narcotic withdrawal score for use with newborn infants. A pragmatic evaluation of its efficacy, Clin Pediatr 14:592, 1975. 103. Little BB, et al: Methamphetamine abuse during pregnancy: outcome and fetal effects, Obstet Gynecol 72:541, 1988. 104. Lumeng JC, et al: Prenatal exposures to cocaine and alcohol and physical growth patterns to age 8 years, Neurotoxicol Teratol 29:446, 2007. 105. Lumley J, et al: Interventions for promoting smoking cessation during pregnancy, Cochrane Database Syst Rev 4: CD001055, 2004. 106. Malanga CJ, Kosofsky BE: Mechanisms of action of drugs of abuse on the developing fetal brain, Clin Perinatol 26:17, 1999. 107. Mansvelder HD, McGehee DS: Cellular and synaptic mechanisms of nicotine addiction, J Neurobiol 53:606, 2002. 108. Mayes L, et al: Visuospatial working memory in school-aged children exposed in utero to cocaine, Child Neuropsychol 13:205, 2007. 109. Mayes LC: Developing brain and in utero cocaine exposure: effects on neural ontogeny, Dev Psychopathol 11:685, 1999. 110. Mayes LC, et al: Developmental trajectories of cocaineand-other-drug-exposed and non-cocaine-exposed children, J Dev Behav Pediatr 24:323, 2003. 111. McElhatton PR, et al: Congenital anomalies after prenatal ecstasy exposure, Lancet 354:1441, 1999. 112. Melvin CL, Gaffney CA: Treating nicotine use and dependence of pregnant and parenting smokers: an update, Nicotine Tob Res 6:S107, 2004. 113. Messinger DS, et al: The maternal lifestyle study: cognitive, motor, and behavioral outcomes of cocaine-exposed and opiate-exposed infants through three years of age, Pediatrics 113:1677, 2004. 114. Metosky P, Vondra J: Prenatal drug exposure and play and coping in toddlers: a comparison study, Inf Behav Dev 18:25, 1995. 115. Miller-Loncar C, et al: Predictors of motor development in children prenatally exposed to cocaine, Neurotoxicol Teratol 27:213, 2005. 116. Morrow CE, et al: Influence of prenatal cocaine exposure on full-term infant neurobehavioral functioning, Neurotoxicol Teratol 23:533, 2001. 117. Morrow CE, et al: Learning disabilities and intellectual functioning in school-aged children with prenatal cocaine exposure, Dev Neuropsychol 30:905, 2006. 118. Obel C, et al: Smoking during pregnancy and babbling abilities of the 8-month-old infant, Paediatr Perinat Epidemiol 12:37, 1998. 119. Oken E, et al: Maternal smoking during pregnancy and child overweight: systematic review and meta-analysis, Int J Obesity 32:201, 2008. 120. Olds D: Tobacco exposure and impaired development: a review of the evidence, Ment Retard Dev Dis Res Rev 3:257, 1997.

121. Ornoy A, et al: The developmental outcome of children born to heroin-dependent mothers, raised at home or adopted, Child Abuse Negl 20:385, 1996. 122. Ostrea EM, et al: Meconium analysis to assess fetal exposure to nicotine by active and passive maternal smoking, J Pediatr 124:471, 1994. 123. Porath AJ, Fried PA: Effects of prenatal cigarette and marijuana exposure on drug use among offspring, Neurotoxicol Teratol 27:267, 2005. 124. Rementeria J, Nunag N: Narcotic withdrawal in pregnancy: stillbirth incidence with a case report, Am J Obstet Gynecol 116:1152, 1973. 125. Richardson GA, et al: Prenatal alcohol, marijuana, and tobacco use: infant mental and motor development, Neurotoxicol Teratol 17:479, 1995. 126. Richardson GA, et al: Effects of prenatal cocaine exposure on growth: a longitudinal analysis, Pediatrics 120:1017, 2007. 127. Richardson GA, et al: The effects of prenatal cocaine use on infant development, Neurotoxicol Teratol 30:96, 2008. 128. Richardson GA, et al: Growth of infants prenatally exposed to cocaine/crack: comparison of a prenatal care and a no prenatal care sample, Pediatrics 104:e18, 1999. 129. Rocha BA, et al: Increased vulnerability to self-administer cocaine in mice prenatally exposed to cocaine, Psychopharmacology 163:221, 2002. 130. Rose-Jacobs R, et al: Intrauterine cocaine exposure and executive functioning in middle childhood, Neurotoxicol Teratol 31:159, 2009. 131. Sampson PD, et al: Prenatal alcohol exposure, birthweight, and measures of child size from birth to age 14 years, Am J Public Health 84:1421-1428, 1994. 132. Sarkar S, Donn SM: Management of neonatal abstinence syndrome in neonatal intensive care units: a national survey, J Perinatol 26:15, 2006. 133. Scher MS, et al: The effects of prenatal alcohol and marijuana exposure: disturbances in neonatal sleep cycling and arousal, Pediatr Res 24:101, 1988. 134. Schuetze P, Eiden RD: The association between maternal cocaine use during pregnancy and physiological regulation in 4-to 8-week-old infants: an examination of possible mediators and moderators, J Pediatr Psychol 31:15, 2006. 135. Schuetze P, Eiden RD: The association between prenatal exposure to cigarettes and infant and maternal negative affect, Infant Behav Dev 30:387, 2007. 136. Sekhon HS, et al: Prenatal nicotine exposure alters pulmonary function in newborn rhesus monkeys, Am J Respir Crit Care Med 164:989, 2001. 137. Sexton M, et al: Prenatal exposure to tobacco: 2. Effects on cognitive functioning at age 3, Int J Epidemiol 19:72, 1990. 138. Sheinkopf SJ, et al: Functional MRI and response inhibition in children exposed to cocaine in utero. Preliminary findings, Dev Neurosci 31:159, 2009. 139. Singer LT, et al: Cognitive and motor outcomes of cocaineexposed infants, JAMA 287:1952, 2002. 140. Singer LT, et al: Prenatal cocaine exposure: drug and environmental effects at 9 years, J Pediatr 153:105, 2008. 141. Skelton MR, et al: Developmental effects of 3 4-methylenedioxymethamphetamine: a review, Behav Pharmacol 19:91, 2008. 142. Slotkin TA: Fetal nicotine or cocaine exposure: which one is worse? J Pharmacol Exp Ther 285:931, 1998.

Chapter 38  Pharmacology 143. Slotkin TA, et al: Heroin neuroteratogenicity: delayed-onset deficits in catecholaminergic synaptic activity, Brain Res 984:189, 2003. 144. Smith LM, et al: The infant development, environment, and lifestyle study: effects of prenatal methamphetamine exposure, polydrug exposure, and poverty on intrauterine growth, Pediatrics 118:1149, 2006. 145. Smith LM, et al: Prenatal methamphetamine use and neonatal neurobehavioral outcome, Neurotoxicol Teratol 30:20, 2008. 146. Smith SM: Alcohol-induced cell death in the embryo, Alcohol Health Res World 21:287, 1997. 147. Sood B, et al: Prenatal alcohol exposure and childhood behavior at age 6 to 7 years: I. Dose-response effect, Pediatrics 108:E34, 2001. 148. Stocks J, Dezateux C: The effect of parental smoking on lung function and development during infancy, Respirology 8:266, 2003. 149. Streissguth AP, et al: Fetal alcohol syndrome in adolescents and adults, JAMA 265:1961, 1991. 150. Streissguth AP, et al: Maternal alcohol use and neonatal habituation assessed with the Brazelton scale, Child Dev 54:1109, 1983. 151. Streissguth AP, et al: Moderate prenatal alcohol exposure: effects on child IQ and learning problems at age 7½ years, Alcohol Clin Exp Res 14:662, 1990. 152. Substance Abuse and Mental Health Services Administration. Results from the 2007 National Survey on Drug Use and Health: National Findings, NSDUH Series H-34, Rockville, MD, 2008, Office of Applied Studies. 153. Sulik K, Johnston M, Webb M: Fetal alcohol syndrome: Embryogenesis in a mouse model, Science 214:936, 1981. 154. Swayze VW, Johnson VP, Hanson JW, et al: Magnetic resonance imaging of brain anomalies in fetal alcohol syndrome, Pediatrics 99:232, 1997.

757

155. Taeusch HW Jr, Carson SH, Wang NS, et al: Heroin induction of lung maturation and growth retardation in fetal rabbits, J Pediatr 82:869, 1973. 156. Testa M, Quigley BM, Eiden RD: The effects of prenatal alcohol exposure on infant mental development: A metaanalytical review, Alcohol 38:295, 2003. 157. The American College of Obstetricians and Gynecologists: At-Risk Drinking and Illicit Drug Use: Ethical Issues in Obstetric and Gynecologic Practice, Obstet Gynecol 112:1449, 2008. 158. Tronick EZ, Frank DA, Cabral H, et al: Late dose-response effects of prenatal cocaine exposure on newborn neurobehavioral performance, Pediatrics 98:76, 1996. 159. U.S. Department of Health and Human Services: The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General—Executive Summary, Rockville, MD, 2006, U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. 160. Warner TD, Behnke M, Eyler FD, et al: Diffusion tensor imaging of frontal white matter and executive functioning in cocaine-exposed children, Pediatrics 118:2014, 2006. 161. Webster WS, Brown-Woodman PD: Cocaine as a cause of congenital malformations of vascular origin: Experimental evidence in the rat, Teratology 41:689, 1990. 162. Yolton K, Dietrich K, Auinger P, et al: Exposure to environmental tobacco smoke and cognitive abilities among U.S. children and adolescents, Environ Health Perspect 113:98, 2005. 163. Zeskind PS, Stephens LE: Maternal selective serotonin reuptake inhibitor use during pregnancy and newborn neurobehavior, Pediatrics 113:368, 2004.

SECTION

VI

Development and Disorders of Organ Systems

759

CHAPTER

39

The Immune System PART 1

Developmental Immunology Reuben Kapur, Mervin C. Yoder, and Richard A. Polin

Developmental immunology can be defined as the study of how adaptive host defense blood cells in an individual sequentially respond to repetitive environmental challenges in a way that promotes the health and survival of the individual throughout maturation from fetal to adult life. According to classic principles, an individual becomes immune, or protected from reinfection, in response to an antigenic encounter during an initial infection. Mature immunologic competence is ultimately achieved through cumulative adaptive changes stimulated by exposure to a large repertoire of foreign antigenic material. Because the in utero fetal environment is sequestered from frequent encounters with microorganisms, the host defense system of the human newborn is inexperienced. This inexperience partially accounts for why newborns are so vulnerable to microbial attack during the first 6 weeks of life. Although many components of the immune system of the fetus are present early in gestation, some are immature and do not become fully functional (compared with the activity of immune defenses of adult subjects) until some time after birth. Despite these limitations, fetal host defense systems are capable of active engagement, and an immune response does occur when the fetus is infected in utero. An understanding of the development of the immune system in a particular fetus or neonate must take into account the specific environment in which that individual is developing. Because more prematurely born neonates with very low birthweight are surviving, it is possible to ask how the immunologic development of these infants (in an extrauterine environment) differs from in utero immunologic development of comparably aged fetuses. This chapter reviews the sequential appearance of the many components of the immune system in neonates and compares the functional activities of these components with those of immunologically competent adults. The immunologic system may be divided into two systems of host defense mechanisms: innate, or nonspecific, immune mechanisms, and acquired, or specific, immune mechanisms.

By definition, innate immunity includes host defense mechanisms that operate effectively without prior exposure to a microorganism or its antigens. Some of these mechanisms include physical barriers, such as intact skin and mucous membranes, and chemical barriers, such as gastric acid and digestive enzymes. Beneath these important protective layers lie phagocytic cells, which constitute the first line of host defense against any microbes that breach the cutaneous and mucosal barriers. Soluble plasma and tissue proteins serve to amplify the function of the phagocytic cells as innate immune effectors. Acquired, or specific, immunity comprises primarily the cell-mediated (T lymphocyte) and humoral (B lymphocyte and immunoglobulin) systems. Innate and acquired immune mechanisms are necessary for an individual to become fully immunocompetent. These systems are intimately interrelated and interdependent. Monocytes and macrophages are important components of natural immunity because these cells function to ingest and clear microbial pathogens from normal tissues. Monocytes and macrophages also play an important role in the processing and presentation of microbial antigenic material to T lymphocytes, a pivotal initiation step in the generation of a specific immune response. Monocytes and macrophages are important components of the innate and the acquired immune systems, and are necessary for an effective immune response.

OVERVIEW OF HEMATOPOIESIS All the cellular components of the immunologic system have limited life spans; the cells must be constantly replenished from a pool of undifferentiated precursor cells derived from a pluripotent stem cell. Hematopoietic stem cells are capable of self-renewal, ensuring maintenance of a lifelong pool of available hematopoietic precursors. Through mechanisms that are not well understood, pluripotent stem cells are stimulated to divide and differentiate into multipotent stem and committed progenitor cells, which mature into circulating blood cells.3 Fibroblasts, endothelial cells, and macrophages present in the microenvironment of the hematopoietic stem cells synthesize extracellular matrix molecules and growth factors that are crucial to the process of progenitorderived blood cell production. Evidence that mice could be protected from the lethal effects of total-body irradiation by exteriorizing and shielding

761

762

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

of the spleen provided some of the first evidence that the hematopoietic system may be maintained through a population of precursor cells. Later, Till and McCulloch50 showed that marrow withdrawn from a donor animal could repopulate the spleen of a lethally irradiated recipient animal with progenitor cells, giving rise to discrete colonies of red blood cells, megakaryocytes, granulocytes, and macrophages. The pluripotent hematopoietic stem cell was first isolated and characterized in mice. An equivalent cell in humans can be inferred from the successful long-term engraftment and proliferation of donor cells in recipient patients after bone marrow transplantation. Development of other monoclonal antibodies that recognize cell surface molecules highly expressed by murine and human hematopoietic stem cells has permitted isolation and characterization of the growth and differentiation requirements of this rare hematopoietic cell population. In vitro colony-forming assays have been used to confirm that the proliferation and survival of hematopoietic progenitor cells are maintained by hematopoietic growth factors. Hematopoietic growth factors not only promote production of blood cells, but also enhance the maturation of cells; granulocyte ingestion and killing of invading microbes are stimulated by several hematopoietic growth factors. Hematopoietic growth factors play an important regulatory role in providing adequate numbers of functionally mature blood cells during systemic infections. Fetal hematopoiesis begins in a primitive state early in gestation, before the development of a mature bone marrow hematopoietic microenvironment. The anatomic location of blood cell production changes several times during gestation as stem cells apparently home from one site to another.61 Early hematopoietic activity is restricted to the yolk sac until 6 to 8 weeks of gestation, when the liver becomes the predominant site. As gestation proceeds to the 20th week, the bone marrow becomes the major site of hematopoiesis, and thereafter remains the primary reservoir for replenishing circulating populations of immune cells.

INNATE IMMUNITY Cellular Components The most primitive host defense mechanism involves the ingestion and killing of bacteria and other microorganisms by phagocytic cells. Polymorphonuclear neutrophils (PMNs), monocytes, and macrophages are the major cell types that accomplish this aspect of host defense. Natural killer (NK) cells are also important components of the innate immune system, but these cells kill invading pathogens by nonphagocytic mechanisms. All these cell types can eliminate pathogens from the host, but do so more efficiently when the pathogens are opsonized, or coated, by complement components and other soluble proteins of the innate immune system. Likewise, the nonphagocytic mechanisms of microbial cytolysis used by NK cells, PMNs, and monocytes are augmented in the presence of specific antibody to the target organism. This section provides an overview of the phagocyte and NK cell systems and highlights the areas in which phagocytes isolated from fetal and neonatal blood seem functionally immature.

POLYMORPHONUCLEAR NEUTROPHIL SYSTEM Kinetics of Production and Circulation The PMN system arises from a progenitor cell called the colony-forming unit granulocyte-macrophage (CFU-GM). As the name implies, this stem cell–derived progenitor cell may differentiate into PMNs or monocytes. The myeloblast is the first identifiable neutrophil precursor. Myeloblasts produce many daughter cells (myelocytes). These nonproliferating cells require 7 days to become fully differentiated PMNs. This postmitotic compartment of differentiating but immature PMNs (metamyelocyte and band forms) and mature bone marrow PMNs constitutes a reserve pool of cells that may be rapidly mobilized into the circulation in response to inflammation. Positive and negative regulators of PMN production have been identified. Positive regulatory factors are interleukin (IL)-3, granulocyte-macrophage colony-stimulating factor (GM-CSF), and granulocyte CSF. Several negative regulators of PMN production have also been identified, including interferons (IFNs), transforming growth factor-a, macrophage inflammatory protein 1a, prostaglandins, and lactoferrin and other iron-binding proteins. Although it is unclear what factors cause PMN release from the bone marrow under normal conditions, IL-1, tumor necrosis factor–a (TNFa), epinephrine, and complement component fragments are known to stimulate PMN release from the bone marrow during inflammatory states. When mature PMNs leave the marrow environment, they circulate for approximately 6 to 8 hours before migrating into tissues, where they live for an additional 24 hours. While in the bloodstream, half of the PMNs are found in a marginated pool (primarily in the pulmonary capillaries), and half are circulating in the peripheral blood. After PMNs emigrate from the blood vessels and enter a tissue, they do not reenter blood vessels, but age and die in the tissue. The actual site of PMN clearance is unknown, but macrophages ingest and degrade senescent apoptotic PMNs in vitro. In the human fetus, PMN precursors are first identified in the yolk sac stage of primitive hematopoiesis.61 Mature PMNs are not identifiable in the fetal liver or bone marrow until approximately 14 weeks of gestation. By 22 to 23 weeks of gestation, the circulating PMN count has increased, but still is only approximately 2% of the circulating PMN concentration measured in the cord blood of term gestation newborns. At the same time that circulating concentrations of PMNs are low, fetal blood contains high concentrations of circulating hematopoietic progenitors. This high concentration of circulating progenitors may not be indicative of a large total body pool of progenitors, however. The actual number of CFUGMs in the human neonate is unknown. In comparative studies of fetal and neonatal rat hematopoiesis, the total-body CFU-GM pool was approximately 10% by weight of that of adult rats. One explanation for the high concentration of circulating progenitors is that these stem cells may be leaving one hematopoietic environment (the liver) for another (the bone marrow). Circulating concentrations of PMNs increase dramatically at birth, peak at 12 to 24 hours, and then decline slowly by 72 hours to values that remain stable during the neonatal period. The PMN count rarely decreases to less

Chapter 39  The Immune System

than 3000/mm3 during the first 3 days and less than 1500/mm3 thereafter in term infants, although PMN counts in preterm infants may be 1100/mm3. When searching for the cause of neutropenia (,1500 PMNs/mm3) in infants, a strong suspicion of infection is warranted, although infants with preeclamptic mothers, premature birth, birth depression, intraventricular hemorrhage, and Rh hemolytic disease may have low peripheral PMN counts. Persistent neutropenia is one feature that is frequently seen in patients who succumb to overwhelming sepsis. Neutropenia in these infants may occur because of increased margination of activated circulating cells or may be related to rapid depletion of circulating and bone marrow storage pool PMNs. Whatever the mechanism, failure to provide adequate numbers of PMNs to areas of microbial invasion in tissues may be a factor that places the infant at increased risk for development of overwhelming sepsis.25

Overview of Polymorphonuclear Neutrophil Functions The PMN is qualitatively and quantitatively the most effective killing phagocyte of host defense. Numerous coordinated steps are required to attract large numbers of PMNs from the circulating blood into a tissue at the site of microbial invasion (Fig. 39-1). When microorganisms penetrate cutaneous or mucosal barriers, macrophages and PMNs appear in the tissue, and along with the complement, kinin, and coagulation systems, initiate a series of events resulting in the release of complement fragments (particularly C5a) and other factors (IL-1, TNFa, and chemokines) that alter vascular endothelial cell morphology and function in the area of infection. The initial recognition of invading microbes by the macrophages and PMNs is facilitated by the presence of pattern

Marrow compartment

Vascular compartment

Maturation

Release Distribution Emigration

763

recognition receptors (Toll-like receptors [TLRs]) on the surface of the phagocytes that are specific for certain molecules distinctly expressed by bacteria, fungi, and viruses. Engagement of TLRs (see later) activates the phagocytes to release inflammatory mediators that affect endothelial cells in the immediate vicinity. These changes in the endothelial cell surface and the presence of numerous soluble inflammatory mediators cause circulating PMNs to withdraw from the circulation and adhere to the “inflamed” endothelium. After tightly adhering to the endothelium, the PMN begins a process of diapedesis through adjacent endothelial cells and the intact underlying basement membrane. When the PMN has penetrated the basement membrane, the cell emigrates from the blood vessel into the area of inflammation. Initial random migration (chemokinesis) becomes highly directed (chemotaxis) by the increasing concentration of chemotactic factors secreted by activated monocytes and macrophages. Other chemotactic agents generated at the site of infection include products of complement activation and some byproducts of the microorganisms. On reaching the area of bacterial invasion, the PMN ingests encountered microbes and begins a process of biochemical activation resulting in generation of reactive oxygen intermediates and other metabolites, which are used to kill the pathogen. Selected aspects of this complex process are described further later on. ADHESION

PMN adhesion to endothelial cells is a crucial step in the recruitment of these leukocytes to inflammatory sites. Although PMNs can adhere to normal or activated vascular endothelium, inflammation is associated with hemodynamic

Tissue compartment

Adherence Deformability Chemotaxis

Phagocytosis

Killing

“Respiratory burst” O2– H2O2 OH

•

Figure 39–1.  Overview of polymorphonuclear neutrophil (PMN) functions. PMNs are produced and mature in the bone marrow over a 2-week period. On release from the marrow, PMNs circulate for 6 to 8 hours before emigrating into tissues. At sites of infection, chemotactic factors enhance PMN adhesion to and emigration through vascular endothelium, and PMNs migrate in a directed fashion (chemotaxis) toward the pathogens. Phagocytosis of the offending organisms stimulates an increase in production of oxygen metabolites (respiratory burst), which facilitates PMN killing of the ingested microbes.

764

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

and biochemical changes in blood vessels that facilitate leukocyte adhesion to the vessel endothelial lining. It is now apparent that a series of changes in the expression of specific glycoproteins occurs on the leukocyte and the endothelial cell surface during inflammation. One family of structurally and functionally related proteins found on the plasma membrane of all myeloid cells seems to mediate many adhesion-related neutrophil functions (i.e., chemotaxis, phagocytosis, degranulation, and aggre­ gation). These cell surface glycoproteins belong to the integrin family of receptors and are designated CD11/CD18 leukocyte adhesion molecules. Each heterodimeric protein consists of one immunologically distinct a subunit (leukocyte functional antigen-1; CD11a, Mac-1; CD11b, P150, 95; CD11c), but all share a common b subunit (CD18). Monoclonal antibodies directed against CD11b/CD18 inhibit PMN aggregation, spread, chemotaxis, and adhesion to endothelial cells. CD11b/CD18 and CD11c/CD18 serve as complement receptors and mediate complement-coated (fragment iC3b) particle ingestion by PMNs. CD11a/CD18 has been shown to be important in PMN killing of target cells through antibodydependent mechanisms. Patients with a heritable deficiency of these leukocyte cell surface glycoproteins have recurrent infections characterized by failure to accumulate granulocytes at sites of infection. Neutrophil adhesion to vascular endothelium and to other phagocytes depends on several specific receptor-ligand interactions at the cell surface membrane (Fig. 39-2). Circulating neutrophils usually roll along vascular endothelial cells with transient interactions between neutrophil selectins on the cell

L-

surface and counter-receptors on the endothelium. When endothelial cells are activated during inflammation, more selectin counter-receptors (see Fig. 39-2) are expressed, and additional proteins (intercellular adhesion molecule [ICAM]-1 and ICAM-2 in Fig. 39-2) of the immunoglobulin supergene family, which serve as ligands for the neutrophil integrin family, are activated. The rolling neutrophil is slowed by the more concentrated interactions of the neutrophil cell surface selectins and the endothelial counter-receptors. Simultaneously, increased interactions of neutrophil cell surface integrins with endothelial ICAM-1 and ICAM-2 (see Fig. 39-2) occur, neutrophil cell surface selectins are shed, and a firm neutrophilendothelial adhesive interaction is established. Neutrophils migrate between endothelial cells and emigrate into the tissue toward the nidus of microorganisms. Adhesion of PMNs to artificial surfaces in vitro is comparable for unstimulated PMNs isolated from neonatal or adult blood. When PMNs from term neonates are stimulated with chemotactic factors, adhesion is greatly diminished, however, compared with cells isolated from adults. More profound deficits are displayed by PMNs from preterm infants. Less stimulated mobilization and overall expression of CD11b/ CD18 glycoproteins on the plasma membrane is one important factor contributing to decreased stimulated adherence. Other causes of diminished stimulated adherence include impaired capacity to upregulate cell surface chemotactic receptors, lower granular content of lactoferrin, less fibronectin binding to the cell surface, and less shape change (failure to increase significantly overall cell surface area on chemotactic factor stimulation). Downregulation of L-selectin

Figure 39–2.  Adhesion of white blood cells to endothelial cells is mediated by several receptor-ligand pair interactions, including the selectin-carbohydrate (sialylated Lewis-x [sLex]) and integrinimmunoglobulin families. ICAM, intercellular adhesion molecule.  (From Bevilacqua MI: Endothelial-leukocyte adhesion molecules, Annu Rev Immunol 11:767, 1993, with permission.)

sLex

Selectins P-

CD11a, b, c/CD18

E-

ICAM-1

ICAM-2

Chapter 39  The Immune System

expression on term newborn cord blood granulocytes and monocytes during acute inflammation has been shown, although the pattern and level of shedding vary from neutrophils isolated from adult subjects.24 Normal PMNs exhibit a biphasic pattern of aggregation, followed by disaggregation when exposed to chemotactic factors. Newborn cord blood PMNs readily aggregate, but do not disaggregate when stimulated. This irreversible aggregation of PMNs would represent an impairment to cellular emigration from inflamed blood vessels if this phenomenon occurred in vivo in newborns. CHEMOTAXIS

Chemotaxis is defined as the directed migration of PMNs toward the origin of a chemoattracting substance (see Fig. 39-1). This movement involves a series of orchestrated events, including PMN binding of the chemoattractant (by cell surface receptors), generation of an intracellular second messenger that is coupled to the receptor-ligand binding, and remodeling of the plasma membrane and cytoskeleton to produce shape changes and proper orientation of the cellular contents toward the highest concentration of the chemoattractant. Morphologically, the PMN orients toward the chemoattractant, with the leading edge (lamellipodium) of the cell becoming more spread and possessing many ruffles. Most of the intracellular organelles remain at the posterior pole of the cell (uropod). As the cell moves, the leading edge adheres to available surfaces, and contraction of cytoskeletal microfilaments (actin and myosin) pulls along the rest of the cell. Many aspects of this process are poorly developed in PMNs isolated from neonatal blood. Some of the chemotactic defects of PMNs present during the neonatal period persist throughout early childhood. PMNs must recognize a concentration gradient of the attracting substance to move in a directed fashion. Specific cell surface receptors at the leading edge of the cell bind the chemoat­ tractant. The number of receptors for a commonly used in vitro chemoattractant, N-formyl-methionyl-leucyl-phenylalanine (f-MLP), seems to be equal on PMNs from neonates and adults. When stimulated with other chemoattractants, fewer receptors for f-MLP are redistributed, however, from intracellular granules to the plasma membrane in PMNs isolated from cord blood. These cells show an apparent defect in transduction of the signals of f-MLP binding. There are significant impairments in the generation of subcellular second messengers and associated alterations in intracellular and extracellular membrane potential that are normally associated with chemoattractant binding to PMNs from adult subjects.60 In addition, chemotactic factor stimulation fails to mobilize sufficient CD11b/CD18 to the plasma membrane, and, as discussed earlier, this complement receptor is important in many adhesion-related processes. The in vivo functional relationship of these intracellular defects to impaired chemotaxis of PMNs from neonatal cord blood remains to be clarified. PMN movement requires dynamic cytoskeletal involvement. Despite similar baseline concentrations of filamentous (F) actin, PMNs from newborns do not generate significant additional levels of F actin after chemotactic factor stimulation compared with PMNs from adults18; this may impair cell movement and orientation. PMNs from neonatal cord blood fail to orient as quickly and do not sustain the bipolar

765

configuration of chemotactic factor–stimulated PMNs isolated from adult blood. PMNs must be deformable to emigrate from the intravascular to the extravascular space and through the extracellular matrix of tissues. PMNs from neonatal cord blood are less deformable than cells from adults. Because the energydependent interaction of the contractile proteins actin and myosin influences cell deformability, differences in the extent or rate of microfilament polymerization between PMNs from neonates and PMNs from adults may be relevant. Decreased deformability may also be due to the failure to generate intracellular second messengers (cyclic adenosine monophosphate) after stimulation, to overall lower intracellular concentrations of adenosine triphosphate (necessary for microfilament contraction), or to differences in plasma membrane fluidity between PMNs isolated from neonates and PMNs from adults.36 PMNs from term neonates are deficient in in vitro chemotactic ability. PMNs from premature infants are even more impaired. Similar deficits in PMN mobility have been shown in human newborns in vivo. When the skin of newborns is inflamed by the skin window technique, leukocyte accumulation is delayed and less intense than in adults. Overall, impaired mobility is the most consistently observed defect in PMN function in human newborns, and it contributes in a large way to the increased infectious susceptibility of neonates. PHAGOCYTOSIS

Phagocytosis is a process of particle ingestion. Most particulate matter must be opsonized (coated) with IgG, complement fragments C3b or iC3b, fibronectin, or other proteins before being recognized and engulfed by PMNs (Fig. 39-3). After binding of the opsonized microbe by an appropriate cell surface receptor, the PMN extends pseudopods to surround the particle and form a phagocytic vacuole. PMNs of term and preterm infants have normal phagocytic activity under most normal in vitro test conditions. These cells express diminished amounts, however, and do not upregulate the complement receptor CR3 (CD11b/CD18) when stimulated with chemotactic factors to the same degree as PMNs from adults. When stressed by a high ratio of microbes to PMNs in vitro, phagocytosis of gram-negative organisms and yeast particles by PMNs of neonates is deficient. In addition, alterations in phagocytosis have been reported for PMNs isolated from clinically stressed premature newborns. PMNs of well newborns show normal phagocytic activity, but under stressful conditions, microbe ingestion may be altered. MICROBICIDAL ACTIVITY

Oxygen-independent and oxygen-dependent mechanisms of microbial killing are used by the PMN. Microbe disruption occurs in the phagolysosome, which is formed when the phagosome fuses with primary and secondary neutrophil granules. Within the phagolysosome, cationic proteins, lysozyme, lactoferrin, and H1 function as oxygen-independent microbicidal mechanisms. Oxygen-dependent killing begins with PMN membrane perturbation during receptor binding of the opsonized particle and phagocytosis. A respiratory burst (an increase in oxygen consumption) ensues as reactive

766

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS Figure 39–3.  Antibody-dependent op­

Macrophage Bacteria

Fc receptors Inefficient phagocytosis

Enhanced phagocytosis

Antibacterial antibody

Opsonization

s­onization and phagocytosis of bacteria. Antibody binding to particles such as bacteria can markedly enhance the efficiency of phagocytosis. Enhancement of phagocytosis in macrophages involves increased attachment of the coated particle to the cell surface membrane and commensurate activation of the phagocyte, both of which are mediated through occupancy of Fc receptors.  (From Abbas AK et al, editors: Cellular and molecular immunology, 2nd ed, Philadelphia, 1994, Saunders, with permission.)

Binding of opsonized bacteria to macrophage Fc receptors

oxygen intermediates are synthesized. First, a membrane oxidase is activated that ultimately leads to production of superoxide (O22), which can lead to hydrogen peroxide (H2O2) synthesis. These intermediate products combine to form highly reactive hydroxyl radicals. In addition, myeloperoxidase present in PMN granules catalyzes the oxidation of halides (iodide, bromide, chloride) to hypohalous acids, which are powerful oxidants that cause bacterial cell wall dissolution. PMNs isolated from well term neonates exhibit more rapid activation of the superoxide-generating system than PMNs from adults. These cells also produce more H2O2 than PMNs from adults; however, there are impairments in the later phases of oxidative metabolism. Decrements in hypochlorous acid production and in quantitative measures of reactive hydroxyl radical formation have been reported. Deficiencies in certain oxygen-detoxifying enzymes such as glutathione peroxidase and catalase may cause oxidative damage to the PMNs, may serve as a source of cell dysfunction, and may decrease neutrophil viability. Overall, microbicidal activity appears normal for PMNs of term and preterm infants, unless these infants are clinically ill. When stressed, defects in killing of group B streptococci, Escherichia coli, and other microorganisms become readily apparent and significant.

Summary PMNs of term and preterm infants are as capable of ingesting and killing microbial pathogens as are PMNs of adults, unless the infants are clinically ill. Under conditions of stress, these PMNs do not function with normal phagocytic and microbicidal activities. PMNs isolated from the blood of neonates consistently display diminished chemotactic and adhesion capacities. It is unclear what role these disturbed PMN functions play in causing the neonate to be so highly susceptible to systemic infections.

A rationale for the use of hematopoietic growth factors to improve neutrophil function qualitatively and quantitatively has been well established.49 In multiple studies involving neonatal animals and human infants, administration of granulocyte CSF or GM-CSF has been documented to increase the neutrophil storage pool, induce neutrophilia, and improve many neutrophil functions. A recommendation for the use of these growth factors to decrease the incidence of morbidity and mortality associated with sepsis in newborns cannot yet be made based on published work, however.

MONONUCLEAR PHAGOCYTE SYSTEM Production and Differentiation The mononuclear phagocyte system comprises bone marrow monocyte precursors, circulating monocytes, and mature macrophages.8 As with the granulocyte lineage, mononuclear phagocytes are derived from CFU-GM progenitor cells. Several hematopoietic growth factors influence mononuclear phagocyte production. CSF-1 or macrophage CSF is the major hematopoietic growth factor influencing maturation and production of mononuclear phagocytes. Cells of this lineage seem to be unique in expressing cell surface receptors for CSF-1. On leaving the bone marrow, monocytes circulate for nearly 72 hours and then migrate into tissues, where the local extracellular milieu strongly influences monocyte-to-macrophage differentiation. Although the ultimate fate of macrophages is unclear, these cells have been observed to live for several months in many human tissues. Monocyte-to-macrophage differentiation has been well characterized, but the control mechanisms involved remain elusive. As monocytes develop into macrophages, certain morphologic changes occur. The cells increase in diameter more than threefold and acquire more cytoplasmic granules and vacuoles. Differentiation is also associated with increased

Chapter 39  The Immune System

expression of cell surface receptors, biosynthesis of numerous biologically active molecules, and improved phagocytic activity.40 Similar to PMNs, all mononuclear phagocytes have a well-developed cytoskeletal apparatus that is important in determining cell mobility and participation in many adhesionrelated functions. Macrophages may be first identified in the fetus in the hematopoietic foci of the yolk sac. Before the liver becomes the major site of hematopoiesis, macrophages constitute nearly 70% of the blood cells present in the liver. Over the next few weeks, erythroblasts increase in number and predominate, with hepatic macrophages declining to 1% to 2% of all differentiated blood cells. Few macrophages are present in the developing fetal lung, but at birth a rapid influx of macrophages is observed. Macrophage populations in the lung and other organs are maintained by local proliferation of macrophages and by influx of circulating bone marrow– derived monocytes. The stimulus for monocyte recruitment into tissues at homeostasis remains unclear. Circulating monocytes do not appear in fetal blood before the fifth month of gestation; however, at 30 weeks’ gestation, monocytes increase to 3% to 7% of the circulating formed blood cells. At birth and throughout the neonatal period, circulating monocyte concentrations exceed 500/mm3, a value considered high for adults.

Mononuclear Phagocyte Functions ADHESION

Mononuclear phagocytes adhere to various cells and tissues, and play central roles in wound healing, inflammation, and immune responses. Similar to neutrophils, mononuclear phagocytes must emigrate from blood vessels at inflammatory sites to eradicate invading microbes. Generally, mononuclear phagocyte recruitment and accumulation lag behind the brisk PMN influx by 6 to 12 hours, but the former process persists for several days. Mononuclear phagocytes are necessary not only for host defense at inflammatory sites, but also for tissue débridement and initiation of wound repair. Inability of PMNs and mononuclear phagocytes to accumulate at sites of inflammation or injury because of impaired adhesion-related functions results in recurring, poorly healing cutaneous abscesses. Monocyte influx into inflammatory sites is impaired in neonates compared with adults. Whether this delay results from inadequate tissue release of monocyte attractants, impaired monocyte mobility, or impaired adhesion and emigration has not been clarified. Most evidence suggests that monocyte adhesion to artificial surfaces is unimpaired during the neonatal period. Deficits in other adhesion-related functions seem to explain the delayed accumulation of monocytes in injured tissues of neonates (see later). CHEMOTAXIS

Monocyte chemotactic activity is decreased during the neonatal period, and this impaired mobility persists throughout early childhood. Although random migration of cord blood and peripheral blood monocytes seems unimpaired at birth, decreased migration in response to chemoattractants is noted soon after. Chemotaxis of cord blood monocytes is nearly normal, but monocytes isolated from the peripheral blood of newborns do not migrate normally. The chemotactic activity

767

of these cells sequentially declines over the first week of life before slowly improving and achieving (by age 6 years) the chemotactic activity of monocytes isolated from the blood of adults. Impaired migration in response to chemoattractants may be a primary factor in the delayed influx of monocytes at inflammatory sites during the neonatal period. PHAGOCYTOSIS

The ability of mononuclear phagocytes to ingest a variety of soluble and particulate matter is vital to the role these cells play in the immune response. Endocytosis is important in microbial elimination, removal of senescent and transformed cells, and antigen processing for initiation of specific immune responses. Mononuclear phagocytes ingest microscopic fluid and soluble matter through a process of pinocytosis, whereas large particulate material is taken up through phagocytosis. Both processes involve cell surface membrane extension, particle enclosure, and translocation of the phagosome into the cell interior. As with neutrophils, macrophage phagocytosis is enhanced when particulate material is coated with IgG or complement fragments, and the opsonized particles engage cell surface Fc and C3b receptors. Monocyte ingestion of Staphylococcus aureus, E. coli, Streptococcus pyogenes, Toxoplasma gondii, and herpes simplex virus type 2 seems quantitatively normal during the neonatal period, although the rate of ingestion may be slower. Similarly, alveolar macrophage ingestion of Candida albicans seems normal in newborns (who were intubated and receiving assisted ventilation) compared with the phagocytic capacity of alveolar macrophages isolated by bronchoalveolar lavage of adult volunteers. Some evidence exists, however, of a selective impairment in phagocytosis of GBS. In these studies, monocytes isolated from neonatal cord blood ingested fewer organisms than monocytes isolated from adult peripheral blood. Phagocytosis of group B streptococci significantly increases when monocytes of newborns are incubated with organisms preopsonized with fibronectin and IgG. MICROBICIDAL ACTIVITY

The generation of reactive oxygen intermediates is only one of several mechanisms used by mononuclear phagocytes to kill microbes. Similar to neutrophils, mononuclear phagocytes undergo a respiratory burst when particulate or soluble stimuli are engaged and ingested. The biosynthesis of O22, H2O2, and hypohalous acids remains an important mechanism for killing of organisms by monocytes, but during the differentiation of these cells into macrophages, the magnitude of the respiratory burst diminishes, and myeloperoxidase-containing granules disappear from the cytoplasm. The reduced microbicidal function of macrophages is important because it permits these cells to remove small numbers of microorganisms from tissues without causing extensive tissue damage. Oxygen-independent micro­ bicidal agents synthesized by macrophages include cationic proteins (defensins), lipid hydrolases, proteases, and nucleases. These agents, in combination with a highly acidified phagolysosome, seem to be cytolytically effective against many pathogenic microbes. The microbicidal activity of mononuclear phagocytes can be regulated by cytokines. IFN-g seems to be the most important

768

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

activating agent; however, GM-CSF and TNFa also play important roles. Activated macrophages exhibit enhanced release of proinflammatory cytokines, upregulation of Fc receptors, and increased production of reactive oxygen intermediates. In contrast, IL-10 and transforming growth factor-a function as potent suppressors of some macrophage functions. The ability of mononuclear phagocytes to generate reactive oxygen intermediates is normal during the neonatal period. Monocytes isolated from neonatal cord blood kill ingested bacteria and parasites as efficiently as cells from adult periph­ eral blood. Alveolar macrophages and monocyte-derived macrophages are also unimpaired in killing S. aureus and C. albicans during the neonatal period. IFN-g production and the response of mononuclear phagocytes to exogenous IFN-g are diminished, however, in newborns. How well fixed-tissue macrophages kill bacteria, parasites, and viruses and the response of these cells to cytokine stimulation remain untested.

Summary The influx of mononuclear phagocytes to sites of inflammation is delayed and attenuated in newborns. This defect is most likely related to the impaired chemotactic activity displayed by the peripheral blood monocytes of these infants. Phagocytosis and microbicidal activity seem equivalent to the level displayed by mononuclear phagocytes of adults. In vivo studies of macrophage function at birth and during the neonatal period are limited, but pulmonary alveolar macrophages seem to function normally in the infants examined to date.

NATURAL KILLER CELL SYSTEM Phenotypic and Functional Characteristics Generally, NK cells are defined by their large granular morphology. These cells make up almost 15% of peripheral blood lymphocytes in adults and are found in several tissues, including liver, peritoneal cavity, placenta, and bone marrow. Human NK cells are defined by the expression of cell surface proteins CD56 and CD16 (low-affinity IgG receptor). NK cells lack the expression of T cell markers such as CD3. Murine NK cells are defined by the presence of the cell surface proteins NK1.1 and DX5. Similar to human NK cells, murine NK cells also lack the expression of CD3. In humans, NK cells have been subdivided further into two subsets on the basis of CD56 cell surface expression. These two NK cell subsets mediate distinct functions. CD56hi NK cells biosynthesize and secrete higher levels of cytokines such as TNFa, IFN-g, GM-CSF, and IL-10 compared with CD56dim NK cells. CD56hi NK cells proliferate more efficiently in response to low doses of IL-2 compared with CD56dim NK cells. In contrast, CD56dim NK cells are more efficient at mediating NK cell cytotoxicity and antibody-dependent cell-mediated cytotoxicity compared with CD56hi NK cells; this is partly due to differences in the level of CD16 expression between these two subsets of NK cells. NK cells play a crucial role in controlling viral infections and in eradicating tumor cells. NK cells recognize infected, transformed, and stressed cells in the body and eliminate them either by directly killing them or by synthesizing cytokines and chemokines, which activate several cellular components of adaptive immunity. NK cells not only play an important role in innate immunity, but also integrate various

components of innate and adaptive immunity to mount an efficient immune response. Mice that lack NK cells are susceptible to viral infections, including cytomegalovirus (CMV) infections. The gene that confers resistance to CMV infection has been mapped to an area on the chromosome that encodes several NK cell receptor genes.62 NK cell receptors recognize virally infected cells and tumor cells and deliver a cytolytic response. Mice expressing defective NK cell cytolytic activity display impaired rejection of tumor cells in vivo. Restoration of NK cell function in NK cell–deficient mice suppresses tumor rejection and growth. Patients with mutations in the gene that encodes the transcription factor nuclear factor-kB (NF-kB) have defective NK cell activity despite normal numbers of NK cells, suggesting that NF-kB activity must regulate a crucial aspect of NK cell cytolytic function. Subsequent studies have shown that NF-kB plays an essential role in regulating the expression of perforin in NK cells. Perforin is known to play a crucial role in regulating NK cell cytolytic functions, and deficiency of perforin in NK cells impairs their cytolytic activity against virally infected host cells and tumor cells. As discussed, NK cells recognize virally infected, stressed, and transformed cells through cell surface receptors. Generally, NK cells survey tissues and cells in the body for the expression of “self” major histocompatibility complex (MHC) class I molecules.6,9 In the absence of adequate levels of MHC class I expression on autologous virally infected, transformed, or stressed cells, NK cells through their activating receptors deliver a “lethal hit,” destroying them. In addition to expressing activating receptors, NK cells also express inhibitory receptors, however, that recognize normal levels of specific self MHC class I molecules on autologous cells and deliver an inhibitory rather than an activating signal, protecting normal cells from the cytotoxic effects of NK cells. The mechanism by which an extracellular signal is converted into a series of intracellular events that eventually result in a cytotoxic response by NK cells is regulated partly by small signaling enzymes and adapter proteins in the cell. Activating NK cell receptors are associated with intracellular signaling molecules that contain a signature amino acid sequence known as the immunoreceptor tyrosine-based activation motif (ITAM).6 ITAMs that activate NK cell receptors are essential for delivering the “lethal hit” to target cells, including virally infected and tumor cells. In contrast, inhibitory NK cell receptors are associated with intracellular signaling molecules that contain motifs crucial for inactivating NK cell cytotoxic functions. These motifs are known as the immunoreceptor tyrosine-based inhibitory motifs, and are closely associated with MHC class I molecules. Together, immunoreceptor tyrosine-based inhibitory motifs and MHC class I molecules deliver a negative “off” signal to NK cells, protecting normal cells from the cytotoxic effects of NK cells. Under normal conditions, a fine balance between activating and inhibitory receptor function is maintained, which is essential for protecting normal autologous cells from the cytolytic activity of NK cells. In addition to using NK cell receptors, NK cells can mediate target cell killing by antibodydependent cellular cytotoxicity on binding to IgG-coated cells through cell surface CD16 receptors. Collectively, these characteristics enable NK cells to act as a first line of defense, playing an important role in natural resistance against cancer and various infectious diseases.

