Inaccuracy of Ballard scores before 28 weeks' gestation. National Institute of Child Health and Human Development Neonatal Research Network.

OBJECTIVE: Ballard scores are commonly used to estimate gestational age (GA). The purpose of this study was to determine the accuracy of the New Ballard Score (NBS) for infants <28 weeks GA by accurate menstrual history and to evaluate NBS as an outcome predictor. METHODS: Infants weighing 401 to 1500 g in 12 National Institute of Child Health and Human Development Neonatal Research Network centers had NBS performed before age 48 hours. Accuracy of NBS estimates of GA was assessed for infants with GA determined by accurate menstrual history. In a larger cohort of infants, NBS was included in regression models of the association of NBS and death, poor outcome, and duration of hospital stay. RESULTS: At each week from 22 to 28 weeks GA by accurate menstrual history, NBS estimates exceeded GA by dates by 1.3 to 3.3 weeks, and estimates varied widely (range of widths of 95% CIs for the observations, 6.8 to 11.9 weeks). NBS did not contribute significantly to regression models of death, poor outcome, or duration of hospital stay. CONCLUSIONS: Inaccuracies in GA determined by the NBS should be considered when treating extremely premature infants, particularly in decisions to forego or administer intensive care. Refinement of GA scoring systems is needed to optimize clinical benefit.

Prognostic significance of cerebrospinal fluid cyclic adenosine monophosphate in neonatal asphyxia.

OBJECTIVE: In piglets prolonged asphyxia resulted in decreased cerebrospinal fluid (CSF) 3;,5;-cyclic adenosine monophosphate (cAMP) during recovery; this was associated with reduced pial arteriolar responses to stimuli that use cAMP as a second messenger. We hypothesized that asphyxia in human neonates results in decreased CSF cAMP and that low CSF cAMP is associated with abnormal outcome. DESIGN: We studied 27 infants with evidence of hypoxic-ischemic insult; 19 were term (group 1) and 8 were preterm (group 2). The normal values of CSF cAMP were determined from 75 infants with no asphyxia; 44 were term (group 3) and 31 were preterm (group 4). CSF cAMP was measured by using radioimmunoassay procedures. RESULTS: CSF cAMP levels in infants with asphyxia (groups 1 and 2) were 12 +/- 9. 5 and 7.9 +/- 7.1 pmol/mL, respectively, significantly lower than those of groups 3 and 4 (control infants), that is, 21.1 +/- 8.7 and 27.1 +/- 9.2 pmol/mL, respectively (P <.0001). Among infants with asphyxia, 3 died and 10 had abnormal neurologic outcome. Univariate analysis showed that abnormal outcomes were significantly related to CSF cAMP levels, phenobarbital use, and multi-organ failure. However, only CSF cAMP was retained in the model by stepwise logistic regression. CSF cAMP of 10.0 pmol/mL discriminated between those with normal and those with abnormal neurologic outcome. Low CSF cAMP concentration was associated with abnormal long-term outcome, estimated odds ratio of 12.4 (95% CI, 2.1-109.3; P <.006), and sensitivity, specificity, and positive and negative predictive values of 85%, 69%, 73%, and 80%, respectively. CONCLUSION: CSF cAMP concentrations were decreased in infants with asphyxia. Low CSF cAMP levels were associated with poor neurologic outcome.

Late-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network.