Chapter 39  The Immune System

Production and Differentiation of Natural Killer Cells NK cells develop in human bone marrow from CD341 stem/ progenitor cell precursors. The development of NK cells in the bone marrow requires the presence of the early-acting cytokines stem cell factor and Fms-like tyrosine kinase-3 (Flt3) ligand.6 In addition to the requirement for stem cell factor and Flt3 ligand during the early phases of NK cell development, IL-15 is necessary for later stages of NK cell development and maturation. Mature NK cells express the IL-15 receptor a chain, the IL-2/15 receptor b chain, and the common cytokine receptor chain (g-c), but do not express the IL-2 receptor a chain that is necessary for the formation of the high-affinity IL-2 receptor. Although high doses of IL-2 can stimulate NK cell proliferation, NK cell development is normal in mice that are deficient in the expression of IL-2 in vivo. In contrast, NK cell development is impaired in vivo in mice that are deficient in the expression of either IL-2/IL-15 receptor b or g-c genes.6 These results suggest that IL-15 is indispensable for NK cell development, and that the a chain of the IL-15 receptor plays a unique role in the development of NK cells. Bone marrow stromal cells secrete IL-15, and exogenously added IL-15 supports NK cell development in vitro from human CD341 progenitor cells in the absence of bone marrow stromal cells. Likewise, in the murine system, IL-15 is necessary for the differentiation of functional NK cells in vitro.

Transcription Factors in the Development and Function of Natural Killer Cells In addition to an essential role for cytokines in the development and maturation of NK cells, transcription factors play a crucial role in regulating the development of NK cells in vivo. Mice that lack the expression of the ID2 gene show a significant reduction in NK cell numbers, although the generation of other lymphoid lineage cells is unaffected.6 Genetic ablation of the Ets family of transcription factors also results in impaired NK cell lineage commitment. In mice, the production of IL-15 from bone marrow–derived stromal cells is partly regulated by the transcription factor, IFN regulatory factor-2. Deficiency of IFN regulatory factor-2 in mice results in impaired production of IL-15, and consequently these mice do not show terminal NK cell maturation. Genetic ablation of IFN regulatory factor2 in mice also results in abnormal NK cell development and function. Taken together, the development of NK cells is complex and depends on several factors, including the regulation of cytokines by transcription factors.

Natural Killer Cell Function in Neonates Most NK cells developing in human neonates express the CD16 and CD56 antigens. The percentage of NK cells in peripheral blood of newborns is similar to that of adult subjects. The absolute number of NK cells at birth is approximately twice that found in adult peripheral blood, presumably because of the higher total lymphocyte count. A small fraction of the neonatal NK cells fail to express CD56 and show poor cytolytic responses to target cells. In addition, compared with CD56expressing NK cells, CD56-deficient NK cells do not respond well to exogenous IFN-g stimulation. NK cell–mediated cytolysis in vitro generally is diminished during the neonatal period and early childhood. Defects in binding and lysis of target cells have been reported and are partly explained by impaired release of NK-derived IFNs and soluble cytotoxic factors. Baseline NK cell activity during the neonatal period is 30% to 80% of that of

769

adult NK cells; however, treatment of NK cells with exogenous IL-2, IL-12, and IL-15 in vitro augments the cytolytic activity of NK cells derived from neonates. Nevertheless, the stimulated NK cell response in neonates remains suboptimal compared with adult NK cells. NK cells derived from neonates are functionally distinct from NK cells found in adults. The functional defects in the development of NK cells in neonates have been proposed to be due partly to reduced expression of the g-c chain on neonate-derived NK cells. The common g-c chain is shared between cytokine receptors for IL-2, IL-4, IL-7, IL-9, and IL-15. Cord blood–derived NK cells express approximately one third of the level of g-c chain expressed by NK cells. In vitro culture of the cord blood– derived NK cells induced a significant increase in the expression of g-c chain, suggesting that reduced expression of g-c chain in neonate NK cells may contribute significantly to infections in neonates. Mutations in the g-c chain in neonates have been linked to severe combined immunodeficiency (SCID) and profound reduction in NK cells. SCID is a rare, fatal syndrome characterized by profound deficiencies in lymphocyte function, including NK cells. Although most mutations resulting in SCID remain unknown, some studies have suggested a link between SCID and mutations in genes encoding adenosine deaminase, the g-c chain of the IL-2 receptor and several other cytokine receptors, and Janus kinase 3, the intracellular signaling molecule that relays receptor-induced signals from the membrane to the nucleus.

Summary NK cells are relatively normal in number in the neonatal period, but surface membrane expression of certain antigens is altered compared with adult NK cells. NK cell cytolytic activity against target cells in vitro is diminished during the neonatal period. The role of NK cell immaturity in contributing to the increased susceptibility of newborns to viral infection remains to be determined.

Humoral Components OVERVIEW OF SERUM OPSONINS The role of humoral factors in the enhancement of leukocyte phagocytosis of bacterial pathogens has been known since the turn of the 20th century. These heat-labile and heat-stable plasma proteins, called opsonins, consist mainly of serum antibodies and components of the complement system, although several other proteins seem to play important roles (see later). The opsonic activity of plasma or serum from newborn term infants is equivalent to activity measured in sera from adults until the test concentrations of plasma or serum are reduced to less than 10%. At low serum concentrations, opsonic activity against various bacteria and fungi is diminished during the neonatal period. Opsonic activity is reduced even more in premature infants and persists at test concentrations of plasma or serum greater than 10%. These deficiencies in opsonic activity may be related partly to lower complement and IgG and IgM concentrations in newborns. The complement system is described in the following section. The opsonic role and other functions of immunoglobulins are reviewed later, when antibodies are discussed as humoral components of specific immunity.

770

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

COMPLEMENT SYSTEM Activation Pathways and Overview of Functional Products The complement system plays an important role as one of the principal humoral effector pathways of immunity.48 Its major function is to facilitate the neutralization of foreign substances either in the circulation or on mucous membranes. This function is accomplished by a series of plasma proteins that are involved in specific and nonspecific host defense mechanisms. The classic pathway of complement activation requires the presence of specific antibodies against a particular antigen and the formation of immune complexes (Fig. 39-4). Two antibody molecules of the immune complex are bridged by the first component of complement, C1, which initiates a chain reaction in which one activated component serves as the enzyme that cleaves the next component in line. The order of component activation in the classic pathway is C1,

Figure 39–4.  Complement system activation cascade. Activation of the classic pathway (left) and the alternative pathway (right) causes generation of soluble factors that amplify phagocyte functions and produces a membrane-bound attack complex that damages cell membranes. (From McLean RH et al: Genetically determined variation in the complement system: relationship to disease, J Pediatr 105:180, 1984, with permission.)

C4, C2, and C3. Peptides with different biologic activities are created and either remain attached to the site of activation or diffuse into the milieu. The third component of complement, C3, is cleaved into membrane-bound C3b and the fluid phase C3a. With the generation of C3b, the classic pathway merges with the other mode of complement activation, the alternative pathway. In contrast to the classic pathway, the alternative pathway may be activated by bacterial or mammalian cell surfaces in the absence of specific antibodies. This activation is possible because small amounts of C3 in the circulation are constantly being converted to C3b. This complement component can bind to cell surfaces, interact with the next alternative pathway components in sequence (factors B and D), and form a potent enzyme for further C3 activation (C3bBb). C3b originally generated by the classic pathway may involve this amplification loop of the alternative pathway and significantly enhance local C3 activation. The central location of C3 in the complement pathway is important not only because the classic and alternative pathways converge at this point, but also because many biologic effects are determined by the interaction of this important molecule with various regulatory systems. The direction in which the complement pathway proceeds depends on the surface to which the C3b molecule is bound. Certain bacteria and other membranes offer a “protective surface” that favors the binding of C3b to factors B and D and the assembly of the enzyme that converts the next component in line, C5, into two biologically active fragments. The smaller product, C5a, acts as an anaphylatoxin to promote many aspects of acute inflammation, including chemoattraction of leukocytes and increasing vascular permeability. The larger fragment, C5b, remains attached to the C5 convertase on the membrane and assembles the components of the membrane attack complex: C5b, C6, C7, C8, and C9. Insertion of this complex into the cell membrane results in loss of membrane functions and cellular integrity. In the absence of a protected site, C3b is exposed to factors I and H, which facilitate cleavage of C3b into iC3b and C3d. All these fragments of C3 can function as ligands for cellular receptors, which are located on erythrocytes and almost all immunocompetent cells. The C3b receptor (CR1) is known to mediate adherence of complement-coated complexes to erythrocytes, neutrophils, and mononuclear phagocytes and plays an important role in the clearance of immune complexes, bacteria, and cellular debris from the circulation. Other receptors also have been identified, such as the C3bbinding protein, CR2, which is found on B lymphocytes and eosinophils. Its function awaits clarification, but there is evidence that adherence of the Epstein-Barr virus is mediated by this receptor. Finally, the iC3b receptor (CR3, CD11b/CD18) seems to be important for the ability of the host to overcome bacterial invasion. As previously mentioned, patients with a genetic deficiency of leukocyte cell surface CR3 experience severe recurrent bacterial infections.

Ontogeny and Analysis of the Complement System in the Neonatal Period Complement proteins are synthesized early in gestation. Synthesis of C2, C3, C4, and C5 can be confirmed between 8 and 14 weeks of gestation. Most evidence, derived through

Chapter 39  The Immune System

several methods, confirms that there is no transplacental passage of complement components. The components of the classic complement system and their functional activity (measured by total hemolytic complement assay) in full-term neonates are nearly comparable with those of normal adults (Table 39-1). In some studies in which neonates were compared with their mothers, significant deficiencies in complement component concentrations were reported (approximately 50% of adult levels). These studies did not take into account that many complement components are acute-phase proteins, and their serum concentration during pregnancy is significantly elevated compared with other normal adults. Preterm neonates have significantly decreased concentrations, however, of C1q, C4, C3, and total hemolytic complement compared with term neonates. There is a statistically significant correlation between increasing birthweight or gestational age and serum concentrations of these components. Alternative pathway activity and individual alternative pathway components are more deficient in concentration than classic pathway activity and components in term and preterm infants (see Table 39-1). Although the concentration of alternative pathway components in some newborn term infants is equivalent to the concentration in adults, most of these infants have lower measurable alternative pathway activity. In contrast to some classic pathway components, there is no correlation between factor B concentration and gestational age. Most serum alternative complement component values are equal to values measured in human adults by the first year of life (6 to 18 months of age).

TABLE 39–1  S  ummary of Published Complement Levels in Neonates MEAN PERCENTAGE OF ADULT LEVELS Complement Component

Term Neonate

Preterm Neonate

CH50

56-90

45-71

AP50

49-65

40-51

C1q

65-90

27-58

C4

60-100

42-91

C2

76-100

96

C3

60-100

39-78

C5

75

—

C6

47

—

C7

67

—

C9

14

—

B

35-59

36-50

P

33-71

13-65

H

61

—

iC3b

55

—

From Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 3rd ed, Philadelphia, 1990, Saunders, p 25.

771

Summary Whether deficiencies in the complement system predispose a neonate to infection has not as yet been established.48 Defects in the complement system and in the alternative pathway, in particular, are likely ultimately to be found to play a role in susceptibility to infection, especially in preterm infants. Newborns have a limited spectrum of antibody transmitted across the placenta; they receive IgG, no IgM, and little antibody to the entire range of gram-negative bacteria. The classic pathway has relatively little value at and shortly after birth. It follows that, in the absence of specific antibody, nonspecific activation of the biologically active fragments and complexes of the complement system through the alternative pathway becomes an extremely important defense mechanism for neonates during the first encounter with many bacteria. In most neonates, the functional deficiency of the alternative pathway, in conjunction with impaired functioning of PMNs, is likely to be clinically relevant.

FIBRONECTIN Fibronectins are a class of multifunctional, high-molecularweight glycoproteins that serve to facilitate cell-to-cell and cell-to-substratum adhesion. As adhesive ligands, fibronectins play an important role in directing cell migration, proliferation, and differentiation. Fibronectins seem essential for certain aspects of embryologic development of the fetus and for hemostasis, hematopoiesis, inflammation, and wound healing. In many pathophysiologic conditions (e.g., sepsis, thrombosis, cancer, organ fibrosis), the normal structure, physiology, and function of fibronectins are altered in a way that contributes to the underlying tissue or organ dysfunction.

Structure and Function Fibronectins exist as soluble and insoluble dimeric molecules. Plasma fibronectin circulates at a mean concentration of 330 mg/mL in human adult peripheral blood. Soluble fibronectins have been identified in nearly every body fluid tested (synovial, ocular, pleural, amniotic, cerebrospinal, and others). Insoluble fibronectins are present in many connective tissues and extracellular matrices throughout the body. It is apparent that structural variants (isoforms) of fibronectin exist, and that expression of these isoforms is highly regulated in a cell-specific and tissue-specific fashion. All fibronectins are capable of binding to multiple ligands simultaneously. The modular design of these glycoproteins translates into a linear array of active globular functional domains. Fibronectins display individual binding sites for some molecules and multiple sites for others. Fibronectins bind certain bacteria, heparin, fibrin, IgG, and DNA in several separate domains, but other bacteria, matrix molecules, actin, complement component C1q, and gangliosides are bound at unique sites in individual domains. In this way, fibronectin facilitates the interaction of cells with other cells, bacteria, tissues, particles, and soluble proteins. Circulating plasma fibronectin concentrations are reduced in fetal cord blood (120 mg/mL) and in term infants (220 mg/mL). Premature infants of 30 to 31 weeks’ gestation have significantly lower plasma concentrations (152 mg/mL) than term infants. Lower plasma concentrations in neonates are correlated with decreased synthetic rates, but plasma clearance of fibronectin also is measurably slower. Plasma fibronectin concentrations are

772

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

reduced further in infants with respiratory distress syndrome, birth depression, sepsis, and intrauterine growth restriction.13 Fibronectin biosynthesis by macrophages in vitro is decreased in the neonatal period.

Role in Immune Responses Fibronectin improves leukocyte function in vitro. Plasma fibronectin and proteolytic fragments of fibronectins promote human adult peripheral blood neutrophil and monocyte chemotaxis, adhesion, and random migration. In addition, fibronectin enhances phagocyte ingestion, reactive oxygen intermediate production, and killing of opsonized (complement or IgG) yeast and bacterial organisms. Fibronectin seems to play an important role as an enhancer of phagocyte function.

Summary Circulating concentrations of fibronectin are decreased in the neonatal period and are correlated with gestational age. Even lower plasma concentrations are measured in infants who are ill. Plasma fibronectin increases phagocyte function in vitro. The role of fibronectin in host immune defense in neonates remains uncertain, but in vitro data suggest a potential role as an enhancer of phagocyte function.

OTHER HUMORAL FACTORS C-Reactive Protein C-reactive protein (CRP) is an acute-phase reactant originally identified by its ability to bind to a pneumococcal polysaccharide antigen. This protein has strong functional similarity to the complement component C1q binding domain of IgG, and similar to IgG is able to activate the classic complement pathway by binding C1q. The polysaccharide antigen recognized by CRP is expressed by many bacteria and some fungi. When CRP binds to the antigen, and the complement system is activated, the organism is effectively opsonized, and rapid clearance occurs through neutrophil, monocyte, or macrophage phagocytosis. CRP is synthesized by the fetus and the newborn. Serum concentrations seem equivalent in uninfected neonates and adults. CRP is one of the most rapidly responsive acute-phase proteins, with increases of 100-fold to 1000-fold (in adults) in the serum concentration detectable during an infection.31 In surveys of infants with proven infections, an elevation of CRP serum concentration has been observed in 50% to 100% of patients. Normal values of CRP are less than 1.6 mg/dL on postnatal days 1 to 2 and less than 1 mg/dL thereafter. Diminished or absent increases in the CRP concentration have been observed during the first 12 to 24 hours of life in infected newborns, particularly in newborns infected with group B streptococci. CRP levels decrease rapidly in infants who clinically respond to antimicrobial therapy and return to normal in 5 to 6 days.

Lactoferrin Lactoferrin is a positively charged iron-binding glycoprotein present in the specific granules of neutrophils. Lactoferrin is released into the phagosome after particle ingestion and is deposited on the cell surface membrane during stimulated degranulation. This glycoprotein seems

to enhance neutrophil-endothelial adhesion interactions and neutrophil aggregation, reactive oxygen intermediate production, and chemotaxis. Deficient neutrophil adherence and directed migration have been reported in studies of patients with a heritable deficiency of lactoferrincontaining neutrophil-specific granules. Neonatal cord blood neutrophils are profoundly deficient in lactoferrin. Stimulation of neutrophils with the synthetic chemoattractant f-MLP results in degranulation of lactoferrin that is quantitatively similar to the concentrations elicited from neutrophils of adults. In contrast, lactoferrin release is diminished when neutrophils are stimulated to spread and move on artificial surfaces. The role of lactoferrin deficiency in contributing to impaired neutrophil chemotaxis is uncertain.

Collectins Collectins are soluble proteins that play a role in innate immunity.53 These molecules include mannose-binding lectin (MBL) and surfactant proteins A (SP-A) and D (SP-D) in humans, and the bovine molecules conglutinin, CL-43, and CL-1. Collectins share several structural features, including a collagen domain, a neck region, and a globular carboxylterminal C-type (calcium-dependent) lectin-binding domain. Collectins specifically recognize the patterns of carbohydrates on the outer walls of microorganisms. In essence, collectins function by binding to microorganisms and enhancing uptake and clearance by phagocytes. MBL can also function to activate the classic complement pathway in much the same fashion as the complement component C1q. MBL, SP-A, and SP-D bind to one or more receptors on phagocytes and stimulate chemotaxis, the respiratory burst, and the opsonophagocytosis of a wide variety of microorganisms, including group B streptococci. SP-A and SP-D are synthesized by type II alveolar cells and by nonciliated bronchiolar epithelial cells. As with other components of surfactant, the biosynthesis and secretion of these collectins increase dramatically during the third trimester of pregnancy. Acute injury or inflammation results in rapid increases in SP-A and SP-D biosynthesis. SP-D protein is also found in gastric mucosal cells, and SP-A is present in small intestinal cells. MBL is synthesized in the liver and secreted into the circulation. MBL serum concentrations vary widely (1000-fold) in adult human subjects primarily related to inherited polymorphisms in the promoter region of the gene. A wide range in cord blood MBL concentrations has been reported. Neonatal serum MBL concentrations generally are lower than the levels measured in adult serum. Relative deficiency of MBL has been associated with increased susceptibility to infections in children and in some adults.

BIOLOGY AND ROLE OF CYTOKINES IN NEWBORNS Overview Newborn infants have an increased susceptibility to infection because of various host defense impairments that exist during the neonatal period. The generation and maintenance of acquired immune responses are controlled by a network of regulatory glycoproteins and phospholipids that mediate the interactions between cells. These cytokines and chemokines are responsible for the generation of the immune response and

Chapter 39  The Immune System

differentiation of a wide variety of immune and nonimmune cells. The infant’s ability to generate the right balance of proinflammatory and anti-inflammatory cytokines when challenged with an infectious agent allows him or her to recover from the encounter with minimal residua. When the balance is not perfectly controlled, morbidity and mortality are increased.23 Unregulated production of cytokines in neonates may contribute to the development of necrotizing enterocolitis, bronchopulmonary dysplasia, and hypoxic-ischemic brain injury. The number of newly discovered cytokines is increasing on a yearly basis. This section focuses on the major cytokines involved in the immune responses to infectious agents and noninfectious stimuli, their developmental patterns in fetuses and newborns, and their coordinated role in neonatal sepsis.

CYTOKINE AND CHEMOKINE BIOLOGY Interleukin-1 Family The IL-1 family comprises three polypeptides with similar tertiary structures (IL-1a, IL-1b, and IL-1 receptor antagonist [IL-1ra]). IL-1a and IL-1b are translated as precursor peptides. Pro-IL-1a is fully active and resides in the cytoplasm, but can be transported to the cell surface where it has a role in cell-cell communication. IL-1a appears in the circulation only during severe disease. IL-1a and IL-1b share little sequence homology, but bind to the same receptor and have the same tertiary structure. Pro-IL-1b (the main circulating member of this family) is inactive and must be cleaved by a cysteine protease (IL-1b-converting enzyme or caspase I) before it is secreted. IL-1ra is a competitive inhibitor of the other members of the family that has no agonist activity. There are two varieties of IL receptors: The type I receptor binds all members of the family; the type II receptor binds only IL-1b. Both receptors are members of the IL-1 receptor/TLR superfamily. IL-1 is synthesized by a wide variety of immune and nonimmune cells, including monocytes, macrophages, neutrophils, endothelial cells, and epithelial cells. Synthesis of these ILs is triggered by microbial products of inflammation, and many of the features of the inflammatory response syndrome can be directly attributed to members of this family. IL-1 production by monocytes and macrophages from term and premature newborns is equivalent to that of adults. In preterm infants with sepsis, monocyte secretion may be diminished, however, during the acute phase of the disease.

Interleukin-6 IL-6 is secreted in a variety of different molecular forms as a result of post-translational modification. IL-6 synthesis is initiated by cytokines (including IL-1 and IL-6), platelet-derived growth factor, epidermal growth factor, viral and bacterial infections, double-stranded RNA, endotoxin, and cyclic adenosine monophosphate. The receptor for IL-6 consists of two subcomponents: (1) a ligand-binding molecule that is not responsible for signal transduction (IL-6R), and (2) a non– ligand-binding signal transducer (gp130). A soluble form of the IL-6 receptor (sIL6Ra) also exists. sIL6Ra can bind to IL-6 and then interact with gp130 on cells that do not express the IL-6 receptor. IL-6 injected intravenously into human patients is less toxic than IL-1b and TNFa, but does result in chills and

773

fever. IL-6 is known to activate T and B cells, stimulate maturation of megakaryocytes, increase the production of acute-phase response proteins, and enhance NK cell activities. Monocytes from term infants produce adequate amounts of IL-6 after stimulation with lipopolysaccharide (LPS), but not IL-1. Cells derived from preterm infants exhibit decreased production no matter what the stimuli. Circulating levels of IL-6 in newborns are lower than corresponding maternal values; however, the percentage of IL-6-positive monocytes (after stimulation with LPS) is higher in preterm and term neonates.

Interleukin-10 IL-10 is a potent anti-inflammatory/immunosuppressive polypeptide synthesized by monocytes, macrophages, T cells, and B cells in response to bacteria, bacterial products, viruses, fungi, and parasites. IL-10 decreases the synthesis of a wide variety of proinflammatory cytokines, and increases the production of naturally occurring proinflammatory cytokine inhibitors (e.g., IL-1ra). The synthesis of IL-10 is enhanced by various other cytokines, including TNFa, IL-1, IL-6, and IL-12. In contrast, IL-10 enhances B cell function and promotes development of cytotoxic T cells. IL-10 receptors are members of the class II cytokine receptor family and consist of two subunits. IL-10 production in neonates is diminished.

Interleukin-12 and Interferon-g IL-12 is a heterodimer consisting of two subunits (p35 and p40) encoded by different genes. The p40 subunit mediates binding to the IL-12 receptor, whereas the p35 subunit is needed for signal transduction. The p40 subunit can also form homodimers with itself that bind to the IL-12 receptor with equal affinity, but without eliciting a cellular effect. The homodimers may help modulate the effects of IL-12. The synthesis of the p35 and p40 subunits is regulated independently. In response to a given stimulus, cells secrete 10 to 100 times more of the homodimer than the heterodimer. IL-12 is produced by phagocytic cells (e.g., monocytes and macrophages) in response to bacteria and bacterial products, intracellular pathogens, and viruses. The IL-12 receptor is a member of the gp130 cytokine receptor superfamily. The IL-12 receptor consists of two subunits, and both must be activated for signal transduction. The principal cellular targets of IL-12 are T cells and NK cells. IL-12 induces the production of interferon (IFN-g), stimulates proliferation, and enhances cytotoxicity. The production of IL-12 by cord blood–derived mononuclear cells (in response to endotoxin) is diminished; however, normal IL-12 synthesis has been observed with heat-killed S. aureus as the stimulus. IFN-g is produced by NK cells, type 1 T-helper (Th) cells, and cytotoxic T cells in response to IL-12, TNFa, IL-1, IL-15, and IL-18. The receptor for IFN-g has two subunits; one is responsible for binding, and the other is responsible for signaling. IFN-g induces class II histocompatibility antigens and activates macrophages, and is important for host defenses against intracellular pathogens, among other functions (Box 39-1). Synthesis of IFN-g is greatly diminished in neonates.

Tumor Necrosis Factor Family The TNF family has two members: TNFa (cachectin) and TNFb (lymphotoxin). TNFa exists as a transmembrane form (prohormone) and a smaller secreted form consisting of three

774

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

BOX 39–1 Functions of Interferon-g n Enhances

macrophage microbicidal activity secretion of inflammatory mediators n Enhances host cell resistance to nonviral intracellular pathogens n Upregulates surface Fc receptors and class II major histocompatibility complex antigens on phagocytes n Inhibits macrophage migration n Promotes formation of giant cells n Induces myelosuppression n Augments functioning of mature neutrophils n Enhances differentiation of cytolytic T cells, natural killer cells, and lymphocyte-activated killer cells n Activates endothelial cells n Suppresses host protein synthesis n Promotes

monomers. The soluble form of TNFa is formed by the cleavage of the prohormone by a matrix metalloproteinase disintegrin. The transmembrane and the secreted forms of TNF can be biologically active. Similar to IL-1, synthesis of TNFa is triggered by microbial products of inflammation (and by IL-1 and TNFa themselves). A wide variety of cells are capable of producing TNFa; however, monocytes and macrophages represent the major sources of this circulating cytokine. There are two varieties of TNF receptors—TNFR-1 and TNFR-2. Stimulation of TNFR-1 reproduces many TNFa functions, including cytotoxicity and upregulation of adhesion molecules. TNFR-1 contains a cytoplasmic sequence of 80 amino acids that regulates programmed cell death. TNFR-2 may facilitate binding of TNF to TNFR-1. TNFs share many of the proinflammatory effects of the ILs, but are more likely to result in leukopenia. The production of TNFa by neonatal monocytes and macrophages is less than that in adults. Endotoxinstimulated cord blood cells from preterm infants secrete signi­ ficantly less TNF than cells derived from term infants or adults. The expression of TNF receptors may also be diminished.

Platelet-Activating Factor Platelet-activating factor (PAF) is a potent phospholipid inflammatory mediator that is rapidly degraded by acetylhydrolase. PAF is produced by many cell types; however, only macrophages and eosinophils exhibit regulated release. PAF primarily resides on the cell surface, where it acts as an intercellular messenger. Intravenous administration of PAF results in systemic hypotension, capillary leakage, pulmonary hypertension, neutropenia, thrombocytopenia, and ischemic intestinal necrosis in animals. At a cellular level, PAF is a potent activator of neutrophils. PAF production is increased by cellular exposure to lipopolysaccharide (LPS), hypoxia, hematopoietic growth factors, TNF, IL-1, thrombin, bradykinin, and leukotriene C4. PAF stimulates the production of other mediators, including TNF, complement, oxygen radicals, prostaglandins, thromboxane, and leukotrienes. There is a strong association (but not a proven etiologic relationship) between neonatal necrotizing enterocolitis and increased concentrations of PAF.

Chemokines Chemokines are the largest family of cytokines and are involved in the activation and recruitment of a wide variety of cell types. Chemokines are 8- to 12-kDa heparin-binding

proteins ranging from 70 to 100 amino acids in length. All chemokines contain four cysteine motifs (forming double bonds) and are classified according to the arrangement of the cysteines at the amino-terminal end. Of the four subgroups (CXC, CC, C, and CX3C), the CC and CXC are the largest (Table 39-2). Chemokine receptors belong to the seven transmembrane–spanning family of G-protein–coupled

TABLE 39–2  H  uman CXC and CC Chemokines and Relative Receptors Chemokines

Systematic Name

Chemokine Receptors

GRO a

CXCL1

CXCR2

GRO b

CXCL2

CXCR2

GRO g

CXCL3

CXCR2

ENA-78

CXCL5

CXCR2

GCP-2

CXCL6

CXCR2, CXCR1

NAP-2

CXCL7

CXCR2

Interleukin-8

CXCL8

CXCR2, CXCR1

Mig

CXCL9

CXCR3

IP-10

CXCL10

CXCR3

I-TAC

CXCL11

CXCR3

SDF-1

CXCL12

CXCR4

BCA-1

CXCL13

CXCR5

MCP-1

CCL2

CCR2

MCP-2

CCL8

CCR3

MCP-3

CCL7

CCR1, CCR2, CCR3

MCP-4

CCL13

CCR2, CCR3

MIP-1a

CCL3

CCR1, CCR5

MIP-1b

CCL4

CCR5

RANTES

CCL5

CCR1, CCR3, CCR5

Eotaxin

CCL11

CCR3

Eotaxin-2

CCL24

CCR3

Eotaxin-3

CCL26

CCR3

LARC

CCL20

CCR6

TECK

CCL25

CCR9

CTACK

CCL27

CCR10

TARC

CCL17

CCR4

MDC

CCL22

CCR4

DC-CK1

CCL18

?

ELC

CCL19

CCR7

SLC

CCL21

CCR7

Adapted from Manzo A et al: Role of chemokines and chemokine receptors in regulating specific leukocyte trafficking in the immune/inflammatory response, Clin Exp Rheumatol 21:501, 2003.

Chapter 39  The Immune System

receptors. Chemokines are constitutively produced in organs where cell attraction is required for maintenance of local homeostasis.12 Most chemokines are inducible by inflammatory cytokines and endotoxin. In inflammatory states, chemokines are responsible for navigation and homing of effector leukocytes. In addition, chemokines induce a wide variety of leukocyte responses, including enzyme release from intracellular stores, oxygen radical formation, shape change through cytoskeletal rearrangement, generation of lipid mediators, and induction of adhesion to endothelium and extracellular matrix proteins. Chemokines are quite diverse in their target cell selectivity. CXC chemokines generally are more selective for neutrophils and T cells, whereas CC chemokines mainly attract monocytes and T lymphocytes. Considerable data suggest that chemotactic factor generation is deficient in newborn infants.

COORDINATED INFLAMMATORY RESPONSE IN NEONATAL SEPSIS In human bacterial sepsis, cytokines are released in a sequential manner, resulting in a cytokine cascade. After a challenge with a low dose of endotoxin, TNFa peaks within 90 minutes. Other proinflammatory cytokines are released shortly afterward, and anti-inflammatory mediators follow in close sequence. The peak in IL-10 production may not occur for hours, however. In general, cytokines are not stored in intracellular compartments; they are synthesized and released in response to an inflammatory stimulus. Regulation of cytokine production occurs at the level of gene transcription. Specific transcription factors (e.g., NF-kB) bind to DNA response elements that either inhibit or promote gene transcription. Proinflammatory and anti-inflammatory cytokines or molecules are produced in response to an inflammatory stimulus. Counterinflammatory molecules include soluble cytokine receptors (resulting from proteolytic cleavage of the extracellular

binding domain), anti-inflammatory cytokines (e.g., IL-10), and cytokine receptor antagonists (e.g., IL-1ra). Antigen-presenting cells (APCs) of the innate immune system (e.g., macrophages, NK cells, neutrophils, mucosal epithelial cells, endothelial cells, and dendritic cells [DCs]) play pivotal roles in the initiation of an inflammatory response to invading pathogens. As depicted in Figure 39-5, the macrophage is activated by endotoxin (LPS) binding to the CD14 receptor (the main LPS binding receptor) by an LPS-binding protein. CD14 also exists in a soluble form that is shed from the macrophage cell surface through the action of serine proteases. Circulating LPS-CD14 complexes can attach to endothelial cells or epithelial cells. Through this mechanism, endothelial cells are activated to produce other cytokines and mediators (e.g., PAF, nitric oxide, and IL-6) that contribute to the proinflammatory response. CD14 is also required for the recognition of other bacterial products, including peptidoglycans and lipoteichoic acid from grampositive bacteria. The formation of the CD14-LPS complex significantly reduces the concentration of LPS needed for activation. CD14 lacks transmembrane and intracellular domains and cannot initiate a cellular response. Another pathway involving TLRs is responsible for cell activation and signal transduction.1 CD14 seems to be able to discriminate between bacterial products and sort their signals to different TLRs. The TLRs are named from Toll, a plasma membrane receptor in Drosophila, which has a cytoplasmic domain homologous to the IL-1 receptor protein. Toll receptors induce signal transduction pathways that lead to the activation of the transcription factor NF-kB. In mammalian species, there are at least 10 TLRs (Table 39-3), which represent type I transmembrane proteins characterized by an extracellular domain, a transmembrane domain, and an intracellular domain. TLRs are pattern recognition receptors that recognize pathogenassociated molecular patterns. Pathogen-associated molecular patterns are shared by many pathogens, but are not

Figure 39–5.  Role of macrophages in

Vasodilation LPS Nitric oxide Bradykinin Prostaglandins

LPS binding protein Cytokines Chemokines

CD14r Activated MØ

PMN

Tissue factor Adhesion molecules

Protease Free radicals

Damaged endothelium

P

M

DIC

Soluble CD14r receptor & LPS

N

Endothelium

775

Cytokines Nitric oxide

mediating the inflammatory response. Lipopolysaccharide (LPS) binding to the CD14 receptor (with or without LPS binding protein) activates the macrophage (MØ) to synthesize and secrete cytokines, chemokines, and other mediators. These soluble factors stimulate polymorphonuclear neutrophils (PMNs) to release proteases and oxygen free radicals that are important for microbial killing, but that also may injure endothelium and cause capillary leak. Binding of soluble LPSCD14 complexes to endothelial cells promotes biosynthesis of other cytokines and small molecular mediators (e.g., nitric oxide, interleukin-6). Endothelial activation also results in upregulation of cell surface adhesive molecules that interact with neutrophil adhesive molecules and promote neutrophil extravasation. DIC, disseminated intravascular coagulation.

776

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–3  R  ole of Toll-like Receptors (TLRs) in Pathogen Recognition and Pathophysiology of Human Disease Toll-like Receptor TLR1

Ligands

Pathogens or Disease State

Signals as a dimer only when combined with TLR2 for all its ligands; recognizes Borrelia burgdorferi OspA; required for adaptive immune response

Lyme disease

Tri-acyl lipopeptides (bacteria, e.g., Mycobacterium tuberculosis) TLR2

Soluble factors (Neisseria meningitidis)

N. meningitidis

Associates with CD11/CD18, CD14, MD-2, TLR1, TLR6, dectin 1; lipoprotein/lipopeptides (various microbial pathogens)

M. tuberculosis

Peptidoglycan

Apoptosis of Schwann cells in leprosy

Lipoteichoic acid Lipoarabinomannan (mycobacteria) Phenol-soluble modulin (Staphylococcus epidermidis) Glycoinositolphospholipids (Trypanosoma cruzi)

Chagas disease

Glycolipids (Treponema maltophilum)

Leptospirosis

Porin (Neisseria) Zymosan (fungi)

Fungal sepsis

Atypical LPS (Leptospira interrogans) Atypical LPS (Porphyromonas gingivalis)

Periodontal disease

HSP70 (host) CMV virions

CMV viremia

Hemagglutinin protein of wild-type measles

Measles

Bacterial fimbriae TLR3

Double-stranded RNA in viruses

Many

TLR4

Gram-negative enteric LPS (requires coreceptors MD-2 and CD14)

Gram-negative bacteria

Additional ligands

Septic shock

Chlamydial HSP60 RSV F protein

Chlamydia trachomatis, Chlamydia pneumoniae Certain viruses (e.g., RSV)

Taxol (plant) M. tuberculosis HSP65

M. tuberculosis

Envelope proteins (MMTV)

Smallpox (vaccinia) blocks TIR domain of TLR4 and others

HSP60 (host) HSP70 (host) Type III repeat extra domain A of fibronectin (host) Oligosaccharides of hyaluronic acid (host) Polysaccharide fragments of heparan sulfate (host) Fibrinogen (host) b-defensin2 TLR5

Flagellin (monomeric) from bacteria

Flagellated bacteria (e.g., Salmonella)

Chapter 39  The Immune System

777

TABLE 39–3  R  ole of Toll-like Receptors (TLRs) in Pathogen Recognition and Pathophysiology of Human Disease—cont’d Toll-like Receptor

Ligands

TLR6

See TLR2 (as dimers with TLR2)

Pathogens or Disease State

Phenol-soluble modulin Di-acyl lipopeptides (mycoplasma) TLR7

Responds to imidazoquinoline antiviral agents (synthetic compounds)

May be useful as adjuvant for cancer therapy

Loxoribine (synthetic compounds)

Viral infections

Bropirimine (synthetic compounds) Endogenous and exogenous ligands unknown Single-stranded RNA TLR8

Imidazoquinoline (synthetic compounds)

Viral infections

Single-stranded RNA TLR9

Bacterial DNA as “CpG” motifs

Viral infections Bacterial and viral infections (e.g., HSV) May be useful as adjuvant for vaccines and cancer therapy HSV type 2

TLR10

Unknown

Unknown

CMV, cytomegalovirus; HSP, heat-shock protein; HSV, herpes simplex virus; LPS, lipopolysaccharide; MMTV, mouse mammary tumor virus; RSV, respiratory syncytial virus; TIR, Toll/interleukin receptor. From Abreu MT, Arditi M: Innate immunity and Toll-like receptors: clinical implications of basic science research, J Pediatr 144:421, 2004.

expressed in host cells. TLR4 is responsible for the recognition of bacterial endotoxin. It is essential for signaling, but requires a small protein (MD-2) to confer responsiveness. Other TLRs bind cell wall products from gram-positive bacteria and fungi. When activated, TLRs initiate a signaling cascade that shares many of the same molecules used by the IL-1 receptor. Activated, macrophages synthesize and secrete the cascade of proinflammatory cytokines, chemokines, and mediators described earlier. Some of these cytokines activate neutrophils to release proteases and free radicals that have the capacity to damage endothelium and promote capillary leak. Upregulation of adhesion molecules on the neutrophils allows them to bind to counter-receptors on the endothelial cells and migrate to sites of inflammation. The proinflammatory cascade is interrupted by the initiation of counter-regulatory mechanisms. Because most patients with life-threatening infections (i.e., systemic inflammatory response syndrome) are not admitted early in the course of their sepsis episode, proinflammatory cytokines are detected in only a subset of patients. Anti-inflammatory cytokines (which appear later in the cascade) are found in most of these infected individuals. The concentration of anti-inflammatory substances (e.g., IL-1ra and soluble receptors) increases substantially with time and has been termed the compensatory antiinflammatory response syndrome. This is a response of the

host to limit the toxicity of proinflammatory substances. Shortly after the onset of an infectious episode, the mononuclear cell becomes refractory and is unable to respond to proinflammatory cytokines. In contrast, the capacity to produce anti-inflammatory substances such as IL-1ra and IL-10 is preserved. This phenomenon has been referred to as monocyte deactivation or immunoparalysis. Although the mechanism responsible for monocyte deactivation is unclear, it probably involves increased production of IL-10. The risk of infection is greatly increased during this period of hyporeactivity.

ACQUIRED IMMUNITY Cell-Mediated and Antibody-Mediated Responses OVERVIEW Lymphocytes constitute almost 20% of blood leukocytes and specialize in the recognition of invading foreign antigens in the context of major histocompatibility antigens. Lymphocytes are of two major types: (1) B lymphocytes that are produced in the bone marrow, mature in secondary lymphoid organs, and subsequently differentiate into antibody secreting plasma cells, and (2) T lymphocytes that mature and differentiate into CD41

778

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

and CD81 subsets in the thymus and subsequently seed the peripheral blood system, including the spleen and the lymph nodes.21,57 T cell functions include helping B cells to make antibody; killing virally infected cells; regulating the level of the immune response; and stimulating the microbial and cytotoxic activity of the immune cells, including macrophages. Each lymphocyte expresses a cell surface receptor that recognizes a particular antigen. In the case of T cells, it is called the T cell receptor (TCR), and in the case of B cells it is called the B cell receptor (BCR). Each lymphocyte is engineered to express a receptor that is specific for only one antigen. In this way, the lymphocyte population as a whole can recognize a wide range of antigens. The antigen receptors are generated during development by a process known as somatic mutation and recombination involving a few germline genes. The antigen receptors used by B cells and T cells are different. The BCR is a surface immunoglobulin, a membranebound form of the antibody that eventually gets secreted. The TCR is generated from a different set of genes that encode only the cell surface receptor. T cells and B cells recognize antigens in different forms: B cells recognize an unmodified antigen molecule, either free in solution or on the surface of other cells. T cells recognize antigen only when it is presented to them in association with molecules encoded by the MHC. The functional consequence of these differences in antigen recognition is that T cells must be presented processed antigens in the form of short peptides by accessory cells such as macrophages or DCs, whereas B cells can directly recognize antigens in tissue fluids. The acquired immune response arises through the process of clonal recognition. An antigen selects the clones of B and T cells that express the cell surface receptors that recognize the antigen. Because the number of different lymphocyte antigen specificities is large, the number of lymphocytes available to recognize each antigen is relatively small—only a few hundred lymphocytes in an adult. In addition, because so few cells are insufficient to eradicate an invading pathogen, the first step toward the generation of a specific immune response involves a rapid expansion of antigen-specific lymphocytes. This step is followed by further differentiation of antigen-specific cells into effector cells. These events underlie the difference between primary and secondary immune responses. During the primary response, the small number of specific cells increases, and the cells undergo differentiation. If the antigen is encountered again (or persists), there is a larger population of specific cells to react with the antigen, and these cells are able to respond more quickly because they have already undergone several steps along their differentiation pathway. These lymphocytes that have been stimulated by antigen (primed) and their progeny either may differentiate fully into effector cells or may form the expanded pool of cells (memory cells) that can respond more efficiently to a future (secondary) challenge with the same antigen.

different polymorphic chains in association with CD3, a complex of polypeptides involved in signaling cellular activation. The antigen-binding portion may consist of an ab chain heterodimer. In humans, the markers CD2 and CD5 are also present on all T cells (Table 39-4). Activated T cells also carry endogenously synthesized MHC class II molecules, although these are absent from resting T cells. Activated T cells may also be induced to express CD25, which forms part of the high-affinity IL-2 receptor and is important in clonal expansion. There are two main subpopulations of T cells, which can be distinguished according to their expression of CD41 or CD81. These molecules act as receptors for class II (CD41) and class I (CD81) MHC molecules and contribute to T cell immune recognition and cellular activation. Most CD41 T cells recognize antigen associated with MHC class II molecules, and these cells act predominantly as T-helper (Th) cells. CD81 T cells recognize antigen-associated MHC class I molecules and are primarily responsible for cytotoxic destruction of virally infected cells. Clones of mature CD41 T cells fall into two major groups, which are functionally defined according to the cytokines they secrete: Th1 cells interact preferentially with mononuclear phagocytes, whereas Th2 cells tend to promote B cell division and differentiation. The balance of activity between these two subsets is related in part to how antigen is presented

TABLE 39–4  C  luster-Designated (CD) Molecules Found on Human T Cells Antigen

Molecular Weight (kDa)

CD1a

49

Expressed on thymocytes

CD1b

45

Expressed on thymocytes

CD1c

43

Expressed on thymocytes

CD2

50

Sheep erythrocyte receptor

CD3

22

Part of T cell antigen receptor complex

CD4

55

MHC class II immune recognition

CD5

67

CD6

120

CD7

40

Possibly IgM Fc receptor

CD8

32

MHC class I immune recognition

CD25

55

Low-affinity interleukin-2 receptor

CD45

180

Expressed on memory T cells (also known as UCHLI)

CD45R

200

Expressed on virgin T cells (also known as Leu-18)

T LYMPHOCYTES Overview T cells develop and differentiate in the thymus before seeding the secondary lymphoid tissues. T cells recognize antigen and MHC molecules through the TCR. This receptor consists of an antigen-binding portion formed by two

Comment

MHC, major histocompatibility complex.

Chapter 39  The Immune System

T Cell Development Lymphocytes destined to become T cells must undergo several maturational steps in the thymus before they become mature effector T cells. In humans, the thymus develops embryologically as an outgrowth from the third and fourth pharyngeal pouches between weeks 6 and 7 of gestation. The cortex and medulla of the thymus begin to differentiate by the 10th week of gestation, and Hassall corpuscles appear the 12th week of gestation. The undifferentiated cells that first enter the thymus (at approximately 7 weeks’ gestation) do not express either the CD41 or the CD81 antigen, but do express the T cell markers CD7 and CD45 (CD7 may be the IgM Fc receptor, and CD45 is the common leukocyte antigen). In the thymus, T cell maturation is accompanied by the sequential appearance of surface phenotypic markers (see Table 39-4). CD1, CD2, and CD5 surface antigens appear soon after CD7

Cell type

Major developmental events

Prothymocyte

Migration into thymus from bone marrow

Type I thymocyte

Proliferation, TCR gene rearrangement

expression, whereas CD3 appears later. As gestation progresses, most cells leaving the thymus express either the CD41 or the CD81 surface antigen. Cells that lack both antigens (double-negative cells) retain stem cell function and possess a receptor (CD25, Tac antigen) for IL-2, which plays an essential role in T cell proliferation. The first thymic cells to express CD41 or CD81 antigens express both antigens simultaneously and appear in the thymic cortex at approximately 10 weeks’ gestation (Fig. 39-6). At this stage, transcribed TCR genes are first expressed on the cell surface. Production of the T cell a chain precedes production of the b chain. By the 12th week of gestation, occasional CD31 cells can be identified in human fetal peripheral blood. As gestation progresses, the percentage of CD31 cells in the peripheral blood increases, coming to represent more than 50% of T lymphocytes by 22 weeks’ gestation. These CD31 cells also express either CD41 or CD81 antigen. By 13 weeks of gestation, CD31 T cells appear in the fetal liver and spleen, and these cells represent more than 50% of the T lymphocytes in those organs by the end of the second trimester.

Role of Cytokine Receptor Signaling in Lymphocyte Development In humans, X-linked SCID (X-SCID) syndrome is characterized by defective T cell development and function. Affected patients lack T cells, but possess near-normal numbers of

CD7

Subcapsular region

CD7

CD4 Type II thymocyte

Selection of the αβ-TCR repertoire

Cortex

TCR/CD3

TCR/CD3 Type III thymocyte

Emigration to periphery

CD4

Thymus

to the cells, and it ultimately determines the type of immune response that develops. The surface phenotypes of T cell populations change during T cell development. In humans, virgin T cells express the cell surface molecule CD45RA, whereas activated cells express the CD45RO isoform and higher levels of adhesion molecules, such as the b1-integrins (CD29). The mechanism resulting in the generation of activated T cells from resting memory T cells is still unclear.

779

CD8 CD7

CD8

Medulla CD7 TCR/CD3

CD4

CD7 CD8

Peripheral CD4+ and CD8+ T cells CD7 CD4+ T Cell

CD7 CD8+ T Cell

Figure 39–6.  Putative stages of human thymocyte development. Prothymocytes from the bone marrow enter the thymus and give rise to three major stages of ab–T cell receptor (ab-TCR) lineage thymocytes. TCR-a and TCR-b chain genes are rearranged during stage I; thymic selection occurs mainly during stage II; and emigration of mature CD41 and CD81 cells occurs in stage III. (From Lewis DB et al: Developmental immunology and role of host defenses in neonatal susceptibility to infection. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 4th ed, Philadelphia, 1995, Saunders, with permission.)

780

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

B cells. B cells from patients with X-SCID tend to secrete predominantly IgM antibody. Altered B cell–mediated antibody production in patients with X-SCID has been attributed to defective T cell help. The mapping of the gene that encodes the common subunit of cytokine receptors, including IL-2, has been linked to the X-SCID locus. Patients with XSCID possess mutations in the common subunit, establishing a correlation between defects in this gene and the cause of X-SCID. Based on these observations, it was predicted that lymphocytes deficient in the expression of the IL-2 receptor would mimic some of the phenotypic abnormalities associated with patients with X-SCID. Deficiency of IL-2 or the IL-2 receptor does not result in any of the defects observed in patients with X-SCID, however. These observations have since been attributed to the presence of the common subunit of multiple cytokine receptors, including the IL-4, IL-7, IL-9, and IL-15 receptors. All these receptors are expressed on T cells, and mutations in the common subunit are likely to affect the function of all these receptors in patients with X-SCID and provide an explanation for the pathophysiologic process of this disease.

T Lymphocyte Activation and Maturation The T cell receptor (TCR) is a multisubunit complex that consists of ab chains noncovalently associated with the invariant CD3-g, CD3-d, CD3-e, and TCR-z chains. T cell activation by APCs results in the activation of protein tyrosine kinases that associate with the CD3 and TCR-z subunits and the coreceptors CD41 or CD81. These activation events result in the transcriptional activation of many genes, resulting in T cell proliferation, differentiation, and acquisition of effector function. Generally, interaction between the TCR and the antigen presenting cells (APC) results in the activation of intracellular enzymes and adapter molecules that consists of recognition sequences known as the Src homology (SH) 2 and SH3 domains containing signaling proteins that belong to the Src family of lymphocyte-specific kinase (LCK).14 Activated LCK associates intracellularly with the TCR and with the CD81 and CD41 coreceptors. Activated LCK also activates the cytoplasmic domain of CD3-z, which contains a conserved amino acid motif, ITAM. ITAM allows the binding of additional signaling molecules such as SYK family kinase ZAP70. When recruited to ITAM, ZAP70 is activated, which subsequently activates a major adapter protein in T cell signaling termed linker for activation of T cells (LAT). Activation of LAT and several other signaling molecules in this cascade results in the activation of nuclear proteins, including transcription factors such as nuclear factor of activated T cells and mitogenactivated protein kinase.14 Taken together, the successive activation of a series of intracellular signaling molecules on TCR engagement with MHC molecules on APCs results in the transcription of important T cell–specific genes that are crucial for development, differentiation, and function of T cells. T cells require two signals for activation: a signal provided through the TCR, and a second, costimulatory signal. The CD28 molecule on T cells binds to B7-1 or B7-2 on APCs, providing the second of the two signals needed for initiation of naive T cell responses. A key feature of the costimulatory signal provided by CD28 is that, in conjunction with a TCR

stimulus, it allows high-level IL-2 production and provides an essential survival signal for T cells. The combined costimulatory signal prevents T cell apoptosis or the induction of anergy (unresponsiveness) that may occur in response to activation of either signal alone. Naive helper T cells can be divided into two subsets, based on the profile of cytokines they produce. Th1 cells produce IL-2, IFN, and TNF, and Th2 cells produce IL-4, IL-5, IL-6, IL-10, 1L-12, and IL-13. This differential pattern of cytokine expression contributes to differences in the function of these two subsets of T cells. Th1 cells are inflammatory cells responsible for mediating cell-mediated immunity. In contrast, Th2 cells act to help B cells produce antibodies. Naive T cells generally do not express mRNA for cytokines such as IFN, IL-4, TNF, or perforin. The decision of a naive T cell to mature into either Th1 or Th2 is partly regulated by the interaction between growth factors and growth factor receptors and transcription factors. For Th1 differentiation, interaction between naive Th cells and APCs results in the synthesis of IL-12. IL-12 in conjunction with transcription factors signals transducer and activator of transcription (STAT4), and T-bet drives the maturation of IFN-producing Th1 cells.41 Likewise, IL-6, STAT6, and the transcription factor GATA-3 are essential for the generation of Th2 cells.41 A balance between cytokines, signaling molecules, and transcription factors is essential for driving the differentiation of Th1 and Th2 cells from naive Th cells.

Role of Cytokines in T Cell Function IL-2 was originally described as a T cell growth factor, but it is now known to have activity on various other cell types, including NK cells, B cells, macrophages, and monocytes. The synthesis and secretion of this cytokine are triggered by the activation of mature T cells. The binding of IL-2 initiates clonal expansion of activated T cells. High-affinity IL-2 receptor expression is induced on B cells by exposure to IL-4 and immunoglobulin receptor binding. The activity of IL-2 on other cell types (NK cells, macrophages, neutrophils, and lymphokine-activated killer cells) is mainly through the intermediate-affinity IL-2 receptor. A key role of the IL-2 receptor g chain (which is also found in the receptors for IL-4, IL-7, IL-9, and IL-13) in the immune response is highlighted by the consequences of its genetic malfunction. Mutations in the IL-2 receptor g chain are responsible for X-SCID in humans. IL-4 is produced by a subpopulation of T cells and by mast cells. Its production follows T cell activation or crosslinkage of FcRI receptors on basophils or mast cells. IL-4 affects many cell types, promoting the growth of T cells, B cells, mast cells, myeloid cells, and erythroid cell progenitors. It promotes class switching in B cells to IgE and augments IgG1 production. Mice in which the IL-4 gene has been knocked out by targeted gene disruption are unable to make IgE. IL-4-overexpressing mice express very high IgE levels. The IL-4 effect on T cell development is to drive T cell differentiation toward a Th2 type at the expense of a Th1 response. The counterbalancing cytokine is IL-12, which cross-regulates IL-4. IL-4 has also been shown to initiate cytotoxic responses against tumor cells. The pleiotropic activity of IL-4 is reflected in the range of cell types

Chapter 39  The Immune System

that express IL-4 receptor. This high-affinity receptor is found on T cells, B cells, mast cells, myeloid cells, fibroblasts, muscle cells, neuroblasts, stromal cells, endothelial cells, and monocytes. IL-7 signaling promotes cell-cycle entry and proliferation of developing T lymphocytes and B lymphocytes. In thymocytes, IL-7 signaling has been implicated in the induction and maintenance of the antiapoptotic protein Bcl-2 and has been shown to inhibit the expression of the proapoptotic factor Bax. These results suggest that one of the principal functions of IL-7 signaling in developing thymocytes is to promote cell survival. IL-10 has important biologic effects on T cell function. Th0 and Th2 subsets of T cells synthesize IL-10, and production of IL-10 is inhibited by IFN-g. IL-12 is made by B cells, monocytes, and macrophages, and acts synergistically with IL-2 to induce IFN production by T cells and NK cells. It is a key factor in the development of Th1 cells, stimulating their proliferation and differentiation. IL-12 enhances the cytotoxic activity of T cells. As noted earlier, IL-12 is a key cytokine for directing the T cell response to Th1.

T Cell Function in Neonates In general, T cell responses in neonates are defective. Cord blood T cells have reduced ability to proliferate and synthesize cytokines such as IL-2, IFN-g, IL-4, and GM-CSF. The defect in T cell functions in neonates has been attributed to qualitative and quantitative differences.3,47 Studies showing reduced numbers of splenic T cells in neonates support the idea that lack of T cell responses in neonates is due to quantitative reduction in the overall numbers of T cells.47 In addition to quantitative reduction, studies have also suggested that neonate T cells are primarily of the immature phenotype and lack the ability to mount a robust immune response. Additional studies have suggested a greater propensity of neonate naive Th cells to differentiate toward the Th2 phenotype rather than Th1 in the presence of endogenous neonatal APCs. The lack of Th1 differentiation from naive Th cells in neonates could result in impaired cell-mediated immunity. Taken together, studies so far point to multiple factors as contributors to neonatal tolerance (nonresponsiveness). More recent studies have shown that neonatal T cells do not lack the ability to undergo proliferation and cytokine production when stimulated in a manner that does not require signaling through their TCR-CD3 complex. Studies comparing TCR-CD3–independent responses between adult and neonatal T cells have shown no differences in the ability of neonatal and adult T cells to synthesize IL-2 and undergo proliferation. Because efficient T cell responses in vivo require the presence of mature and functional APCs, studies using the ability of endogenous neonatal APCs to sustain proliferation in neonatal T cells showed that providing additional costimulatory signals to neonatal T cells in the context of endogenous APCs corrects the ability of these cells to induce proliferation and IL-2 production to adult levels. Neonatal T cells apparently do not possess intrinsic defects in T cell function; rather, the machinery necessary to provide costimulatory signals to neonatal T cells is defective.

781

Neonatal APCs such as dendritic cells (DCs) have been shown to express low levels of costimulatory molecules, including CD40, CD80, and CD86, compared with adult DCs.10,28,44 The inability of neonatal DCs to deliver adequate levels of costimulatory signals to neonatal T cells is likely to be the cause of neonatal tolerance, in addition to lack of cytokine production, such as IL-12 by APCs and IFN-g by naive Th cells, preventing the differentiation of these cells into Th1 cells. In addition to reduced CD40 signaling, the expression of MHC class II is also reduced on neonatal APCs, and may contribute to the reduced immune response of neonates. As mentioned earlier, neonatal APCs show reduced expression of CD86 and CD40 costimulatory molecules. Treatment of these cells with a combination of IFN and CD40 ligand does not upregulate the expression of CD86 or CD40.16 Neonatal APCs synthesize low levels of proinflammatory cytokines, including TNFa, IL-1b, or IL-12, in response to LPS stimulation (i.e., TLR4 ligand) and in response to other TLR ligands.5,16,19,52 Upregulation of TLR4 and CD14 expression is not observed in neonatal APCs on LPS stimulation, and the expression of My88, a crucial adapter molecule in TLR signaling, is also reduced in neonatal APCs.27,59 TLR4 expression is reduced on APCs derived from premature infants compared with full-term infants.11 It is conceivable that premature infants are more susceptible to infections because of impaired TLR expression. Examples of additional complexity in this process have also been documented. Studies have shown a differential response to TLR ligands on neonate APCs. TNFa synthesis is reduced in neonatal monocytes that are stimulated with LPS, whereas this response seems to be normal when these same cells are stimulated with R-848, a ligand for TLR7 and TLR8.27 Neonatal APCs also show reduced IFN-g responsiveness,33,34 which is associated with defects in phagocytosis. The ability of neonatal monocytes to differentiate into DCs is significantly modulated, as reflected by differences in the morphologic characteristics of cord blood–derived DCs compared with adult DCs.30 Taken together, studies so far implicate significant defects in TLR activation in neonatal monocyte/macrophage APCs. Most studies related to neonatal DCs have been performed using in vitro generated monocyte-derived DCs (mDCs). These cells have several defects relative to adult mDCs. They tend to be immature based on low expression of MHC class II, CD80, and CD40. Even on LPS stimulation, the phenotype of these cells remains immature as reflected by continued lack of upregulation of MHC and costimulatory molecules.26,58 The immaturity of neonatal DCs can be attributed to impaired TLR signaling, a phenomenon also observed in monocyte/macrophage–derived APCs discussed earlier. Additionally, neonatal DCs synthesize reduced levels of IL-12 in response to LPS.15 This defect can be rescued, however, by providing exogenous IFN-g to cultures.15 Consistent with the above-described defects associated with neonatal DCs, these cells also show significant reduction in the priming of allogeneic cord blood–derived T cells relative to adult DCs.15,29,30 Neonatal DCs apparently require additional mechanisms of activation to achieve the activation status of adult DCs.

782

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

When activated, they seem to be relatively competent, however. In addition to monocyte-derived, cord blood–derived DCs, studies have suggested the presence of another group of DCs in cord blood that are characterized on the basis of lack of CD11c (a marker for myeloid DCs) expression and presence of CD123. CD123 is a marker used to identify plasmacytoid DCs (pDCs).45 Cord blood contains a higher frequency of pDCs compared with peripheral blood. An enhanced ratio of pDC to mDC is also observed in cord blood compared with adults. Similar to mDCs, neonatal pDCs also show functional defects. They produce reduced levels of IFN-a/b.7 Although cord blood pDCs respond to R-848 by upregulating the expression of CD40, CD80, CD86, and MHC antigens, the extent of upregulation is less in these cells compared with adult pDCs. Likewise, R-848-stimulated TNFa synthesis in cord blood pDCs is also reduced compared with adult pDCs. Neonatal mDCs and pDCs share similar functional characteristics. In adults, Flt3 ligand treatment of cells results in significant growth of cultured mDCs and pDCs derived from peripheral blood.4 Treatment of hematopoietic progenitors from human fetal tissues and cord blood with Flt3 ligand induces differentiation of these cells into pDCs that synthesize significant amounts of IFN-a/b in response to viral stimulation.39 Flt3 ligand could likely be used clinically to induce the expression of IFN in neonatal pDCs. Neonatal DCs express reduced levels of MHC and costimulatory molecules such as CD80 and CD86. In addition, these cells express less IL-12 compared with adult DCs on activation of TLRs. IFN expression in cord blood pDCs is significantly reduced compared with adult pDCs. Recombinant Flt3 ligand given in vitro to cultures consisting of cord blood pDCs significantly enhances the number of pDCs, which are fully competent in secreting adult amounts of IFN subsequently.54,56 Compared with CD41 neonatal Th cells, the status of CD81 neonatal cytotoxic cells is poorly understood and remains controversial. Compared with neonatal CD41 cells, neonatal CD81 cells have a normal frequency of precursors; they are free of cytokine promoter modifications, such as the hypermethylation that is commonly associated with neonatal CD41 Th cells.55 Neonates show the presence of functional CD81 cytotoxic T cells against congenital human CMV and Rous sarcoma virus. They also show the presence of functional memory CD81 cells against congenital CMV.20,32 These results suggest that under certain conditions of stimulation, despite the absence of adequate CD41 T cell help, long-lived, functional memory CD81 T cells can be generated by an in utero CMV infection. This type of antiviral immunity in neonates persists partly because of persistent and prolonged viral secretion associated with some viral infections, such as CMV infection. This type of infection is likely to result in sustained and continuous stimulation of CD41 and CMV-specific, long-lived memory CD81 T cells in utero. Whether in utero antigen stimulation could be used to develop infant vaccines for protecting neonates from viral infections is an area of active research. In addition to defects in cell-mediated immunity, neonates manifest defects in humoral immunity. Generally,

neonatal antibody responses to viral and bacterial infections are quantitatively and qualitatively different from adults. Although human neonates do mount an antibody response to pathogens, they predominantly produce IgM, and the magnitude of antibody response is significantly less than in adults.

B Cell Responses in Neonates Several distinctions between neonatal and adult splenic B cells have been identified.2 Neonatal B cells express reduced levels of costimulatory molecules CD40, CD80, and CD86, which results in defective T cell responses via CD40 ligand (CD40L) and IL-10. Marginal zone B cells derived from spleens of neonates tend to express reduced levels of CD21.51 In addition, expression of a critical costimulatory receptor, TACI, is impaired in neonatal B cells, in particular in premature infants.22 Although these cell intrinsic factors contribute significantly to defects in antibody responses in neonates, B cell responses in neonates are also affected by external factors. Maternal derived antibodies can bind to vaccine antigens and inhibit neonatal B cells from recognizing them.42 Additionally, serum complement levels, in particular, C3 levels, in neonates are reduced; this results in reduction in the formation of antigen-antibody complexes. The number of marginal zone macrophages in neonate spleens is also reduced compared with adults.51 These cells play an essential role in trapping antigens. One also observes defects in the maturation of follicular DCs in neonates. Follicular DCs are crucial for attracting B cells and provide signals that result in somatic hypermutation and class switching of antibodies. Additional defects include a lack of prolonged antibody response; this is due partly to impairment in the establishment and maintenance of antibody-secreting plasma cells in the bone marrow of neonates.37 Although plasma cells in neonates do migrate to the bone marrow, they are impaired in their ability to thrive as long-lasting cells largely because of lack of survival and differentiation signals from bone marrow–derived stromal elements.38 Finally, studies have also shown that in some instances administration of vaccines to neonates preferentially leads to the generation of memory B cells, rather than plasma cells. This situation has been largely attributed to reduced overall B cell receptor affinity of neonatal naive B cells for antigens resulting in an overall reduction in the strength of the intracellular signal, which may preferentially lead to the generation of memory B cells rather than plasma cells. In neonates, a significant number of B cell intrinsic factors and microenvironmental factors cooperate to regulate the development and maintenance of antibody-producing plasma cells and the preferential generation of memory B cells.43

Summary Neonatal T cells are impaired in producing cytokines, notably IL-2 and Th1-type IFN-g, under conditions of physiologic stimulation, including in response to TCR-CD3 stimulation. Neonatal T cells are not intrinsically incapable of mature function, however. Adult-level cytokine production can be elicited in human neonatal T cells by increasing the magnitude of Th1-promoting costimulatory signals. In contrast,

Chapter 39  The Immune System

neonatal cytotoxic T lymphocytes are capable of generating long-lasting memory effectors against several viral infections, including CMV and Rous sarcoma virus. These findings could be exploited in the future to generate vaccines for the treatment of viral infections. Qualitative differences in neonatal T cells and APCs compared with adult cells might contribute to these deficient T cell–mediated responses of neonates. Compared with adult T cells, neonatal T cells seem to require more costimulation to achieve robust Th1 responses in vitro and in vivo. Neonatal APCs might be poorly functional in vivo and normally unable to promote vigorous Th1 responses. If APC function is augmented or supplemented, however, mature Th1 re­s­ponses are promoted.

B LYMPHOCYTES AND ANTIBODY PRODUCTION Overview of B Cell Function and Phenotypic Appearance B lymphocytes represent 5% to 15% of circulating lymphocytes in the peripheral blood and are characterized by the presence of cell surface immunoglobulin.21 Most B lymphocytes simultaneously express endogenously synthesized IgM and IgD and various other cell surface receptors (Table 39-5). Relatively few cells express cell surface IgG, IgA, or IgE. Cell surface immunoglobulin is similar in structure to secreted

783

antibody and consists of four polypeptide chains (two identical heavy chains and two identical light chains) joined by disulfide bonds. Surface immunoglobulin is inserted into the lymphocyte membrane at the constant region of the immunoglobulin molecule. Immune responses to foreign antigens can be classified as primary or secondary (Table 39-6). The primary immune response results in an increase in the titer of antibody, which plateaus and then is catabolized. In the secondary immune response, the antibody titer is usually greater, appears more quickly, and consists almost entirely of IgG antibody (versus IgM in the primary immune response). Most important, after the primary response, the host acquires an immunologic memory of that foreign antigen by expanding the population of antigen-specific T cells and B cells. Memory B cells are prone to making IgG earlier and exhibit higher affinity antigen receptors. Although a few foreign antigens are considered T cell independent (i.e., they do not require the help of T cells), the antibody response to most antigens requires the coordinated response of T cells, B cells, and APCs (B cells, macrophages, and DCs). Two kinds of signals are required to activate B cells: The first signal is provided by the interaction of the foreign antigen with surface immunoglobulin, and the second signal originates from Th cells, which are needed for amplification of the immune response.

B Cell Development

TABLE 39–5  C  luster-Designated (CD) Molecules Found on Human B Cells* Antigen

Molecular Weight (kDa)

Comment

CD5

67

B cell subset marker

CD10

100

Pre-B cell marker

CD19

95

CD20

95

CD21

35

CD22

140

CD23

45

IgE low-affinity receptor on activated B cells

CD25

55

Interleukin-2 receptor (low-affinity) chain

CDw32

40

Fc receptor (FcR11)

CD35

220

Complement receptor CR1 (C3d receptor)

CD45

180-220

Leukocyte common antigen

CD45R

220, 205

Restricted leukocyte common antigen

Complement receptor CR2 (C3b receptor)

*All human B cells express surface immunoglobulin, and most B cells express class I and class II major histocompatibility antigens.

B cell maturation occurs in two stages. In the first stage, undifferentiated stem cells mature into cells identifiable as B lymphocytes; this is an antigen-independent phase that occurs in the fetal liver and bone marrow in humans. The second stage of lymphoid differentiation is antigen dependent; during this phase, B lymphocytes are transformed into plasma cells. The first recognizable cell in the B cell lineage is the pre-B cell (Fig. 39-7). This cell can be detected in fetal liver by 7 to 8 weeks of gestation and is characterized by cytoplasmic staining for the heavy chain of IgM (m chain). As gestation progresses, pre-B cells can be detected in the fetal bone marrow. Clonal diversity is generated at the pre-B cell stage of development. Intact immunoglobulin genes are formed by the rearrangement of gene segments composing each heavy-chain and light-chain family. Genes encoding the human heavy chain are located on chromosome 14; the k light-chain genes are located on chromosome 2, and the l light-chain genes are located on chromosome 22. The heavy-chain family consists of several hundred variable genes (VH), a smaller number of diversity genes (DH), and six joining genes (JH). The JH genes are linked to constant genes, which encode for the heavy-chain classes. The lightchain genes are similarly constructed. Antibody variable regions are generated from multiple smaller gene segments; the potential number of different antigen combining sites is the product of the number of VH, DH, and JH genes. Antibody diversity is additionally increased by the random addition of nucleotides at the splice site junctions of V, D, and J segments, and by point mutations in variable-region gene segments. Pre-B cells give rise to immature B lymphocytes, which express surface IgM and complement receptors, but no other

784

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–6  Features of Primary and Secondary Antibody Responses Feature

Primary Response

Secondary Response

Lag after immunization

Usually 5-10 days

Usually 1-3 days

Peak response

Smaller

Larger

Antibody isotype

Usually IgM . IgG

Relative increase in IgG and, under certain circumstances, in IgA or IgE

Antibody affinity

Lower average affinity, more variable

Higher average affinity (“affinity maturation”)

Induced by

All immunogens

Only protein antigens

Required immunization

Relatively high doses of antigens, optimally with adjuvants

Low doses of antigens, with adjuvants usually not necessary

From Abbas AK et al, editors: Cellular and molecular immunology, 2nd ed, Philadelphia, 1994, Saunders, p 188.

Stage of maturation

Stem cell

Pre-B cell

Immature B cell

Mature B cell

Activated B cell

Antigen-independent

Role of antigen

Antibody-secreting cell

Antigen-induced

Bone marrow Anatomic site(s)

Pattern of immunoglobulin production

Functional status

Periphery

None

Cytoplasmic µ chain of IgM with low membrane µ (with surrogate light chains)

Membrane IgM (with k or l light chain)

Membrane IgM, IgD

Low rate Ig secretion; heavy chain isotype switching; affinity maturation

High rate Ig secretion, reduced membrane Ig

Precursor

Antigenunresponsive

Sensitive to tolerance induction

Antigenresponsive

Early stage of antibody response

Established primary and secondary antibody responses

Figure 39–7.  Sequence of B lymphocyte maturation and antigen-induced differentiation. Ig, immunoglobulin. (From Abbas

AK et al, editors: Cellular and molecular immunology, 2nd ed. Philadelphia, 1994, Saunders, with permission.)

Chapter 39  The Immune System

immunoglobulin classes. These cells can be detected in the fetal liver at 8 to 9 weeks of gestation. Immature B lymphocytes have a unique functional property; when exposed to an antigen or ligand in the absence of activated T cells, they are rendered tolerant to additional stimulation with the same antigen; this process, called clonal anergy, accounts for the tolerance of B lymphocytes to self-antigens. Immunoglobulin class diversity occurs by a process of isotype switching, during which cells that express surface IgM with a particular specificity generate daughter cells that express another immunoglobulin class (Fig. 39-8). Cells that express other membrane immunoglobulin isotypes (IgG, IgA) can be shown by the 12th week of gestation. These other immunoglobulin classes almost always appear on cells that express membrane IgM concurrently. At a later stage (the mature B cell stage), cells express membrane-bound IgG or IgA in association with membrane IgM and IgD. Cells that express surface IgD are incapable of being deactivated by antigen. In cells that express three heavy-chain classes, all three isotypes exhibit the same specificity and express the same variableregion genes. During the antigen-dependent phase of B lymphocyte differentiation, most B lymphocytes express only a single isotype, and plasma cells do not express surface immunoglobulin (see Fig. 39-7). By the 15th week of gestation, a normal fetus has levels of circulating B lymphocytes that are equal to or higher than the levels of adults. Fetal B lymphocytes can be shown in highest proportions in the spleen (30%), blood (35%), and lymph nodes (13%).

B Cell Activation by B Cell Antigen Receptors and Coreceptors The BCR serves multiple functions on B cells that can differ substantially through various stages of their development. Proliferative expansion and differentiation are triggered

through the pre-BCR and BCR complexes in pre-B cells and mature resting cells. At the opposite extreme, apoptosis is triggered in immature B cells on excessive clustering of newly expressed BCRs as a mechanism to eliminate autoreactive membrane IgM-expressing B cells. Strong BCR ligation in mature B cells has also been shown to inhibit V(D)J recombination at the membrane immunoglobulin locus, whereas weak ligation promotes recombination, presumably to generate higher affinity antibodies. Finally, the BCR serves as a specific receptor to internalize antigen efficiently for processing and peptide presentation to Th cells. The B cell has devised a complex network of BCR signaling cascades to perform such a diverse array of functions.

Role of Cytokines in B Cell Function The cytokines IL-1, IL-2, and IL-4 affect B cell and T cell function. IL-2 stimulates the growth and differentiation of B cells. The ability of IL-4 to induce class switching in B cells has been documented earlier. This section deals in more detail with other cytokines having more specific effects on B cells. IL-5 is produced by activated T cells. On B cells, IL-5 functions as a late-acting B cell differentiation factor, playing a major role in the production of IgA. IL-5 is perhaps better known, however, for its effect on eosinophil differentiation. It not only induces the generation of eosinophils from human bone marrow precursors, but also upregulates expression of CD11b on human eosinophils and activates IgA-induced eosinophil degranulation. T cell production of this cytokine is stimulated by parasitic infections. IL-6 is a pleiotropic cytokine produced by many cell types, including T cells. Its effect on B cells is to promote growth and facilitate maturation of the B cells, causing immunoglobulin secretion. Resting B cells do not express the IL-6 receptor, but are induced to express the IL-6 receptor after activation.

Figure 39–8.  Development of B cell

IgM

IgM + IgD

IgG

IgM

IgM

IgM + IgG + IgD

IgM

IgG IgA STEM CELL

PRE–B CELL

IMMATURE B CELL

IgM + IgA + IgD MEMORY CELL

MATURE B CELL

ANTIGEN–INDEPENDENT PHASE

785

IgA

PLASMA CELL

ANTIGEN–DEPENDENT PHASE

surface immunoglobulin expression. Immunoglobulin class diversity occurs by a process of isotype switching. Immature B lymphocytes express surface IgM, but as cells mature, surface IgM and IgD in association with IgA or IgG are expressed. During the antigendependent phase of B cell maturation, B cells express a single immunoglobulin isotype, and mature plasma cells are devoid of surface immunoglobulin.

786

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Immunoglobulin Structure and Function

IL-7 has been described previously in the T cell section, but this cytokine, which is secreted by stromal cells, has a great effect on the development of progenitor B cells. IL-10 also affects T cells and B cells, acting synergistically with other cytokines on the growth of hematopoietic lineages, including those giving rise to B cells. Antibody depletion of IL-10 in vivo results in a reduced IgM and IgA response with an increase in IgG2 and a depletion of certain B cell subsets. IL-13 is predominantly expressed in activated Th2 cells and regulates human B cell and monocyte activity. It acts as a costimulant with CD40 receptor engagement of human B cells. Interaction between CD40 and CD40 ligand in the presence of IL-13 induces isotype switching and IgE synthesis, as does IL-4. IL-14 is believed to play a role in the development of B cell memory. IL-14 enhances the proliferation of activated B cells and inhibits the synthesis of immunoglobulin. It is produced by follicular DCs and activated T cells. IL-14 expression in reactive lymph nodes suggests that, during a secondary immune response, surface IgD B cells migrating through the lymph node encounter antigen, become activated, and express IL-14 receptor; after binding of IL-14, the increased Bcl-2 expression prevents apoptosis and permits B cell memory development.

Specific antibody may be produced in response to direct microbial exposure or through immunization by an almost infinite spectrum of antigens (e.g., proteins, carbohydrates, bacteria, viruses, fungi, and drugs). Antibodies are synthesized and secreted by B lymphocyte–derived plasma cells that reside in the lymph nodes, spleen, mucosal linings of the gastrointestinal and respiratory tracts, and bone marrow. Antibodies comprise a unique family of glycoproteins called immunoglobulin, which in humans consists of five major classes: IgG, IgA, IgM, IgD, and IgE. The basic structure of the IgG molecule is depicted in Figure 39-9. The IgG molecule is composed of four polypeptide chains—two heavy chains and two light chains—held together by covalent disulfide bonds and noncovalent forces. In a given IgG molecule, the two heavy chains and two light chains have identical amino acid sequences. The different immunoglobulin classes are distinguished by antigenic and amino acid sequence differences in their heavy chains. Some of the immunoglobulin classes are composed of subclasses; IgG has four subclasses— IgG1, IgG2, IgG3, and IgG4—that arise from antigenic differences in the heavy chains. In addition to these differences between the immunoglobulin classes and subclasses, there

N

N

N

VL

N

S

S

S

S

CL S

S

S

S

S

S

Antigen-binding site

S

S

S

S

C C

S

CH1

S

Heavy chain

S

S

VH

S

S

Light chain

Hinge

CH2

CH3

S

S

S

S

S

S

S

S

S

S

S

S

C

C

Fc receptor- and complement-binding sites

Carbohydrate group S

S Disulfide bond

Figure 39–9.  Schematic diagram of an immunoglobulin molecule. In this drawing of an IgG molecule, the antigenbinding sites are formed by the juxtaposition of light-chain (VL) and heavy-chain (VH) variable domains. The locations of complement-binding and Fc receptor-binding sites within the heavy-chain constant (CH) regions are approximations. S-S refers to intrachain and interchain disulfide bonds; N and C refer to amino and carboxyl termini of the polypeptide chains.  (From Abbas AK et al, editors: Cellular and molecular immunology, 2nd ed. Philadelphia, 1994, Saunders, with permission.)

787

Chapter 39  The Immune System

are antigenic and amino acid sequence variations in the amino-terminal portions of the heavy chains and light chains in the so-called variable (Fab) regions. The structural variability in this region permits different antibody molecules to react specifically with different antigens. In contrast, the amino acid sequences of the carboxyl-terminal portions of the heavy (Fc region) and light chains do not vary between molecules of the same immunoglobulin class or subclass; these are called the constant regions. The variable regions of immunoglobulin molecules confer the antigen-binding specificity, and the constant regions differ among the various immunoglobulin classes; these regions’ biologic properties and functions vary as well. Only IgM and IgG activate the complement system through the classic pathway, and only IgG can be actively transported across the placenta. The chemical characteristics and biologic properties of the immunoglobulin classes are summarized in Table 39-7. IgG is the major immunoglobulin in the serum and interstitial fluid and has a long half-life of approximately 21 days. It is responsible for immunity to bacteria (particularly grampositive bacteria), bacterial toxins, and viral agents. IgG antibodies can neutralize viruses and toxins, and facilitate the phagocytosis and destruction of bacteria and other particles to which they are bound. IgG can also activate the complement pathway and amplify the inflammatory response by increasing leukocyte chemotaxis and complement-mediated opsonization. IgA is the second most abundant immunoglobulin in serum, but it is the predominant one in the gastrointestinal and respiratory tracts and in human colostrum and breast milk. In secretions, IgA occurs as a dimer joined by a J chain and bears an additional polypeptide chain, called the secretory component. This moiety endows the molecule with resistance against degradation by proteolytic enzymes. Secretory IgA is uniquely suited for functioning in the secretions of the respiratory and gastrointestinal tracts. IgA provides local mucosal immunity against viruses and limits bacterial overgrowth on mucosal surfaces. It also may limit absorption of antigenic dietary proteins. IgA does not activate the classic

complement pathway, but can activate the alternative pathway. IgM antibodies exist in serum primarily as pentamers joined together by a J chain. IgM provides protection against blood-borne infection. It occurs only in small quantities in interstitial fluids and secretions. IgM antibodies are potent bacterial agglutinins and activate the classic complement pathway. Through activation of the complement system, IgM antibodies cause deposition of C3b on bacterial cell surfaces and facilitate phagocytosis. There are no phagocytosispromoting receptors for IgM. Most serum antibodies to gramnegative bacteria are of the IgM type. Intrauterine and neonatal infections elicit the formation of predominantly IgM antibodies. Because IgM does not cross the placenta, the presence of specific IgM antibody in cord blood to spirochetes, rubella, CMV, or other microorganisms can be taken as reliable evidence of intrauterine infection with these agents. The absence of IgM does not exclude the possibility of congenital intrauterine infection, however. IgE antibodies are present in extremely small quantities in serum and secretions. These antibodies play a major role in allergic reactions of the immediate hypersensitivity type. IgE antibodies bind to the cell membranes of basophils (mast cells) by a receptor for the carboxyl-terminal portion of the heavy chain. Binding of antigen (allergen) to the IgE fixed to the basophil results in the liberation of histamine, leukotrienes, and other pharmacologic mediators of immediate allergic reactions. IgD, which occurs in low concentration in serum, is present on the surface of B lymphocytes. The role of circulating IgD is unclear.

Antibody Production in Fetuses and Neonates The fetus acquires the ability to produce serum immunoglobulins early in gestation. In vitro studies have shown the ability of fetal cells to produce antibody (IgM) by 8 weeks’ gestation. IgG synthesis occur slightly later, and IgA synthesis begins at approximately 30 weeks’ gestation. For numerous reasons (i.e., the sterile environment in utero, the inability of the fetus to respond to certain kinds

TABLE 39–7  Chemical and Biologic Properties of Human Immunoglobulin Classes Variable

IgG

IgA

IgM

IgD

IgE

Heavy chains

g

a

m

d

e

Molecular weight (Da)

150,000

160,000 400,000*

900,000

180,000

190,000

Biologic half-life (days)

21

5

5

2

2

Adult serum concentration (mg/dL)

1000

250

100

2.3

0.01

Placental transfer

1

0

0

0

0

Binds complement (classic pathway)

1

0

1

0

0

Reaginic activity

0

0

0

0

1

Mucosal immunity

1

111

1

0

11

*Secretory IgA. 0-111 indicates increasing ability to protect mucosal surfaces from pathogen invasion.

788

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

of antibody, T cell suppression of B cell differentiation), the fetus makes little antibody before the time of birth. As discussed later, at the time of birth, most of the circulating antibodies are IgG antibodies that have been transported across the placenta from the maternal circulation. Low levels of “fetal” IgM (,10% of adult levels) are present at term gestation and reach adult levels by 1 to 2 years of age. The concentration of IgG decreases postnatally (because of the catabolism of maternal IgG) and reaches a nadir (physiologic hypogammaglobulinemia) at approximately 3 to 4 months of age. Adult concentrations of IgG are reached by 4 to 6 years of age, and adult levels of IgA are attained near puberty (Table 39-8).

Summary The ability to mount cell-mediated or antibody-mediated immune responses to specific antigens is acquired sequentially during the course of embryonic development. Early in gestation, fetuses can respond to certain antigens, whereas other antigens elicit antibody production or cell-mediated immune reactions only after birth. The well-known inability of children younger than 18 months of age to induce antibodies to polysaccharides from Pneumococcus and Haemophilus organisms is a clinically important example of the sequential acquisition of antigen-specific immunocompetence in humans. Antibody responses of fetuses and premature and fullterm newborns differ from responses of children and adults. Fetuses and newborns do not respond to some antigens (e.g., pneumococcal polysaccharide), and the antibody responses to other antigens (e.g., rubella, CMV, Toxoplasma) are predominantly of the IgM type. T lymphocytes are less experienced in neonates and may tend to suppress rather than stimulate B cell differentiation. In addition, B cells of newborns differentiate predominantly into IgM-secreting plasma cells, whereas activated adult B cells produce IgG-secreting

TABLE 39–8  N  ormal Values for Immunoglobulins at Various Ages Age

IgG (mg/dL)

IgA (mg/dL)

IgM (mg/dL)

Newborn

600-1670

0-5

5-15

1-3 mo

218-610

20-53

11-51

4-6 mo

228-636

27-72

25-60

7-9 mo

292-816

27-73

12-124

10-18 mo

383-1070

27-169

28-113

2y

423-1184

35-222

32-131

3y

477-1334

40-251

28-113

4-5 y

540-1500

48-336

20-106

6-8 y

571-1700

52-535

28-112

14 y

570-1570

86-544

33-135

Adult

635-1775

106-668

37-154

From Buckley RH et al: Serum immunoglobulins, I: levels in normal children and in uncomplicated childhood allergy, Pediatrics 41:600, 1968.

and IgA-secreting plasma cells as well. The adult pattern of B cell differentiation develops during the first year of life.

PASSIVE IMMUNITY Passive immunity is the acquisition of specific antibody or sensitized lymphocytes from another individual, and it represents a means by which specific immunity can be acquired without previous exposure to antigen or the mounting of a specific immune response. Such immunity is transient, but nonetheless may provide sufficient antimicrobial protection during a vulnerable period of life. The development of an antibody response to an antigen seen for the first time requires 7 to 14 days. Active antibody-mediated immunity is of little value during the crucial first few days of an infection with a new microorganism. The presence of circulating specific antibody to that organism (i.e., passive immunity) permits the mobilization of multiple host defense mechanisms (e.g., complement system, neutrophils) to eliminate the invading microorganism and limit its colonization. In humans, the major avenues for the acquisition of passive immunity are the transfer of IgG across the placenta and the transfer of secretory IgA through colostrum and breast milk.

Placental Transport of Antibodies Although B lymphocytes are present in a fetus by the end of the first trimester, there is little active fetal immunoglobulin production because this process depends on exposure to antigens. Serum immunoglobulin levels in fetuses are extremely low until 20 to 22 weeks of gestation, at which time an accelerated active transport of IgG across the placenta begins. Only maternal IgG is transported. The specificity of this transport process is due to the presence of specific placental receptors for the heavy chain (Fc region) of the IgG molecule. The transport of IgG is an active placental process, and the neonate’s serum IgG concentration at birth is 5% to 10% higher than that of the mother. Prematurely delivered infants have lower IgG levels than infants delivered at term; infants who are very premature (24 to 26 weeks) may have acquired little maternal antibody before delivery. Infants who are small for gestational age have lower IgG levels than infants who are an appropriate size at any gestational age. The placental dysfunction that reduces the nutrient supply to these poorly growing infants may also limit transfer of IgG. This phenomenon is manifested further by the progressive decrease in IgG transport after 44 weeks of gestation, a period when the placenta is known to become increasingly dysfunctional. Elevated levels of IgM or IgA in cord blood usually indicate that the infant has been exposed to antigen in utero and has synthesized antibody itself. Congenital infections with syphilis and rubella characteristically produce elevation of the cord blood IgM concentration, and specific fetal antibody of the IgM type directed against the infecting agent can be detected. Elevated levels of IgM and IgA also may be found if maternal-to-fetal transplacental bleeding has occurred.

Immunologic Properties of Human Breast Milk Maternal transfer of immunity to the newborn infant is also possible through breast milk.17 Table 39-9 lists specific and nonspecific protective factors that are transferred to the

Chapter 39  The Immune System

789

TABLE 39–9  I mmunologically and Pharmacologically Active Components and Hormones Found in Human Colostrum and Milk Soluble

Cellular

Hormones and Hormone-like Substances

Immunologically specific

Immunologically specific

Epidermal growth factor

Immunoglobulin: secretory IgA (11S), 7S IgA, IgG, IgM

T lymphocytes

Prostaglandins

IgE, IgD, secretory component

B lymphocytes

Relaxin

T cell products

Accessory cells

Neurotensin

Histocompatibility antigens

Neutrophils

Somatostatin

Nonspecific factors

Macrophages

Bombesin

Complement

Epithelial cells

Gonadotropins

Chemotactic factors

Ovarian steroids

Properdin

Thyroid-releasing hormone

Interferon

Thyroid-stimulating hormone

a-fetoprotein

Thyroxine and triiodothyronine

Bifidus factor

Adrenocorticotropic hormone

Antistaphylococcal factors

Corticosteroids

Antiadherence substances

Prolactin

Epidermal growth factor

Erythropoietin

Folate uptake enhancer

Insulin

Antiviral factors Migration inhibition factor Carrier proteins Lactoferrin Transferrin Vitamin B12-binding protein Corticoid-binding protein Enzymes Lysozyme Lipoprotein lipase Leukocyte enzymes From Ogra PL et al: Human breast milk. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 4th ed, Philadelphia, 1995, Saunders, p 114.

neonate by breast milk. Although all immunoglobulin classes can be detected in colostrum, secretory IgA constitutes most of the immunoglobulin in human breast milk. Secretory IgA consists of two “serum” IgA subunits, a J chain and a secretory component, which render it resistant to digestion by trypsin and pepsin and to hydrolysis by gastric acid. There is no evidence that immunoglobulins present in breast milk enter the systemic circulation of the human neonate. The levels of IgA, IgM, and IgG have been studied serially in milk. The IgG concentration is relatively constant during the first 180 days of lactation, whereas IgM and IgA

are highest in colostrum, decrease during the first 5 days of lactation, and remain relatively constant during the next 175 days. Breast milk contains antibodies to a broad spectrum of enteric bacteria and viruses (e.g., poliovirus, echovirus, coxsackievirus), and the antibody titers to these agents decrease in parallel to the decrease in concentrations of IgA in the milk. Most immunoglobulins present in human breast milk are believed to be produced by plasma cells located in the breast itself, and very little milk immunoglobulin is derived from maternal serum immunoglobulins. Milk antibodies are directed predominantly against enteric bacterial and viral

790

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

antigens; the concentration of such antibodies is much higher in colostrum than in maternal serum. The antibody composition of breast milk compensates partly for the deficiency of antibodies directed against enteric antigens in placentally transferred IgG. This unique spectrum of antibody specificity is achieved by the “homing” of B lymphocytes, sensitized in the mother’s gastrointestinal tract, to her mammary glands. B cells stimulated by enteric bacterial or viral antigens in the Peyer patches of the small intestine migrate to the mucosal linings of the lactating mammary gland, where they differentiate into plasma cells and secrete their antibodies. Antigens introduced into the gastrointestinal tract stimulate the development of antibodies in breast milk; however, antibodies to these antigens do not develop in the serum. A mother transfers to her infant through breast milk antibodies specific for microbial agents present in her own gastrointestinal tract. As the neonate is being freshly colonized by primarily maternal flora, these antibodies limit bacterial growth in the gastrointestinal tract and protect against overgrowth. The cellular components of human breast milk consist of macrophages, T and B lymphocytes, neutrophils, and epithelial cells. The cellular content of colostrum or early milk is higher than that of later milk and varies greatly among women. Neutrophils are present in significant numbers early in lactation, and their presence may be related to breast engorgement during the initial days of lactation. Although the function of the breast milk neutrophil is unclear, its presence does not imply infection. Epithelial cells occasionally present in milk may originate from the skin of the nipple. Breast milk macrophages are mononuclear phagocytic cells that constitute approximately 80% of the leukocytes in milk. This is an active phagocytic cell that contains large amounts of intracytoplasmic lipid and IgA; bears cell surface receptors for IgG and C3b; and synthesizes several important host resistance factors, including lysozyme, C3 and C4 complement components, and lactoferrin. Milk macrophages are capable of phagocytizing and killing gram-positive and gram-negative bacteria and apparently interact with the lymphocytes present in breast milk. The host defense factors that the breast milk macrophage synthesizes provide important nonspecific antimicrobial protection for neonates. Lysozyme is capable of lysing the cell walls of many bacteria. This enzyme is synthesized by the milk macrophage, and its concentration in human milk is 300 times that found in cow’s milk. It is stable in an acid environment comparable with that of the gastric contents. Lactoferrin is synthesized by the milk macrophage, and its concentration in breast milk is higher than in any other body fluid. Lactoferrin antimicrobial activity derives from its ability to chelate iron, depriving bacteria of a cofactor important for their growth. The growth of Staphylococcus organisms and E. coli is limited by lactoferrin. The C3 and C4 complement components are also actively synthesized by breast macrophages. The function of these proteins in milk is unclear because there is little IgG and IgM or the early complement components, C1 and C2, which are necessary for activation of C3 to its biologi-

cally active forms. IgA may activate C3 through the alternative pathway; however, it is found in highest concentrations in the breast macrophage itself. This macrophage-associated IgA is not synthesized, but rather is ingested by the macrophage. Viable in situ macrophages have been shown to release this IgA slowly, and this has led to the hypothesis that the breast milk macrophage may represent a vehicle for immunoglobulin transport down the neonate’s gastrointestinal tract. Viable T and B lymphocytes are present in human breast milk. T lymphocytes represent 50% of the milk lymphocytes early in lactation, declining to less than 20% as lactation progresses. The spectrum of responses of breast milk T cells differs from that of peripheral blood T cells from the same donor. Milk T cells are often unresponsive to C. albicans antigen, whereas peripheral blood T cells from the same donor are highly reactive. In contrast, milk T cells respond well to the K1 capsular antigen of E. coli, whereas peripheral blood lymphocytes exhibit minimal or no response to K1. This phenomenon may be related to the previously described homing of lymphocytes sensitized in the gastrointestinal tract to mammary glands. The transfer of cell-mediated immunity from tuberculinsensitive mothers to their breast-fed infants has been reported. If these reports are substantiated, this form of passive transfer of cellular immunity could be of major clinical significance. It is unlikely that intact T lymphocytes are passing from the mother’s milk across the mucous membranes of the infant’s gastrointestinal tract. A soluble T cell growth factor or lymphokine more likely is involved. The B lymphocytes present in milk have IgG, IgA, IgM, and IgD on their surfaces. These cells synthesize IgA almost exclusively, however. The contribution of B cells as passive immune effectors in breast milk is as yet unclear.