OBJECTIVE: Late-onset sepsis (occurring after 3 days of age) is an important problem in very low birth weight (VLBW) infants. To determine the current incidence of late-onset sepsis, risk factors for disease, and the impact of late-onset sepsis on subsequent hospital course, we evaluated a cohort of 7861 VLBW (401 to 1500 gm) neonates admitted to the 12 National Institute of Child Health and Human Development (NICHD) Neonatal Research Network centers during a 32-month period (1991 to 1993). METHODS: The NICHD Neonatal Research Network maintains a prospectively collected registry of all VLBW neonates cared for at participating centers. Data from this registry were analyzed retrospectively. RESULTS: Of 6911 infants who survived beyond 3 days, 1696 (25%) had one or more episodes of blood culture-proven sepsis. The vast majority of infection (73%) were caused by gram-positive organisms, with coagulase-negative staphylococci accounting for 55% of all infections. Rate of infection was inversely related to birth weight and gestational age. Complications of prematurity associated with an increased rate of infection included intubation, respiratory distress syndrome, prolonged ventilation, bronchopulmonary dysplasia, patent ductus arteriosus, severe intraventricular hemorrhage, and necrotizing enterocolitis. Among infants with bronchopulmonary dysplasia, those with late-onset sepsis had a significantly longer duration of mechanical ventilation (45 vs 33 days; p <0.01). Late-onset sepsis prolonged hospital stay: the mean number of days in the hospital for VLBW neonates with and without late-onset sepsis was 86 and 61 days, respectively (p <0.001). Even after adjustment for other complications of prematurity, including intraventricular hemorrhage, necrotizing enterocolitis, and bronchopulmonary dysplasia, infants with late-onset sepsis had a significantly longer hospitalization (p <0.001). Moreover, neonates in whom late-onset sepsis developed were significantly more likely to die than those who were uninfected (17% vs 7%; p <0.000 1), especially if they were infected with gram-negative organisms (40%) or fungi (28%). Deaths attributed to infection increased with increasing chronologic age. Whereas only 4% of deaths in the first 3 days of life were attributed to infection, 45% of deaths after 2 weeks were related to infection. CONCLUSIONS: Late-onset sepsis is a frequent and important problem among VLBW preterm infants. Successful strategies to decrease late-onset sepsis should decrease VLBW mortality rates, shorten hospital stay, and reduce costs.

Early-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network.

OBJECTIVE: Early-onset sepsis (occurring within 72 hours of birth) is included in the differential diagnosis of most very low birth weight (VLBW) neonates. To determine the current incidence of early-onset sepsis, risk factors for disease, and the impact of early-onset sepsis on subsequent hospital course, we studied a cohort of 7861 VLBW neonates (401 to 1500 gm) admitted to the 12 National Institute of Child Health and Human Development (NICHD) Neonatal Research Network centers during a 32-month period (1991-1993). METHODS: The NICHD Neonatal Research Network maintains a prospectively collected registry on all VLBW neonates born or cared for at participating centers. Data from this registry were analyzed retrospectively. RESULTS: Blood culture-proven early-onset sepsis was uncommon, occurring in only 1.9% of VLBW neonates. Group B streptococcus was the most frequent pathogen associated with early-onset sepsis (31%), followed by Escherichia coli (16%) and Haemophilus influenzae (12%). Decreasing gestational age was associated with increased rates of infection. Antibiotic therapy for suspected sepsis is frequently initiated at birth in VLBW neonates. Almost half of the infants in this cohort were considered to have clinical sepsis and continued to receive antibiotics for 5 or more days, despite a negative blood culture result in 98% of cases. These findings underscore the difficulty of ruling out sepsis in the symptomatic immature neonate and the special concern for culture-negative clinical sepsis in the face of maternal antibiotic use. Neonates with early-onset sepsis were significantly more likely to have subsequent comorbidities, including severe intraventricular hemorrhage, patent ductus arteriosus, and prolonged assisted ventilation. Although 26% of VLBW neonates with early-onset sepsis died, only 4% of the 950 deaths that occurred in the first 72 hours of life were attributed to infection. For those infants discharged alive, early-onset sepsis was associated with a significantly prolonged hospital stay (86 vs 69 days; p <0.02). CONCLUSIONS: Early-onset sepsis remains an important but uncommon problem among VLBW preterm infants. Improved diagnostic strategies are needed to enable the clinician to distinguish between the infected and the uninfected VLBW neonate with symptoms and to target continued antibiotic therapy to those who are truly infected.

A multicenter randomized trial comparing two surfactants for the treatment of neonatal respiratory distress syndrome. National Institute of Child Health and Human Development Neonatal Research Network.