Summary Maternal transfer of IgG antibodies across the placenta provides a newborn with a measure of immune protection. Antibodies to viral agents, diphtheria, and tetanus antitoxins, which are usually of the IgG class, are efficiently transported across the placenta and attain protective levels in the fetus. In contrast, antibodies to agents that evoke primarily IgA or IgM antibody responses are transported poorly or not at all, leaving the neonate unprotected against those organisms. An infant cannot be protected against agents to which the mother has not made significant amounts of antibody. Breast milk constituents may interact with the neonate’s immune system in ways other than those already mentioned. A significant increase in the secretory IgA content of nasal and salivary secretions has been observed in breast-fed versus formula-fed newborns during the first few days of life. It is postulated that this increase may reflect the influence of a soluble factor in milk that acts to stimulate the mucosal immune system of breast-fed infants. Factors that enhance IgA synthesis by B cells and promote epithelial cell growth are also secreted by milk macrophages.

Chapter 39  The Immune System

EVALUATION OF HOST DEFENSES IN NEONATES Some disorders of immunologic function may become clinically apparent in the neonatal period or first year of life.46 Because of differences in the developmental status of the newborn’s host defense mechanisms and the lack of vast exposure to antigens, evaluation of the function of host defense mechanisms in the infant is different from the evaluation performed for older children and adults. Infants who have experienced two or more significant bacterial or fungal infections should be suspected of having a defect in host defense mechanisms. Patients with unusual infections (e.g., from Pneumocystis jiroveci [formerly Pneumocystis carinii]) or infections that respond incompletely to therapy

and recur are also suspect. Growth failure, chronic diarrhea, chronic dermatitis, hepatosplenomegaly, and recurrent abscesses commonly occur in infants with immunologic deficiency. Primary immunodeficiency disorders are relatively uncommon, however, and numerous other predisposing conditions should be considered that may be predisposing an infant to multiple infectious episodes. In evaluating a patient for a possible host defense mechanism defect, natural (cellular and humoral components) and acquired (cellular and humoral components) immune mechanisms should be considered and investigated. The evaluation should be divided into initial screening tests and definitive tests that allow one to establish a specific diagnosis (Box 39-2). Screening tests should be obtainable by physicians at all hospitals, but definitive tests may be available only at major medical centers.

BOX 39–2 Tests to Evaluate Neonatal Host Defense Mechanisms NATURAL IMMUNITY Cellular Components n White blood cell count and differential* n Nitroblue tetrazolium (NBT) dye reduction test* n Random mobility and chemotaxis assay n Phagocytosis assay n Quantitative bactericidal assay n Flow cytometric analysis of cell surface receptor expression n Analysis of oxidative metabolism and enzyme activity Humoral Components (Complement) n Total hemolytic complement assay (CH50)* n Determination of C3 and C4 concentrations* n Assay of individual components of classic and alternative pathways n Functional measurement of alternative pathway n Functional assay for C3a and C5a ACQUIRED IMMUNITY B Lymphocytes and Antibody Production n Quantitation of serum IgG, IgA, and IgM* n Measurement of specific antibodies after immunization* n Isoagglutinin titer (anti-A and anti-B)* n Determination of IgE and IgD concentrations n Flow cytometric enumeration and analysis of B cell phenotype n Tests of B cell-to-plasma cell maturation and antibody production in vitro T Lymphocytes n Total lymphocyte count and morphology* n Delayed hypersensitivity skin tests to common antigens* n Chest radiograph for thymic size* n Proliferative responses to mitogens, antigens, and allogeneic cells n Flow cytometric enumeration and analysis of T cell subsets n Cytotoxicity assays n Lymphokine production assays *Considered one of the initial screening tests.

791

792

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

REFERENCES 1. Abreau M, Arditi M: Innate immunity and Toll-like receptors: clinical implications of basic science research, J Pediatr 144:421, 2004. 2. Adkins B et al: Neonatal adaptive immunity comes of age, Nat Rev Immunol 4:553, 2004. 3. Adkins B: T-cell function in newborn mice and humans, Immunol Today 20:330, 1999. 4. Blom B et al: Generation of interferon alpha-producing predendritic cell (Pre-DC)2 from human CD34(1) hematopoietic stem cells, J Exp Med 192:1785, 2000. 5. Chelvarajan RL et al: Defective macrophage function in neonates and its impact on unresponsiveness of neonates to polysaccharide antigens, J Leukoc Biol 75:982, 2004. 6. Colucci F et al: Natural killer cell activation in mice and men: different triggers for similar weapons? Nat Immunol 3:807, 2002. 7. De Wit D et al: Blood plasmacytoid dendritic cell responses to CpG oligodeoxynucleotides are impaired in human newborns, Blood 103:1030, 2004. 8. Douglas S et al: The mononuclear phagocytic, dendritic cell, and natural killer cell systems. In Steihm E et al, editors: Immunologic disorders of infants and children, 5th ed., Philadelphia, Saunders, 2004, p 129. 9. Farag SS et al: Biology and clinical impact of human natural killer cells, Int J Hematol 78:7, 2003. 10. Flamand V et al: CD40 ligation prevents neonatal induction of transplantation tolerance, J Immunol 160:4666, 1998. 11. Forster-Waldl E et al: Monocyte toll-like receptor 4 expression and LPS-induced cytokine production increase during gestational aging, Pediatr Res 58:121, 2005. 12. Gale L, McColl S: Chemokines: extracellular messengers for all occasions? Bioessays 21:17, 1999. 13. Gerdes J et al: Tracheal lavage and plasma fibronectin: relationship to respiratory distress syndrome and development of bronchopulmonary dysplasia, J Pediatr 108:601, 1986. 14. Germain RN: T-cell development and the CD4-CD8 lineage decision, Nat Rev Immunol 2:309, 2002. 15. Goriely S et al: Deficient IL-12(p35) gene expression by dendritic cells derived from neonatal monocytes, J Immunol 166:2141, 2001. 16. Han P et al: Potential immaturity of the T-cell and antigenpresenting cell interaction in cord blood with particular emphasis on the CD40-CD40 ligand costimulatory pathway, Immunology 113:26, 2004. 17. Hanson L et al: The transfer of immunity from mother to child, Ann N Y Acad Sci 987:199, 2003. 18. Harris M et al: Diminished actin polymerization by neutrophils from newborn infants, Pediatr Res 33:27, 1993. 19. Hodge S et al: Cord blood leucocyte expression of functionally significant molecules involved in the regulation of cellular immunity, Scand J Immunol 53:72, 2001. 20. Holt PG: Functionally mature virus-specific CD8(1) T memory cells in congenitally infected newborns: proof of principle for neonatal vaccination? J Clin Invest 111:1645, 2003. 21. Insel R, Looney R: The B-lymphocyte system: fundamental immunology. In Steihm ER et al, editors: Immunologic disorders in infants and children, 5th ed., Philadelphia, 2004, Saunders, p 53.

22. Kaur K et al: Decreased expression of tumor necrosis factor family receptors involved in humoral immune responses in preterm neonates, Blood 110:2948, 2007. 23. Kobayashi K, Flavell R: Shielding the double-edged sword: negative regulation of the innate immune response, J Leukoc Biol 75:428, 2004. 24. Koenig J et al: Diminished soluble and total cellular L-selectin in cord blood is associated with its impaired shedding from activated neutrophils, Pediatr Res 39:616, 1996. 25. Koenig J, Yoder M: Neonatal neutrophils: the good, the bad, the ugly, Clin Perinatol 31:39, 2004. 26. Langrish CL et al: Neonatal dendritic cells are intrinsically biased against Th-1 immune responses, Clin Exp Immunol 128:118, 2002. 27. Levy O et al: Selective impairment of TLR-mediated innate immunity in human newborns: neonatal blood plasma reduces monocyte TNF-alpha induction by bacterial lipopeptides, lipopolysaccharide, and imiquimod, but preserves the response to R-848, J Immunol 173:4627, 2004. 28. Li L et al: Neonatal immunity develops in a transgenic TCR transfer model and reveals a requirement for elevated cell input to achieve organ-specific responses, J Immunol 167:2585, 2001. 29. Liu E et al: Tolerance associated with cord blood transplantation may depend on the state of host dendritic cells, Br J Haematol 126:517, 2004. 30. Liu E et al: Decreased yield, phenotypic expression and function of immature monocyte-derived dendritic cells in cord blood, Br J Haematol 113:240, 2001. 31. Malik A et al: Beyond the complete blood count and C-reactive protein, Arch Pediatr Adolesc Med 157:511, 2003. 32. Marchant A et al: Mature CD8(1) T lymphocyte response to viral infection during fetal life, J Clin Invest 111:1747, 2003. 33. Marodi L: Deficient interferon-gamma receptor-mediated signaling in neonatal macrophages, Acta Paediatr Suppl 91:117, 2002. 34. Marodi L et al: Cytokine receptor signalling in neonatal macrophages: defective STAT-1 phosphorylation in response to stimulation with IFN-gamma, Clin Exp Immunol 126:456, 2001. 35. Newton J et al: Effect of pentoxifylline on developmental changes in neutrophil cell surface mobility and membrane fluidity, J Cell Physiol 140:427, 1989. 36. Pihlgren M et al. Delayed and deficient establishment of the long-term bone marrow plasma cell pool during early life, Eur J Immunol 31:939, 2001. 37. Pihlgren M et al: Reduced ability of neonatal and early-life bone marrow stromal cells to support plasmablast survival, J Immunol 176:165, 2006. 38. Pulendran B et al: Flt3-ligand and granulocyte colony-stimulating factor mobilize distinct human dendritic cell subsets in vivo, J Immunol 165:566, 2000. 39. Schibler K: Mononuclear phagocyte system. In Polin R et al, editors: Fetal and neonatal physiology, 3rd ed., Philadelphia, 2004, Saunders, p 1523. 40. Seder RA, Ahmed R: Similarities and differences in CD41 and CD81 effector and memory T cell generation, Nat Immunol 4:835, 2003. 41. Siegrist CA: Mechanisms by which maternal antibodies influence infant vaccine responses: review of hypotheses and definition of main determinants, Vaccine 21:3406, 2003.

Chapter 39  The Immune System 42. Siegrist CA, Aspinall R: B-cell responses to vaccination at the extremes of age, Nat Rev Immunol 9:185, 2009. 43. Simpson CC et al: Impaired CD40-signalling in Langerhans’ cells from murine neonatal draining lymph nodes: implications for neonatally induced cutaneous tolerance, Clin Exp Immunol 132:201, 2003. 44. Sorg RV et al: Identification of cord blood dendritic cells as an immature CD11c– population, Blood 93:2302, 1999. 45. Steihm E et al: Immunodeficiency disorders: general considerations. In Steihm E et al, editors: Immunologic disorders of infants and children, 5th ed., Philadelphia, 2004, Saunders, p 289. 46. Stockinger B: Neonatal tolerance mysteries solved, Immunol Today 17:249, 1996. 47. Sullivan K, Winkelstein J: Deficiencies of the complement system. In Steihm ER et al, editors: Immunologic disorders of infants and children, 5th ed., Philadelphia, 2004, Saunders, p 652. 48. Suri M et al: Immunotherapy in the prophylaxis and treatment of neonatal sepsis, Curr Opin Pediatr 15:155, 2003. 49. Till J, McCulloch E: A direct measurement of the radiation sensitivity of normal mouse bone marrow cells, Radiat Res 14:213, 1961. 50. Timens W et al: Immaturity of the human splenic marginal zone in infancy: possible contribution to the deficient infant immune response, J Immunol 143:3200, 1989. 51. Upham JW et al: Development of interleukin-12-producing capacity throughout childhood, Infect Immun 70:6583, 2002. 52. van de Wetering J et al: Collectins: players of the innate immune system, Eur J Biochem 271:1229, 2004. 53. Velilla PA et al: Defective antigen-presenting cell function in human neonates, Clin Immunol 121:251, 2006. 54. White GP et al: Differential patterns of methylation of the IFN-gamma promoter at CpG and non-CpG sites underlie differences in IFN-gamma gene expression between human neonatal and adult CD45RO-T cells, J Immunol 168:2820, 2002. 55. Willems F et al: Phenotype and function of neonatal DC, Eur J Immunol 39:26, 2009. 56. Wilson C, Edelmann K: The T-lymphocyte system. In Steihm ER et al, editors: Immunologic disorders in infants and children, 5th ed., Philadelphia, 2004, Saunders, p 20. 57. Wong OH et al: Differential responses of cord and adult bloodderived dendritic cells to dying cells, Immunology 116:13, 2005. 58. Yan SR et al: Role of MyD88 in diminished tumor necrosis factor alpha production by newborn mononuclear cells in response to lipopolysaccharide, Infect Immun 72:1223, 2004. 59. Yang K, Hill H: Neutrophil function disorders: pathophysiology, prevention and therapy, J Pediatr 119:354, 1991. 60. Yoder MC: Embryonic hematopoiesis. In Christensen RD, editor: Hematologic problems of the neonate. Philadelphia, 2000, Saunders, p 3. 61. Yokoyama WM, Plougastel BF: Immune functions encoded by the natural killer gene complex, Nat Rev Immunol 3:304, 2003.

793

PART 2

Postnatal Bacterial Infections Morven S. Edwards

NEONATAL SEPSIS Despite the development of newer, more potent antimicrobial agents, infections are still an important cause of neonatal morbidity and mortality. Part 2 of this chapter focuses on bacterial infections; however, viral, fungal, and parasitic infections must be considered in the differential diagnosis of neonatal sepsis.

Incidence and Mortality The term sepsis neonatorum is used to describe any systemic bacterial infection documented by a positive blood culture in the first month of life. Neonatal sepsis has distinct presentations based on the postnatal age at onset (Table 39-10): Earlyonset sepsis occurs in the first 7 days of life and is acquired by vertical transmission from the mother. It usually has a fulminant onset with multisystem involvement, and has a higher case-fatality rate than late-onset sepsis (see Chapter 23). Lateonset sepsis usually is more insidious, but can have an acute onset. Late, late-onset sepsis occurs after 3 months of life and affects premature infants who are of very low birthweight (VLBW). Late, late-onset sepsis is often caused by Candida species or by commensal organisms such as coagulasenegative staphylococci (CONS), and often is associated with prolonged instrumentation, such as indwelling intravascular lines and endotracheal intubation. The incidence of neonatal sepsis ranges from 1 to 5 cases per 1000 live births. In the preantibiotic era, neonatal sepsis was almost uniformly fatal. Mortality rates decreased dramatically after the introduction of antimicrobial agents and with technological advances in neonatal care. Over the past two decades, the case-fatality rate has declined to approximately 5% to 10%.

Microbiology The bacterial pathogens responsible for sepsis neonatorum tend to shift over time.9 In the United States, gram-positive cocci, including group A Streptococcus, were common pathogens before the introduction of antibiotics, but this predominance shifted to gram-negative enteric bacilli after antimicrobial agents came into common use. In the 1950s and early 1960s, Staphylococcus aureus and Escherichia coli predominated as neonatal pathogens. In the late 1960s, group B Streptococcus (GBS) emerged as a perinatal pathogen, and it continues to be an important organism in newborn infections, as is E. coli. This predominance is reflected in the distribution of bacteria causing early-onset neonatal sepsis in active surveillance conducted over a multistate area during 1998-2000 (Table 39-11).28 In neonates of VLBW with early-onset sepsis, GBS was the most frequent pathogen, followed by E. coli and Haemophilus influenzae.65 An increase in the incidence of E. coli and a decrease in GBS has been noted over the last decade.64

794

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–10  Characteristics of Neonatal Sepsis Early Onset (,7 Days)

Late Onset ($7 Days to 3 Months)

Late, Late Onset (.3 Months)

Intrapartum complications

Often present

Usually absent

Varies

Transmission

Vertical; organism often acquired from mother’s genital tract

Vertical or through postnatal environment

Usually postnatal environment

Clinical manifestations

Fulminant course, multisystem involvement, pneumonia common

Insidious or acute, focal infection, meningitis common

Insidious

Case-fatality rate

5%-20%

5%

Low

TABLE 39–11  Etiologic Agents in Neonatal Sepsis Organism

No. (%)

Bacteria Causing Early-Onset Neonatal Sepsis*

Group B Streptococcus

166 (41)

Escherichia coli

70 (17)

Viridans streptococci

67 (16)

Enterococcus species

16 (4)

Staphylococcus aureus

15 (4)

Group D Streptococcus

12 (3)

Pseudomonas species

9 (2)

Other gram-negative enteric bacilli

16 (4)

Other

37 (9)

Total

408 (100)

Etiologic Agents in Late-Onset Neonatal Sepsis†

Coagulase-negative Staphylococcus

629 (48)

Staphylococcus aureus

103 (8)

Candida albicans

76 (6)

Escherichia coli

64 (5)

Klebsiella

52 (4)

Candida parapsilosis

54 (4)

Enterococcus species

43 (3)

Pseudomonas

35 (3)

Group B Streptococcus

30 (2)

Other bacteria

197 (15)

Other fungi

30 (2)

Total

1313 (100)

*Early onset, ,7 days of age. Adapted from Hyde TB et al: Trends in incidence and antimicrobial resistance of early-onset sepsis: population-based surveillance in San Francisco and Atlanta, Pediatrics 110:690, 2002. † Infants were all very low birthweight (,1500 g). Blood cultures were obtained after 72 hours of life. Adapted from Stoll BJ et al: Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network, Pediatrics 110:285, 2002.

Other pathogens, including CONS, Candida species, and S. aureus, are more commonly etiologic agents in late-onset infection, as shown by data from a multicenter prospective registry in neonates with VLBW (see Table 39-11).66 Community-acquired methicillin-resistant S. aureus (MRSA) strains have emerged as a significant cause of sepsis in neonates hospitalized in the neonatal intensive care unit since birth.27 As history suggests, further changes in the etiologic agents of neonatal sepsis are likely and warrant continued observation.

Transmission Infectious agents can be transmitted to a neonate in many ways. Transplacental transmission is well documented for congenital viral infections, but not for perinatal bacterial infections, with the exceptions of infections caused by Treponema pallidum and Listeria monocytogenes. Ascending intra-amniotic infection followed by aspiration of infected amniotic fluid can result in systemic neonatal infection, especially in the setting of prolonged rupture of membranes (ROM). Approximately 1% to 4% of neonates born to mothers with intra-amniotic infection develop systemic infection. Neonatal infection can also be acquired during vaginal delivery from bacteria colonizing the mother’s lower genital tract. Inadequate hand washing by the nursery staff can promote the spread of microorganisms from an infected to an uninfected infant or from the hands of colonized caregivers to the newborn. The use of instrumentation, including endotracheal tubes, nasogastric feeding tubes, umbilical catheters, central venous catheters, and urinary catheters, also increases the risk of neonatal infection. Early-onset infection is most often transmitted vertically by ascending amniotic fluid infection and by delivery through an infected or colonized birth canal. The pathogens that cause later onset disease can be acquired vertically in the peripartum period or horizontally from fomites in the environment or from colonized caregivers after delivery.

Risk Factors MATERNAL RISK FACTORS Maternal factors can influence the development of systemic bacterial infection in the neonate. (See also Chapter 23.) For reasons that remain unclear, the overall incidence of neonatal

Chapter 39  The Immune System

GBS infections is higher among blacks than in other racial groups.45 Maternal factors such as malnutrition and sexually transmitted diseases can also increase the risk of infection. The rates of prematurity and low birthweight, both of which predispose to neonatal infection, are inversely related to socioeconomic status. Maternal colonization with GBS is well documented as a risk factor for neonatal sepsis. Colonization during the third trimester in an uncomplicated pregnancy carries approximately a 1% risk of infection if acquisition by the neonate is not prevented by intrapartum antibiotic prophylaxis; this risk is increased if colonization is associated with prematurity, maternal fever, or prolonged ROM. Asymptomatic bacteriuria has been associated with premature birth. Colonization with genital mycoplasmas has been associated with low birthweight.

PERIPARTUM RISK FACTORS Some peripartum factors associated with an increased risk of neonatal infection are untreated or incompletely treated focal infections of the mother (including urinary tract, vaginal, or cervical infections) and systemic infections, such as maternal septicemia or maternal fever without a focus. (See also Chapter 23.) Uncomplicated ROM lasting longer than 24 hours carries a 1% risk of neonatal sepsis above the baseline rate of 0.1% to 0.5%. The risk of infection increases fourfold if chorioamnionitis and prolonged ROM coexist. Prematurity and low birthweight are associated with an increased incidence of sepsis. In a populationbased cohort study of neonatal GBS disease, attack rates of 5.99 per 1000, 2.51 per 1000, and 0.89 per 1000 live-born infants were noted for infants with birthweights of less than 1500 g, 1500 to 2500 g, and more than 2500 g.56 Rates of first episode of late-onset infection per 1000 hospital days also decreased with increasing birthweight and gestational age, so that infants with a birthweight of 750 g or less had

795

a fourfold higher rate than infants with a birthweight of 1000 g or more.66 Another peripartum risk factor is the use of fetal scalp electrodes. Deliveries using electronic fetal monitoring can be associated with subsequent development of neonatal scalp abscesses. Cephalhematomas rarely may be complicated by sepsis, meningitis, and osteomyelitis. Perinatal asphyxia, defined as a 5-minute Apgar score of less than 6, in the presence of prolonged ROM has been associated with an increased incidence of sepsis.

NEONATAL RISK FACTORS Although no significant gender difference has been documented for infections acquired in utero, it was noted in the 1960s that male infants had a higher incidence of neonatal sepsis than female infants, possibly related to X-linked immunoregulatory genes. Immaturity of the immune system of the newborn host, as shown in Table 39-12, can also have a role in the predisposition to infection (see Part 1 of this chapter).47 Metabolic disorders can predispose to infection. Among neonates of VLBW with late-onset sepsis, complications of prematurity associated with an increased rate of infection included mechanical ventilation, umbilical vessel catheterization, other vascular catheters, hyperalimentation, and duration of hospital stay.66

OTHER RISK FACTORS It has been suggested that bottle feeding can predispose to infection. Prepared formulas lack several important biologic factors found in colostrum, such as bacterial agglutinins and iron-binding proteins, which have a local gastrointestinal protective effect against gram-negative enteric bacilli. Breast milk also contains immunoglobulins, macrophages, and lymphocytes, all of which play a role in immunologic defense.

TABLE 39–12  Host Responses to Bacterial Infection in Neonates Component

Function

Status in Neonate

Clinical Significance

Complement

Opsonization, chemoattraction

Decreased complement components, especially in preterm infants

Decreased production of chemotactic factors; decreased opsonization of bacteria

Antibody

Opsonization, complement activation

IgG concentration decreased in preterm infants; term infants have higher concentration than adults; IgA absent from secretions

Lack of antibody to specific pathogens results in increased risk of infection; increased risk of mucosal colonization with potential pathogens

Neutrophil

Chemotaxis

Impaired migration; impaired binding to chemotactic factors

Decreased inflammatory response; inability to localize infection

Phagocytosis

Normal with sufficient quantities of opsonin

Bacterial killing

Normal in healthy neonates; diminished in stressed neonates

Monocyte

Chemotaxis

Decreased

Decreased inflammatory response

Phagocytosis

Controversial

Uncertain

Bacterial killing

Controversial

Uncertain

Adapted from Polin RA et al: Neonatal sepsis, Adv Pediatr Infect Dis 7:25, 1992.

796

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Factors that may affect late-onset neonatal infection include prior antimicrobial use, prematurity, a high infant-to-nurse ratio in the neonatal intensive care unit, and the presence of foreign materials such as endotracheal tubes and ventriculoperitoneal shunts. Contaminated parenteral fluids, such as lipid emulsions, also have been associated with systemic infections. Infants who acquire early-onset disease often have at least one major risk factor associated with pregnancy and delivery, such as prolonged ROM, preterm delivery, low birthweight, perinatal asphyxia, or maternal peripartum infection. By contrast, late-onset disease is seldom associated with obstetric complications.

Pathology Histologic findings can be minimal in fulminant cases of neonatal sepsis.58 When findings are present, they often reflect coexisting septic shock. Such findings may include renal medullary hemorrhage, renal cortical necrosis or acute tubular necrosis, adrenal cortical and medullary hemorrhage and necrosis, hepatic necrosis, intraventricular hemorrhage, and periventricular leukomalacia. Evidence of disseminated intravascular coagulation can be observed as well. In late-onset disease, pathologic changes consistent with the particular focal infection can be shown, including meningitis, pneumonia, hepatic abscesses, and arthritis or osteomyelitis. (Specific organ system infections are discussed later in the chapter.)

Diagnosis SYMPTOMS AND SIGNS The signs and symptoms of neonatal sepsis often are nonspecific. The temperature of an infant with sepsis may be elevated, depressed, or normal. Most, but not all, infants with sepsis have respiratory signs, including cyanosis or apnea. Other signs such as feeding difficulties or lethargy are nonspecific and may be subtle or insidious. A high index of suspicion is required to identify and evaluate at-risk infants.

CLINICAL MANIFESTATIONS Bonadio and colleagues10 evaluated prospectively a clinical observation scoring method for febrile infants younger than 2 months to determine its predictive value in distinguishing infectious from noninfectious illness. The variables examined were affect, feeding pattern, level of activity, level of alertness, respiratory status or effort, muscle tone, and peripheral perfusion. Inclusion criteria required that the infants had a rectal temperature of equal to or greater than 38°C ($100.4°F) and had received no antibiotics within the previous 72 hours. Infants underwent a complete sepsis evaluation, including lumbar puncture for blood cell count, glucose and protein determinations, and Gram stain and culture; blood and urine cultures; and urinalysis from a catheterized sample. The mean score for infants with serious bacterial infection (meningitis, sepsis, or urinary tract infection) was significantly higher than the mean score for infants with aseptic meningitis or for infants with negative cultures and normal cerebrospinal fluid (CSF). Affect, peripheral

perfusion, and respiratory status were the variables that best differentiated infected from noninfected febrile infants. Two of the 29 infants with serious bacterial infections had normal observation scores; however, both had abnormal clinical or laboratory findings suggestive of infection. The negative predictive value of this scoring system was 96%. Although observational findings alone cannot replace the physical examination and laboratory studies, they are an important aspect of the evaluation of a febrile neonate. Although medical attention often is sought for young infants with fever, it is not a finding specific for infection. Many noninfectious processes can result in pyrexia, including dehydration, drug withdrawal, and extensive hematomas. Palazzi and associates44 came to the following conclusions pertaining to the presentation of neonatal sepsis: n Temperature

elevation in full-term infants is uncommon. elevation is infrequently associated with systemic infection when only a single elevated temperature occurs. n Temperature elevation that is sustained longer than 1 hour is frequently associated with infection. n Temperature elevation without other signs of infection is infrequent. n The infant’s temperature can be helpful in suggesting an infectious or noninfectious process if the trend of temperatures and the newborn’s clinical status is considered, but a single reading should not be taken alone as an indication of infection. n Temperature

Septic infants can present with neurologic findings such as seizures and full fontanelle even in the absence of meningitis. Gastrointestinal symptoms include hepatomegaly, abdominal distention, vomiting, diarrhea, guaiac-positive stools, and jaundice. Cutaneous manifestations other than jaundice can be present, but are uncommon findings in neonatal sepsis. Focal infections can precede or accompany neonatal sepsis; these can include cellulitis, impetigo, soft tissue abscesses, omphalitis, conjunctivitis, otitis media, meningitis, and osteomyelitis. The presence of certain focal infections can suggest the causative agent, such as streptococci with cellulitis, staphylococci with abscesses, and Pseudomonas aeruginosa with necrotic skin lesions. Complications of neonatal sepsis include metastatic foci of infection, disseminated intravascular coagulation, congestive heart failure, and shock.

DIFFERENTIAL DIAGNOSIS Because the signs and symptoms of neonatal sepsis are nonspecific, noninfectious etiologies should be considered in the differential diagnosis. Sepsis with or without pneumonia can manifest as respiratory distress, transient tachypnea of the newborn, and meconium aspiration. Central nervous system (CNS) symptoms can be caused by sepsis and meningitis, and by intracranial hemorrhage, drug withdrawal, and inborn errors of metabolism. Intestinal obstruction, gastric perforation, and necrotizing enterocolitis can manifest with some of the gastrointestinal signs and symptoms seen with sepsis. Some nonbacterial infections, such as disseminated herpes simplex virus (HSV) infection, can

Chapter 39  The Immune System

be indistinguishable from bacterial sepsis and must be considered in the differential diagnosis.

MICROBIOLOGIC TESTS Cultures A definitive diagnosis of neonatal sepsis can be made only with a positive blood culture. Blood, urine, and CSF should be obtained from all infants suspected to have sepsis. The optimal number of blood cultures and volume of blood per culture have not been established for neonates. Obtaining more than one blood culture can be helpful in distinguishing blood culture contaminants from true pathogens. A minimum of 0.5 mL of blood per bottle is recommended. To avoid contamination, blood should not be obtained from a capillary stick or from umbilical catheters except perhaps immediately after insertion. Several manual and automated methods are available for detecting growth in the blood culture medium. Many new automated radiometric techniques can detect growth 8 hours after collection, and almost always within 24 to 48 hours after collection. The yield from a urine culture is low in early-onset sepsis and most often reflects spread to the bladder in the setting of bacteremia. Urine for culture should be obtained in a sterile manner, such as by catheterization or suprapubic bladder aspiration, and sent for chemical and microscopic analyses and culture before antimicrobial therapy is started for infants with suspected late onset of infection. CSF should be obtained before antibiotic administration and sent for blood cell count, differential, and chemistry determinations, and for Gram stain and culture (see “Meningitis,” later, and Appendix B for normal CSF indexes). Although controversial, some authorities believe that lumbar puncture may be postponed or excluded from the evaluation of an infant with suspected early-onset disease manifested by pneumonia. Meningitis accompanies sepsis, however, in approximately 10% of infants with early-onset disease and more often with late-onset disease. Meningitis cannot be diagnosed or excluded solely on the basis of symptoms, and blood cultures can be sterile in 10% to 15% of infants with earlyonset meningitis and in one third of infants of VLBW with late-onset meningitis.67 Additional cultures should be obtained as indicated by clinical findings. Cultures of tracheal aspirates should be obtained in intubated neonates with a clinical picture suggestive of pneumonia, or when the quality and volume of the secretions change substantially and are consistent with a pneumonic process. Indiscriminate cultures of tracheal secretions can often be difficult to interpret, however. Aspirates or biopsy specimens of skin and soft tissue lesions can be sent for stains and cultures. If a bone or joint infection is suspected, evaluation of aspirated material is invaluable to establishing the diagnosis and determining susceptibility of the infecting pathogen. Stool cultures assist in the diagnosis of neonatal septicemia caused by enteric pathogens such as Shigella, Salmonella, and Campylobacter, but most often the bacteria recovered from stool cultures reflect gastrointestinal colonization, rather than infection. Cultures of gastric aspirates obtained on the first day of life reflect amniotic fluid infection and do not predict the development of neonatal infection.

797

Buffy Coat Examination Leukocyte smears made from the buffy coat layer of centrifuged, anticoagulated blood can be stained with Gram stain and methylene blue or with acridine orange, then examined microscopically for intracellular bacteria. A positive buffy coat smear supports the diagnosis of sepsis and identifies the morphologic and Gram stain characteristics of the organism, but does not identify the infectious agent or include or exclude other foci of infection. Buffy coat examination is used infrequently since the advent of automated blood culture monitoring systems.

ANTIGEN DETECTION ASSAYS Immunoassays that detect bacterial cell wall or capsule carbohydrate antigens in body fluids are an adjunct to diagnosis. Multiple studies have shown that antigen tests are an inappropriate substitute for properly performed bacterial cultures in the diagnosis of neonatal sepsis. With the available radiometric blood culture technology, rapid antigen testing is now infrequently required or indicated. In addition, these tests can provide results that are misleading. The only specimens recommended for testing with these devices are serum and CSF. A positive result should be taken to indicate the presence of antigen and not the presence of viable organisms.

OTHER LABORATORY TESTS Leukocyte Counts Many different aspects of the leukocyte count have been examined for their predictive value in diagnosing sepsis. Leukocytosis and leukopenia, defined as more than 20,000/mm3 (leukocytosis) and less than 5000/mm3 (leukopenia), have proven insensitive and nonspecific. A single leukocyte count obtained shortly after birth is not adequately sensitive for diagnosing sepsis. Manroe and colleagues37 established reference ranges for the absolute total neutrophil count, absolute total immature neutrophil count, and the ratio of immature to total neutrophils (I:T) for neonates during two time periods: the first 60 hours of life and from 60 hours to 28 days of life (Table 39-13). Mouzinho and associates,40 from the same institution, observed that neonates with VLBW (weighing ,1500 g and of #30 weeks’ gestation) often had neutrophil indexes that did not fall within these ranges. Based on the results of combined retrospective and prospective study of infants with VLBW, the values for absolute total immature neutrophil count and I:T ratio were not significantly different from the ranges established by Manroe and colleagues,37 but new absolute total neutrophil count ranges for infants with VLBW were proposed (Table 39-14). These ranges were based on small numbers of “normal, uninfected” neonates with VLBW. The latter reference range has higher specificity than the reference ranges of Manroe and colleagues,37 suggesting that these ranges may be clinically useful in determining length of therapy in infants in whom cultures remain sterile. The total neutrophil count has been examined for its value in predicting the presence of infection. Neutropenia, especially if it occurs in the first hours of life and is associated with respiratory distress, has a strong association with early-onset GBS sepsis. Many noninfectious processes are associated either with neutropenia or with neutrophilia

798

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–13  Reference Ranges for Neutrophil (per mm3) Indexes in Neonates Index

Birth

12 Hours

24 Hours

48 Hours

72 Hours

$120 Hours

ANC

1800-5400

7800-14,400

7200-12,600

4200-9000

1800-7000

1800-5400

INC

#1120

#1440

#1280

,800

,500

,500

I:T

,0.16

,0.16

,0.13

,0.13

,0.13

,0.12

ANC, absolute neutrophil count (mature and immature forms); INC, immature neutrophil count (all neutrophils except segmented ones); I:T, ratio of INC divided by ANC. Adapted from Manroe BL et al: The neonatal blood count in health and disease, I: reference values for neutrophilic cells, J Pediatr 95:89, 1979.

TABLE 39–14  P  roposed Reference Range for Absolute Total Neutrophil Count in Infants with Very Low Birthweight* ABSOLUTE TOTAL NEUTROPHIL COUNT

†

Age

Minimum

Maximum

Birth

500

6000

18 h

2200

14,000

60 h

1100

8800

120 h

1100

5600

*Defined as #1500 g. † Neutrophils/mm3. Adapted from Mouzinho A et al: Revised reference ranges for circulating neutrophils in very-low-birth-weight infants, Pediatrics 94:76, 1994.

(Table 39-15 and Box 39-3).71 Many infants with documented sepsis have normal total neutrophil counts at the time of the initial evaluation. The absolute total immature neutrophil count, defined as the absolute number of all neutrophils excluding the segmented neutrophils, has also been extensively studied. All newborns, but especially premature infants, have a relatively large number of immature neutrophils in the first few days of life. Infected neonates can have an increase above the upper limits of normal in immature cells released from the bone marrow in response to infection, but this response is inconsistent and sometimes delayed, and is an insensitive marker for the early diagnosis of infection. It is unusual, however, for uninfected infants to have an elevated absolute total immature neutrophil count above the reference ranges; if such a finding is present, further evaluation for occult infection should be considered. The I:T ratio has been investigated as an early predictor of sepsis. The maximal I:T ratio in uninfected neonates is 0.16 in the first 24 hours, decreasing to 0.12 by 60 hours. The upper limit of normal for neonates of 32 weeks’ gestation or less is slightly higher, at 0.2. The test has a good negative predictive value, that is, there is a high likelihood that infection is absent if the I:T ratio is normal. Most infected neonates have an elevated I:T ratio some time during the infection, so repeatedly normal I:T ratios can be reassuring. The usefulness of this test is limited because many noninfectious processes, including prolonged induction

with oxytocin, stressful labor, and even prolonged crying, are associated with increased I:T ratios.

Acute-Phase Reactants The acute-phase response is a response of the body to infection or trauma clinically manifested by malaise, anorexia, fever, leukocytosis, negative nitrogen balance, and hepatic production of acute-phase proteins. Acute-phase reactants (APRs) are proteins produced by hepatocytes in response to inflammation. The inflammation can be secondary to infection, trauma, or other processes of cellular destruction. There are many different APRs, including CRP, fibrinogen, a1-acid glycoprotein, a1-antitrypsin, and elastase a1-proteinase inhibitor. These APRs have different plasma half-lives and different incremental responses to inflammation. The method for the detection of APRs has improved with the development of rapid, automated, quantitative specific immunoassays. Numerous studies have evaluated APRs as early indicators of neonatal septicemia; an elevated APR does not distinguish between infectious and noninfectious causes of inflammation.

C-Reactive Protein CRP is a globulin so named because it forms a precipitate in the presence of the C-polysaccharide of Streptococcus pneumoniae. It is believed to be a carrier protein involved in removing potentially toxic material. There is minimal transplacental passage of maternal CRP, and concentrations are unaffected by gestational age. Current methods are fully automated and provide a quantitative assessment of protein concentration. Normal concentrations in neonates are assay-dependent, but the upper limit of normal usually is 1 mg/dL. CRP is significantly elevated at the time of the initial evaluation in 50% to 90% of infants with systemic bacterial infections, but CRP has a low positive predictive value and should not be used alone to diagnose sepsis. An increasing CRP value is usually detectable within 6 to 18 hours, and the peak CRP is seen at 8 to 60 hours after onset of an inflammatory process. The serum half-life is short, so the CRP usually declines to a normal concentration within 5 to 10 days in most infants with infection who have a favorable response to therapy. Serial determinations of CRP are valuable in excluding serious infections. The very high negative predictive value of several normal CRP values in sequence can allow for early discontinuation of empirical antibiotics in selected clinical settings. Nonbacterial infections can elicit a variable CRP response, with normal values in culture-positive viral meningitis and increased values in minor viral infections. Noninfectious processes, including

Chapter 39  The Immune System

799

TABLE 39–15  Clinical Factors Affecting Neutrophil Counts TOTAL IMMATURE NEUTROPHILS

TOTAL NEUTROPHILS

I:T RATIO*

Decreased

Increased

Increased

Increased

Duration (Hours)

Maternal hypertension

1111

0

1

1

72

Maternal fever, healthy neonate

0

11

111

1111

24

$6 hours of intrapartum oxytocin

0

11

11

1111

120

Stressful labor

0

111

1111

1111

24

Asphyxia (5-min Apgar score ,5)

1

11

11

111

24-60

Meconium aspiration syndrome

0

1111

111

11

72

Pneumothorax with uncomplicated respiratory distress syndrome

0

1111

1111

1111

24

Periventricular hemorrhage‡

111

1

11

1111

120

Seizures

0

111

111

1111

24

Prolonged crying ($4 min)

0

1111

1111

1111

1

Asymptomatic hypoglycemia (#30 mg/dL)

0

11

111

111

24

Hemolytic disease

11

11

111

11

7-28 days

Surgery

0

1111

1111

111

24

High altitude

0

1111

1111

0

6¶

†

§

*I:T, immature neutrophils/total neutrophils. † Duration $18 hours, midforceps rotation, breech extraction, 10-minute second stage of labor. ‡ No associated seizures. § Seizures in the absence of hypoglycemia, asphyxia, or central nervous system hemorrhage. ¶ Not tested after 6 hours. 1, 0%-25%; 11, 25%-50%; 111, 50%-75%; 1111, 75%-100%. From Weinberg GA et al: Laboratory aids for diagnosis of neonatal sepsis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infants, 6th ed, Philadelphia, Saunders, 2006, p 1210.

BOX 39–3 Clinical Neonatal Factors Having No Effect on Neutrophil Counts n Race n Sex n Maternal

diabetes mellitus n Route of delivery* n Premature amniorrhexis, mother afebrile n Meconium staining, no lung disease n Uncomplicated respiratory distress syndrome n Uncomplicated transient tachypnea of the newborn n Hyperbilirubinemia, physiologic, unexplained n Phototherapy n Brief (#3 minutes) crying n Diurnal variation *The total neutrophil count of the cord blood of infants delivered vaginally or by cesarean section after labor (2 to 14 hours) is twice that of infants delivered by cesarean section without labor. From Weinberg GA, Powell KR: Laboratory aids for diagnosis of neonatal sepsis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed, Philadelphia, 2006, Saunders, p 1210; data adapted from Manroe BL et al: The neonatal blood count in health and disease, I: reference values for neutrophilic cells, J Pediatr 95:89, 1979.

meconium aspiration pneumonitis, can have an elevated CRP 10 times the normal concentration.

Erythrocyte Sedimentation Rate The erythrocyte sedimentation rate (ESR) is not a direct measure of an APR, but rather reflects changes in many serum protein APRs. A micro-ESR has been developed for use in infants. An approximation of the maximal normal rate in the first 2 weeks of life can be obtained by adding 3 to the age of the newborn in days. Beyond 2 weeks of life, the maximal rate varies between 10 and 20 mm/h. Owing to interlaboratory variation, each laboratory must develop its own reference range. The ESR is limited in that other factors unrelated to inflammation (e.g., anemia, hyperglobulinemia) can affect the rate. Micro-ESR values vary inversely with the hematocrit, but are affected little, if at all, by birthweight or gestational age. Slightly elevated micro-ESR values can occur with superficial infections and with noninfectious processes, including asphyxia, aspiration pneumonia, and respiratory distress syndrome. Markedly elevated values in the absence of infection are unusual, but have been observed with Coombs-positive hemolytic disease and physiologic hyperbilirubinemia. The long delay

800

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

after onset of the inflammatory process before the peak ESR is reached and its long half-life render it of limited usefulness in monitoring the progress of bacterial infections in neonates.

Fibrinogen Plasma fibrinogen concentrations are known to increase in association with infection, although some factors can result in a low fibrinogen level despite severe infection, including disseminated intravascular coagulation, exchange transfusion, and respiratory distress syndrome. Measurement of fibrinogen concentrations is not useful in the early diagnosis of infection because there is a large overlap in values between infected and healthy infants.

Fibronectin Fibronectin is a multifunctional, high-molecular-weight glycoprotein produced primarily by the liver and endothelial cells, and widely distributed in the body, including in plasma and body fluids, on cell surfaces, and in the extracellular matrix. Fibronectin is involved in hemostasis, vascular integrity, and wound healing. It is important in embryogenesis, directing cell migration, proliferation, and differentiation. Fibronectin aids in the immune response by augmenting macrophage and neutrophil phagocytosis and acting as a nonspecific opsonin for the reticuloendothelial system. The plasma fibronectin concentration varies with age. In healthy neonates, it is approximately half that found in adults, whereas healthy premature infants have approximately one third of the amount in normal adults. After birth, the plasma concentration gradually increases, reaching adult values by 2 months of age. Fibronectin has been found to be decreased in neonates with infection and in neonates with asphyxia, respiratory distress syndrome, and bronchopulmonary dysplasia.

Cytokines Cytokines such as IL-1b, IL-6, IL-8, TNFa, and others are endogenous mediators of the immune response to inflammation. There is evidence that measuring cytokine concentrations can be helpful in the early diagnosis of neonatal sepsis, but the study design employed and the method for each assay can affect the reported performance.52 In some reports, such as that of Girardin and colleagues,22 assay performance is excellent. These investigators found that 8 of 9 infants with systemic infection, but only 1 of 60 uninfected infants, had elevated serum TNFa. Confounding effects of maternal complications can affect findings, however. IL-6 and IL-8 can be elevated in infants with sepsis and in the setting of chorioamnionitis. Timing can have an impact on assay performance. Buck and coworkers12 evaluated prospectively the use of IL-6 and CRP measurements in the diagnosis of early-onset sepsis in 222 neonates. Elevated IL-6 concentrations were observed in 73% of newborns with culture-positive infection and in 87% of infants with a clinical diagnosis of sepsis but negative cultures. Most newborns (78%) without evidence of infection had normal IL-6 concentrations. Of the infected newborns with normal IL-6 concentrations, 55% had elevated CRP measurements, suggesting that the peak IL-6 concentration was missed because of its short half-life. Many investigators have evaluated colony-stimulating factors in the neonatal period. Data are conflicting, but suggest

that concentrations of granulocyte colony-stimulating factors vary with gestational age and are influenced by mode of delivery, nutritional status, maternal hypertension, maternal glucocorticoid therapy, and infection.4 A peak in granulocyte colonystimulating factor concentration was observed approximately 7 hours after birth in healthy newborns, with a corresponding increase in the total neutrophil count 7 to 12 hours after birth. Infected newborns had a much higher peak concentration at 7 hours than uninfected infants.

Screening Panels Excluding cultures, none of the previously mentioned tests, when used alone, is sensitive or specific enough to diagnose or exclude neonatal sepsis reliably. Numerous investigators have evaluated the predictive values of panels of tests for diagnosing sepsis. (Table 39-16 presents definitions of terms used in evaluating such tests.) Krediet and colleagues32 measured daily CRP values and I:T ratios in all newborns admitted to the nursery—185 patients during the first 4 days of life and 107 infants after the fourth day of life. A sepsis workup, including cultures, complete blood count, and radiologic studies, was performed as clinically indicated. For early-onset disease, the positive predictive value of either test alone was 18% to 23%, and the negative predictive value was 95% to 98%. When the tests were used together, the positive predictive value increased to 32%, and the negative predictive value decreased to 95%. For late-onset sepsis, the positive predictive value and negative predictive value for the tests, when used alone, ranged from 59% to 63% and 90% to 94%. When the tests were used together to screen for late-onset infection, the positive predictive value increased to 65%, and the negative predictive value decreased to 88%. These researchers concluded that a screening panel comprising CRP and the I:T ratio had limited value. Tegtmeyer and coworkers69 assessed a screening panel in 74 infected neonates. Their panel comprised CRP, I:T ratio, granulocyte count, and Ea1-PI. The sensitivity of each test alone ranged from 36% to 70% except for Ea1-PI, which

TABLE 39–16  T  erms Used for Analyzing Accuracy and Reliability of Tests Term

Definition

Sensitivity

Percentage of patients with infection who have an abnormal test result

Specificity

Percentage of patients without infection who have a normal test result

Positive predictive value

If test result is abnormal, percentage of patients with infection

Negative predictive value

If test result is normal, percentage of patients with no infection

From Weinberg GA et al: Laboratory aids for diagnosis of neonatal sepsis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed. Philadelphia, Saunders, 2006, p 1208.

Chapter 39  The Immune System

ranged from 87% (early-onset sepsis) to 100% (late-onset disease). When all the tests were used together, the sensitivity increased to 100%. Because the authors assessed only infected infants, they were unable to determine the specificity or predictive accuracy of their screening panel. Philip and colleagues46 evaluated the predictive accuracy of a three-part screen (Ea1-PI, I:T ratio, and CRP) in more than 300 infected and uninfected infants admitted to the neonatal intensive care unit. When used alone, Ea1-PI was more sensitive than the other two tests for diagnosing sepsis, but had a lower positive predictive value. When the three tests were used as a panel, the sensitivity was 23%, the specificity was 99.7%, the positive predictive value was 87.6%, and the negative predictive value was 99.7%. The use of screening panels does not significantly improve the positive predictive accuracy; however, the predictive accuracy of a negative panel often increases to 98% to 100%, and a panel with this accuracy of performance in excluding disease can provide useful information.

Treatment EMPIRICAL ANTIMICROBIAL THERAPY Although signs and symptoms of neonatal sepsis are often subtle and nonspecific, when the clinical diagnosis is obvious, the infant is often critically ill. Empirical antimicrobial therapy should be instituted immediately after obtaining samples for culture rather than waiting for the culture results. The choice of therapy should be based on several factors, including the timing and setting of the disease (e.g., early onset, late onset community acquired, health care associated late or very late onset), the microorganisms most frequently encountered, the susceptibility profiles for those organisms, the site of the suspected infection and the penetration of the specific antibiotic to that site, and the safety of the antibiotic. Caregivers should be aware of the susceptibility profiles for the most common neonatal pathogens isolated in their community and in their neonatal unit. Nonbacterial infectious etiologies must be considered in the differential diagnosis and, if deemed a significant possibility, appropriate antiviral or antifungal therapy must be started in addition to antibiotic therapy. For early-onset disease in the United States, the antimicrobial regimen should provide coverage against GBS, E. coli and other gram-negative enteric bacilli, and L. monocytogenes. The combination of ampicillin and an aminoglycoside is frequently used. Listeria and GBS are uniformly susceptible to ampicillin, whereas the susceptibility of E. coli to ampicillin is less reliable. The aminoglycoside provides coverage against most gram-negative enteric bacilli and, with gentamicin specifically, has been found to act synergistically with ampicillin against GBS and Listeria organisms in vitro and in animal models. The choice of aminoglycoside should be based on susceptibility patterns for gram-negative enteric bacilli in the community. Gentamicin is used most frequently, with tobramycin and amikacin reserved for treatment of multidrugresistant bacteria. If meningitis is suspected, especially gramnegative bacillary meningitis, many clinicians add or replace the aminoglycoside with the third-generation cephalosporin, cefotaxime, for better CNS penetration.

801

Empirical therapy for late-onset disease acquired in the community should provide coverage for the same neonatal pathogens discussed earlier and for potential communityacquired pathogens, such as S. pneumoniae and Neisseria meningitidis. Because meningitis frequently is a component of late-onset sepsis, antibiotics with good CNS penetration should be selected. Ampicillin and a third-generation cephalosporin (e.g., cefotaxime) are commonly recommended. Health care–associated late-onset disease can be caused by the usual neonatal pathogens, but also by coagulase negative staphyloccus (CONS), enterococci, gram-negative enteric bacilli (including drug-resistant strains), and fungi. Therapy depends on the presence of risk factors for commensal organisms such as CONS (use of catheters or shunts) and the susceptibility profiles for common nosocomial pathogens isolated from the nursery. Virtually all staphylococci produce penicillinase and are resistant to ampi­cillin and penicillin. Vancomycin and gentamicin are commonly used for initial therapy. Vancomycin generally is active against all staphylococcal species, streptococci, and most enterococci, whereas the spectrum of activity of a penicillinase-resistant penicillin includes streptococci and only the methicillin-susceptible strains of CONS and S. aureus. Cefotaxime or ceftriaxone can be included when gram-negative meningitis is a concern. Cefotaxime does not have activity against L. monocytogenes, enterococci, or Pseudomonas species. Ceftazidime with an aminoglycoside should be used if Pseudomonas infection is suspected. Meropenem can be required when multidrug-resistant enteric gram-negative organisms are isolated. Many other combinations of antimicrobial agents can be effective therapy for nosocomial lateonset sepsis; however, it is prudent to use agents that have proven with experience to be safe. The risks and benefits should be considered thoroughly if routine use of third-generation cephalosporins for lateonset disease is contemplated. One disadvantage is an increased risk for development of gastrointestinal colonization and perhaps subsequent infection with fungi and drugresistant bacteria. With ceftriaxone, in particular, there is a theoretical risk of bilirubin displacement from albumin because of the drug’s high protein-binding capacity. There are no clinical data to substantiate this effect in neonates. These cephalosporins are safe and effective against many of the common neonatal pathogens, and there is extensive experience with cefotaxime use during the neonatal period. Because third-generation cephalosporins do not cause the dose-related toxicities seen with agents such as aminoglycosides, monitoring of serum concentrations is unnecessary. Empirical therapy usually may be narrowed to one drug when final culture identification and susceptibility results are available. The dosage and frequency of administration of antimicrobial agents vary with the newborn’s gestational age, postnatal age, birthweight, and status of hepatic and renal function. Recommendations for antibiotic use in neonates are given in Tables 39-17 through 39-19 and in Appendix A.41 These are guidelines only; specific doses and intervals may change frequently, especially in critically ill infants. Serum concentrations of aminoglycosides and, in some circumstances, vancomycin, should be monitored to ensure therapeutic efficacy while minimizing toxicity. The duration of therapy depends on the site of infection (see later).

802

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–17  R  ecommended Dosage Schedule for Antimicrobial Agents Frequently Used to Treat Neonatal Sepsis DOSAGE (mg/kg/d) AND INTERVALS OF ADMINISTRATION BODY WEIGHT #2000 g

BODY WEIGHT .2000 g

Antibiotic

Route

0-7 Days

.7 Days

0-7 Days

.7 Days

Ampicillin

IV, IM

100 div q12h

150 div q8h

150 div q8h

200 div q6h

Aztreonam

IV, IM

60 div q12h

90 div q8h

90 div q8h

120 div q6h

Cefotaxime

IV, IM

100 div q12h

150 div q8h

100 div q12h

150 div q8h

Ceftazidime

IV, IM

100 div q12h

150 div q8h

150 div q8h

150 div q8h

Ceftriaxone

IV, IM

50 once daily

50 once daily

50 once daily

75 once daily

Clindamycin

IV, IM, PO

10 div q12h

15 div q8h

15 div q8h

20 div q6h

Erythromycin

IV, PO

20 div q12h

30 div q8h

20 div q12h

30 div q8h

Metronidazole

IV, PO

15 div q12h

15 div q12h

15 div q12h

30 div q12h

Mezlocillin

IV, IM

150 div q12h

225 div q8h

150 div q12h

225 div q8h

Nafcillin

IV

50 div q12h

75 div q8h

75 div q8h

150 div q6h

Penicillin G

IV

100,000 U div q12h

200,000 U div q8h

150,000 U div q8h

200,000 U div q6h

Benzathine penicillin G

IM

50,000 U (one dose)

50,000 U (one dose)

50,000 U (one dose)

50,000 U (one dose)

Procaine penicillin G

IM

50,000 U q24h

50,000 U q24h

50,000 U q24h

50,000 U q24h

Ticarcillin

IV, IM

150 div q12h

225 div q8h

225 div q8h

300 div q6h

div, divided. Adapted from Nelson JD: Antibiotic therapy for newborns. In: Pocketbook of pediatric antimicrobial therapy, Baltimore, 1998, Williams & Wilkins.

TABLE 39–18  R  ecommended Dosage Schedule for Aminoglycosides and Vancomycin in the Treatment of Neonatal Sepsis DOSAGE FOR WEEKS’ GESTATION OR POSTCONCEPTIONAL AGE Antibiotic

Route

,30

30-37

.37

IV, IM

15 mg/kg 24h

15 mg/kg q18h

15 mg/kg q12h

15 mg/kg q18h

15 mg/kg q12h

15 mg/kg q8h

3 mg/kg q24h

3 mg/kg q18h

2.5 mg/kg q12h

3 mg/kg q18h

2.5 mg/kg q12h

2.5 mg/kg q8h

20 mg/kg q24h

15 mg/kg q18h

15 mg/kg q12h

20 mg/kg q18h

15 mg/kg q12h

15 mg/kg q8h

Amikacin*

#7 days .7 days Tobramycin†‡

IV, IM

#7 days .7 days Vancomycin

§

#7 days .7 days

IV

*Desired serum concentrations: peak 20-40 mg/mL, trough ,10 mg/mL. † Desired serum concentrations: peak 5-10 mg/mL, trough #2 mg/mL. ‡ Gentamicin dosing: for ,35 weeks’ postmenstrual age, 3 mg/kg/dose q24h; for $35 weeks’ postmenstrual age, 4 mg/kg/dose q24h. § Desired serum concentrations: peak 20-40 mg/mL, trough ,15 mg/mL.

803

Chapter 39  The Immune System

TABLE 39–19  Indications, Pharmacology, and Toxicity of Antibiotics Commonly Used in Newborns Antibiotic

Indications

Pharmacology

Toxicity

Comments

Amikacin

Aerobic gram-negative infections; use should be limited to treatment of gentamicin-resistant organisms

Renal excretion; activity in CSF low; not absorbed from GI tract

Possible ototoxicity, nephrotoxicity, and neuromuscular blockade

Toxicity rare if appropriate dosage is used, and blood concentration is monitored

Ampicillin

Initial treatment of sepsis and meningitis; grampositive organisms except staphylococci; gram-negative organisms if susceptible (Salmonella, Shigella, Haemophilus, Escherichia coli)

Renal excretion

Seizures when high dosages are given

Cefotaxime

Sepsis, meningitis caused by susceptible gramnegative organisms

Primarily renal excretion; good penetration into CSF

Ceftazidime

Can be used in combination with aminoglycoside for treatment of Pseudomonas infection

Renal excretion; penetrates blood-brain barrier

Ceftriaxone

Sepsis, meningitis, soft tissue and bone/joint infections caused by susceptible organisms; not effective against staphylococci, Listeria species, enterococci, or Pseudomonas species

30%-65% excreted by kidneys, the remainder excreted in bile; penetrates bloodbrain barrier

Clindamycin

Treatment of susceptible anaerobic infections

Chloramphenicol

Treatment of infections caused by bacteria resistant to all other antibiotics (e.g., Salmonella species)

Metabolized by liver, small amount excreted unchanged in urine; good penetration through blood-brain barrier

Gray baby syndrome (vasomotor collapse) related to immature hepatic function and associated with elevated concentrations of unconjugated chloramphenicol

Erythromycin

Chlamydia, Pertussis species, minor staphylococcal or streptococcal skin infections

Excreted in urine, stool, and biliary system; crosses blood-brain barrier poorly

No significant toxicity

Active against streptococci; routine use can result in emergence of resistant gramnegative organisms

Potential gallbladder sludging

May displace bilirubin from albuminbinding sites in neonates

Pseudomembranous colitis in older children, but rare in neonates Dose-related reversible bone marrow suppression; idiopathic irreversible aplastic anemia (rare); monitoring of blood concentration mandatory

Continued

804

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–19  Indications, Pharmacology, and Toxicity of Antibiotics Commonly Used in Newborns—cont’d Antibiotic

Indications

Pharmacology

Toxicity

Comments

Gentamicin

Can be used for initial treatment of neonatal sepsis; not effective alone, but can be synergistic when used with ampicillin against group B streptococci, enterococci, and Listeria species

Renal excretion; activity low in CSF; not absorbed from GI tract in normal host

Possible ototoxicity, nephrotoxicity, and neuromuscular blockade

Toxicity rare if appropriate dosage is used, and blood concentrations are monitored

Nafcillin, oxacillin

Penicillin-resistant Staphylococcus aureus infections; active against streptococci, but not a first-line agent

Excretion is renal and hepatic for nafcillin and oxacillin; nafcillin and oxacillin are highly protein bound

Neomycin

Bacterial diarrhea, enteropathogenic E. coli

Not absorbed by GI tract

Ototoxic, nephrotoxic if absorbed

Do not use parenterally

Penicillin G

Most streptococci, Treponema pallidum, Bacteroides species (except Bacteroides fragilis), Neisseria meningitidis

Renal excretion; fair penetration of inflamed meninges

Streptomycin

Mycobacterium tuberculosis

Renal excretion

Vestibular and auditory damage, nephrotoxicity

Must be given IM

Ticarcillin

Expanded gram-negative activity; can be used to treat susceptible Pseudomonas infections

Renal excretion

Platelet dysfunction

Can be associated with hypernatremia and hypokalemia; electrolytes are monitored

Tobramycin

Broad coverage of gramnegative organisms

Renal excretion; low activity in CSF

Possible ototoxicity and nephrotoxicity

Blood concentrations are monitored

Vancomycin

Effective against coagulasenegative staphylococci, methicillin-resistant S. aureus; most grampositive aerobic organisms are susceptible

Renal excretion

Possible ototoxicity; previous preparations associated with nephrotoxicity

Flushing or hypotension may result from rapid infusion

Sulfonamides

Contraindicated in newborns

Displaces bilirubin from albumin-binding sites

Tetracyclines

Contraindicated in newborns

Permanent discoloration of teeth, enamel hypoplasia; inhibits bone growth in premature infants

CSF, cerebrospinal fluid; GI, gastrointestinal.

Can be used to treat infections caused by susceptible organisms

Chapter 39  The Immune System

SUPPORTIVE THERAPY Although appropriate antimicrobial therapy is crucial, supportive care is equally important. Ventilatory support may be necessary, particularly for infants with fulminant early-onset disease. Intravenous hydration and perhaps parenteral nutrition, with close monitoring of electrolytes and glucose, should be considered. Septic shock, if present, should be treated appropriately with fluids and inotropes as indicated by the clinical situation.

IMMUNOTHERAPY The newborn’s immune system is compromised in many aspects, including the neutrophil’s chemotactic response to pathogens, T cell production of proinflammatory cytokines, and functional complement activity (see Part 1 of this chapter). Preterm infants are compromised further by hypogammaglobulinemia because significant transfer of maternal IgG does not begin until 32 to 34 weeks’ gestation. Even term infants can lack specific antibodies to the most common pathogens in early-onset neonatal infections because most adults have low concentrations of antibody against GBS and E. coli. With the increased risk of overwhelming bacterial infection in preterm neonates, researchers have studied the effect of intravenous immune globulin (IVIG) in preventing and treating neonatal infections.

Intravenous Immune Globulin There have been numerous studies evaluating IVIG for treatment of infected neonates. A beneficial effect from IVIG and appropriate antibiotics was observed in several studies compared with antibiotics used alone for the treatment of sepsis. These studies were limited, however, by small numbers of patients, nonblinded investigators, lack of a placebo control, or lack of bacteria-specific analysis of the IVIG preparation used. Meta-analysis of studies of IVIG for the treatment of neonates with sepsis showed a significant decrease in the mortality rate compared with standard therapies.30 Further studies are warranted before IVIG use in infections can be recommended as routine therapy.

Other Agents in Development Other agents that may be beneficial in the treatment of neonatal infections include human monoclonal IgM antibodies and pathogen-specific hyperimmune globulins. An intravenous S. aureus polyclonal immune globulin was well tolerated in VLBW infants, but further investigation is needed to assess efficacy.5 Fibronectin administration may be useful in the prevention and treatment of neonatal sepsis because of its multifunctional roles, including nonspecific opsonization that aids in clearing debris in the reticuloendothelial system, augmentation of phagocytic activity, and hemostasis. Fibronectin has been used in adults as a topical agent to improve wound healing with persistent corneal ulcerations and intravenously in uncontrolled studies of patients with multiorgan system failure. Preliminary results seem encouraging, but large, prospective, controlled studies are required before any recommendations about its use can be made. Early postnatal prophylactic granulocyte-macrophage colony-stimulating factor corrected neutropenia in a singleblind, multicenter, randomized controlled trial in preterm

805

infants, but it did not reduce sepsis or improve survival and short-term outcomes.13

PREVENTION Intrapartum antibiotic prophylaxis (IAP) for prevention of early-onset GBS disease has been in widespread use since 1996. The guidelines issued in 1996 recommended screening of pregnant women for GBS colonization either by lower vaginal and rectal cultures obtained at 35 to 37 weeks’ gestation or by assessing clinical risk factors to identify candidates for IAP. A comparison of the two prevention strategies showed that culture-based screening was 50% more effective than a riskbased strategy in preventing early-onset disease in neonates.55 An 80% reduction in early-onset GBS disease, to approximately 0.3 to 0.4 per 1000 live births, has been achieved in the IAP era.45 The 2002 revised guidelines from the Centers for Disease Control and Prevention (CDC) recommend that all pregnant women be screened during each pregnancy for GBS carriage. Lower vaginal and rectal cultures obtained at 35 to 37 weeks’ gestation should be placed in a non-nutritive transport medium, incubated in selective broth medium, and subcultured into 5% sheep blood agar for isolation of GBS. All pregnant women identified as GBS carriers should receive IAP at the time of labor or ROM. Penicillin (5 million U initially and then 2.5 million U every 4 hours until delivery) is recommended.15 Ampicillin (2 g initially and then 1 g every 4 hours until delivery) is an acceptable alternative. Penicillin-allergic women who are not at high risk for anaphylaxis should receive cefazolin (2 g initially and 1 g every 8 hours until delivery). Vancomycin is recommended for women with a high risk for anaphylaxis. Because of increasing rates of resistance, clindamycin and erythromycin are less effective as alternative agents. These drugs should be reserved for women with risk of anaphylaxis in whom testing has indicated that the isolates are susceptible. Delivery of a previous infant with invasive disease and GBS bacteriuria is always an indication for IAP. IAP is not indicated for planned cesarean section before ROM and onset of labor. Risk factors (labor onset or ROM before 37 weeks’ gestation, ROM $18 hours before delivery, or intrapartum fever) should be used only when the results of cultures are unknown at the onset of labor. An algorithm for IAP for women with threatened preterm delivery is included in the current recommendations.15 The management of infants born to women receiving IAP depends on the infant’s status at birth, the duration of prophylaxis, and the gestational age of the infant. If a woman receives IAP for suspected chorioamnionitis, her infant should have a full diagnostic evaluation and empirical therapy pending culture results based on the infant’s exposure to established infection. Symptomatic infants should undergo full diagnostic evaluation and empirical therapy. A lumbar puncture, if feasible, should be performed. Limited evaluation consisting of complete blood count with differential and blood culture and observation for at least 48 hours is indicated for asymptomatic infants of less than 35 weeks’ gestation and for infants whose mothers received chemoprophylaxis for less than 4 hours before delivery. Observation is appropriate for asymptomatic infants of at least 35 weeks’ gestation whose mothers received chemoprophylaxis at least 4 hours before delivery.

806

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Exposure to antibiotics during pregnancy has not changed the clinical spectrum of GBS disease or the onset of clinical signs of infection within 24 hours of birth for term infants with early-onset GBS infection.11 In reports that document increases in non-GBS sepsis, the increase has occurred only in infants born prematurely or with low birthweight. Surveillance in two cities found stable rates of sepsis caused by other organisms, but an increase in ampicillin-resistant E. coli among preterm but not term infants.28 Intrapartum ampicillin exposure has been shown to be an independent risk factor for ampicillin-resistant, early-onset sepsis.8 These trends indicate the importance of ongoing surveillance. A comprehensive program for prevention of neonatal bacterial infections awaits the development and licensure of vaccines, some of which are now under testing, and of strategies to prevent premature delivery.

INFECTION OF ORGAN SYSTEMS Meningitis INCIDENCE The contemporary incidence of bacterial meningitis in neonates is approximately 0.3 per 1000 live births. Bacterial meningitis can occur in 15% of neonates with bacteremia. For GBS, a higher proportion of late-onset cases than of early-onset cases identified from 2003-2005 manifested as meningitis (27% versus 7%; P ,.001).45 Low birthweight (,2500 g) and preterm birth (,37 weeks of gestation) are associated with an increased risk for neonatal meningitis. Meningitis has a predilection for young age; the incidence is higher in the first month of life than at any other age.

ETIOLOGY The causative agents for neonatal meningitis generally mirror the causative agents associated with neonatal sepsis (see Table 39-11). One review of 101 infants with a gestational age of at least 35 weeks found that GBS accounted for 50%, E. coli accounted for 25%, other enteric gram-negative rods accounted for 8%, and Listeria accounted for 6% of cases. Among the remaining patients, S. pneumoniae, group A Streptococcus, and nontypable H. influenzae each accounted for less than 5% of cases.31 Among premature infants, and especially VLBW infants, Enterococcus, CONS, and a-hemolytic streptococci also must be considered potential pathogens.

PATHOGENESIS There are three mechanisms by which the meninges can become infected: (1) primary sepsis with hematogenous seeding; (2) focal infection outside the CNS, with either secondary bacteremia and resulting hematogenous dissemination or direct extension (e.g., from an infected sinus); and (3) direct inoculation after head trauma or neurosurgery, or from an open congenital defect, such as myelomeningocele or dermal sinus. The most common mechanism in neonates is hematogenous spread associated with a primary bacteremia, so the perinatal risk factors that predispose an infant to neonatal sepsis also influence the risk of acquiring meningitis. Certain microbial factors

are associated with increased risk of meningeal invasion. Despite the presence of more than 100 different capsular polysaccharide K antigens in E. coli, the K1 capsular type is isolated in 70% of neonates with meningitis. Similarly, type III strains are responsible for 70% of all early-onset or late-onset GBS meningitis. The IVb serotype of L. monocytogenes is more frequently isolated in late-onset Listeria meningitis than are the other serotypes.

PATHOLOGY Berman and Banker7 reviewed clinical data from 29 neonates with bacterial meningitis, including 25 with postmortem examinations. All but two of the infections were caused by gram-negative enteric bacilli. The pathologic characteristics closely resembled characteristics found in older infants and children. In the acute stage, brain edema was prominent, but had not led to herniation in any of the infants. A subarachnoid exudate located at the base of the brain or evenly distributed over the cerebral hemispheres or, less commonly, isolated to the convexity was observed. In the acute stage of meningitis, a neutrophilic exudate predominated, which progressed to a mononuclear exudate by the chronic stage. Ventriculitis, resulting from destruction of the epithelial lining of the ventricles after inflammatory cell infiltration, was common. Vasculitis and radiculopathy caused by inflammatory cell infiltration were seen. Parenchymal changes included gliosis, encephalopathy, and infarction. Microglial proliferation was observed in the cerebral and cerebellar hemispheres, in the marginal white matter of the spinal cord and brainstem, and in the ependymal lining of the fourth ventricle and aqueduct. This proliferation sometimes resulted in narrowing or obliteration of the foramen of Luschka and aqueduct of Sylvius. A diffuse encephalopathy of uncertain etiology occasionally was present, believed to be metabolic in origin. Many of the known pathologic sequelae of meningitis were shown at autopsy. Hydrocephalus was observed in 14 of the 25 infants. Communicating hydrocephalus, defined as normal flow of CSF from the lateral ventricles through the fourth ventricle, but obstruction of flow from impaired resorption in the arachnoid villi, was noted in nine infants. Five infants had noncommunicating hydrocephalus, defined as obstructed CSF flow at the level of the aqueduct of Sylvius or the foramina secondary to purulent exudate or gliosis. Hydrocephalus ex vacuo also can be observed when the ventricles are enlarged owing to loss of cerebral cortical tissue and secondary white matter degeneration as a result of encephalopathy; however, CSF flow is not impeded. Infants with ventriculitis can have prolonged symptoms with delayed sterilization of CSF. Thrombosis and cerebral infarcts can develop as a result of the vasculitis. Brain abscesses and subdural effusions were not observed in the series by Berman and Banker.7 When brain abscess formation does occur, it tends to be associated with certain organisms, such as Citrobacter koseri (previously Citrobacter diversus) or Enterobacter sakazakii.

CLINICAL MANIFESTATIONS The early signs and symptoms of neonatal meningitis are nonspecific. The signs most frequently observed include lethargy, reluctance to feed, emesis, respiratory distress, irritability, and temperature instability. Signs suggestive of a CNS

807

Chapter 39  The Immune System

process are less commonly observed in neonates with meningitis than in older children. A review of the clinical signs of bacterial meningitis in 255 newborns from six medical centers found that 40% had convulsions, 28% had a full or bulging fontanelle, and only 15% had nuchal rigidity.44 In preterm infants, the signs and symptoms can be subtle. Because of the high mortality rate and risk of serious sequelae in survivors, and the nonspecificity of signs and symptoms in meningitis, the clinician must have a high index of suspicion when evaluating ill neonates.

DIAGNOSIS The diagnosis of neonatal meningitis requires growth of the microorganism from the CSF. A presumptive diagnosis can be made when the CSF indexes suggest a bacterial process, and a pathogen is isolated from a blood culture. A thorough evaluation of CSF for the presence of infection includes bacterial culture and Gram-stained smear, cell count with leukocyte differential, and glucose and total protein determinations. Normal values for CSF indexes in the neonate are different from normal values in older infants or children and can be difficult to interpret if the age of the patient is not considered. As a guide to interpretation, a CSF leukocyte count exceeding 20 to 30/mm3 is a threshold for the presence of meningeal inflammation and should warrant consideration of meningitis as a possible diagnosis. Sarff and colleagues53 analyzed the CSF findings in 117 neonates with signs and symptoms suggestive of infection, but with sterile CSF cultures and no laboratory evidence consistent with a viral or bacterial process. These neonates had obstetric or neonatal factors that increased their risk for infection, including maternal toxemia, prolonged ROM, chorioamnionitis, maternal fever, unexplained jaundice, and prematurity. The CSF protein and glucose determinations and leukocyte counts in these neonates were elevated compared with older infants and children (Table 39-20). The CSF cell count was observed to decrease during the first week of life in term infants, but to increase in preterm infants. An alteration in blood-brain barrier permeability was cited as a potential mechanism for these findings. Rodriguez and colleagues50 evaluated CSF indexes during the first 12 weeks of life in high-risk infants with VLBW. These infants had sterile CSF bacterial cultures, weighed less than 1500 g at birth with weights appropriate for gestational age, and had no ultrasound evidence of intracranial hemorrhage. The CSF values were analyzed by chronologic age of the infant and by postconceptual age, defined by the authors as gestational age plus the chronologic postnatal age. The CSF glucose and protein determinations were higher in infants 26 to 28 weeks of age than in other age groups (Table 39-21; see also Appendix B). CSF findings in 135 neonates with bacterial infections, including 98 with gram-negative bacillary meningitis, 21 with GBS meningitis, and 16 with septicemia without meningitis, were analyzed by Sarff and colleagues53 using the reference ranges derived from 117 high-risk infants without documented infection. Only 4% of neonates with gramnegative bacillary meningitis had CSF leukocyte counts within the normal range, whereas 23% and 15% had normal protein and normal CSF and blood glucose determinations. E. coli was isolated from a CSF sample with normal cell

TABLE 39–20  C  erebrospinal Fluid Indexes in High-Risk Neonates without Meningitis Term

Preterm*

White Blood Cells (cells/mm ) 3

Mean

8.2

Range

0-32

Percentage PMNs

61.3

9.0 0-29 57.2

Protein (mg/dL)

Mean

90

Range

20-170

115 65-150

Mean

52

50

Range

34-119

24-63

Glucose (mg/dL)

CSF-to-Blood Glucose Ratio (%)

Mean

81

74

Range

44-248

55-105

*Preterm infant defined as 28 to 38 weeks’ gestational age. CSF, cerebrospinal fluid; PMNs, polymorphonuclear neutrophils. Adapted from Sarff LD et al: Cerebrospinal fluid evaluation in neonates: comparison of high-risk infants with and without meningitis, J Pediatr 88:473, 1976.

count, glucose, and protein determinations in one infant. The indexes on repeat CSF evaluation from this patient were consistent with bacterial meningitis. Because CSF may be obtained early in the course of meningitis, before the inflammatory process is evident, normal CSF parameters do not exclude a diagnosis of meningitis. Despite overlap in CSF findings from uninfected neonates and neonates with meningitis, most neonates have at least one abnormal finding on the initial CSF examination. Other tests can be helpful in supporting or establishing a diagnosis of neonatal meningitis. The Gram-stained smear is the most useful. Antigen detection should not be used as a substitute for bacteriologic culture in the diagnosis of GBS meningitis. Results of antigen detection assays should be confirmed by culture. Indications for obtaining CSF in the evaluation of infection are controversial. Some investigators have questioned the usefulness of lumbar puncture in the evaluation of asymptomatic newborns with maternal risk factors. In a retrospective study of 284 asymptomatic infants younger than 7 days of age with a sepsis workup that included lumbar puncture performed in response to maternal risk factors, Fielkow and associates19 found no episodes of meningitis. Many authorities maintain that CSF should be obtained in any neonate with suspected sepsis neonatorum before antibiotics are given. Relying on a positive blood culture to determine when to evaluate CSF can miss 10% to 15% of neonates with meningitis and sterile blood cultures. The diagnosis of meningitis would have been missed or delayed in 16 of the 43 infants

808

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TABLE 39–21  Cerebrospinal Fluid Values in Infants with Very Low Birthweight* by Postconceptional Age† Postconceptional Age in Weeks (No. Infants)

Leukocytes/mm3 6 SD (Range)

PMN (%) 6 SD (Range)

Glucose (mg/dL) 6 SD (Range)

Protein (mg/dL) 6 SD (Range)

26-28 (17)

6 6 10 (0-44)

6 6 13 (0-50)

85 6 39‡ (41-217)

177 6 60‡ (108-370)

29-31 (23)

5 6 4 (0-14)

10 6 19 (0-66)

54 6 18 (33-94)

144 6 40 (84-227)

32-34 (18)

4 6 3 (1-11)

4 6 11 (0-36)

55 6 21 (29-109)

142 6 49 (54-260)

35-37 (8)

6 6 7 (2-22)

5 6 14 (0-40)

56 6 21 (31-90)

109 6 53 (45-187)

38-40 (5)

9 6 9 (0-23)

16 6 23 (0-48)

44 6 10 (32-57)

117 6 33 (67-148)

*,1500 g. † Postconceptional age 5 gestational age 1 chronologic postnatal age. ‡ Group at 26 to 28 weeks had significantly higher glucose and protein levels than the other groups (P , .04). PMN, polymorphonuclear neutrophil; SD, standard deviation. Data from Rodriguez AF et al: Cerebrospinal fluid values in the very low birth weight infant, J Pediatr 116:971, 1990.

reported by Wiswell and associates73 if lumbar puncture had been omitted as part of the early neonatal sepsis evaluation. In a multicenter study of infants with VLBW, one third of 134 infants with late-onset meningitis had negative blood cultures.66 Compared with noninfected infants, infants with meningitis had extended requirement for mechanical ventilation and hospitalization, and were more likely to have seizures and to die. These data support the need to include a lumbar puncture in the diagnostic evaluation of infants with VLBW with suspected late-onset infection. Lumbar puncture can be delayed until a critically ill infant’s status has stabilized or until correction of a coagulopathy, but omitting lumbar puncture as part of the evaluation would cause the diagnosis of meningitis to be delayed or missed completely in some infants.

TREATMENT When initiating antimicrobial therapy for neonatal meningitis, potential pathogens and their susceptibility patterns and the achievable concentrations of the drugs in CSF should be considered. In the United States, empirical therapy should provide coverage for GBS, gram-negative bacillary enteric organisms, and L. monocytogenes. Ampicillin reliably provides coverage for GBS and Listeria organisms, but its activity against gram-negative enteric organisms is variable. When used for meningitis, ampicillin must be given in a higher dosage than that used for infections outside the nervous system to achieve adequate CSF concentrations. Although gentamicin is active against most gram-negative enteric organisms, aminoglycosides in general do not achieve satisfactory activity in CSF and are rarely used as monotherapy for treatment of meningitis. When gram-negative bacillary meningitis is suspected, many clinicians favor using ampicillin, gentamicin, and cefotaxime pending final culture and susceptibility reports. Therapy for hospitalized neonates with late-onset meningitis should include coverage against nosocomial pathogens. Vancomycin and an aminoglycoside are frequently chosen to provide activity against CONS in addition to GBS and enterococci. If GBS is suspected, ampicillin should be included in the regimen. If enteric gram-negative organisms are suspected, cefotaxime should be added to the

regimen. Antimicrobial therapy often can be simplified when the etiologic agent and its susceptibility profile have been determined. The duration of therapy for culture-proven bacterial meningitis depends on the etiologic agent recovered and the infant’s clinical response. The minimal duration in uncomplicated GBS or Listeria meningitis is 14 days. Gram-negative bacillary organisms are more difficult to eradicate from CSF. These infants often have persistently positive CSF cultures 3 to 4 days after the initiation of appropriate antimicrobial agents. In an uncomplicated case, the duration of therapy should be a minimum of 3 weeks, or 2 weeks after documented sterilization of CSF, whichever is longer. Examination of CSF, Gram-stained smears, and culture should be repeated routinely at 24 to 48 hours after initiation of therapy to document CSF sterilization and again near completion of therapy to document adequacy of treatment. Supportive care of a newborn with meningitis is similar to that described for sepsis. In addition, treatment of seizures and inappropriate secretion of antidiuretic hormone may be necessary (see Chapter 40, Part 5). Neuroimaging is useful to assess ventricular size and to assist in delineating potential intracranial complications of meningitis.

PROGNOSIS AND OUTCOME The mortality rate of neonatal meningitis has declined from almost 50% in the 1970s to contemporary rates of less than 10%. Survivors are at a high risk for neurologic sequelae.26 A poor early outcome after GBS meningitis correlates with coma or semicoma, need for pressor support, total peripheral leukocyte count less than 5000/mm3, absolute neutrophil count less than 1000/mm3, and CSF protein greater than 300 mg/dL at the time of presentation.33 Seizures at presentation, the burden of bacteria observed on Gram stain, and the severity of hypoglycorrhachia on the initial CSF sample are not predictive of sequelae at hospital discharge. Factors associated with poor outcome in gram-negative bacillary meningitis include CSF protein level greater than 500 mg/dL, CSF leukocyte count greater than 10,000/mm3, persistence of positive CSF cultures, and presence and persistence of elevated IL-1a and TNF.39 For E. coli meningitis, the presence and persistence of K1 capsular polysaccharide antigen and

Chapter 39  The Immune System

the concentration of endotoxin in the CSF correlated with poor outcome. One fourth to one third of the survivors of gram-negative and gram-positive neonatal meningitis experience permanent sequelae resulting in significant neurologic damage, including hearing loss, language disorders, mental retardation, motor impairment, seizures, and hydrocephalus.63 Clinical follow-up and neurodevelopmental assessment are appropriate after this disease.

Meningoencephalitis Infection of the brain and meninges caused by viruses, fungi, or protozoal parasites can occur in the neonatal period. Transplacental infection with CMV, rubella virus, and Toxoplasma gondii and perinatal infection with HSV are discussed in other parts of this chapter. Poliovirus, coxsackievirus, echoviruses, varicella-zoster virus, and Epstein-Barr virus can be acquired before, at, or after delivery. Disorders caused by these agents, tuberculosis, and syphilis are discussed in greater detail later. When conventional bacterial cultures of CSF are sterile in an infant with clinical evidence of CNS infection or evidence of inflammation on examination of CSF, or both, these agents and Mycoplasma hominis should be considered. In young infants, it is usually impossible to distinguish clinically between meningitis, meningoencephalitis, and encephalitis. Identification of the etiologic agent and specialized radiologic studies can be helpful in this regard.

Pneumonia INCIDENCE Pneumonia is the most common form of neonatal infection and one of the most important causes of perinatal death worldwide.42 (See also Chapter 44.) In the 1920s and 1930s, pneumonia was found at autopsy in 20% to 30% of stillborn and newborn infants. More recently, intrauterine pneumonia has been reported in 5% to 35% of autopsies.

ETIOLOGY Neonatal pneumonia is caused by many of the same pathogens associated with neonatal sepsis. When the disease is acquired in utero or at the time of delivery, GBS is a common bacterial pathogen, although infections with E. coli, other gram-negative enteric organisms, Listeria species, H. influenzae, T. pallidum, staphylococci, and S. pneumoniae have been described. Other causative agents include Chlamydia trachomatis, Mycoplasma pneumoniae, CMV, HSV, enterovirus, adenovirus, and rubella. When disease is acquired in the nursery or at home, the same organisms can be responsible, but infection with S. aureus, Pseudomonas, Serratia, Bordetella pertussis, and Candida species should be considered, especially in premature infants with VLBW who receive care in intensive care units. Viral pneumonia can be noted at or shortly after birth, and reports of epidemic disease in the nursery have occurred with respiratory syncytial virus, echovirus, and adenovirus. Epidemic bacterial pneumonia has occurred after contamination of respiratory equipment and from personnel with suppurative staphylococcal lesions.

809

PATHOGENESIS AND PATHOLOGY Pneumonia can be acquired (1) transplacentally, (2) by aspiration of contaminated amniotic fluid before or at the time of delivery, (3) by aspiration of infected materials during or after delivery, (4) by inhalation of aerosols from infected personnel or contaminated equipment in the newborn nursery, or (5) hematogenously during the course of septicemia or from another focus of infection. Congenital (intrauterine) pneumonia frequently follows prolonged ROM. Gasping from asphyxia can result in the aspiration of infected amniotic fluid. Pathologic findings of congenital pneumonia reveal a diffuse inflammatory process with neutrophils present in the alveoli, but bacteria often absent. Amniotic debris and maternal leukocytes can be found, suggesting that disease has followed aspiration rather than hematogenous dissemination of microorganisms. Features commonly observed in bacterial pneumonia in older patients are not seen, such as infiltration of bronchopulmonary tissue, fibrinous alveolar exudate, and pleural reaction. Aspiration during or after delivery can result in bronchopneumonia, which is sometimes associated with hemorrhage and with evidence of pleural inflammation. Pathologic studies reveal areas of densely cellular exudate with bacteria present. Vascular congestion, hemorrhage, and necrosis can be observed. Microabscesses and empyema can be found when the infection is caused by S. aureus or Klebsiella pneumoniae. Pneumatocele formation, commonly associated with infection by S. aureus, has been described in neonatal disease caused by E. coli and Klebsiella species. Fatal cases of GBS pneumonia have had evidence of hyaline membranes similar to those found in respiratory distress syndrome.

CLINICAL MANIFESTATIONS Pneumonia should be suspected when respiratory distress develops in any infant, but suspicion should be heightened if antecedent problems referable to pregnancy and delivery have been noted. Infants with congenital pneumonia often die in utero or are critically ill at birth. Spontaneous respirations may not occur or may be established with difficulty. If respirations are established, tachypnea, moderate retractions, and grunting may be observed. Fever may or may not be noted, but is often a prominent sign of neonatal herpes simplex and enteroviral disease. In most cases, there is no cough. Cyanosis can be constant or intermittent, and respirations may be irregular and punctuated by periods of apnea. Congestive heart failure, manifested by cardiac enlargement, hepatomegaly, and tachycardia, can complicate the pneumonic process. Infants who acquire pneumonia during delivery or postnatally can have systemic signs, including fever, reluctance to feed, and lethargy. Respiratory signs can occur early or late in the course and include cough, grunting, costal and sternal retractions, flaring of the alae nasi, tachypnea or irregular respirations, rales, decreased breath sounds, and cyanosis. Congestive heart failure can be present in severe disease. Postnatally acquired pneumonia can develop at any age. Infants with chlamydial pneumonia typically present at 4 to 11 weeks of age with a prodrome of nasal congestion followed by tachypnea and a paroxysmal, staccato cough. Conjunctivitis is present or occurred earlier in approximately half

810

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

of these infants. Inspiratory rales may be present, but expiratory wheezes are not commonly seen. These infants are afeb­ rile and frequently gain weight slowly.

DIAGNOSIS Chest radiographs are necessary to support the diagnosis of pneumonia and to exclude other causes of respiratory distress. In some patients, no abnormality is found if imaging is performed soon after the onset of symptoms, but by 24 to 74 hours, a diagnosis should be possible. Radiographic findings can reveal bilateral homogeneous consolidation when pneumonia is acquired in utero or diffuse bronchopneumonia with postnatally acquired disease. Pneumatoceles can be observed, usually later in the course of illness, and occasionally hemothorax or pleural effusions are seen. Pneumonia caused by GBS can be difficult to distinguish radiographically from respiratory distress syndrome. This can be a diagnostic dilemma, especially in premature infants. Bilateral, symmetric interstitial infiltrates with hyperexpansion of the lungs are the typical radiographic findings seen in C. trachomatis pneumonia (see Chapter 37). The radiographic pattern of segmental or lobar atelectasis can be difficult to distinguish from that of massive meconium aspiration. In meconium aspiration, opacities may be distributed in the same manner as in bronchopneumonia, but the radiologic changes tend to be maximal early and disappear rapidly during the ensuing days. In contrast, the patchy opacifications noted with bronchopneumonia tend to be minimal early and become more impressive during the subsequent days. Blood cultures should be obtained in infants with pneumonia; positive cultures establish an etiologic diagnosis. If obtained before 8 hours of age, Gram-stained smears of tracheal secretions showing neutrophils and bacteria can help identify neonates with pneumonia. Bacterial cultures of tracheal secretions may not be helpful in establishing an etiologic diagnosis because they often reflect the flora present in the vaginal canal. In newborns with prolonged ventilation, tracheal aspirate cultures frequently fail to identify the cause of respiratory deterioration.70 A definitive microbiologic diagnosis can be made if organisms are isolated from cultures of pleural fluid, material aspirated from lung abscesses, or open lung biopsy tissue, but these interventions often are not performed. Tracheal secretions should be processed for viral isolates when nonbacterial pneumonia is suspected. A diagnosis of Chlamydia pneumonia can be assumed if the organism is isolated from nasopharyngeal samples or from secretions obtained by deep tracheal suctioning, but special culturing techniques are required. Culture of B. pertussis requires inoculation of nasopharyngeal secretions onto appropriate media for isolation. Cultures tend to grow slowly. Polymerase chain reaction of nasopharyngeal or endotracheal secretions provides a rapid and preferable means for diagnosis of pertussis.

TREATMENT When a diagnosis of pneumonia is suspected, antimicrobial therapy should be initiated promptly. If the diagnosis is supported by historical and clinical findings, but initial radiographs fail to confirm the clinical suspicion, treatment should not be delayed and is continued until results of a repeat chest

radiograph, taken at 48 to 72 hours, are available. Empirical antimicrobial therapy is the same as that for neonatal sepsis. Ampicillin and either an aminoglycoside or cefotaxime are indicated for early-onset or late-onset, community-acquired pneumonia, and vancomycin and an aminoglycoside are indicated for late-onset nosocomial infection (see Tables 39-17 and 39-18).41 Pneumonia caused by Chlamydia or Pertussis organisms responds best to treatment with azithromycin. Acyclovir therapy should be administered promptly if HSV pneumonia is suspected. Treatment for bacterial pneumonia is continued for 10 to 14 days or longer as dictated by the clinical course of the patient. Adequate fluids and oxygen, as required to treat hypoxemia, are useful adjunctive measures.

PROGNOSIS The outcome of intrauterine pneumonia is variable, but more critically ill infants are likely to die in utero or within the first 2 days of life despite meticulous care and administration of appropriate antibiotics. The prognosis is worse for premature than for term infants. In the absence of other problems, the prognosis for term infants who acquire pneumonia postnatally seems to be good.

Otitis Media INCIDENCE Acute otitis media is the most common infection diagnosed in young children, and it can occur in neonates. In one prospective study, 9% of infants had an episode of otitis media by age 3 months.68 The exact incidence is uncertain. It is estimated that otitis media develops in a minimum of 0.6% of all live births during the first month of life, and that the rate may reach 2% to 3% in premature infants. Otitis media occurs more often in male than in female infants, in infants with cleft palate, and in infants requiring prolonged intubation.

ETIOLOGY The microbiology of acute otitis media in neonates differs from acute otitis media seen in older infants and children. Although organisms from the respiratory tract are also the most commonly isolated pathogens in neonatal disease, GBS, E. coli, and other gram-negative bacteria can be causative agents in the first 2 to 6 weeks of life. Studies of infants younger than 6 to 8 weeks have shown that S. pneumoniae was the pathogen isolated in 19% to 30% of cases, H. influenzae was recovered in 14% to 25% of cases, and b-hemolytic streptococci (groups A and B) were recovered in 5% of cases. Gram-negative bacilli (E. coli and Enterobacter, Klebsiella, and Pseudomonas species) were found in 7% to 18% of neonates or very young infants with otitis media.6,57 Of cultures obtained by tympanocentesis in this age group, 40% to 50% have been sterile or judged nonpathogenic. Included among the nonpathogenic bacteria are organisms such as S. aureus, Staphylococcus epidermidis, and Moraxella catarrhalis, which likely are responsible for purulent otitis media in some newborns. Because strict anaerobic culture techniques were not used in most studies, many sterile isolates may have contained anaerobic microorganisms. Viral infection also has been associated with neonatal middle ear disease.