OBJECTIVE: To compare the efficacy of two surfactants, Exosurf Neonatal (Burroughs Wellcome Co.) and Survanta (Ross Laboratories), for the treatment of neonatal respiratory distress syndrome. DESIGN: Multicenter randomized trial. SETTING: Eleven tertiary care university neonatal intensive care units participating in the National Institute of Child Health and Human Development Neonatal Research Network. PATIENTS: Newborn infants (n = 617) weighing 501 to 1500 gm with respiratory distress syndrome who were receiving assisted ventilation with 30% oxygen or more within 6 hours of birth were enrolled between January 1991 and January 1992. INTERVENTIONS: Infants were randomly assigned to receive up to four intratracheal doses of either Exosurf Neonatal (n = 309) or Survanta (n = 308). MAIN OUTCOME MEASURES: The occurrence of death or bronchopulmonary dysplasia 28 days after birth and the average fraction of inspired oxygen (FIO2) and mean airway pressure (MAP) during the first 72 hours after treatment. RESULTS: Death or bronchopulmonary dysplasia occurred in 67% of the infants in the Exosurf group and 62% of those in the Survanta group (adjusted relative risk, 1.07; 95% confidence interval, 0.96 to 1.20). During the 72 hours after the first surfactant dose, the average FIO2 (+/- SEM) was 0.50 +/- 0.01 for Exosurf and 0.42 +/- 0.01 for Survanta (difference, 0.08; 95% confidence interval, 0.05 to 0.11); the average MAP (+/- SEM) was 7.64 +/- 0.21 cm H2O for Exosurf and 6.93 +/- 0.21 cm H2O for Survanta (difference, 0.71 cm H2O; 95% confidence interval, 0.13 to 1.29 cm H2O). There was no difference between the groups in the incidence of other neonatal morbidities or in the duration of hospitalization, assisted ventilation, or supplemental oxygen administration. CONCLUSION: We found no difference between treatment groups in the incidence of death or bronchopulmonary dysplasia, although we did observe a difference in the initial response to treatment as measured by FIO2 and MAP.

Asymptomatic syndrome of polycythemic hyperviscosity: effect of partial plasma exchange transfusion.

We determined the cerebral hemodynamic changes in infants with asymptomatic polycythemic hyperviscosity syndrome and whether treatment with partial plasma exchange transfusion (PPET) would affect hemodynamics as well as outcome. From a routine cord blood hematocrit screening, 71 babies were identified as needing to be tested for polycythemic hyperviscosity. In addition to clinical evaluation, each infant had radial artery hematocrit and viscosity determinations, blood gas determinations, cerebral blood flow velocity studies, cranial ultrasonography, and noninvasive intracranial pressure determination. Babies with symptomatic hyperviscosity (n = 17) were treated by PPET, whereas those with asymptomatic hyperviscosity (n = 28) were randomly selected to have PPET (n = 14) or to be observed (n = 14). The remaining babies (n = 26) with normal viscosity served as control subjects. Both hematocrit and viscosity decreased after PPET but remained unchanged in babies with hyperviscosity who were merely observed. Reversal of cerebral blood flow velocity abnormalities was observed after PPET in the infants with symptomatic hyperviscosity, whereas those who had no symptoms had normal results on Doppler studies at the outset, and no significant changes occurred with either PPET or observation. There were two deaths in the group with symptoms. A total of 46 babies returned for follow-up evaluation at a mean age of 30 +/- 7.7 months. Outcome of the control group was no better than that of those who had hyperviscosity, and outcomes did not differ between the babies with symptomatic and those with asymptomatic hyperviscosity, nor between those treated with PPET and those who were only observed. Multivariate analysis revealed that other perinatal risk factors and race rather than polycythemia or PPET, significantly influenced long-term outcome.

Necrotizing enterocolitis in very low birth weight infants: biodemographic and clinical correlates. National Institute of Child Health and Human Development Neonatal Research Network.

We studied the occurrence of necrotizing enterocolitis in 2681 very low birth weight infants during an 18-month period to characterize the biodemographic and clinical correlates. Proven necrotizing enterocolitis (Bell stage II and beyond) occurred in 10.1% of study infants; necrotizing enterocolitis was suspected in 17.2% of study infants. Positivity of blood cultures was related to necrotizing enterocolitis staging. The mortality rate increased only for stage III necrotizing enterocolitis (54% died). Logistic regression identified medical center of birth, race, gender, birth weight, maternal hemorrhage, duration of ruptured membranes, and cesarean section as significant risk factors. For one center the odds ratio was 3.7, whereas for another center it was only 0.3. For black boys, the odds ratio was 2.3 relative to nonblack boys; for girls, race did not affect prevalence of necrotizing enterocolitis. Age at onset was related to birth weight and gestational age. Intercenter differences in necrotizing enterocolitis prevalence were related to time required to regain birth weight and other indicators of fluid management. Gram-positive organisms predominated in positive blood cultures for stage I and II necrotizing enterocolitis; enteric bacteria were isolated more frequently in infants with stage III disease. We conclude that necrotizing enterocolitis prevalence varies greatly among centers; this may be related to early clinical practices of neonatal care.