Chapter 39  The Immune System

PATHOGENESIS AND PATHOLOGY Otitis media is more common in premature than in term infants. This increased risk of infection may be related to the small size of the eustachian tube, with resultant obstruction and secondary infection. Aspiration of infected amniotic fluid is probably a leading cause of otitis media in newborns. The disease increases in frequency in bottle-fed infants, a finding that may be related to use of the supine position during feeding or to the lack of local immunity that may be conferred by ingredients in breast milk. Secretory IgA and other components in breast milk may exert protection by preventing the attachment of otitis media–causing bacterial strains to retropharyngeal or buccal epithelial cells.

CLINICAL MANIFESTATIONS The most common presenting symptoms are respiratory complaints such as cough or rhinorrhea and fever.6 Irritability, lethargy, vomiting, poor feeding, or diarrhea can also be present. Young infants with otitis media often are asymptomatic. The ear is difficult to examine adequately in newborn infants. Erythema, dullness, and bulging of the pars flaccida of the tympanic membranes can be noted, and pneumatic otoscopy can reveal decreased mobility of the tympanic membrane. Infants with typical facial or submandibular GBS cellulitis often have ipsilateral otitis media.

DIAGNOSIS The nonspecificity of clinical signs and symptoms combined with difficulties in visualizing the tympanic membrane early in life probably account in part for the discrepancy between incidence data derived from prospective clinical studies and data from postmortem studies. In infants who appear ill or fail to respond to initial therapy, a specific diagnosis should be made by culture and Gram-stained smears of purulent middle ear fluid obtained by myringotomy or tympanocentesis. Cultures of the nasopharynx can yield the same organism in some cases, but cannot be routinely relied on to guide selection of antibiotics. Blood cultures can be helpful in patients with concomitant septicemia. Lumbar puncture should be performed to exclude meningitis in ill infants. Meningitis that was not suspected before examination of CSF has been noted in some newborns with otitis media, but the frequency with which the two occur simultaneously is unknown.

TREATMENT The empirical antibiotics chosen and the route administered depend on the age and the appearance of the infant. For neonates with acute otitis media in the first 2 weeks of life, a combination of ampicillin and either an aminoglycoside or cefotaxime provides effective initial treatment. Hospitalized premature infants require parenteral therapy as well. A regimen such as vancomycin and gentamicin is one option for empirical therapy. Oral therapy with amoxicillin can be considered for well-appearing term infants older than 3 weeks who have had an uncomplicated intrapartum and neonatal course. If the infant is toxic-appearing or has evidence of infection elsewhere, cultures of middle ear fluid, blood, and CSF should be obtained, and the infant should be hospitalized for broad-spectrum parenteral antimicrobial therapy.

811

When culture results are known, antibiotic therapy can be altered appropriately. Treatment should be continued for 10 days or longer, depending on the etiologic agent, the associated condition (e.g., sepsis or meningitis), and the clinical response to therapy.

PROGNOSIS Recurrent or persistent disease can occur. Infants with onset of otitis media before 2 months of age are at high risk for development of chronic otitis media with effusion. In addition to young age at the onset of the first episode, other risk factors for recurrent infections include low socioeconomic status and the presence of smokers in the household. Careful follow-up evaluation for infection and hearing loss is important. Insertion of tympanostomy tubes may be indicated to ensure adequate drainage of the middle ear and to reduce the likelihood of recurrent middle ear infections.

Conjunctivitis INCIDENCE Neonatal conjunctivitis, also referred to as ophthalmia neonatorum, is the most common ocular disease in neonates.23 Previously, gonococcal conjunctivitis was the leading cause of infant blindness. The epidemiology of ophthalmia neonatorum in the United States has changed significantly, however, in the past century with the introduction of ocular prophylaxis by Credé in 1881 and the advent of prenatal screening and treatment of maternal gonococcal infections. Gonococcal conjunctivitis remains an important neonatal disease in less developed countries. In developed countries, ophthalmia neonatorum can be caused by chemical irritation after prophylactic instillation of silver nitrate into the conjunctival sac. Chemical conjunctivitis can occur in 10% to 90% of infants so treated. Chemical conjunctivitis caused by topical therapy with erythromycin or tetracycline occurs less frequently. The incidence of other forms of conjunctivitis is difficult to assess, but has been estimated at approximately 8 cases of chlamydial infection per 1000 live births and 0.3 cases of gonococcal disease per 1000 live births.

ETIOLOGY The differential diagnosis of infectious conjunctivitis in a newborn includes bacterial, chlamydial, and viral agents. C. trachomatis and Neisseria gonorrhoeae are the most important causes of conjunctivitis in the newborn period with respect to frequency of the former and severity of complications in untreated cases for both. Other bacteria, including S. aureus, GBS, Pseudomonas species, S. pneumoniae, and Haemophilus species, can also be causative. Conjunctivitis is the most common ocular manifestation of neonatal herpes simplex infections.

PATHOGENESIS Conjunctivitis caused by N. gonorrhoeae, C. trachomatis, GBS, or HSV usually is initiated by infection with microorganisms acquired during passage through the birth canal and reflects the prevalence of sexually transmitted diseases in the community. Premature ROM and prematurity increase the risk of

812

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

disease by N. gonorrhoeae. Infection with S. aureus or other organisms more commonly reflects infection acquired postnatally. Pseudomonas conjunctivitis occurs most often in hospitalized premature infants and often is associated with systemic complications.

CLINICAL MANIFESTATIONS The typical age at onset of conjunctivitis varies depending on the etiology, although there is much overlap. Chemical conjunctivitis usually is noted soon after birth and becomes less prominent within 48 hours. Bacterial conjunctivitis is most prominent during the first week of life, but can occur at any time during the neonatal period. Chlamydial conjunctivitis usually develops during the second week of life. The signs of bacterial conjunctivitis in a newborn are similar to the signs observed in older patients. A purulent ocular discharge, erythema and edema of the lids, and injection or suffusion of the conjunctiva can be found. Gonococcal disease typically manifests with abrupt onset of profuse, purulent ocular discharge. Conjunctivitis caused by Haemophilus species and S. pneumoniae has been associated with dacryocystitis. When C. trachomatis is involved, minimal inflammatory changes may be detected, or there can be intense inflammation, swelling, and a yellow discharge associated with pseudomembrane formation. The cornea is affected rarely. Chlamydial infections usually are bilateral, whereas unilateral disease is common when staphylococcal disease is present. Complications are rare, but can be devastating, with loss of vision. If bacterial conjunctivitis is left untreated, corneal involvement with punctate epithelial erosions can occur. In the case of disease with N. gonorrhoeae, large, punctate, superficial lesions can be seen that can coalesce and progress to corneal perforation. Gonococcal conjunctivitis should be considered a medical emergency. Untreated disease can progress rapidly from purulent conjunctivitis without corneal involvement to corneal ulceration and perforation. Meningitis or arthritis can complicate gonococcal ophthalmia neonatorum, but systemic spread of infection occurs rarely when appropriate therapy is given promptly before corneal involvement develops. When staphylococcal conjunctivitis is complicated by corneal involvement, the lower portion of the cornea is infected more frequently than the upper half, and marginal corneal infiltrates with peripheral vascularization can be seen. When chlamydial conjunctivitis is not treated, inflammation persists for 1 or 2 weeks. A subacute phase follows with slight injection of the conjunctiva and an accumulation of purulent material along the lid margins. Necrosis of the cornea has been observed, followed by fibrosis and scarring of the lid margin. Occasionally, chronic disease develops. Untreated chlamydial conjunctivitis, especially when associated with nasopharyngeal colonization, can progress to pneumonia in some infants. Even without invasive eye disease, systemic complications such as bacteremia and meningitis often develop in infants with P. aeruginosa conjunctivitis.

DIAGNOSIS A presumptive diagnosis of chemical or infectious conjunctivitis can be made based on the history and physical examination. There is such overlap in clinical appearance and age at

presentation that a definitive etiologic diagnosis cannot be determined reliably without the aid of laboratory testing. Purulent material must be sent for Gram stain and bacterial culture. Conjunctival scrapings should be sent when evaluating for Chlamydia infection because the organism resides in the epithelial conjunctival cells and not in neutrophils. When gonococcal disease is suspected, the purulent material should be placed in special transport media and sent to the bacteriology laboratory. Other organisms are less fastidious, and a swab of purulent material, cultured promptly, is sufficient to establish a diagnosis. The presence of intracellular gram-negative diplococci on a Gram-stained smear suggests a diagnosis of gonococcal disease. Characteristic intracytoplasmic inclusions can be identified by examining a Giemsa-stained smear from material obtained by gently but firmly scraping the lower portion of the palpebral conjunctiva with a blunt spatula or loop. Care must be taken not to injure the conjunctiva during the process. Leber cells (macrophages containing cellular debris) and inclusion-bearing epithelial cells (hence the name inclusion blennorrhea or conjunctivitis) are characteristic. The Giemsa stain is a specific, but not a sensitive, method for detection of Chlamydia conjunctivitis. Culture is the gold standard, but many laboratories do not have the capability of performing cell cultures. Numerous antigen detection assays, such as nucleic acid amplification and direct fluorescent antibody tests, are also options with a high degree of accuracy for detection of Chlamydia from conjunctival specimens.

TREATMENT Therapy of gonococcal conjunctivitis involves frequent irrigation of the conjunctival sac with sterile isotonic saline until the discharge has resolved and parenteral antimicrobial therapy. The recommended antimicrobial therapy for ophthalmia neonatorum is ceftriaxone (25 to 50 mg/kg, not to exceed 125 mg) administered intravenously or intramuscularly one time. Evaluation for disseminated infection should include cultures of blood and CSF. For disseminated neonatal gonococcal infection, the recommended therapy is ceftriaxone (25 to 50 mg/kg) given once daily for 7 days or cefotaxime (50 to 100 mg/kg daily) given in two doses for 7 days. If meningitis is documented, treatment should be continued for 10 to 14 days. Persistence of conjunctivitis after treatment may indicate coinfection with C. trachomatis and requires additional therapy directed against this agent. Because there are reports of high rates of C. trachomatis and N. gonorrhoeae coinfection in adults, empirical treatment with erythromycin in addition to cephalosporin often is indicated. Parenteral treatment with an aminoglycoside and an antipseudomonal penicillin in addition to topical treatment is indicated for conjunctivitis caused by Pseudomonas organisms. Antibiotics containing various combinations of bacitracin, neomycin, and polymyxin, administered topically as an ophthalmic ointment or solution several times a day for 7 to 10 days, have been recommended for other forms of bacterial conjunctivitis. An association between orally administered erythromycin and infantile hypertrophic pyloric stenosis has been reported in infants younger than 6 weeks. The efficacy of a single erythromycin course for treatment of Chlamydia conjunctivitis was

Chapter 39  The Immune System

approximately 80%. Preliminary data indicate that a short course of azithromycin treatment is an effective alternative to erythromycin,25 and azithromycin now is the treatment of choice for infants with chlamydial conjunctivitis. Topical treatment is ineffective and is not indicated.

PREVENTION The CDC and the American Academy of Pediatrics Committee on Infectious Diseases recommend erythromycin (0.5%) ophthalmic ointment, tetracycline (1%) ophthalmic ointment, or a 1% silver nitrate solution for prophylaxis of ophthalmia neonatorum. The prophylactic agent should be dispensed in a single-dose tube or ampule and given within 1 hour of birth. In a large study by Hammerschlag and colleagues24 in which the efficacy of three topical prophylactic agents was studied in 230 infants born to women diagnosed with cervical chlamydial infection during pregnancy, 20% of infants given silver nitrate at birth, 14% of infants given erythromycin, and 11% of infants given tetracycline topical prophylaxis developed neonatal chlamydial conjunctivitis. This observation and findings from several other studies suggest that the most effective way to prevent chlamydial conjunctivitis (and other chlamydial disease) in neonates is to screen and treat pregnant women harboring the infection. Infants born to mothers known to have untreated chlamydial infection are at high risk for infection; however, prophylactic antimicrobial treatment is not indicated because the efficacy is unknown. Ceftriaxone, 25 to 50 mg/kg (maximum 125 mg), should be administered intravenously or intramuscularly as a single dose to infants born to mothers with untreated gonococcal infection. Cefotaxime, 100 mg/kg, is an alternative. Administration of one of the topical agents used for prophylaxis does not eradicate established gonococcal infection that is not yet clinically apparent. Caution should be exercised in providing ceftriaxone to premature infants of low birthweight with hyperbilirubinemia.

Gastroenteritis INCIDENCE Most neonatal diarrheal disease is sporadic. It is usually selflimited and brief, but may cause significant morbidity in some infants. Failure to take appropriate steps to control the spread of infection to other infants can result in an epidemic outbreak of the disease in the nursery. The incidence varies from nursery to nursery and within the same nursery from year to year.

ETIOLOGY Neonatal diarrheal disease is more likely to be associated with noninfectious processes. The organisms that have been associated with epidemic and sporadic infectious diarrheal disease include certain strains of E. coli; Salmonella, Shigella, Yersinia, and Campylobacter species49; and rarely Pseudomonas, Klebsiella, and Enterobacter species and Candida albicans. An epidemic of enterocolitis has been attributed to infection by group A streptococci. Enteric types of adenovirus, rotavirus, and calicivirus have been identified as causative agents

813

for gastroenteritis in neonates. In many cases, the etiologic agent is not identified, and viruses are assumed to be responsible despite negative viral and serologic studies. There are at least five pathotypes of diarrhea-producing E. coli. Enteropathogenic E. coli (EPEC) was associated classically with severe outbreaks of infantile diarrhea in the developed world through the 1960s, and now is a common cause of diarrhea in infants in the developing world. There is a predilection for this infection in infants who are not breastfed. Enterotoxigenic E. coli also has been associated with neonatal diarrheal disease, but infection usually is delayed until the introduction of foods to supplement human milk. All Salmonella organisms are closely related and are now considered a single species, Salmonella enterica. There are more than 2000 Salmonella types that cause human disease. Three surface antigens determine reactions to antisera: O (somatic), H (flagellar), and Vi (capsular) antigens. Serogrouping is based on O antigens and designates groups A through E. Salmonella serotype typhi is in group D. Commonly reported isolates in the United States include Salmonella serotype typhimurium (group B) and Salmonella serotype enteritidis (group D). Shigella comprises four serogroups with more than 40 serotypes. In the United States, Shigella sonnei (group D) followed by Shigella flexneri (group B) are the most common isolates from patients of any age with shigellosis. The central event in pathogenesis for Shigella is its ability to cross the colonic epithelium and invade the colonic mucosa. All virulent Shigella organisms have virulent plasmid-associated antigens that provide the means for intracellular and intercellular spread. Hemolytic uremic syndrome, associated with production of Shiga toxin, can complicate bacillary dysentery secondary to Shigella dysenteriae, serotype 1. Campylobacter, a comma-shaped, gram-negative bacillus, is recognized increasingly as a common cause of gastroenteritis. Campylobacter fetus causes prenatal and neonatal infections that result in abortion, premature delivery, bacteremia, and meningitis. Campylobacter jejuni and Campylobacter coli usually cause neonatal gastroenteritis. Health care–associated outbreaks in nurseries, although uncommon, have been described. Although it is likely that some diarrheal episodes in neonates can be caused by Clostridium difficile, the diagnostic criteria used in older children and adults are inadequate to establish a definitive etiologic diagnosis in this age group. C. difficile is rarely isolated from the stools of healthy children older than 1 year or healthy adults, but the organism and its cytotoxin can be shown in the stools of more than 50% of healthy neonates. The pathogenicity of this organism in neonates has not been well defined.

PATHOGENESIS The gastrointestinal tract is colonized initially by organisms that enter the oropharynx at delivery. Maternal carriers of Salmonella, Shigella, Campylobacter, and EPEC can serve as the initial source of these pathogens in the nursery. Person-toperson transmission also has occurred in neonates of infected mothers and has resulted in health care–related outbreaks in nurseries. Asymptomatic health care workers who are carriers of these organisms have proved to be the source of spread.

814

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Fomite transmission has been observed, and in rare instances infection has followed ingestion of contaminated milk, water, or materials used for radiographic evaluation of the gastrointestinal tract.

CLINICAL MANIFESTATIONS Gastrointestinal tract bacterial infection can manifest without clinical symptoms; as a watery diarrhea without other findings; or as a severe diarrhea with fever, vomiting, and abdominal distention. It is difficult to predict the etiology on the basis of clinical findings. Bloody diarrhea containing mucus suggests Shigella, Yersinia, or Campylobacter gastroenteritis, but these findings have been observed occasionally in neonates with gastroenteritis caused by Salmonella or EPEC. Infection with Salmonella is frequently asymptomatic or mild. Neonates with diarrhea usually have a self-limited infection with loose, mucoid stools and, rarely, hematochezia. Salmonella gastroenteritis in the first month of life can be associated with bacteremia or extraintestinal manifestations, including meningitis. Neonatal shigellosis is rare. The classic picture is that of dysentery, but asymptomatic infection occurs as well. Complications include hemolysis, hemolytic uremic syndrome, colonic perforation, and pseudomembranous colitis. Seizures associated with Shigella bowel infection are uncommon in the first 6 months of life. Sepsis occurs uncommonly, but the case-fatality rate is significant for infants with systemic infection. Infections caused by C. fetus are the most common type of Campylobacter infection in the first 3 weeks of life and are associated with a high incidence of fetal and neonatal mortality. Systemic C. jejuni or C. coli infection in neonates is rare. A blue-green discoloration of the stools or diaper can be found when Pseudomonas species predominate. The isolation of Pseudomonas organisms does not represent disease. Dehydration, acidosis, electrolyte disturbances, and hypotension can occur during the course of disease caused by any of these enteric pathogens.

DIAGNOSIS Stool for bacterial culture should be obtained from infants with gastroenteritis. Recovery of Shigella, Yersinia, Campylobacter, or Salmonella from the stool of an infant with diarrhea is presumptive evidence that disease is caused by that organism. The diagnosis of EPEC is difficult because most clinical laboratories cannot differentiate diarrhea-associated strains of E. coli from normal stool flora E. coli strains. Isolates of E. coli suspected to cause a case or outbreak should be sent to a state health or other reference laboratory for serotyping. A stool smear for polymorphonuclear leukocytes can be valuable. Fecal leukocytes can be observed in stool smears from patients with diarrhea caused by EPEC, Yersinia, Campylobacter, Salmonella, or Shigella, but usually are not seen in patients with nonspecific or viral gastroenteritis. Blood cultures should be obtained from any neonate with isolation of Salmonella, Shigella, Campylobacter, or Yersinia enterocolitica from the stool. Lumbar puncture for examination and culture of CSF should be considered if the neonate is febrile or otherwise “toxic” in appearance, or if Salmonella or other potential pathogens are isolated from the blood culture.

TREATMENT Supportive care and correction of fluid and electrolyte abnormalities are the most important aspects of management. Because many infections are self-limited, antimicrobial therapy may not be indicated except in certain situations. It may be unnecessary to treat an infant with diarrhea caused by EPEC. During nursery outbreaks, infants with EPEC isolated from stool should be treated independent of the symptoms, however. When antimicrobial therapy is given, nonabsorbable orally administered antibiotics (neomycin, colistin sulfate) are appropriate options, and the duration of treatment is 5 days. Infants with Salmonella isolated from stool should be treated with cefotaxime after obtaining blood and CSF cultures. When bacteremia or meningitis is present, therapy with cefotaxime or ampicillin for susceptible strains should be given for 10 days (bacteremia) to 4 to 6 weeks (meningitis). Antimicrobial therapy is effective in shortening the duration of diarrhea caused by Shigella and eliminating organisms from the feces, and is recommended for patients with bacillary dysentery. For cases in which susceptibility is unknown, ceftriaxone or cefotaxime should be administered. For susceptible strains, ampicillin or trimethoprim-sulfamethoxazole are effective. Treatment should be administered for 5 days. Campylobacter enteritis may be treated with azithromycin. Benefit from antimicrobial therapy for patients with Yersinia enterocolitis has not been established. Patients with septicemia and compromised hosts, including premature infants, with enterocolitis should receive treatment, however. Cefotaxime and aminoglycosides are appropriate antimicrobials.

PREVENTION AND PROGNOSIS In addition to standard precautions, contact precautions are indicated for infants with diarrhea. Clusters of infections should be investigated appropriately. During outbreaks of disease caused by EPEC, contact precautions should be maintained until tests of stools collected after completion of antimicrobial therapy are negative for the infecting strain.

Osteomyelitis and Septic Arthritis INCIDENCE Osteomyelitis occurs rarely in newborns and can be difficult to diagnose. Septic arthritis is frequently concomitant, probably reflecting spread of infection through blood vessels that penetrate the epiphyseal plate (see Chapter 54).

ETIOLOGY The most frequent causative agents for community-acquired osteomyelitis and septic arthritis are S. aureus, E. coli, and GBS. Community-associated MRSA strains and health care– associated MRSA are common pathogens. Uncommon etiologic agents include N. gonorrhoeae, H. influenzae, group A streptococci, and T. pallidum. Various staphylococcal and streptococcal species and anaerobic bacteria have been cultured from scalp electrode–related cases of osteomyelitis. Candida species and gram-negative enteric bacilli have assumed major roles for nosocomially acquired bone and joint infections, especially in preterm infants with low birthweight. These infections are most commonly isolated in nursery outbreaks.

Chapter 39  The Immune System

815

PATHOLOGY Osteomyelitis usually originates in the metaphysis of long bones. Involvement of the small bones of the hands and feet, the vertebrae, and the ribs has been noted, however. Multiple foci of osteomyelitis are seen more frequently in newborns than in older individuals; 10% to 40% of newborns have more than a single focus of infection. The cortical bone is thin, permitting infection to extend readily into soft tissues. Infection extends readily from an affected bone into the adjacent joint via transphyseal vessels. Destruction of cartilage, dislocations, and pathologic fractures can occur and permanently affect bone growth.

PATHOGENESIS Osteomyelitis almost invariably follows hematogenous dissemination of microorganisms. Infection of the skin, omphalitis, umbilical catheterization, and occasionally improperly performed femoral venipunctures are predisposing factors in cases principally caused by S. aureus. Occasionally, osteomyelitis can follow direct extension of a soft tissue infection to bone, especially if the underlying bone is fractured, as illustrated in cases of osteomyelitis of the skull associated with infected cephalhematomas. Transplacental acquisition with dissemination to bone has been observed with T. pallidum.

CLINICAL MANIFESTATIONS Destructive changes in bone caused by osteomyelitis can be advanced by the time a definitive diagnosis is made. In some patients, the disease is not accompanied by systemic signs or symptoms. Fever is absent or is a very late, intermittent finding. Findings of swelling, tenderness, and decreased motion of an extremity usually prompt initial appraisal. In a patient with osteomyelitis of the clavicle, an incomplete Moro reflex can be the only finding on physical examination. Localized tenderness and erythema can be seen, and heat and fluctuance can be noted. Focal findings overlying affected sites occur commonly when S. aureus is the pathogen. Less often, infants have an acute onset of systemic illness with fever, icterus, hemorrhagic manifestations, and hypotension. GBS osteomyelitis commonly manifests during the third and fourth weeks of life. In contrast to the fulminant onset and poor outcome that occur in some patients with late-onset meningitis, GBS bone and joint infection often has an indolent nature and an almost uniformly good outcome.

DIAGNOSIS When the history and physical examination suggest the possibility of osteomyelitis, imaging studies should be obtained. Plain radiographic bony changes are delayed for 7 to 10 days after onset of infection, but a plain radiograph provides a helpful baseline. The earliest finding is deep soft tissue swelling. Later in the course of infection, periosteal thickening, cortical destruction, irregularities of the epiphysis, and periosteal new bone formation are seen. Additional findings include localized areas of cortical rarefaction, periosteal elevation, a widened joint space, and subluxation or dislocation of the bone from the joint (e.g., the hip) (Fig. 39-10).

Figure 39–10.  Radiographic evidence (posterior view) of

multiple sites of osteomyelitis in an infant, the result of infection with Staphylococcus aureus. Clinical evidence of infection became apparent at 2 weeks of age. Osteolytic lesions can be seen in the right femur, right tibia and fibula, and left tibia. Periosteal new bone formation is visible in many areas.

Magnetic resonance imaging (MRI) is the diagnostic modality of choice for neonatal osteomyelitis. The diagnosis can be established within 24 to 48 hours after the onset of symptoms. The anatomic detail provided by MRI optimizes the definition of structural abnormalities and can guide surgical drainage. A technetium 99m phosphate scan can show areas of increased blood flow in the affected area, but this modality lacks the structural clarity of MRI and is rarely indicated. Aspiration of soft tissues and bone is indicated, and samples should be sent for Gram stain and culture. Blood cultures are essential and have been reported to yield the causative agent in 60% of cases. Synovial fluid obtained from patients with joint involvement should be sent for Gram stain and culture. Because osteomyelitis frequently follows bacteremia, caregivers should consider obtaining CSF to exclude concomitant meningitis. Normal peripheral white blood cell counts are commonly seen in neonates with bone and joint infection, and ESR values less than 40 mm/h are reported in

816

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

25% of cases. Syphilis, tuberculosis, scurvy, rickets, and deep cellulitis must be considered in the differential diagnosis.

TREATMENT Until the specific etiology has been clarified by culture, combination therapy should be initiated with vancomycin and either an aminoglycoside or an extended-spectrum cephalosporin for gram-negative coverage. When the infecting isolate and its susceptibility pattern have been identified, treatment should be continued with the most appropriate antibiotic. The dosages used should be the same as the dosages used for treatment of septicemia (see Tables 39-17 and 39-18). Parenteral therapy should be continued for 3 to 4 weeks or longer until clinical and radiographic findings indicate healing, and CRP and ESR are normal. Intra-articular administration of antibiotics is not indicated. Surgical drainage of purulent material is a key component of treatment, and consultation should be sought with a pediatric orthopedic surgeon in the management of this infection. When the hip or shoulder joints are involved, prompt surgical decompression and drainage are crucial.

PROGNOSIS The degree of residual damage depends on the extent of disease before effective treatment. Permanent disability is more common when joint involvement has occurred. Growth disturbances can result from destruction of the epiphyseal plate. The number of foci of osteomyelitis in a patient who is effectively treated does not seem to correlate specifically with the extent or likelihood of permanent disability. Chronic osteomyelitis occurs infrequently. Overall, residual effects can be noted in 25% of neonates with infection of the bones and joints.

Urinary Tract Infection The incidence of neonatal urinary tract infection varies from 0.1% to 1% of all infants, with a higher frequency in neonates with low birthweight. (See also Chapter 51.) In contrast to bacteriuria in all other age groups, neonatal disease is more commonly seen in male infants. This preponderance of male infants may reflect the increased risk of urinary tract infection in uncircumcised boys.72 The presence of virulence factors in some bacteria also influences the development of urinary tract infections. The organisms associated with neonatal urinary tract infection mirror those that cause neonatal sepsis. For communityacquired neonatal disease, infection with E. coli is most common. Although GBS can be isolated from the urine of infants with GBS sepsis, primary urinary tract infection is rare. The frequency of health care–associated urinary tract infection has increased with the survival of infants with VLBW. Pathogens causing health care–associated infections include E. coli, other gram-negative enteric bacilli, Enterococcus, Candida, and CONS. Symptoms referable to the genitourinary system are rare. In most infants, the symptoms are nonspecific, similar to that of neonates with septicemia. Reluctance to feed, weight loss, and diarrhea can be the presenting features in some patients. An unexplained fever should prompt an investigation of the urinary tract in addition to an evaluation for sepsis and meningitis.

Urine culture need not be performed routinely in evaluation of infants with early-onset infection because isolation of a pathogen from the urine usually is a manifestation of sepsis, rather than an indicator of urinary tract infection. Definitive diagnosis of neonatal urinary tract infection for an infant evaluated beyond the first few days of life can be made only with culture of properly obtained urine. Samples collected by voiding into a bag are inappropriate. Catheterized specimens or samples obtained by suprapubic bladder aspiration are suitable. Isolation of any bacteria from a bladder aspirate is considered significant, whereas counts of 103 or higher colony-forming units per milliliter of catheterized urine are meaningful. Blood and CSF should be obtained as described in the evaluation for sepsis. Although the presence or absence of pyuria on examination of urine sediment is not conclusive evidence for diagnosing or excluding urinary tract infection, the presence of bacteria on Gram-stained smears of the sediment does support the diagnosis. Empirical therapy for neonatal urinary tract infection should include ampicillin and an aminoglycoside in dosages used for sepsis. Administration should be parenteral because of the high incidence of sepsis in association with urinary tract infections in newborns, and because of the often erratic oral antibiotic absorption in infants. Vancomycin and an aminoglycoside should be considered for empirical therapy of health care–associated urinary tract infections. Sterilization of the urine should be documented by repeat culture after 48 hours of therapy. Treatment duration is usually 10 days, but can be longer if there is persistent bacteriuria, anatomic obstruction, or a perinephric abscess. In uncomplicated cases of primary urinary tract infection, parenteral therapy can be given for 5 to 7 days, followed by oral antibiotic therapy to complete the course of treatment. Imaging studies, including renal ultrasound and voiding cystourethrogram or renal scan, should be performed to diagnose any anatomic or physiologic urinary tract anomalies.

Infections of the Skin INCIDENCE Infection of the skin is common and can be a result of bacterial, viral, or fungal disease. (See also Chapter 52.) The lesions may be localized or the cutaneous manifestation of systemic disease.

ETIOLOGY Colonization of the skin begins at birth, and the organisms that initially are acquired reflect those present in the mother’s birth canal. Subsequently, infants can acquire organisms from the environment or from the hands of family members or caregivers. S. aureus is the organism most often associated with neonatal skin infections. Application of triple dye, iodophor ointment, or chlorhexidine powder to the umbilical stump has been used to delay or prevent S. aureus colonization. Virtually any organism that causes disease in infancy can result in cutaneous infection, including group A streptococci and GBS, Listeria, P. aeruginosa, T. pallidum, HSV, and C. albicans.

Chapter 39  The Immune System

817

Histologic examination of the skin of patients with bullous impetigo can show intraepidermal bullae filled with polymorphonuclear leukocytes. Ritter disease, a severe generalized form of staphylococcal scalded skin syndrome, is caused by a toxin-elaborating strain of S. aureus. It is characterized by a severe bullous eruption with shedding of the epidermis. Histologic examination of the skin reveals an intraepidermal blister, cellular death and acantholysis, and a striking absence of inflammation. Pseudomonas or Aeromonas hydrophila infections cause ecthyma gangrenosum, which is characterized by the development of yellow-green pustules that progress rapidly to form hemorrhagic and necrotic ulcers. In most neonates with ecthyma gangrenosum, these lesions have developed as a consequence of bacteremia. Skin abscesses with minimal inflammation may be caused by Candida species; their occurrence should be considered diagnostic of invasive infection.3

noted at birth, particularly on the palms and soles. The blisters develop on an erythematous base and may consist of cloudy or hemorrhagic fluid that contains spirochetes. When the blisters rupture, a macerated area remains on which crusts form. A diagnosis of Pseudomonas or Aeromonas infection should be suspected in patients with ecthyma gangrenosum. Candida infection may be localized to the oral cavity or the diaper area, or both. Early cutaneous lesions are vesicular with a surrounding area of erythema. Pustules may develop, and a confluent erythematous moist erosion is soon noted. Cutaneous manifestations of congenital candidiasis characteristically are present at birth or within hours of delivery. A diffuse, burnlike dermatitis has been described in several infants with low birthweight (,1500 g) as an early manifestation of systemic candidiasis. In the early stages of infection with S. aureus, Listeria, GBS, or Pseudomonas, vesicles may be present. Herpes simplex and varicella zoster viruses produce vesicular lesions that are similar in appearance clinically and histologically.

PATHOGENESIS

DIFFERENTIAL DIAGNOSIS

Infection usually develops as a result of a small break in the skin and is caused by organisms that have colonized the skin and nares. Potential risk factors include abrasions after forceps use, scalp wounds associated with intrapartum fetal monitoring, the process of degeneration of the umbilical stump after cord clamping, intravascular catheter use, and circumcision. Infectious complications after severing the umbilical cord or circumcision were common before aseptic surgical techniques were introduced.

Noninfectious processes that can be mistaken for the cutaneous manifestation of infection include milia, transient pustular melanosis, erythema toxicum, and sclerema neonatorum. (See Chapter 52.) Purpuric and bullous lesions can result from trauma, inherited skin disorders such as epidermolysis bullosa, acrodermatitis enteropathica, dermatitis herpetiformis, and mastocytosis. Coagulopathies and congenital leukemia can be associated with purpura.

PATHOLOGY

CLINICAL MANIFESTATIONS Skin and soft tissue infections should be discovered promptly because clinical findings usually are apparent on casual observation. Lesions include pustules, vesicles, bullae, maculopapules, cellulitis, impetigo, abscesses, petechiae, purpura, and erythema multiforme. The most frequently observed lesion is a maculopapular rash that is nonspecific and can be seen with bacterial infections (staphylococcal and streptococcal), fungal infections, and viral infections (measles, enteroviruses, rubella). Bullous lesions that appear after the second day of life usually suggest staphylococcal disease, in contrast to noninfectious blistering disease that is present at birth. Blisters can range in size from small vesicles to larger bullae filled with straw-colored fluid. When these lesions rupture, an erythematous, weeping, denuded area is present. In staphylococcal scalded skin syndrome, the epidermis can be shed in sheets, and intact bullae frequently are sterile. HSV, T. pallidum, P. aeruginosa, and GBS less frequently are associated with bullous lesions. Pustules are a common manifestation of staphylococcal disease, but can be seen with listeriosis as well. Impetigo and erysipelas should suggest streptococcal infection. Less commonly, impetigo can be caused by S. aureus or E. coli. GBS has been associated with the triad of cellulitis, adenitis, and bacteremia. Cellulitis and soft tissue abscesses are the hallmark of infection with S. aureus, but occasionally can be caused by streptococci or E. coli. Erythema multiforme can occur during neonatal sepsis with S. aureus, streptococci, or Pseudomonas. Congenital syphilis may be suggested by the presence of maculopapular lesions, and in some cases bullae have been

DIAGNOSIS A diagnosis should be based on stains and bacterial, fungal, and viral cultures of the lesion. In patients with staphylococcal scalded skin syndrome, microorganisms may not be present in the intact bullous lesions. Candida can be shown by use of a potassium hydroxide preparation that shows pseudohyphae. Skin biopsy may be required to document the depth of invasiveness. Dark-field examination of bullous lesions usually reveals T. pallidum in an infant with congenital syphilis. When this diagnosis is suspected, appropriate serologic studies must be performed, and contact precautions must be undertaken. Fluid obtained by aspiration of vesicles may permit identification of disease caused by various viral agents. Blood cultures should be obtained and may be positive in neonates with systemic infection or with staphylococcal or streptococcal disease.

TREATMENT When lesions are localized and superficial, local therapy can suffice for neonatal staphylococcal pustulosis in an otherwise healthy term infant. Other manifestations of staphylococcal infections and most other bacterial infections of the skin should be treated by parenteral administration of antibiotics. Nafcillin is appropriate for methicillin-susceptible strains of staphylococci, and vancomycin should be used for MRSA. Streptococcal infections can be treated with penicillin G, ampicillin, or a third-generation cephalosporin. An aminoglycoside, an extended-spectrum penicillin, or ceftazidime should provide coverage for ecthyma gangrenosum caused by P. aeruginosa. Treatment of cutaneous fungal infections is discussed in Part 3 of this chapter.

818

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Parenteral fluid therapy may be required to maintain hydration in patients with staphylococcal scalded skin syndrome. Specific attention should be paid to maintenance of adequate hemoglobin and serum albumin concentrations and of the body temperature. Incision and drainage are required if abscesses are found. Full evaluation for disseminated disease should be carried out, and treatment should be initiated promptly with acyclovir for vesicular lesions presumed to be caused by HSV. Early initiation of treatment may prevent dissemination of localized infection and optimize outcome of generalized herpes infection heralded by cutaneous lesions. (See also Part 4.)

Omphalitis INCIDENCE Omphalitis was previously an important cause of neonatal disease and death worldwide. Since the introduction of simple umbilical cord hygienic measures, it is rarely seen in developed countries, but it remains an important cause of neonatal illness elsewhere. The incidence is unknown. Umbilical phlebitis and arteritis may develop spontaneously or may follow catheterization of umbilical vessels. Tetanus may follow contamination of the umbilical stump, but is rarely encountered in societies that practice routine immunization for tetanus and asepsis at the time of delivery. Necrotizing fasciitis is a life-threatening complication resulting from rapidly spreading destruction of the fascia and subcutaneous tissue around the umbilicus.

ETIOLOGY Organisms that are found on the skin or that are introduced into the umbilical vessel by catheterization can produce omphalitis. S. aureus and E. coli are frequent pathogens, but group A streptococci, anaerobic bacteria, and polymicrobial infections may occur.54

PATHOGENESIS Direct bacterial invasion of the umbilical cord and surrounding skin is common. Bacteria can invade the umbilical artery and spread across its lumen, causing necrosis of the loose connective tissue of the arterial wall. If the umbilical and iliac ends are occluded, a septic, loculated focus of infection may be found. When the umbilical end remains patent, purulent material may drain through the umbilicus. If the connective tissue of the artery is extensively involved, peritonitis can develop, or the infection can extend along its course and manifest as a scrotal or deep thigh abscess. If the iliac end of the artery is patent, but the umbilical end is sealed, septicemia can ensue.

CLINICAL MANIFESTATIONS Purulent drainage can be noted from the umbilical stump at its base of attachment to the abdominal wall or from the navel after the cord has separated. The discharge can be foul smelling. Periumbilical erythema and induration may be noted. If extensive periumbilical edema or involvement of the abdominal wall is noted, the complication of necrotizing fasciitis should be considered.

DIAGNOSIS Gram stain and culture for aerobic and anaerobic bacteria should be performed on the purulent material from the umbilical stump. Whenever septic umbilical arteritis is suspected, blood cultures should be obtained.

TREATMENT Otherwise well-appearing infants who have moist or “smelly” cords, but no periumbilical erythema, edema, or exudate often respond to local measures. The exception is omphalitis caused by group A streptococci, which requires parenteral penicillin therapy. Parenteral administration of antibiotics is indicated if a neonate presents with periumbilical erythema, edema, and tenderness with or without purulent drainage. Combination therapy should be administered to provide broad-spectrum coverage. Vancomycin should be provided for gram-positive coverage. An aminoglycoside, or perhaps a third-generation cephalosporin for better tissue penetration, can be given for gram-negative coverage. The presence of crepitus or black discoloration of the periumbilical tissues suggests an anaerobic or mixed infection. In that case, consideration should be given to adding clindamycin or metronidazole for anaerobic coverage. Necrotizing fasciitis requires extensive surgical débridement and pathogen-directed antibiotic therapy and supportive care.

COMPLICATIONS Omphalitis complicated by necrotizing fasciitis can be associated with bacteremia, coagulopathy, and shock, and frequently progresses to death despite heroic surgical and supportive measures. Septic embolization with metastasis to the lungs, kidneys, and skin can occur. Other infections complicating or occurring in the absence of cutaneous signs of omphalitis include pyelophlebitis (suppurative thrombophlebitis of portal or umbilical veins), liver abscess, septic umbilical arteritis, endocarditis, and subacute necrotizing funisitis.20

Mastitis (Breast Abscess) Neonatal mastitis is most commonly caused by S. aureus. Infection with other organisms has been described, including E. coli, Pseudomonas, Proteus, Salmonella, GBS, and anaerobes. The incidence is low; in one center, only 21 cases of neonatal mastitis or breast abscess were seen in a 7-year period.17 Mastitis rarely develops during the first week of life, but is seen more frequently during the second and third weeks after birth. A preponderance of female infants has been observed when mastitis occurs beyond the second week of life. Breast abscesses are rare in premature neonates, presumably because of underdevelopment of the mammary glands. Infection probably results from invasion of the duct system of the breast by skin flora. The breast can be enlarged, erythematous, and tender, but is more likely to be firm than fluctuant. Systemic signs and symptoms are uncommon; only one fourth of infants have a low-grade fever. The disorder must be distinguished from enlargement that is physiologic or hormonally induced; hormonally induced cases more likely produce swelling bilaterally.

Chapter 39  The Immune System

Aspiration of the abscess for Gram stain and culture and blood culture are indicated. If gram-positive cocci are seen on the Gram stain, vancomycin or clindamycin should be given until the results of susceptibility testing are available. An aminoglycoside or cefotaxime is indicated when gramnegative bacilli are present. If no organisms are noted on Gram stain, vancomycin or clindamycin and either an aminoglycoside or cefotaxime should be given pending culture results. Surgical incision and drainage may be necessary and should be performed by an experienced surgeon to minimize scarring of breast tissue.

Parotitis Suppurative parotitis is a rare infection in newborns; only 32 cases have been described in English since 1970.59 More than one third of the infants were born prematurely, and the causative organism in most cases was S. aureus. Infection is likely caused by the ascent of microorganisms from the oral cavity through the Stensen duct in the setting of salivary stasis or by invasion with blood-borne organisms. Dehydration has been noted before the onset of infection in some patients. Typically, infants present in the second week of life with parotid gland swelling and fever. The swollen parotid gland is accompanied by an area of erythema in the overlying skin. The gland may be tender, warm, firm, or fluctuant. Purulent material can drain spontaneously or be expressed from the Stensen duct. Septicemia may accompany localized infection. The diagnosis is made by Gram-stained smear and culture of purulent material expressed from the duct or obtained by percutaneous aspiration of a fluctuant area of the parotid gland. A blood culture should obtained. Incision and drainage should be performed if the lesions are fluctuant. Empirical systemic antibiotic treatment should provide coverage for S. aureus, E. coli, and Pseudomonas. If the cultures are sterile, or the infant fails to respond to antimicrobial therapy, or both, addition of an antibiotic with anaerobic activity should be considered. Surgical drainage may be required if clinical improvement is not apparent within 48 hours. Complications include salivary fistulas, facial palsy, mediastinitis, and extension into the external auditory canal. Involvement of another salivary gland (most often the other parotid) or septicemia occasionally complicates cases of acute parotitis. The disease must be differentiated from infections of the maxilla and from lymphangiomas and hemangiomas.

819

causing neonatal sepsis from 1989-2003 in another review.9

ETIOLOGY With the exceptions of Clostridium botulinum and Clostridium tetani, most anaerobic bacteria are part of the normal flora of the gastrointestinal tract, the genital tract, or the skin, and can be potential neonatal pathogens. Gram-positive anaerobes associated with clinical disease in patients of any age include Peptococcus, Peptostreptococcus, microaerophilic streptococci, Clostridium species, Propionibacterium, and Eubacterium. Clinically significant gram-negative anaerobes include Bacteroides fragilis, other Bacteroides species, Fusobacterium, Veillonella, and Prevotella melaninogenica.

PATHOGENESIS Anaerobic infections occur in association with aspiration, gastrointestinal ischemia or perforation, trauma, and tissue necrosis—events that have been associated with complicated deliveries or premature infants who are critically ill. Obstetric and neonatal factors associated with increased risk of sepsis neonatorum, including prolonged ROM, maternal chorioamnionitis, prematurity, and respiratory distress, are also risk factors for anaerobic sepsis.

CLINICAL MANIFESTATIONS

MISCELLANEOUS BACTERIAL INFECTIONS Anaerobic Infection

The signs and symptoms of anaerobic septicemia or peritonitis are indistinguishable from those described for other forms of neonatal septicemia or peritonitis. Neonatal ana­ erobic bacteremia occurs in two distinct settings. In the first 3 days of life, anaerobic bacteremia is associated with perinatal sepsis and chorioamnionitis, suggesting vertical transmission. In older neonates, anaerobic bacteremia is associated with a gastrointestinal process, including appendicitis, necrotizing enterocolitis, or other causes of ruptured viscus. The bacteria isolated vary, depending on the clinical setting. Gram-positive anaerobic bacteria susceptible to penicillin G were isolated from blood cultures from 8 of 12 neonates with sepsis in the first 48 hours of life, whereas 11 of 17 neonates older than 2 days of age with clinical evidence of necrotizing enterocolitis had penicillinresistant, gram-negative anaerobes isolated from blood cultures.43 This study suggests that gram-positive anaerobic bacteria are more frequently associated with early-onset sepsis and gram-negative anaerobes with gastrointestinal disease. The exception was observed in congenital pneumonia occurring in the absence of chorioamnionitis; Bacteroides was frequently isolated from blood cultures from these infants. Anaerobic infection caused by Clostridium can manifest as systemic illness or localized infection, such as cellulitis, omphalitis, or necrotizing fasciitis.

INCIDENCE

DIAGNOSIS

The true incidence of neonatal anaerobic bacteremia is difficult to ascertain. Surveys in the 1960s and 1970s suggested that anaerobic bacteria were the causative agents in one fourth of all neonatal bacteremias. A later retrospective review suggests that neonatal anaerobic bacteremia is less common even in the setting of significant gastrointestinal disease.43 Anaerobes accounted for only 1% of isolates

The diagnosis of anaerobic infection should be considered in neonates with clinical signs suggestive of sepsis associated with prolonged ROM, chorioamnionitis, intestinal perforation, or tissue necrosis. Anaerobic culture media are necessary to optimize recovery of these organisms. Use of an anaerobic transport tube or a sealed syringe is recommended for collection of clinical specimens.

820

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TREATMENT The choice of antibiotic therapy depends on the etiologic agent. Susceptibility testing for anaerobic bacteria is technically difficult and not readily available. Generally, penicillin is active against most gram-positive and some gramnegative anaerobic microorganisms. Bacteroides species of the gastrointestinal tract usually are penicillin-resistant. Because B. fragilis is commonly isolated from peritoneal fluid in newborns with intestinal perforation or necrotizing enterocolitis, anaerobic coverage is appropriate in those settings. Bacteroides species are predictably susceptible to metronidazole and sometimes to clindamycin. Metronidazole is favored over clindamycin for treatment of meningitis or when clindamycin resistance is a concern. A beta-lactam penicillin combined with a beta-lactamase inhibitor, such as piperacillin-tazobactam, can be useful for treatment in some infants.

PROGNOSIS The reported case-fatality rate for neonatal anaerobic septicemia varies from 4% to 45%.