Mean arterial blood pressure changes in premature infants and those at risk for intraventricular hemorrhage.

Bedside microcomputer-derived, minute-to-minute mean arterial pressure (MAP) values during the first 48 hours of life were studied in 100 preterm babies with birth weight less than or equal to 1500 gm. In those babies (n = 72) with no periventricular-intraventricular hemorrhage (PV-IVH) or with grade 1 PV-IVH, the MAP values increased during the study period, with minute-to-minute variation and interval undulation. The MAP values in those with birth weight greater than 1000 gm were higher than in those of lower birth weight. Infants in whom grades 2 to 4 PV-IVH developed (n = 28) had consistently lower MAP values during the study period. Minute-to-minute variability, expressed as the average of the coefficients of variation at 15-minute intervals, did not differ between birth weight groups, nor did they differ between the PV-IVH group and their matched control subjects. However, those with PV-IVH spent a greater percentage of time, with a coefficient of variation greater than or equal to 13% or less than 3%, than their matched control subjects spent (p less than 0.005). This study provides reference data for MAP changes in premature babies. The observed MAP changes in those with PV-IVH lend support to a significant role for MAP alterations in the pathogenesis of PV-IVH.

Frequent handling in the neonatal intensive care unit and intraventricular hemorrhage.

The association between periventricular-intraventricular hemorrhage (PV-IVH) and frequent handling resulting from various neonatal intensive care procedures and routine interventions was evaluated in a prospective clinical study. Inborn premature babies with birth weight less than or equal to 1500 gm (n = 156) who did not have PV-IVH or who had grade 1 PV-IVH at less than or equal to 1 hour were randomly assigned to the reduced manipulation protocol (n = 62) or to standard care (n = 94). A bedside microcomputer-based data acquisition system was used to monitor the duration of rest or the number of interventions per day. Infants assigned to receive reduced manipulation spent a significantly higher percentage of time each day at rest than did those who received standard manipulation (p less than 0.006). However, the incidence of grades 2 to 4 PV-IVH did not differ significantly (30% in the study vs 37% in the standard manipulation group). When we analyzed the effect of manipulation in relation to risk of PV-IVH, while taking into account other perinatal variables, standard manipulation was not associated with increased risk of grades 2 to 4 PV-IVH. However, low birth weight, maternal smoking, general anesthesia, early grade 1 PV-IVH, low hematocrit, lowest arterial oxygen pressure within the first 6 hours of life, and large base deficit at 6 hours of age all increased the relative risk of grades 2 to 4 PV-IVH.

Indomethacin reduces the risks of severe intraventricular hemorrhage.

A prospective, random selection, double-blind clinical trial was carried out to determine the efficacy of indomethacin in preventing periventricular-intraventricular hemorrhage (PV-IVH). Babies who were born in our institution, had birth weights less than or equal to 1500 gm, and had no PV-IVH or grade 1 PV-IVH were given either placebo (n = 70) or indomethacin (n = 71), 0.2 mg/kg intravenously at 6 hours of age and 0.1 mg/kg at 18 and 30 hours. Two major outcomes were determined: the development of grades 2 to 4 PV-IVH and the development of severe PV-IVH (i.e., hemorrhages with blood filling greater than 50% of the ventricles and in some cases with associated parenchymal echodensities). Grades 2 to 4 PV-IVH occurred in 16 (23%) of the indomethacin group and 27 (39%) of the placebo group (p less than 0.03). The incidence of severe PV-IVH was 3% in the indomethacin-treated babies and 14% in the control group (p less than 0.02). The influence of other perinatal factors on the incidence of grades 2 to 4 or severe PV-IVH was determined by stepwise logistic regression. Placebo use, early grade 1 PV-IVH, lower birth weight, and higher fraction of inspired oxygen at 6 hours of life were associated with higher estimated odds of the development of grades 2 to 4 PV-IVH. Placebo use, male gender, lower 5-minute Apgar score, and a large base deficit were predictive of severe PV-IVH. Estimated odds ratios of severe PV-IVH with placebo use and male gender were 11.25:1 and 9:1, respectively. Thus indomethacin prophylaxis reduced the relative risk of grades 2 to 4 PV-IVH and severe PV-IVH, but other perinatal variables contributed significantly to the overall risk of PV-IVH.