Infant Botulism INCIDENCE Infant botulism was first recognized in the United States in 1976, and more than 1500 cases have been confirmed since then. Cases have been reported from many countries and from most regions of the United States. The greatest proportion of cases have occurred in California and the eastern Pennsylvania–New Jersey–Delaware area.34 The onset of illness peaks between 2 and 4 months of age, although disease occurring in the second week of life has been described. More than 90% of patients are younger than 6 months of age. The actual incidence is difficult to determine because most mildly ill infants go unrecognized, and many severely affected infants die suddenly and can be misdiagnosed as having sudden infant death syndrome.

synapses, preventing release of acetylcholine. Clinical manifestations include motor weakness in peripheral and cranial nerve distributions and autonomic instability. Studies in a murine model suggest that infants at particular risk for development of botulism transiently lack competitive microbial intestinal flora, or that alterations in motility or pH permit overgrowth of vegetative forms of ingested spores. Infants who have been exclusively breastfed and recently weaned are an at-risk group for botulism. Formula-fed infants with disease are hospitalized at a younger age than are breast-fed infants (7.6 weeks versus 13.7 weeks).

CLINICAL MANIFESTATIONS The clinical features associated with infant botulism range from mild disease to sudden death. Most recognized cases require hospitalization. The onset can be insidious or fulminant. The symptoms for which most parents seek medical attention are lethargy, poor feeding, and progressive weakness. In retrospect, most parents acknowledge, however, that the infant has been constipated. Typically, the infant is afebrile unless a secondary infection has occurred. Other than hypotonia and weakness, the results of physical examination can be normal early in the course. Cranial nerve palsies soon develop and can manifest as poor head control, ptosis, expressionless facies, and weak cry. Airway protection becomes compromised if gag, swallow, and suck reflexes are impaired. Progressive descending symmetric flaccid paralysis can ensue. Deep tendon reflexes can be normal, but often become diminished or absent as the paralysis progresses. Paralysis of respiratory muscles can be complicated by aspiration pneumonia. Generalized weakness and hypotonia can persist for 1 to 3 weeks without evidence of improvement.

DIAGNOSIS

Infant botulism is caused by C. botulinum, a ubiquitous, gram-positive, spore-forming, toxin-elaborating obligate anaerobe. Seven serologically distinct types of toxin have been identified (types A to G), but disease in the United States has been caused almost exclusively by toxin A or B. Clostridium barati and Clostridium butyricum, other species of Clostridium, elaborate neurotoxins similar to botulinum toxin and have been associated with infant botulism in rare cases. C. botulinum spores have been found worldwide in soil, water, agricultural products, and honey. Although extensive epidemiologic studies have been performed, no single source has been identified.

The diagnosis of infant botulism should be considered in any infant presenting with hypotonia, constipation, and poor feeding. Confirmation of the diagnosis requires isolation of C. botulinum or its toxin from the stool. The organism and occasionally the toxin can be isolated from stool for prolonged periods. Special culture techniques using enrichment and selective media are necessary to isolate C. botulinum. Toxin neutralization bioassay in mice performed on stool filtrate at a state laboratory or the CDC is the only reliable confirmatory test. A stool specimen for toxin assay is the test of choice for infant botulism.34 While awaiting the results of stool studies, a presumptive diagnosis can be made in infants with clinical features suggesting disease if typical findings are noted on electrodiagnostic studies. Nerve conduction study results are normal, but electromyography reveals abnormal spontaneous activity at the motor endplate with abundant brief, small-amplitude motor unit action potentials.

PATHOGENESIS

TREATMENT

Classic botulism follows ingestion of botulinal spores. Infant botulism occurs when ingested C. botulinum spores germinate in the intestine, releasing bacteria that colonize the colon and produce botulinum neurotoxin. Systemically absorbed toxin binds irreversibly to ganglionic and postganglionic cholinergic

A single dose of human-derived botulinum antitoxin, known as botulism immune globulin intravenous (BIG-IV) is efficacious in reducing hospitalization from 5.5 to 2.5 weeks and reducing by two thirds the rate of intubation required.34 The U.S. Food and Drug Administration (FDA) licensed

ETIOLOGY

Chapter 39  The Immune System

BIG-IV for treatment of infant botulism in 2003. It now is under the proprietary name of BabyBIG and is available through the California Department of Health Services (510540-2646) for treatment of botulism caused by types A or B C. botulinum.35 BIG-IV should be administered as early as is feasible in the course of the disease to interrupt neuromuscular blockade. Meticulous supportive care is also a mainstay of therapy. The goal is to provide nutritional and respiratory support, while avoiding potential complications. Patients should not be fed by mouth until adequate gag reflex and swallowing are observed. Gavage feeding can supply the necessary nutritional intake. Parenteral hyperalimentation may be necessary because normal bowel motility may not return for several weeks. Respiratory and cardiovascular status should be monitored closely. Intubation may be required for airway protection and respiratory failure. The average duration of hospitalization is approximately 1 month. Antibiotics are not beneficial and can exacerbate the disease process by causing the release of neurotoxin into the gut when bacteria are killed. Antimicrobial therapy is indicated for treatment of pneumonia or urinary tract infection that may develop. The use of aminoglycoside antibiotics should be avoided because of the possibility of potentiation of neuromuscular blockade.

PROGNOSIS Because the toxin binds irreversibly to the motor endplates, regeneration of nerve endings and their neuromuscular junctions is necessary for recovery. This often requires several weeks. Some centers have described relapse after initial improvement. A small percentage of infants have recurrence of symptoms after discharge home. No predictors of relapse are recognized. The case-fatality rate is estimated to be less than 2% for hospitalized patients.

821

PATHOGENESIS Listeria has been recovered from soil, sewage, and decayed vegetation. It is a well-known cause of sheep and cattle epizootics. Most non-neonatal human infections are acquired after ingestion of contaminated foods such as unpasteurized milk, soft cheeses, raw meat, and vegetables. Several epidemics of listeriosis in the general population have been reported in pregnant women and their offspring in association with consumption of Mexican-style soft cheese. Family contacts of patients with listeriosis can be colonized in their gastrointestinal tract. Maternal disease is most frequently documented in the third trimester of pregnancy, probably in association with the decline in cell-mediated immunity that occurs at 26 to 30 weeks of gestation.48 Infection acquired early in pregnancy can lead to abortion and, if acquired later, to stillbirth or premature labor and delivery. Transplacental transmission is believed to be the most significant mechanism for acquiring early-onset disease, although ingestion or aspiration of infected amniotic fluid before delivery and aspiration of infected secretions during delivery are possible. The development of late-onset disease usually is not associated with symptomatic maternal infection. Transmission can be vertical from a colonized but asymptomatic mother or from other colonized or infected caregivers, or it can be nosocomial. The distribution of Listeria serotypes from neonates with early-onset and late-onset disease differs. Early-onset disease is predominantly associated with serotypes Ia and Ib and occasionally IVb, whereas late-onset disease is associated most frequently with serotype IVb. Serotyping is useful in epidemiologic studies, but is not important clinically.

CLINICAL MANIFESTATIONS

Listeria monocytogenes can produce disease in healthy individuals, but infection is more prevalent in pregnant women, neonates, and immunocompromised hosts. A populationbased surveillance study by the CDC showed a geographic variation in the incidence of neonatal listeriosis. In the year after a large outbreak in California, the incidence was 24.3 per 100,000 live births in Los Angeles County, whereas in the rest of the United States, the incidence was 7.5 per 100,000 live births.

The signs and symptoms of early-onset and late-onset disease are indistinguishable from signs and symptoms seen in other neonatal bacterial infections. In early-onset disease, the mother often has symptoms of a flulike illness before onset of labor. Listeria is frequently isolated from blood cultures obtained from febrile mothers. Evidence of neonatal infection is apparent in most infants at or soon after birth. Many appear meconium stained, even infants born before 32 weeks’ gestation, and this is the result of a green-brown staining of the amniotic fluid. Sepsis and pneumonia are the most frequently observed clinical syndromes. Neonates can present with anorexia, lethargy, vomiting, respiratory distress, apnea, cyanosis, and a papular, pustular, or petechial rash. Late-onset listeriosis most commonly manifests as meningitis. Fever and irritability can be present. The onset is often insidious.

MICROBIOLOGY

DIAGNOSIS

Listeriosis is caused by L. monocytogenes, a nonsporulating, b-hemolytic, short, gram-positive bacillus. It is an intracellular pathogen and can be observed in polymorphonuclear leukocytes on Gram-stained smears of infected body fluids. Listeria is similar morphologically to diphtheroids and can be overlooked as a contaminant. Listeria organisms decolorize readily during the Gram-staining procedure and have been mistakenly identified as gram-negative organisms, including Haemophilus.

A maternal history of stillbirth or repeated spontaneous abortions should prompt a high index of suspicion for listeriosis. Blood and CSF cultures should be obtained in all infants. Listeria monocytogenes isolated from cultures of the maternal amniotic fluid or placental tissue when infants have earlyonset sepsis supports the diagnosis. Aspirates or biopsy specimens of the rash can reveal the organism. CSF findings in cases of Listeria meningitis are characteristic of bacterial meningitides, with a predominance of

Listeriosis INCIDENCE

822

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

polymorphonuclear leukocytes, an elevated protein concentration, and a depressed CSF glucose concentration. An increased number of mononuclear cells can be seen in CSF from some infants. On Gram stain, the organisms may be gram-variable and may look like diphtheroids, cocci, or diplococci. A peripheral leukocytosis with left shift, neutropenia, or thrombocytopenia can be observed. Anemia, possibly related to the production of hemolysin, is occasionally present.

TREATMENT Cephalosporins have no activity against Listeria. Ampicillin is the first line of therapy. An aminoglycoside together with ampicillin is synergistic in vitro and in animal models of infection, and is suggested for initial treatment of neonatal infection. After clinical response occurs, ampicillin alone can be given for less severe infections. Treatment should be continued for approximately 14 days in uncomplicated infections. Repeat lumbar puncture should be performed at 24 to 48 hours after the start of treatment to document CSF sterilization.

PROGNOSIS The case-fatality rate can approach 50% in early-onset infection with Listeria, but is less than 10% if disease occurs after the fifth day of life. In the absence of meningitis or other CNS complications, the prognosis is good even in premature infants. CNS sequelae can be observed in some infants after meningitis, but the prognosis has not been extensively studied.

Syphilis INCIDENCE The incidence of congenital syphilis parallels that of primary and secondary disease in women. The last national syphilis epidemic, which was followed by a congenital syphilis epidemic, occurred during the late 1980s and early 1990s. Congenital syphilis rates have declined yearly since 1991. During 2000-2002, the rate of congenital syphilis decreased 21% to 11.2 cases per 100,000 live births.14 Lack of prenatal care and limited prenatal care are risk factors for congenital syphilis.

MICROBIOLOGY Syphilis is caused by T. pallidum, a tightly coiled, motile spirochete. It is too narrow to be visualized by conventional light microscopy (0.09 to 0.18 mm wide 3 5 to 15 mm long), but can be detected on dark-field microscopy. It can be propagated in animal models of infection, but has not been cultured in vitro.

PATHOGENESIS AND PATHOLOGY Transmission of congenital syphilis is most frequently transplacental, although infection can be acquired by contact with genital lesions during delivery. Transplacental infection can occur at any stage of pregnancy and during any stage of maternal syphilis. Infants are more likely to become infected, however, if the mother had primary or secondary syphilis during pregnancy than if she acquired syphilis months or years before conception. The risk of fetal infection also increases with advancing gestational age. If early

gestational infection is not recognized, potential outcomes include miscarriage, stillbirth, premature delivery, or neonatal death. Approximately 30% to 40% of fetuses with congenital syphilis are stillborn. Mothers of infants with congenital syphilis should have testing performed for other sexually transmissible infections, such as hepatitis B and hepatitis C, Chlamydia, human immunodeficiency virus (HIV), and gonococcal infection. Previously, it was believed that fetal infection did not occur before 18 weeks’ gestation. It is now known that early gestational infection can occur, but because of fetal immunoincompetence, the pathologic changes in fetal tissues are not observed until after the fifth month of gestation. The infection disseminates hematogenously from the placenta to the fetus, so diffuse involvement is common. The pathologic changes observed in congenital syphilis resemble the changes seen in acquired disease. Fibrosis and obliterative endarteritis with histiocytic, plasmacytic, and lymphocytic perivascular infiltrates are typical findings present in most affected organs. Examination of the placenta and the umbilical cord can assist in the diagnosis of congenital disease. Histologic examination reveals focal villositis with endovascular and perivascular proliferation in the placenta and necrotizing funisitis on umbilical cord sections. Spirochetes may be visible on silver-stained specimens of the placenta or umbilical cord.

CLINICAL MANIFESTATIONS Congenital syphilis is classified into early and late stages based on age at onset. Early congenital syphilis occurs before 2 years of age and represents active infection and inflammation. Approximately two thirds of infected infants are asymptomatic at birth. A wide spectrum of clinical manifestations can be seen in symptomatic infants (Box 39-4). Most symptomatic infants have hepatosplenomegaly and bone changes on radiographs, and many are anemic. The enlarged liver is believed to be related to extramedullary hematopoiesis. A diffuse hepatitis with elevated aminotransferases and direct and indirect hyperbilirubinemia can be seen. Skeletal manifestations include metaphyseal osteochondritis, periostitis, and osteitis. Osseous des­ truction of the proximal medial tibial metaphysis (Wimberger sign) can be observed (Fig. 39-11). Symptomatic infants are more likely to have spirochetal invasion of the CNS, but laboratory evidence of CNS involvement is not always apparent. Late manifestations reflect the body’s response to early infection or to persistent inflammation. Cutaneous, dental, skeletal, ocular, auditory, and CNS involvement can occur (Box 39-5). The most frequently observed signs include frontal bossing, saddle nose deformity, short maxilla with high-arched palate, and Hutchinson triad.

DIAGNOSIS As noted by Ingall and colleagues,29 congenital syphilis should be considered in the differential diagnosis of any newborn with unexplained prematurity, hydrops fetalis of unknown etiology, or placental enlargement. An infant with persistent rhinitis; failure to thrive; or unexplained anemia, thrombocytopenia, jaundice, or hepatosplenomegaly should prompt consideration of congenital syphilis.

Chapter 39  The Immune System

BOX 39–4 Clinical Findings of Early Congenital Syphilis* n Nonimmune

hydrops fetalis growth restriction n Failure to thrive n Generalized lymphadenopathy n Bone abnormalities Periostitis Osteochondritis (Wimberger sign) Osteitis n Hepatosplenomegaly (with or without elevated aminotransferases, jaundice) n Mucocutaneous lesions Pemphigus syphiliticus (vesiculobullous eruption, contagious) Maculopapular eruption Mucous patches of palate, perineum, intertriginous areas Condyloma latum n Persistent rhinitis (snuffles, contagious) n Pneumonitis (pneumonia alba) n Nephrotic syndrome n Neurologic abnormalities Syphilitic leptomeningitis n Hematologic abnormalities Leukocytosis Leukopenia Anemia, Coombs-negative hemolytic Thrombocytopenia n Ocular abnormalities Chorioretinitis Uveitis n Intrauterine

823

BOX 39–5 Clinical Manifestations of Late Congenital Syphilis* DENTAL Hutchinson teeth (notched, peg-shaped upper central incisors)† Mulberry molars (multiple small cusps) BONE Frontal bossae of Parrot Saddle nose deformity Short maxillae High-arched palate Higouménakis sign (sternoclavicular thickening) Flaring scapulae Saber shins (anterior bowing of tibia) Clutton joints (painless synovitis, hydrarthrosis) CUTANEOUS Rhagades (linear scars radiating from mouth, nares, and anus) OCULAR Interstitial keratitis† NEUROLOGIC Eighth cranial nerve deafness† Mental retardation Hydrocephalus Cranial nerve palsies Seizure disorder *Defined as age .2 years. † Features of Hutchinson triad.

*Defined as age ,2 years.

Figure 39–11.  Osteochondritis and periostitis of the bones in both lower extremities of a patient with congenital syphilis. The radiolucent area seen at the medial aspect of the proximal tibial metaphyses is known as the Wimberger sign; the zone of translucency visible proximal to the epiphyseal ends of both tibias is called the Wegner sign.

824

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

T. pallidum cannot be cultured in vitro. Treponemes can be detected by dark-field examination or direct fluorescent antibody staining of material from mucocutaneous lesions, the umbilical cord or placenta, amniotic fluid, nasal discharge, or postmortem tissue.51 More commonly, the diagnosis is suggested by the history or clinical and radiographic findings, or both, and confirmed by serologic testing. Serologic tests for syphilis (STS) include nontreponemal and treponemal antibody tests. Nontreponemal tests, such as the rapid plasma reagin test or the Venereal Disease Research Laboratory (VDRL) test, use a purified cardiolipin and lecithin antigen to detect reagin, a nonspecific antibody produced in response to T. pallidum infection. Biologic false-positive results can be seen with nontreponemal STS in patients with autoimmune disorders. Quantitative titers of nontreponemal tests can be followed to evaluate response to therapy. Treponemal serologic tests include the fluorescent treponemal antibody absorption test and the T. pallidum particle agglutination test. Because both tests remain positive after initial infection, serial titers are not helpful for evaluating response to therapy. A fluorescent treponemal antibody absorption IgM test and a polymerase chain reaction technique for detection of T. pallidum in tissues and body fluids have been developed, but are not commercially available. An infant should be evaluated for congenital syphilis when the maternal titer has increased fourfold, if the infant’s titer is fourfold greater than the mother’s, or if the infant is symptomatic. In addition, if born to a mother with positive nontreponemal and treponemal tests, the infant should be evaluated further if the mother had untreated, inadequately treated, or undocumented treatment for syphilis; if erythromycin was used for treatment during pregnancy; if treatment during pregnancy occurred less than 1 month before delivery; if the expected decrease in nontreponemal antibody has not been documented; or if there is insufficient serologic follow-up to document the response to treatment. The evaluation of an infant with suspected congenital syphilis should include (1) physical examination; (2) quantitative nontreponemal test of serum (not cord blood) for syphilis;

(3) routine CSF evaluation and CSF VDRL; (4) long bone radiographs; (5) complete blood count and platelet count; and (6) other tests, such as chest radiography, as clinically indicated. Criteria for the diagnosis of proven or highly probable congenital syphilis include any of the following: (1) physical, laboratory, or radiographic evidence of active infection; (2) placenta or umbilical cord is positive for treponemes using specific direct fluorescent antibody staining or dark-field test; (3) active CSF VDRL result; or (4) an infant quantitative serum nontreponemal STS titer fourfold or more than that of the mother. A negative nontreponemal STS at delivery does not exclude recent maternal infection that has not yet elicited an antibody response, or has elicited only IgM antibodies that do not cross the placenta.

TREATMENT Serologic tests on the infant and on the mother during pregnancy and at delivery, maternal antepartum antimicrobial therapy, and the infant’s clinical status must be considered before treatment plans can be made. Treatment should be completed before nursery discharge for infants with proven or highly probable disease. It should be considered in infants who are asymptomatic with normal CSF and radiographic examination results under the conditions dictated by the maternal treatment history (Table 39-22).1 Jarisch-Herxheimer reactions, manifested by fever, tachypnea, tachycardia, hypotension, prominence of cutaneous lesions, and death, have been reported after initiation of treatment. The cause of this reaction is unknown, although release of endotoxin from the spirochetes has been suggested.

FOLLOW-UP Infants with reactive nontreponemal STS must have serial quantitative tests after discharge home. Uninfected infants with reactive STS from transplacental transfer of maternal antibody should have a decrease in titer by 3 months of age and a nonreactive test at 6 months. To confirm therapeutic response, treated infants with suspected or proven disease

TABLE 39–22  Recommended Therapy for Neonates (,4 Weeks Old) with Congenital Syphilis* Clinical Status

Treatment

Proven or highly probable disease

Aqueous crystalline penicillin G for 10 days†

Normal examination and nontreponemal test the same or less than four-fold the maternal result with maternal treatment history: None, inadequate penicillin treatment‡

Aqueous crystalline penicillin G IV for 10-14 days† or Clinical, serologic follow-up, and benzathine penicillin G IM, single dose§

Adequate therapy given .4 wk before delivery; mother has no evidence of reinfection or relapse

Clinical, serologic follow-up, and benzathine penicillin G IM, single dose¶

*If .1 day is missed, the course should be restarted. † Aqueous crystalline penicillin G 50,000 U/kg IV given every 12 hours for the first 7 days of life and every 8 hours thereafter. Alternatively, procaine penicillin G 50,000 U/kg IM single daily dose for 10 days can be given. ‡ Maternal treatment is termed inadequate when her penicillin dose is unknown, is undocumented, or was inadequate; if she received erythromycin or other nonpenicillin regimen; if treatment was given ,4 weeks before delivery; or the response to treatment was not documented by showing a fourfold decrease in titer of a nontreponemal test for syphilis. § Benzathine penicillin G 50,000 U/kg IM. Some experts recommend aqueous crystalline penicillin G for proven or highly probable disease. ¶ Some experts would not treat the infant, but would provide close serologic follow-up. Adapted from American Academy of Pediatrics: Syphilis. In Pickering LK et al, editors: Red Book: 2009 report of the Committee on Infectious Diseases, 28th ed, Elk Grove Village, IL, American Academy of Pediatrics, p 645.

Chapter 39  The Immune System

should have quantitative nontreponemal STS titers at 2 to 4 months, 6 months, and 12 months after completion of therapy, or until the test becomes nonreactive or the titer has decreased by fourfold. Retreatment should be considered for any child with persistent, unchanging nontreponemal STS titers. CSF evaluations at 6-month intervals, in addition to serial serum quantitative nontreponemal titers, are recommended for infants with congenital neurosyphilis. Retreatment should be undertaken if the CSF VDRL is positive at 6 months of age. Treatment of asymptomatic infected infants in the first 3 months of life usually prevents development of late stigmata. Late congenital syphilis can develop if congenital infection goes unrecognized and untreated. Stigmata of late congenital syphilis can develop in infants symptomatic at birth despite appropriate treatment in the neonatal period.

PREVENTION Routine prenatal screening and penicillin therapy of infected women and their partners can prevent congenital syphilis. Prenatal screening consisting of nontreponemal STS in the first trimester and at delivery is recommended. Another test at the beginning of the third trimester should be considered in areas with a high prevalence of syphilis. Screening of the mother at delivery, in addition to the infant, is recommended to avoid potential false-negative infant serologic test results related to low antibody concentrations.

Tetanus Neonatorum INCIDENCE Tetanus neonatorum is rare in the United States, but remains a significant cause of neonatal morbidity and mortality in developing countries.

MICROBIOLOGY Clostridium tetani is a slender, anaerobic, spore-forming, gram-positive bacillus. C. tetani is present in soil and can be found in human and animal feces. Risk factors associated with the higher incidence in developing countries include lack of maternal immunization against tetanus, childbirth under unhygienic conditions, use of contaminated umbilical stump dressings, and rituals involving the application of manure to the umbilicus.

PATHOLOGY No specific pathologic lesions are attributed to this infectious agent, although changes in the brain and spinal cord have been described in patients dying of tetanus.

PATHOGENESIS The signs and symptoms of tetanus are the result of toxin production, rather than invasive infection. The umbilical stump contaminated by soil or feces can serve as a portal of entry for the spores. When the spores germinate, the bacteria produce two toxins: tetanolysin, which hemolyzes red blood cells in vitro, but does not seem to be important in vivo, and tetanospasmin. Tetanospasmin is second only to botulinum toxin in potency. By inhibiting release of acetylcholine, it interferes with neuromuscular transmission, resulting in

825

muscular contractions. Tetanospasmin affects the motor endplates of skeletal muscle, the spinal cord and brain, and occasionally the sympathetic nervous system. When toxin becomes fixed to nervous tissue, it cannot be neutralized by antitoxin. The mechanism by which the toxin reaches the CNS is unclear. Transportation within the nerves is likely. The incubation period ranges from 2 to 21 days.

CLINICAL MANIFESTATIONS The clinical syndrome differs from that seen in older children and adults. Most infants are irritable and show diminished suck and cessation of crying. Fever is often noted at presentation. The disease progresses to generalized rigidity with muscle spasms and seizures. Flexor spasms are exacerbated by stimulation. Some infants become cyanotic if the spasms are prolonged. Testing of deep tendon reflexes usually shows hyperreflexia. Extreme flexion of the toes is frequently observed. The hallmark of tetanus in older patients is lockjaw, also referred to as risus sardonicus. Lockjaw does not develop in most infants, but severe reflex spasms of the masseter muscles do occur if the jaw is moved for feeding. Cardiorespiratory difficulties, including tachycardia, tachypnea, cyanosis, or apnea, can be observed. Laryngoglottal spasm can predispose to aspiration pneumonia. Pulmonary processes such as bronchopneumonia or hemorrhage are a frequent cause of death in neonatal tetanus.

DIFFERENTIAL DIAGNOSIS Cultures of the blood and the wound (umbilical stump) are invariably sterile. Diagnosis is made by clinical presentation and exclusion of other possibilities. Infants can be misdiagnosed as having neonatal seizures.

TREATMENT Supportive care is essential. Maintenance of a clear airway and provision of adequate ventilation are most important. Meticulous care must be exercised to decrease the likelihood of secondary bacterial pulmonary infections. The bladder may need to be catheterized if it does not empty spontaneously. Because stimulation can precipitate spasms and seizures, timing of medical interventions should be coordinated, and exposure to nonessential external stimuli should be limited. Diazepam is an effective agent in controlling the tonic spasms of tetanus. Barbiturates may be helpful as an adjunct to therapy and may help sedate the patient. Metronidazole should be given for a 10- to 14-day course to decrease the number of vegetative forms of C. tetani and is the treatment of choice. Penicillin G is an alternative antibiotic. Human tetanus immune globulin given intramuscularly in a single dose is recommended to neutralize circulating unbound toxin. Débridement of the entry site, an essential component of therapy in adult disease, is performed in some cases of neonatal tetanus, but wide excision is not indicated.

PREVENTION A well-immunized mother offers the best protection against tetanus neonatorum. Because antitoxin crosses the placenta, infants born to mothers whose tetanus immunizations are current should have adequate antitoxin antibody concentrations. Aseptic obstetric and neonatal care minimizes the likelihood of

826

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

introduction of C. tetani spores. Infants who recover from disease require routine tetanus immunizations because disease does not usually confer immunity.

Tuberculosis INCIDENCE Since the early 1990s, the incidence of tuberculosis in the United States has increased, especially among women of childbearing age. Perinatal tuberculosis is uncommon, but management of an infant born to a mother with tuberculous infection is common.

MICROBIOLOGY Mycobacterium tuberculosis is the causative agent of tuberculosis. M. tuberculosis is a slow-growing, obligately aerobic, acid-fast bacillus (AFB). Neonatal infections with other strains of mycobacteria are uncommon. Mycobacterium bovis infections were occasionally seen before pasteurization of milk became standard. Neonatal infection with atypical or environmental mycobacteria is rare.

PATHOGENESIS AND PATHOLOGY Infected mothers can transmit infection to their fetus transplacentally if primary tuberculous infection (asymptomatic or miliary) is acquired during pregnancy, or if the mother has tuberculous endometritis.60,62 Postpartum transmission can occur if the mother has active pulmonary tuberculosis with cavitation. Neonatal tuberculosis can be acquired intrapartum from ingestion or aspiration of infected secretions and postnatally by inhalation of infected droplets. Before aseptic technique was routinely used, transmission from contamination of the skin or mucous membranes during circumcision was described. The most common mode of transmission for perinatal disease is inhalation of infected droplets (Table 39-23). Coinfection with HIV increases the risk of extrapulmonary disease, and mothers with extrapulmonary tuberculosis, including pleural effusions, and endometrial, miliary, or meningeal disease are more likely to transmit infection to the fetus. Many infants born to infected mothers do not contract disease, however.

Primary complex formation in the porta hepatis of the liver occurs with congenital infection transmitted hematogenously. Miliary tubercles can be found in the spleen, bone marrow, lung, kidney, adrenal glands, and brain. Gross or microscopic placental lesions may be detected. Primary complex formation in the lung can reflect hematogenous spread, aspiration of infected amniotic fluid or vaginal secretions, or postnatal inhalation.

CLINICAL MANIFESTATIONS Signs and symptoms of congenital tuberculosis are vague. Some infants may be symptomatic at birth, whereas others do not develop symptoms until late in the first month of life.38 The appearance on chest radiographs frequently is abnormal. Many infected newborns are premature and have hepatosplenomegaly, respiratory distress, fever, reluctance to feed, and lethargy. Miliary tuberculosis, lymphadenopathy, and otitis media with tympanic membrane perforations, otorrhea, and facial nerve palsy can be seen as well. Meningitis, pleural effusions, and pulmonary cavitations are unusual. Neonatal tuberculosis is manifested by fever, vomiting, cough, tachypnea, and weight loss. Hepatosplenomegaly occasionally occurs. Unless hepatic primary complex formation is present, it is often difficult to distinguish congenital from neonatal tuberculosis.

DIAGNOSIS The diagnosis of tuberculosis can be established by demonstration of organisms on AFB smears and growth of M. tuberculosis in special culture media. The infant should have a tuberculin skin test (TST), a chest radiograph, and AFB stains and cultures of CSF and gastric and tracheal aspirates. The infant’s TST is usually nonreactive even in the presence of active disease. Because treatment of an infected infant does not prevent the skin test from becoming reactive, repeat skin testing should be performed 3 months later if the initial skin test is negative. Other sites that can yield AFB growth include lung or liver tissue, lymph node, middle ear fluid, urine, and bone marrow aspirate.

TREATMENT Because perinatal tuberculosis is uncommon, the safety and efficacy of antituberculous agents in neonates have not been determined. Standard therapy for older infants consists of

TABLE 39–23  Perinatal Transmission of Tuberculosis Maternal Focus of Infection

Mode of Spread

Timing

Relative Frequency

Pneumonia with cavitary lesion*

Inhalation of infected droplets

Postnatal

Most common

Amniotic infection after rupture of placental caseous lesion

Aspiration or ingestion of infected fluid

Congenital or intrapartum

Less common

Placentitis after miliary or endometrial tuberculosis

Hematogenous through umbilical vein

Congenital

Rare

Cervicitis

Direct contact, aspiration, or ingestion

Intrapartum

Rare

Mastitis

Ingestion of infected milk

Postnatal

Extremely rare

*Any caregiver with cavitary pulmonary tuberculosis can transmit infection to the infant.

Chapter 39  The Immune System

isoniazid, rifampin ethambutol, and pyrazinamide daily for 2 months followed by twice-weekly therapy with isoniazid and rifampin alone (Table 39-24).61 An aminoglycoside is also initiated when treating potentially life-threatening disease, such as meningitis, until susceptibility testing is completed. The duration of therapy is 6 months in older children without meningitis. The optimal duration has not been determined for neonates, but prolonged treatment has been suggested because of potential neonatal immunoincompetence. Corticosteroids have been used in tuberculous meningitis to reduce inflammation and decrease intracranial pressure and in endobronchial disease to reduce tracheal compression. Appropriate antituberculous agents must be administered concurrently with steroid therapy to avoid dissemination of infection. The case-fatality rate for congenital tuberculosis is approximately 50%. Most deaths are related to undiagnosed and untreated disease. The prognosis is worse for premature infants than for term infants with congenital infection.

MANAGEMENT OF INFANTS BORN TO MOTHERS WITH POSITIVE PURIFIED PROTEIN DERIVATIVE SKIN TESTS Regardless of the mother’s clinical status, symptomatic newborns should have a thorough investigation for bacterial sepsis and tuberculous disease. Chest radiograph; CSF examination; blood and CSF bacterial cultures; and mycobacterial stains and cultures of CSF, gastric aspirates, and tracheal aspirates should be obtained. Decisions on antituberculous treatment of the infant should be based on the mother’s status and on the results of the infant’s evaluation. For asymptomatic neonates, investigation and treatment are guided by the mother’s clinical status. If a mother with a

827

positive TST is clinically well, has a negative chest radiograph, and is not believed to have active disease, no intervention is necessary, provided that the household contacts have been investigated, and active disease has been excluded. Mothers with an abnormal appearance on chest radiograph must have an investigation for tuberculosis, including AFB stains and cultures of sputum and testing for HIV infection. Separation of the mother and the infant is recommended until the mother and infant have been evaluated; if tuberculosis disease is suspected, until the mother and infant are receiving appropriate antituberculous therapy, the mother wears a mask and understands and expresses her willingness to comply with infection control measures. If there is no evidence of neonatal infection, the infant should be given isoniazid prophylaxis and reevaluated in 3 to 4 months with another TST. Isoniazid can be discontinued at 3 months if the second TST is negative, if the infant is well, and if there is no active infection in the household. Three- or four-drug antituberculous therapy should be started after cultures are obtained if there is evidence of disease in the infant. Bacille Calmette-Guérin immunization of the infant should be considered only in limited and select circumstances, such as if the mother or another household contact has possible multidrug-resistant tuberculosis or has poor adherence to treatment.

PREVENTION OF INFECTION IN THE NEONATAL INTENSIVE CARE UNIT Barrier Nursing Technique The use of masks and gowns by nursery personnel does not significantly alter staphylococcal colonization or infection rates in infants. Hand washing, if performed adequately, is

TABLE 39–24  Commonly Used Drugs for Treatment of Tuberculosis in Children Drug*

Daily (mg/kg/d)

Isoniazid

10-15; maximum 300 mg

Rifampin

Twice Weekly† (mg/kg per Dose)

Side Effects

Comments

20-40; maximum 900 mg

Mild hepatic enzyme elevation, hepatitis‡, peripheral neuritis, hypersensitivity

Monitor liver function tests monthly

10-20; maximum 600 mg

10-20; maximum 600 mg

Hepatitis‡, vomiting, thrombocytopenia, orange discoloration of secretions

Monitor liver function tests monthly

Pyrazinamide

20-40; maximum 2 g

50-70; maximum 2 g

Hepatotoxicity, hyperuricemia

Monthly uric acid and liver function tests

Streptomycin

20-40; maximum 1 g

20-40; maximum 1 g

Ototoxicity, nephrotoxicity

Monitor renal function, consider hearing test

Ethambutol

15-25; maximum 2.5 g

50; maximum 2.5 g

Optic neuritis, color blindness, decreased visual acuity

Monitor visual fields, visual acuity, and color discrimination frequently

†

*All drugs may be given PO except streptomycin, which must be given IM. † See text for details and duration. ‡ Incidence of hepatotoxicity increases when isoniazid and rifampin are used together, especially when the isoniazid dosage exceeds 10 mg/kg/d. Adapted from Starke JR: Perinatal tuberculosis, Semin Pediatr Infect Dis 5:20, 1994.

828

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

effective in reducing the incidence of nosocomial nursery infections. A 3% solution of hexachlorophene soap is effective in reducing infection by gram-positive bacteria, and various iodinated preparations, including iodinated chlorhexidine (0.5%), are effective against gram-negative organisms. Studies suggest that alcoholic chlorhexidine hand washing agents are effective against drug-resistant strains of Enterococcus faecium and Enterobacter cloacae. Physicians should be aware that some organisms, including P. aeruginosa, can survive in many antiseptic solutions.

Bathing of Infants Historically, infants were bathed with hexachlorophene to decrease the risk of gram-positive infections. In the 1970s, infants in France developed a toxic encephalopathy attributed to excessive amounts of hexachlorophene in talcum powder. Subsequent neuropathologic studies in the United States revealed an association between brainstem vacuolar lesions and hexachlorophene baths especially in premature infants. Routine bathing with hexachlorophene was discontinued in the early 1970s after this association was recognized. The Committee on the Fetus and Newborn of the American Academy of Pediatrics recommends that the first bath be postponed until a newborn is thermally stable. Nonmedicated soap and water should be used. Sterile sponges soaked in warm water can be used. During nursery outbreaks of S. aureus infection, hexachlorophene bathing of the diaper area can be undertaken to prevent staphylococcal disease.21 Surveillance of neonatal infections, prompt isolation and treatment of infected infants, and adequate hand washing should be used to prevent nosocomial infections. Crowding should be avoided, and cohort nursing techniques should be applied whenever possible. Hexachlorophene can still be used as an antibacterial agent by nursery personnel.

Cord Care The umbilicus is a direct portal of entry to the bloodstream, and serves as a site from which other areas of the skin may become contaminated or colonized. No single method of cord care has proved superior in preventing colonization and the development of omphalitis. Options for cord care include application of alcohol, triple dye (brilliant green, proflavin hemisulfate, and crystal violet), or antimicrobial agents such as bacitracin ointment. Alcohol hastens drying of the cord, but probably is ineffective in preventing cord colonization or omphalitis.

Resuscitation and Ventilatory Equipment Gram-negative microorganisms, particularly Pseudomonas, Aeromonas, and Serratia, have been associated with sporadic and epidemic infection in the nursery. Routine cultures of medications, nebulizers, and inhalation therapy equipment should be performed, and equipment should be changed frequently to prevent such infections. Use of umbilical, arterial, and venous catheters for parenteral alimentation also has been accompanied by an increased risk

of health care–associated infections. Meticulous care is required in the insertion and care of catheters and in the preparation of intravenous solutions for use in total parenteral alimentation. Administration of fluids should be discontinued if signs of inflammation, thrombosis, or purulence are observed. All apparatus used for intravenous administration should be replaced at intervals to decrease the hazard of extrinsic contamination.

Antibiotic Prophylaxis The effectiveness of intravenous antimicrobial prophylaxis in newborns has not been proven. Indiscriminate use of antibiotics can result in colonization or infection with drug-resistant strains of bacteria or with fungi. The only role for antimicrobial prophylaxis in newborns is for the prevention of gonococcal ophthalmitis or tuberculous disease.

Immune Globulin There have been numerous in vitro, animal model, and human studies evaluating the use of IVIG in the prevention of infections. Results of several well-designed human trials using IVIG to prevent nosocomial infections in preterm neonates have been published.2,36 The results of the large, multicenter, randomized, controlled trial led by Fanaroff and associates18 revealed no significant decrease in the incidence of nosocomial infections, in the number of days hospitalized, or in mortality rate in the premature infants given IVIG. One possible explanation for the lack of protection against health care–associated infections could be that pooled IVIG does not contain adequate amounts of specific antibody against the most common neonatal nosocomial pathogens—CONS and Candida species. Even specific antibody may not be protective against health care–associated infections, however, especially if they are associated with foreign bodies such as indwelling catheters. Administration of an IVIG derived from donors with high titers of antibody to surface adhesins of S. aureus and S. epidermidis failed to reduce the incidence of staphylococcal bacteremia in premature infants.16 In the future, development of specific, high-titered immunoglobulin preparations against the specific pathogens may be beneficial in prevention of some neonatal infections.

REFERENCES 1. American Academy of Pediatrics: Syphilis. In Pickering LK et al, editors: Red Book: 2006 report of the Committee on Infectious Diseases, 27th ed, Elk Grove Village, IL, 2006, American Academy of Pediatrics, p 631. 2. Baker CJ et al: Intravenous immune globulin for the prevention of nosocomial infection in low-birth-weight neonates. The Multicenter Group for the Study of Immune Globulin in Neonates, N Engl J Med 327:213, 1992. 3. Baley JE, Silverman RA: Systemic candidiasis: cutaneous manifestations in low birth weight infants, Pediatrics 82:211, 1988.

Chapter 39  The Immune System 4. Bedford Russell AR et al: Plasma granulocyte colony-stimulating factor concentrations (G-CSF) in the early neonatal period, Br J Haematol 86:642, 1994. 5. Benjamin DK et al: A blinded, randomized, multicenter study of an intravenous Staphylococcus aureus immune globulin, J Perinatol 26:290, 2006. 6. Berkun Y et al: Acute otitis media in the first two months of life: characteristics and diagnostic difficulties, Arch Dis Child 93:690, 2008. 7. Berman PH, Banker BQ: Neonatal meningitis: a clinical and pathological study of 29 cases, Pediatrics 38:6, 1966. 8. Bizzarro MJ et al: Changing patterns in neonatal Escherichia coli sepsis and ampicillin resistance in the era of intrapartum antibiotic prophylaxis, Pediatrics 121:689, 2008. 9. Bizzarro MJ et al: Seventy-five years of neonatal sepsis at Yale: 1928-2003, Pediatrics 116:595, 2005. 10. Bonadio WA et al: Reliability of observation variables in distinguishing infectious outcome of febrile young infants, Pediatr Infect Dis J 12:111, 1993. 11. Bromberger P et al: The influence of intrapartum antibiotics on the clinical spectrum of early-onset group B streptococcal infection in term infants, Pediatrics 106:244, 2000. 12. Buck C et al: Interleukin-6: a sensitive parameter for the early diagnosis of neonatal bacterial infection, Pediatrics 93:54, 1994. 13. Carr R et al: Granulocyte-macrophage colony stimulating factor administered as prophylaxis for reduction of sepsis in extremely preterm, small for gestational age neonates (the PROGRAMS trial): a single-blind, multicentre, randomized controlled trial, Lancet 373:226, 2009. 14. Centers for Disease Control and Prevention: Congenital syphilis—United States, 2002, MMWR Morb Mortal Wkly Rep 53:716, 2004. 15. Centers for Disease Control and Prevention: Prevention of perinatal group B streptococcal disease: revised guidelines from CDC, MMWR Morb Mortal Wkly Rep 51:1, 2002. 16. DeJonge M et al: Clinical trial of safety and efficacy of IHN-A21 for the prevention of nosocomial staphylococcal bloodstream infection in premature infants, J Pediatr 151:260, 2007. 17. Efrat M et al: Neonatal mastitis—diagnosis and treatment, Isr J Med Sci 31:558, 1995. 18. Fanaroff AA et al: A controlled trial of intravenous immune globulin to reduce nosocomial infections in very-low-birthweight infants, N Engl J Med 330:1107, 1994. 19. Fielkow S et al: Cerebrospinal fluid examination in symptomfree infants with risk factors for infection, J Pediatr 119:971, 1991. 20. Fraser N et al: Neonatal omphalitis: a review of its serious complications, Acta Paediatr 95:519, 2006. 21. Freeman RK et al, editors: Infection control. In American Academy of Pediatrics Committee on the Fetus and Newborn: Guidelines for perinatal care, 3rd ed, Elk Grove Village, IL, 1992, American Academy of Pediatrics Press, p 3. 22. Girardin EP et al: Serum tumour necrosis factor in newborns at risk for infections, Eur J Pediatr 149:645, 1990. 23. Hammerschlag MR: Neonatal conjunctivitis. Pediatr Ann 22:346, 1993.

829

24. Hammerschlag MR et al: Efficacy of neonatal ocular prophylaxis for the prevention of chlamydial and gonococcal conjunctivitis, N Engl J Med 320:769, 1989. 25. Hammerschlag MR et al: Treatment of neonatal chlamydial conjunctivitis with azithromycin, Pediatr Infect Dis J 17:1049, 1998. 26. Harvey D et al: Bacterial meningitis in the newborn: a prospective study of mortality and morbidity, Semin Perinatol 23:218, 1999. 27. Healy CM et al: Emergence of new strains of methicillinresistant Staphylococcus aureus in a neonatal intensive care unit, Clin Infect Dis 39:1460, 2004. 28. Hyde TB et al: Trends in incidence and antimicrobial resistance of early-onset sepsis: population-based surveillance in San Francisco and Atlanta, Pediatrics 110:690, 2002. 29. Ingall D et al: Syphilis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed., Philadelphia, 2006, Saunders, p 545. 30. Jenson HB, Pollock BH: The role of intravenous immunoglobulin for the prevention and treatment of neonatal sepsis, Semin Perinatol 22:50, 1998. 31. Klinger G et al: Predicting the outcome of neonatal bacterial meningitis, Pediatrics 106:477, 2000. 32. Krediet T et al: The predictive value of C-reactive protein and I:T ratio in neonatal infection, J Perinat Med 20:479, 1992. 33. Levent F et al: Early outcomes from group B streptococcal (GBS) meningitis in the intrapartum antibiotic prophylaxis (IAP) era. Paper presented at Pediatric Academic Society Annual Meeting, 2009, Philadelphia, PA; Abstract 2848.470. 34. Long SS: Infant botulism, Concise Rev Pediatr Infect Dis 20:707, 2001. 35. Long SS: Infant botulism and treatment with BIG-IV (BabyBIG), Pediatr Infect Dis J 26:261, 2007. 36. Magny JF et al: Intravenous immunoglobulin therapy for prevention of infection in high-risk premature infants: report of a multicenter, double-blind study, Pediatrics 88:437, 1991. 37. Manroe BL et al: The neonatal blood count in health and disease, I: reference values for neutrophilic cells, J Pediatr 95:89, 1979. 38. Mazade MA et al: Congenital tuberculosis presenting as sepsis syndrome: case report and review of the literature, Pediatr Infect Dis J 20:439, 2001. 39. McCracken GH Jr et al: Cerebrospinal fluid interleukin-1b and tumor necrosis factor concentrations and outcome from neonatal gram-negative enteric bacillary meningitis, Pediatr Infect Dis J 8:155, 1989. 40. Mouzinho A et al: Revised reference ranges for circulating neutrophils in very-low-birth-weight neonates, Pediatrics 94:76, 1994. 41. Nelson JD: Antibiotic therapy for newborns. In Nelson JD, editor: Pocketbook of pediatric antimicrobial therapy, Baltimore, 1998, Williams & Wilkins, p 14. 42. Nissen MD et al: Congenital and neonatal pneumonia, Paediatr Resp Rev 8:195, 2007. 43. Noel GJ et al: Anaerobic bacteremia in a neonatal intensive care unit: an eighteen-year experience, Pediatr Infect Dis J 7:858, 1988.

830

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

44. Palazzi DL et al: Bacterial sepsis and meningitis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed., Philadelphia, 2006, Saunders, p 247. 45. Phares CR et al: Epidemiology of invasive group B streptococcal disease in the United States, 1999-2005. JAMA 299:2056, 2008. 46. Philip AGS et al: Neutrophil elastase in the diagnosis of neonatal infection, Pediatr Infect Dis J 13:323, 1994. 47. Polin RA et al: Neonatal sepsis, Adv Pediatr Infect Dis 7:25, 1992. 48. Posfay-Barbe KM, Wald ER: Listeriosis, Pediatr Rev 25:151, 2004. 49. Rennels MB, Levine MM: Classical bacterial diarrhea: perspectives and update: Salmonella, Shigella, Escherichia coli, Aeromonas, and Plesiomonas. Pediatr Infect Dis J 5(suppl):S91, 1986. 50. Rodriguez AF et al: Cerebrospinal fluid values in the very low birth weight infant, J Pediatr 116:971, 1990. 51. Sánchez PJ: Congenital syphilis, Adv Pediatr Infect Dis 7:161, 1992. 52. Santana Reyes C et al: Role of cytokines (interleukin-1b, 6, 8, tumour necrosis factor-alpha, and soluble receptor of interleukin-2) and C-reactive protein in the diagnosis of neonatal sepsis, Acta Paediatr 92:221, 2003. 53. Sarff LD et al: Cerebrospinal fluid evaluation in neonates: comparison of high-risk infants with and without meningitis, J Pediatr 88:473, 1976. 54. Sawardekar KP: Changing spectrum of neonatal omphalitis, Pediatr Infect Dis J 23:22, 2004. 55. Schrag SJ et al: A population-based comparison of strategies to prevent early-onset group B streptococcal disease in neonates, N Engl J Med 347:233, 2002. 56. Schuchat A et al: Population-based risk factors for neonatal group B streptococcal disease: results of a cohort study in metropolitan Atlanta, J Infect Dis 162:672, 1990. 57. Shurin PA et al: Bacterial etiology of otitis media during the first six weeks of life, J Pediatr 92:893, 1978. 58. Singer DB: Infections of fetuses and neonates. In Wigglesworth JS et al, editors: Textbook of fetal and perinatal pathology, Boston, 1998, Blackwell Scientific, p 454. 59. Spiegel R et al: Acute neonatal suppurative parotitis: case reports and review, Pediatr Infect Dis J 23:76, 2004. 60. Starke JR: Tuberculosis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed. Philadelphia, Saunders, 2006, p 581. 61. Starke JR: Perinatal tuberculosis, Semin Pediatr Infect Dis 5:20, 1994. 62. Starke JR: Tuberculosis: an old disease but a new threat to the mother, fetus, and neonate, Clin Perinatol 24:107, 1997. 63. Stevens JP et al: Long term outcome of neonatal meningitis, Arch Dis Child Fetal Neonatal Ed 88:F179, 2003. 64. Stoll BJ et al: Changes in pathogens causing early-onset sepsis in very-low-birth-weight infants, N Engl J Med 347:240, 2002. 65. Stoll BJ et al: Early-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network, J Pediatr 129:72, 1996. 66. Stoll BJ et al: Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network, Pediatrics 110:285, 2002.

67. Stoll BJ et al: To tap or not to tap: high likelihood of meningitis without sepsis among very low birth weight infants, Pediatrics 113:1181, 2004. 68. Teele DW et al: Epidemiology of otitis media during the first seven years of life in children in greater Boston: a prospective cohort study, J Infect Dis 160:83, 1989. 69. Tegtmeyer FK et al: Elastase a1-proteinase inhibitor complex, granulocyte count, ratio of immature to total granulocyte count, and C-reactive protein in neonatal septicaemia, Eur J Pediatr 151:353, 1992. 70. Thureen PJ et al: Failure of tracheal aspirate cultures to define the cause of respiratory deteriorations in neonates, Pediatr Infect Dis J 12:560, 1993. 71. Weinberg GA et al: Laboratory aids for diagnosis of neonatal sepsis. In Remington JS et al, editors: Infectious diseases of the fetus and newborn infant, 6th ed. Philadelphia, 2006, Saunders, p 1207. 72. Wiswell TE et al: Declining frequency of circumcision: implications for changes in the absolute incidence and male to female sex ratio of urinary tract infections in early infancy, Pediatrics 79:338, 1987. 73. Wiswell TE et al: No lumbar puncture in the evaluation for early neonatal sepsis: will meningitis be missed? Pediatrics 95:803, 1995.

PART 3

Fungal and Protozoal Infections Morven S. Edwards

FUNGAL INFECTIONS Disseminated candidiasis is a frequent infection in infants with very low birthweight (VLBW). Other fungal infections are considerably less common in neonates. The increasing incidence of candidiasis is attributable in part to advances in the life-sustaining care provided by neonatal caregivers. Early recognition of infection is important because untreated infections, especially in infants with VLBW, are associated with considerable morbidity and mortality.

Candida INCIDENCE Mucocutaneous, cutaneous, and disseminated candidiasis have been reported in newborns. Systemic candidiasis occurs more frequently in premature infants with VLBW.

MICROBIOLOGY Candida organisms are saprophytic yeasts that are ubiquitous and are constituents of the normal microbial flora of humans. The yeast, or blastospore form of the fungus, is round or eggshaped and is important in tissue colonization. All Candida species form pseudohyphae that are important in invasion. Candida albicans is the predominant species associated with maternally acquired neonatal disease. Candida parapsilosis is a common species that can account for one quarter of all cases of invasive fungal infection in infants with VLBW.8

Chapter 39  The Immune System

Other species, such as Candida glabrata, Candida lusitaniae, Candida krusei, and Candida stellatoidea, are reported less frequently. No special medium is required for growth of Candida in the laboratory.

PATHOGENESIS AND PATHOLOGY C. albicans can be acquired from the vaginal flora of the mother at delivery or from person-to-person contact after birth. Candida usually has low pathogenicity for humans. Congenital cutaneous candidiasis seems to be more common if the pregnancy has been complicated by maternal vaginitis, the presence of an intrauterine device, or cervical cerclage.42 Fungal colonization occurs on the skin and in the gastrointestinal tract before colonization of the respiratory tract.22 A prospective study in intubated infants with VLBW suggested that there was an increased risk of systemic infection when infants became colonized in the respiratory tract in the first week of life.40 Factors that alter host defense or that allow proliferation of the organism can increase the risk of invasion. Broad-spectrum antimicrobial use can expedite overgrowth of Candida in the gastrointestinal tract. Newborns in general, and infants with VLBW in particular, have qualitative and quantitative deficiencies in humoral and cellular immunity. The organism produces multiple virulence factors, such as adhesins, proteinases, and phospholipases, that promote attachment and invasion. After penetration of epithelial or endothelial barriers, in this setting, Candida can penetrate into lymphatics, blood vessels, and deep tissues, resulting in disseminated infection. Disseminated candidiasis can cause disease in any organ system. Candida microabscesses have been described in the liver, spleen, kidneys, heart, brain, eyes, bones, and joints.

CLINICAL MANIFESTATIONS Clinical manifestations depend on the location and extent of the infection. Acute pseudomembranous candidiasis (thrush) manifests with white, curdlike patches that cover the buccal mucosa, gingiva, and tongue. These membranes can be adherent to underlying tissue and, when removed, reveal a denuded, erythematous, painful base. Severe thrush sometimes results in difficulty feeding, but usually no systemic symptoms are apparent. Thrush is frequently associated with cutaneous perineal infection. Candida diaper dermatitis begins with an erythematous vesiculopapular eruption that coalesces, producing large areas with satellite lesions that are surrounded by a fine, white, scaly collarette. These lesions usually are restricted to the intertriginous areas in the groin, but can be noted in any warm, moist area, including the neck, axilla, and antecubital and popliteal fossae. The peak incidence occurs at 3 to 4 months of age (see Chapter 52). After an ascending in utero infection, generalized cutaneous lesions have been noted at birth in congenital cutaneous candidiasis. This condition typically manifests with intensely erythematous maculopapular lesions on the trunk and extremities that rapidly become pustular and rupture, leaving denuded skin with well-defined, raised, scaling borders.42 Congenital cutaneous candidiasis is usually a benign, self-limited infection, unless it occurs in an infant weighing less than 1500 g, or is associated with respiratory symptoms, or both. In those circumstances, parenteral antifungal therapy is indicated.

831

Catheter-associated and systemic candidiasis occur later, at a mean of 30 days of age.2 Infants with invasive neonatal candidiasis usually have one or more risk factors predisposing to infection (Box 39-6). These risk factors relate to immunologic immaturity, to bypassing of natural barriers, to the use of agents such as broad-spectrum antimicrobials that favor the growth of Candida species, and to the supportive care required by infants of VLBW. Prior fungal colonization is an important factor in the development of candidiasis. A blood culture obtained through an indwelling catheter, or peripherally while a catheter is in place, suggests the diagnosis. In catheterassociated fungemia, the blood culture becomes sterile concurrently with initiation of amphotericin B therapy and removal of the indwelling catheter. There is no involvement of distant organs, such as the liver, kidneys, or eye. A brief course of amphotericin B usually suffices for treatment. Catheter removal is mandatory. Failure to remove the colonized catheter promotes dissemination and increases the risk for an adverse outcome, including death. The presentation of systemic candidiasis in an infant with VLBW is similar to that of bacterial sepsis. Infants also can have an insidious onset of symptoms with respiratory deterioration, enteral feeding intolerance, abdominal distention associated with guaiac-positive stools, temperature instability, hypotension, hyperglycemia, and glucosuria. In contrast to disseminated candidiasis in older children and adults, multiple foci of infection are common in infants with systemic disease. Meningitis occurs in approximately 40% of affected infants. Examination of cerebrospinal fluid (CSF) is mandatory. Renal involvement may manifest as candiduria, multiple renal abscesses, or fungus balls that can cause obstruction to the flow of urine. The latter complication may not become evident until later in the treatment course; renal ultrasonography should be performed at the initiation of treatment for all infants suspected to have disseminated candidiasis and repeated as clinically indicated. The ultrasound

BOX 39–6 Risk Factors for Invasive Candidiasis in Neonates n Gestational

age ,32 weeks #1500 g n Male gender n Apgar score ,5 at 5 minutes n Intubation/mechanical ventilation n Placement of indwelling devices Umbilical catheters Peripheral or central venous catheters Urinary catheters Cerebrospinal fluid shunt devices n Abdominal surgery n Lack of enteral feeding n Fungal colonization n Use of intralipid n Total parenteral nutrition use n Corticosteroid use n Use of histamine type 2 receptor blockers n Administration of cephalosporins n Prolonged administration of antibiotics n Birthweight

832

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

appearance of fungus balls may persist long after clinical resolution of infection. Surgical management can be required for decompression of obstructive candidiasis and drainage of abscesses.19 Endophthalmitis is more likely with prolonged candidemia; it most commonly manifests as retinitis, with fluffy or hard white infiltrates. A dilated retinal examination should be part of the diagnostic evaluation for all infants. Involvement of the liver or spleen with microabscesses; endocarditis among infants with indwelling catheters; and extension of infection to the lungs, bones, or joints occur less frequently. The presentation of bone or joint disease is usually indolent, and involvement may not be evident when amphotericin B is initiated, so a high index of suspicion is important. Percutaneous aspiration of an involved joint or bone at the bedside usually is sufficient to establish the diagnosis, and an open drainage procedure is rarely, if ever, required.

DIAGNOSIS The diagnosis of systemic candidiasis is established by growth of Candida species in cultures from sites that are normally sterile. The isolation of Candida species from nonsterile sites, such as endotracheal tube secretions or the skin or mucous membranes, is an indication of colonization, and does not establish the diagnosis of disseminated infection. In contrast to adults, in whom the diagnosis of disseminated candidiasis is documented by blood culture in only one third of cases, 80% or more of neonates with invasive candidiasis have documented fungemia.6 The use of special media for the isolation of fungi is not required. Candida species grow robustly in the routine blood culture media used in most clinical laboratories, and these media usually yield the organism within 48 to 72 hours of incubation.44 If catheter-associated candidemia is suspected, a blood culture obtained through the catheter and one obtained peripherally should be collected before catheter removal and the initiation of therapy. The peripheral culture should be repeated to document that fungemia has resolved with catheter removal. If disseminated candidiasis is suspected, or if a positive peripheral blood culture is obtained, blood cultures should be repeated at intervals until sterility is documented. Urine culture should be obtained by suprapubic tap or catheter. The CSF should be evaluated. CSF evaluation usually reveals a modest pleocytosis, with several hundred cells or fewer, lymphocyte predominance, and mildly elevated protein. The inflammatory response in CSF can be minimal, and one half of infants with Candida meningitis can have normal CSF parameters.9 Routine cultures yield yeast within 2 to 3 days.34 Infants with disseminated candidiasis require a baseline abdominal ultrasound examination; ophthalmologic examination; and, if central venous access catheters have been in place, echocardiographic examination of the heart and the great vessels. Subsequent examinations can be conducted if there is evidence of focal infection of these organ systems, or if the clinical status indicates. Additional laboratory tests include a baseline complete blood count (CBC), blood urea nitrogen, creatinine, potassium, and liver enzymes. The CBC can reveal thrombocytopenia, and initial renal function abnormalities can be detected, particularly in infants with VLBW with disseminated candidiasis. These usually resolve as the infection is controlled.

TREATMENT Treatment varies by the location and extent of infection and the age of the patient. Thrush usually responds to nystatin (Mycostatin) suspension (1 to 2 mL) given orally four times daily for 5 to 10 days. Nystatin cream or ointment three times daily for 7 to 10 days is usually effective for Candida diaper dermatitis, although secondary bacterial infection may require treatment with another topical or systemic antibiotic. Nystatin with a corticosteroid (Mycolog-II cream or ointment) can be useful in severe cases of Candida dermatitis. Congenital cutaneous candidiasis requires no therapy unless the infant has evidence of pneumonia, or weighs less than 1500 g at birth, or both, at which time parenteral antifungal therapy is indicated.42 Amphotericin B is the mainstay of treatment for systemic candidiasis in a newborn infant (Table 39-25). It is extremely well tolerated by infants with VLBW. Pharmacokinetic studies in neonates suggest that a single daily infusion administered over 4 hours is necessary to achieve detectable serum concentrations.1 An initial dose of 0.5 mg/kg, administered over 1 to 2 hours, can be advanced to the 1 mg/kg per day dose after 12 to 24 hours. The duration of therapy varies, but a typical course for disseminated candidiasis is a 20 to 25 mg/kg cumulative dose. For catheter-associated candidiasis, the usual course of therapy is a cumulative dose of 10 to 15 mg/kg. This duration usually is required to ascertain that the sequential blood cultures obtained after catheter removal are sterile. Infants who have a delay in removal of central catheters are at higher risk of death and neurodevelopmental delay compared with infants whose catheters are removed promptly.3 Amphotericin B rarely causes nephrotoxicity in infants. Initially abnormal renal function usually improves as treatment is initiated, suggesting that renal dysfunction was caused by the fungemia, rather than precipitated by the treatment. Infants receiving the initial dose of amphotericin B should be monitored for cardiac arrhythmias, but this is a rare side effect. Because amphotericin B causes renal tubular wasting of potassium, the potassium should initially be monitored daily and supplemented as required to maintain the concentration at greater than 3 mEq/dL. If daily potassium, blood urea nitrogen, and creatinine levels are stable after the first week of therapy, these can be monitored twice weekly, and the CBC and liver enzymes can be determined weekly, for the duration of the treatment course. The combination of flucytosine and amphotericin B has been used successfully to treat systemic candidiasis, especially when the meninges are involved. Flucytosine should not be used as monotherapy because resistance can develop. The dosage range is 50 to 150 mg/kg daily, given orally in divided doses every 6 hours. Bone marrow suppression, hepatotoxicity, and gastrointestinal symptoms associated with flucytosine administration are more common when serum concentrations exceed 70 to 100 mg/mL. The CBC and liver enzymes should be monitored, and serum concentrations of flucytosine should be measured. Available data suggest that liposomal amphotericin B and amphotericin B lipid complex are safe and effective when used as initial therapy18,27 or for therapy in neonates with systemic candidiasis who were intolerant of or refractory to conventional antifungal therapy.49 Pharmacokinetic studies

Chapter 39  The Immune System

833

TABLE 39–25  Antifungal Agents Used for Treatment of Systemic Infections in Neonates Drug

Dosage* (mg/kg/d) Interval/Route

Toxicity

Comments

Amphotericin B

1†

q24h IV

Anemia, hypokalemia

Monitor blood urea nitrogen, creatinine nephrotoxicity, and K1 daily initially and twice weekly if stable after 1 wk; hold dose until K1 ,3 mEq/dL is corrected

Lipid formulations

3-7‡

q24h IV

Less nephrotoxic than amphotericin B

Monitor renal function and K1 as above for amphotericin B

Flucytosine

50-150

Divided q6h PO

Bone marrow suppression, hepatotoxicity, and gastrointestinal symptoms

Good penetration into CSF; must reduce dosage in patients with renal failure; monitor serum concentrations

Fluconazole

3-6

q24h PO, IV

Adjustment of dosage needed for renal impairment

Good penetration into CSF Limited experience

Itraconazole

5‡

q24h PO

Occasional hepatotoxicity

Limited experience

Caspofungin

1-2

q24h IV

Hepatotoxicity

Limited experience

‡

*See text for details. † Initial dose of 0.5 mg/kg should be followed 12-24 h later by 1 mg/kg dose administered daily. Dose should be increased to 1.5 mg/kg/d for neonates with invasive aspergillosis. ‡ Optimal dose has not been established. CSF, cerebrospinal fluid.

have not been performed in neonates, and randomized trials are lacking to establish the optimal and cumulative dosages, and whether these preparations should be used in combination with other antifungal agents.43 Nephrotoxicity is less often observed with lipid formulations than with amphotericin deoxycholate. Lipid formulations have poor penetration of the kidney, however, and therapeutic failure has occurred in neonates with renal or systemic candidiasis. Until comparative data are available, the use of these products should be reserved for infants with dose-limiting toxicity attributable to conventional amphotericin B for whom renal fungal involvement has been excluded. Among the echinocandin antifungal agents, limited clinical experience with caspofungin suggests that it is an effective, safe, and well-tolerated alternative agent to ampho­tericin B for neonates who have persistent fungemia, and are unresponsive to or are intolerant of amphotericin B deoxycholate. A dose of 1 mg/kg per day throughout the course of treatment and initial dosing with 1 mg/kg per day advanced to 2 mg/kg per day have been employed.35,36 Potential toxicity in association with a daily dose of 6 mg/kg per day has been noted.46 Early experience with micafungin revealed linear plasma pharmacokinetics over a broad dose range in experimental neonatal candi­ diasis, and indicated that a dose exceeding 2 mg/kg would be required to penetrate most compartments of the central nervous system.15 Doses of 2 to 3 mg/kg have been well tolerated in premature infants.14,37 The efficacy and safety of fluconazole for the treatment of Candida fungemia have been studied in a few neonates. Treatment of infants with VLBW who had documented C. albicans infection has been carried out with fluconazole for 6 to 48 days.16 An 80% cure rate was achieved, but four infants relapsed despite at least 14 days of therapy. Similarly, itraconazole has been used in a few cases, with a generally good

outcome. Until an efficacy trial is available that compares amphotericin B with fluconazole or itraconazole in a direct comparison, however, amphotericin B should remain the standard treatment for neonatal candidiasis.

PREVENTION In 2001, administration of fluconazole during the first 6 weeks of life was shown to be an effective intervention to prevent fungal colonization and invasive candidiasis in infants with birthweights of less than 1000 g.20 Fluconazole was administered intravenously at a dose of 3 mg/kg every third day for the first 2 weeks, every other day during weeks 3 and 4, and daily during weeks 5 and 6. The efficacy and safety of fluconazole prophylaxis in preventing invasive Candida infections in infants weighing less than 1000 g have subsequently been shown in additional multicenter, randomized trials, which, taken together, also show a significant decrease in invasive Candida infection-related mortality.23,29 Longitudinal analysis during 4 years found continued benefit without the development of fluconazole-resistant Candida species.13 This evidence base supports the use of fluconazole prophylaxis in high-risk infants weighing less than 1000 g at birth, and suggests that each neonatal intensive care unit should determine its incidence of infection and institute prevention in preterm infants at high risk for infection.21

Coccidioidomycosis INCIDENCE Coccidioidomycosis is endemic in the San Joaquin Valley in California and in other areas of the southwestern United States, Mexico, Argentina, Venezuela, and Paraguay. Most susceptible individuals living in endemic areas acquire

834

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

asymptomatic infection within 5 years. Disease can occur in any geographic location after reactivation of infection. Despite the high incidence of infection in endemic areas, perinatal coccidioidomycosis rarely occurs.

MICROBIOLOGY Coccidioides immitis, the causative agent of coccidioidomycosis, is a biphasic fungus. Highly contagious mycelia grow on culture media and soil, whereas the less infectious spherules grow in tissues. The spherule contains hundreds of endospores that, when released, can become spherules.

TREATMENT In older children and adults, primary infection is often selflimited and requires no therapy. The frequency of disseminated disease in young infants and the high case-fatality rate warrant treatment of all neonatal infections, however. Amphotericin B is the drug of choice for the initial treatment of infection (see Table 39-25). The duration of therapy is prolonged, and anecdotal therapy suggests that fluconazole administered orally can be given to complete the course of treatment.7,26

PATHOGENESIS The infectious arthrospores can become airborne, or can be transferred from inanimate objects contaminated with dust. Infection is acquired from inhalation of arthrospores and less commonly from direct inoculation into cutaneous lacerations or abrasions. The incubation period is 7 to 21 days. Infection occurring in infants younger than 1 week of age has been described, suggesting vertical transmission.4 Despite several reports of disseminated infection during pregnancy, placental and perinatal infections are rare.7,26 After arthrospore inhalation, mature spherules develop within several days. Granuloma formation with tracheobronchial lymph node involvement can follow. In extensive disease, polymorphonuclear leukocytes infiltrate the lung, similar to the process seen with bacterial pneumonia. Hematogenous dissemination occurs frequently in infants.

CLINICAL MANIFESTATIONS In contrast to older children and adults, whose primary infection is often asymptomatic, pneumonia is usually present in most infants with recognized infection. Chest radiographs can show focal consolidation or diffuse nodular infiltrates, hilar adenopathy, and pleural effusions. Fever, anorexia, and respiratory distress often accompany neonatal coccidioidomycosis. If the disease becomes disseminated, lesions can develop in the skin, bone, lymph nodes, liver, spleen, and meninges.

DIAGNOSIS Clinical and radiographic features of coccidioidomycosis can resemble features seen with histoplasmosis, pulmonary tuberculosis, and viral pneumonitis. If meningitis develops, an elevated protein concentration and hypoglycorrhachia can be seen on CSF examination. Early in the course, a neutrophilic pleocytosis can be seen, but this quickly progresses to a lymphocytic predominance. The diagnosis is best established using serologic and histologic methods. An IgM response is usually detectable 1 to 3 weeks after the onset of symptoms. A high IgG titer indicates severe disease, and decreasing titers suggest improvement. Transplacental passage of complement-fixing antibody occurs, so an increase in the infant’s titer must be shown to document neonatal infection. Antibodies are detectable in CSF in patients with meningitis. Spherules occasionally can be observed on silver-stained tissue samples. Culture of the organism is feasible, but is potentially hazardous to laboratory personnel. Coccidioidin skin tests are not helpful in the neonatal period, and these skin tests are not currently available in the United States.

Cryptococcosis INCIDENCE Cryptococcosis occurs worldwide. Many infections are likely to be asymptomatic, and most symptomatic infections occur in individuals older than 30 years. Infection can occur in otherwise healthy individuals, but is more common in immunocompromised hosts, including patients with acquired immunodeficiency syndrome (AIDS), malignancies, and diabetes mellitus. Cryptococcal infection in the newborn period is an extremely rare disease.

MICROBIOLOGY Cryptococcus neoformans is an encapsulated yeast that reproduces by budding. Its natural habitat is soil, and it has been commonly found in soil contaminated with pigeon excreta.

PATHOGENESIS AND PATHOLOGY Cryptococcosis is acquired from inhalation of the yeast with resulting primary pulmonary infection. Ingestion or cutaneous inoculation is possible; a central venous catheter may have been the source of cryptococcemia in one neonate.11 Case reports of disease with onset shortly after birth suggest that in utero or intrapartum transmission is possible. Otherwise, there has been no evidence of human-to-human transmission. Pulmonary infection can remain localized or can disseminate hematogenously to any organ, with the meninges most commonly affected. In adults, the infection is usually subacute or chronic, with large, solitary pulmonary nodules frequently observed. Diffuse pulmonary infiltration or miliary disease with dissemination is more common in infants. Pathologic findings vary from minor inflammatory responses to abscess formation. Noncaseating granulomas, hepatitis, and cirrhosis of the liver are common findings. Meningitis frequently leads to obstructive hydrocephalus. Granulomas of the brain have been reported.

CLINICAL MANIFESTATIONS Infantile cryptococcosis is a multisystemic infection. The signs and symptoms are similar to congenital infection caused by T. pallidum, Toxoplasma, cytomegalovirus, and rubella, and include failure to thrive, jaundice, hepatosplenomegaly, chorioretinitis, rash, and intracranial calcifications. Other symptoms suggestive of a central nervous system process, such as lethargy, irritability, vomiting, and seizures, can be observed. Respiratory symptoms are minimal, but occasionally interstitial pneumonitis can be present. Infants with meningeal involvement can show a pleocytosis ranging from

Chapter 39  The Immune System

40 to 1000 leukocytes/mm3 with a lymphocytic predominance, elevated protein, and slightly decreased glucose concentrations on CSF examination.

DIAGNOSIS Definitive diagnosis of cryptococcosis requires isolation of the organism from body fluids or tissue specimens. Fungal cultures of CSF, blood, sputum or tracheal aspirates, material from abscess cavities, and bone marrow can yield growth of Cryptococcus. The presence of encapsulated, budding yeast on India ink-stained CSF or respiratory tract samples is usually indicative of cryptococcal disease. Serologic investigation includes antibody and antigen detection tests, but antigen detection is the preferred technique. Latex agglutination and enzyme immunoassay for detection of cryptococcal antigen in serum or CSF are excellent rapid diagnostic tests.

TREATMENT In healthy adults, pulmonary cryptococcosis is usually selflimited and requires no therapy. Pulmonary disease frequently disseminates in immunocompromised patients if left untreated, however. Before the introduction of amphotericin B, cryptococcosis was almost uniformly fatal. The case-fatality rate has significantly decreased with amphotericin B therapy, but at least one third of adults fail to respond to therapy, and another one fourth have relapse after discontinuation of therapy. The combination of amphotericin B and flucytosine or fluconazole is indicated as initial therapy for patients with meningeal and other serious manifestations of cryptococcal infection. Because only six cases have been reported in neonates, only anecdotal data exist for treatment. One infant with VLBW survived cryptococcemia after receiving a 6-week course of therapy with amphotericin B.11

Histoplasmosis INCIDENCE Histoplasmosis is one of the most common pulmonary fungal infections in immunocompetent humans. It occurs worldwide in temperate climates and is endemic in the Ohio and Mississippi river valleys of the United States. In older children and adults, most infections are asymptomatic; however, in young infants, infections are often apparent and frequently disseminate. Despite the high incidence of infection in endemic areas, neonatal histoplasmosis is rare.

MICROBIOLOGY Histoplasma capsulatum is the causative agent. It is not encapsulated, but artifacts resembling a capsule can be seen with some staining techniques. It is a thermally dimorphic fungus, growing as mycelia (mold) with microconidia and macroconidia (spores) in soil and converting to yeast form at human body temperatures. Soil is its natural habitat, and it thrives in moist soil contaminated with avian or bat excreta.

PATHOGENESIS Inhalation of conidia is the most common mode of transmission. Other routes, such as ingestion or cutaneous inoculation, occur rarely. Human-to-human transmission, if possible, has not been well described. Several days after inhalation, the

835

spores germinate in the alveoli, releasing yeast forms. An inflammatory response, initially neutrophilic, followed by an influx of lymphocytes and macrophages, ensues. The yeasts are phagocytosed, but not killed, and begin to proliferate within macrophages and can spread hematogenously to the liver, spleen, and other organs or to regional lymph nodes through the lymphatics. The inflammatory response can result in discrete granulomas resembling sarcoid, or caseating lesions that frequently calcify during resolution.

CLINICAL MANIFESTATIONS The clinical spectrum of histoplasmosis in children includes asymptomatic infection, pulmonary disease that can be complicated by mediastinal lymphadenopathy and subsequent tracheobronchial obstruction, primary cutaneous infection, and disseminated infection. Asymptomatic infection occurs in most older children and adults, whereas acute disseminated infection with pulmonary disease is most common in young infants.24 The most frequently observed signs and symptoms in disseminated disease are prolonged fever, hepatosplenomegaly, anemia, and thrombocytopenia. Because the infection can disseminate to the lymph nodes, adrenal glands, gastrointestinal tract, bone marrow, central nervous system, kidneys, heart, and bones, many other signs and symptoms can be present.

DIAGNOSIS Fungal cultures from patients with acute, self-limited pulmonary infection rarely yield the organism, but frequently are positive from patients with disseminated disease. Histoplasma can be isolated from lower respiratory tract specimens, blood, bone marrow, hepatic and splenic biopsy specimens, and CSF, but growth may not be detected for 8 to 12 weeks. Lysiscentrifugation of blood samples submitted for fungal culture has increased the sensitivity and reduced the interval until cultures become positive. Demonstration of intracellular yeast forms supports the diagnosis when the clinical picture is compatible. Detection of H. capsulatum polysaccharide antigen in serum, urine, or bronchoalveolar lavage specimens is a rapid and specific diagnostic method. Complement fixation titers greater than 1:32 or a fourfold increase in yeast-phase or mycelial-phase titers suggests acute infection. Chest radiographs in pulmonary histoplasmosis often are negative or reveal nodular infiltrates with mediastinal and hilar adenopathy. In disseminated disease, chest radiographs most frequently are negative, but hilar adenopathy, bronchopneumonia, or miliary nodules can be present. Pulmonary and splenic calcifications can be seen on subsequent radiographs after recovery from infection. Findings common in adult disease, including pleural effusions or cavitary formation associated with chronic pulmonary infection, are not seen in infantile disease.

TREATMENT Primary pulmonary histoplasmosis in a healthy child usually does not require treatment. Therapy is recommended for disease in infancy because the risk of dissemination is higher, and untreated disseminated disease is almost uniformly fatal.24 Amphotericin B is effective for initial treatment of disseminated disease and other serious infections. In patients older than neonates, itraconazole is effective, either as initial treatment or after clinical improvement has been observed.

836

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

Other Fungi Malassezia yeasts are dimorphic, lipophilic fungi that are the causative agent of pityriasis versicolor (formerly tinea versicolor). Neonates acquire Malassezia through direct contact with their mothers or hospital personnel.50 Malassezia furfur is the most commonly isolated of the several Malassezia species, and has been associated with fungemia and occasionally pneumonia in premature infants with low birthweight receiving lipid emulsion alimentation through central venous catheters.38,48 Sporadic bloodstream infections have been de­ s­cribed, with transmission probably by the hands of medical personnel. Because it requires high concentrations of mediumchain and long-chain fatty acids to grow, Malassezia cannot be cultivated on Sabouraud medium without the addition of sterile olive oil. Most infections respond to temporary cessation of lipid infusions and removal of the central venous catheter. Amphotericin B should be used for treatment until a clinical response and negative blood culture are documented. Blastomyces dermatitidis is a dimorphic fungus endemic to the midwestern, southeastern, and Appalachian areas of the United States. Blastomycosis occurs more commonly in adults than in children and is rare in neonates. Infections can be asymptomatic, limited to the respiratory tract, or disseminated. Cutaneous and skeletal lesions are the most common extrapulmonary sites of infection. Cutaneous and genital disease have been documented during pregnancy and, if inadequately treated, can result in neonatal blastomycosis.30 The diagnosis should be considered in infants with reticulonodular pneumonia born to mothers with chronic skin infections who live in endemic areas. Candida diaper dermatitis is the most common superficial cutaneous fungal infection in young infants; however, tinea capitis and tinea corporis have been described.12,47 Dermatophyte infections, caused by Microsporum, Trichophyton, or Epidermophyton, are acquired postnatally from contact with contaminated soil, infected animals, or household members. Because lesions associated with dermatophytosis can resemble the lesions seen with psoriasis, seborrhea, or impetigo, they can easily be misdiagnosed. Invasive fungal dermatitis is an entity that is recognized increasingly with the survival of infants with VLBW. These infants have the same risk factors for invasive fungal infection as described for neonatal candidiasis. It is believed that the skin serves as the portal of entry for colonizing species. Diffuse involvement with a buff-colored crust suggests Candida species, but Aspergillus and Trichosporon have been seen.41 Discrete erythematous lesions that become purple or black suggest Curvularia or one of the Zygomycetes.39 The diagnosis is established by a full-thickness skin biopsy showing fungal invasion of the dermis. Treatment should be initiated with amphotericin B while awaiting the histopathologic and culture results from skin biopsy.

PROTOZOAL INFECTIONS Malaria INCIDENCE Worldwide, there are 300 to 500 million cases of malaria annually and greater than 2 million deaths, most of which occur in children. Malaria remains an important cause of

abortion, stillbirth, and neonatal death in many parts of the world. The 81 cases of congenital malaria reported in the United States from 1966-2005 occurred almost exclusively in infants of foreign-born women who were exposed within the year before the infant’s delivery.25

MICROBIOLOGY Malaria is caused by an obligate, intracellular protozoan of the genus Plasmodium. Congenital malaria has been recorded with each species, including Plasmodium vivax, Plasmodium malariae, Plasmodium ovale, and Plasmodium falciparum. Most congenital cases have been caused by P. vivax and P. falciparum.

PATHOLOGY AND PATHOGENESIS Malaria is transmitted by the bite of infected female Anopheles mosquitoes or from transfusions (maternal-fetal, blood products, or contaminated needles) of infected blood. Anemia is the result of hemolysis. Parasites can be found in other organs of the body, including the intestinal tract, liver, spleen, lung, and brain. Because P. falciparum and P. malariae have no persistent exoerythrocytic phase, relapses do not occur. P. vivax and P. ovale are associated with relapses from dormant exoerythrocytic organisms. Transfusion-related malaria and congenital malaria have no exoerythrocytic phase. The placenta is involved in most women who acquire malaria during pregnancy. It is unclear whether transmission to the infant is transplacental or from direct contact with maternal blood during labor or parturition. Most pregnancies resulting in congenital malaria are associated with a malaria attack during pregnancy; however, congenital infection has been described after uncomplicated asymptomatic pregnancies. Host factors that can decrease the risk of malarial infections include abnormal hemoglobin and malaria-specific antibodies. Erythrocytes with fetal hemoglobin or hemoglobin S are less likely to become infected than erythrocytes with hemoglobin A. Women living in endemic areas are continuously exposed to malaria and develop antimalarial antibodies. Maternal antibody is believed to exhibit a protective effect for the fetus. One survey found that 7% of infants born to women evaluated at seven African sites had congenital malaria.10

CLINICAL MANIFESTATIONS Most infants with congenital malaria have onset of symptoms by the eighth week of life, with an average age at onset of 10 to 28 days. Occasionally, onset has been documented at several months after birth. The most common clinical findings are fever, anemia, and splenomegaly. Approximately one third of infants have jaundice and direct or indirect hyperbilirubinemia. Hepatomegaly can be present. Nonspecific symptoms include irritability, failure to thrive, loose stools, and reluctance to feed. Congenital malaria often is complicated by bacterial illnesses in developing countries. In rare cases, malaria can be complicated by hypoglycemia; central nervous system infection; splenic rupture; renal failure; and, in P. falciparum infections, blackwater fever (severe hemolysis, hemoglobinuria, and renal failure). Untreated P. falciparum infection is associated with a high case-fatality rate.

Chapter 39  The Immune System

DIAGNOSIS Although a maternal history of a febrile illness during pregnancy can be elicited in most cases, congenital disease after an asymptomatic pregnancy in women has been reported. The diagnosis of congenital malaria should be considered in any infant presenting with fever, anemia, and hepatosplenomegaly born to a mother who at any time resided in an endemic area. The diagnosis depends on demonstration of parasites in the bloodstream. Thin and thick smears should be prepared and examined on several different occasions to maximize the possibility of parasite detection.

TREATMENT Several types of antimalarial drugs act at different stages of the Plasmodium life cycle. Tissue schizonticides such as primaquine are effective against exoerythrocytic forms, whereas blood schizonticides such as chloroquine, quinine, and quinidine act only on parasites in the erythrocytic phase. Because transfusion-acquired disease, including congenital infection, does not have an exoerythrocytic phase, primaquine is not required. Chloroquine, given in an initial dose of 10 mg of chloroquine base per kilogram of body weight administered orally, followed by doses of 5 mg base/kg at 6, 24, and 48 hours after the initial dose, is frequently used. For treatment of congenital disease caused by chloroquine-resistant P. falciparum, quinidine gluconate or the combination of quinine administered orally and clindamycin has been suggested. Mefloquine is effective against most P. falciparum strains, but is not approved for use in infants. Experience with newer treatment options, such as artemisinin derivatives, in young infants is limited. Severe congenital malaria can require intensive care, and exchange transfusion can be necessary for high-grade parasitemia. Current recommendations regarding treatment of congenital malaria can be obtained from the malaria branch of the Centers for Disease Control and Prevention in Atlanta, Georgia.

Pneumocystis INCIDENCE Pneumocystis pneumonia occurs in patients with congenital immune defects or hematologic malignancies, or patients receiving immunosuppressive medications for organ transplantation. Pneumocystis is an unusual cause of pneumonia in the first year of life, but it can be observed as epidemic disease, first recognized during World War II and presumed to be related to malnutrition, and as sporadic disease associated with congenital immunodeficiency or AIDS. Some surveys suggest that rare cases can develop in healthy infants.

MICROBIOLOGY Because of difficulties in laboratory cultivation, the taxonomy of Pneumocystis jiroveci (formerly Pneumocystis carinii) is inconclusive; it has features in common with protozoan parasites and fungi. Pneumocystis is a unicellular organism that can be found in three forms: a thick-walled cyst, a thin-walled trophozoite,

837

and an intracystic sporozoite. Each cyst can contain up to eight sporozoites. Organisms can be detected on tissue samples stained with Grocott-Gomori methenamine silver nitrate.

PATHOLOGY AND PATHOGENESIS Serologic studies suggest that asymptomatic infection with Pneumocystis is widespread and commonly occurs in the first years of life. The organisms are believed to persist in a latent stage until impairment in host defense mechanisms permits their reactivation. Disease associated with Pneumocystis usually is limited to the lungs. There is now evidence that person-to-person transmission is the most likely mode of acquiring new infections. Airborne transmission from mother to infant has been proposed.33 Postmortem examination reveals a diffuse process, with the posterior and dependent regions of the lung most significantly affected. Microscopic examination reveals an eosinophilic, foamy, honeycombed intra-alveolar exudate composed of cysts. In epidemic infantile disease, a plasma cellular infiltrate is observed (hence the name interstitial plasma cell pneumonia), whereas in sporadic disease associated with immunodeficiency or immunosuppression, there is hyperplasia of the cells lining the alveoli and minimal cellular infiltrate with a paucity of lymphocytes.

CLINICAL MANIFESTATIONS The clinical characteristics of sporadic Pneumocystis pneumonia are different from the characteristics observed in epidemic infantile disease. In sporadic cases associated with congenital or acquired immunodeficiency, there is an abrupt onset of high fever, coryza, nonproductive cough, and tachypnea. There is a quick progression to dyspnea and cyanosis. Radiographic findings most commonly reveal diffuse infiltrative disease. Concurrent infections, most commonly with cytomegalovirus, occur in more than 50% of immunocompromised infants with Pneumocystis pneumonia. In epidemic infantile Pneumocystis pneumonia, a rare disease in developed countries, the onset is slow and insidious, with nonspecific symptoms such as anorexia, diarrhea, and restlessness.28 Cough is not prominent initially, and fever is absent. In the subsequent weeks, infants become tachypneic, cyanotic, and dyspneic, with sternal retractions and flaring of the nasal alae. Auscultatory findings are minimal, consisting of fine, crepitant rales on deep inspiration. The chest radiograph can be negative early in the course, or can reveal a perihilar or diffuse haziness that progresses to a finely granular, interstitial pattern. Coalescent nodules can form in the periphery. Pneumothorax with interstitial and subcutaneous emphysema and pneumomediastinum can occur.

DIAGNOSIS Typical findings on radiographic studies can suggest P. jiroveci pneumonia, but are not diagnostic. Demonstration of Pneumocystis cysts or extracystic trophozoite forms establishes the diagnosis. P. jiroveci can be detected in induced sputum or in tracheal or gastric aspirates, but the yield is low, and detection does not imply disease.28 Bronchoalveolar lavage or open lung biopsy can yield the organism. Serologic tests are not useful.

838

SECTION VI  DEVELOPMENT AND DISORDERS OF ORGAN SYSTEMS

TREATMENT Trimethoprim-sulfamethoxazole is the drug of choice for treatment of P. jiroveci pneumonia in infants. It must be used with caution in young infants because of its potential for displacement of bilirubin from albumin-binding sites. The therapeutic dose is 15 to 20 mg/kg per day of the trimeth­ oprim component in divided doses every 6 to 8 hours. The intravenous route of administration is preferable in infants with moderate or severe disease. Treatment is usually continued for 3 weeks. In infants who do not respond to trimethoprimsulfamethoxazole or in whom adverse reactions develop, pen­ tamidine isethionate, 4 mg/kg given parenterally as a single daily dose for 14 days, can be used. Although radiographic improvement can take several weeks, clinical improvement is usually seen within 4 to 6 days after beginning therapy.28 Local reactions at the injection site, tachycardia, hypotension, pruritus, hypoglycemia, and nephrotoxicity have been associated with pen­ tamidine administration. Trimethoprim-sulfamethoxazole is the standard for prophylaxis against Pneumocystis pneumonia and has been used for that purpose in infants 2 months old and older with human immunodeficiency virus (HIV) infection.

Toxoplasmosis INCIDENCE Among 22,845 pregnant women analyzed by the Collaborative Perinatal Research Study in the United States, 38% had Toxoplasma IgG antibodies, reflecting past infection, and the incidence of acute maternal infection during pregnancy was estimated to be 1.1 per 1000 women.45 A higher incidence is seen in women born in Cambodia or Laos.17 In the United States, 1 to 3 infants per 1000 live births have Toxoplasma-specific IgM antibody. Worldwide, 3 to 8 infants per 1000 live births are infected in utero. An estimated 400 to 4000 cases of congenital toxoplasmosis occur in the United States each year.

MICROBIOLOGY Toxoplasmosis is caused by Toxoplasma gondii, an obligate intracellular protozoan parasite. This ubiquitous organism exists in three forms: an oocyst excreted by infected cats that produces sporozoites, a proliferative form (trophozoite or tachyzoite), and a cyst (cystozoite) found in tissues of infected animals. Toxoplasma can be propagated in tissue cell cultures or by animal inoculation in research laboratories.

TRANSMISSION AND PATHOGENESIS The cat is the only definitive host, but other mammals can be infected incidentally. Farm animals (cattle, pigs, sheep) can acquire infection after ingestion of food or water contaminated with infected cat feces that contain oocysts. Humans can acquire infection by ingestion of raw or poorly cooked meat containing the Toxoplasma cysts or by ingestion of food or water contaminated with oocysts. Risk factors include any exposure to cat feces, such as changing cat litter boxes, playing in sandboxes, or gardening in areas used by cats.

Congenital toxoplasmosis occurs almost exclusively as a result of primary maternal infection during pregnancy. Rarely, reactivation of infection in immunocompromised women during pregnancy can result in congenital toxoplasmosis. Most maternal infections are asymptomatic or result in mild illnesses. Fatigue and lymphadenopathy involving only a single posterior cervical node or generalized lymphadenopathy may be the only manifestation. Less commonly, acute maternal infection can manifest as an infectious mononucleosis-like syndrome with fever, nonsuppurative lymphadenopathy, headache, fatigue, sore throat, and myalgias. The general risk of transmission of acute infection from mother to fetus is estimated to be 40%; however, the actual risk and the severity of congenital infection vary with gestational age. The risk of transmission increases with increasing gestational age, but the earlier during pregnancy that fetal infection is acquired, the more severe the manifestations of congenital disease.

CLINICAL MANIFESTATIONS The classic clinical presentation of congenital toxoplasmosis is the triad of hydrocephalus, chorioretinitis, and intracranial calcifications, but there is a wide spectrum of manifestations, and more than 75% of infected newborns are asymptomatic in early infancy. As described by McAuley and colleagues,31 the four most common presentations include (1) a healthy-appearing term infant with subclinical infection in whom symptoms develop later in childhood, (2) a healthy-appearing term infant in whom clinical evidence of disease develops in the first few months of life, (3) an infant with generalized disease at birth, and (4) an infant with predominantly neurologic involvement at birth. Many infants with subclinical infection who were believed to be normal at birth have evidence of infection on closer evaluation, including CSF abnormalities, such as lymphocytic pleocytosis, hypoglycorrhachia, and elevated protein concentrations. If the disease goes unrecognized and untreated, these infants can present with chorioretinitis, late-onset seizures, mental retardation, developmental delay, and hearing loss later in infancy or childhood. The second group of healthy-appearing, infected infants present with hydrocephalus and chorioretinitis in the first few months of life. Manifestations of generalized infection at birth include prematurity and intrauterine growth restriction, jaundice, hepatosplenomegaly, pneumonitis, temperature instability, lymphadenopathy, and cutaneous lesions (e.g., exfoliative dermatitis, petechiae, ecchymoses, and maculopapular lesions). Other signs of generalized infection include myocarditis; nephrotic syndrome; and gastrointestinal symptoms such as vomiting, diarrhea, or feeding difficulties. The fourth group of infected infants has predominantly neurologic disease, but can have systemic manifestations as well. Subtle neurologic deficits, obstructive hydrocephalus, or acute encephalopathy can be seen. Unilateral or bilateral macular chorioretinitis frequently occurs, and infants often have rash, hepatosplenomegaly, thrombocytopenia, granulocytopenia, and typical CSF findings. The most common manifestations of congenital toxoplasmosis during the first few months of life are compared with the manifestations of congenital rubella and cytomegalovirus infection in Figure 39-12.

Chapter 39  The Immune System Manifestation: 0

20

40

60

80

100

Low birthweight (