Incidence, course, and prediction of hyperbilirubinemia in near-term and term newborns.

OBJECTIVE: In this study, we investigated prospectively the incidence of significant hyperbilirubinemia and demographic and laboratory characteristics and pattern of serum bilirubin levels of near-term newborns (35-37 weeks' [245-265 days'] gestation) by comparing them with those of term newborns (38-42 weeks' [266-294 days'] gestation) longitudinally in the first 7 days of life; we also aimed to determine the value of an early (6th-hour) serum bilirubin measurement in predicting the development of significant hyperbilirubinemia later during the first week of life in near-term newborns.
METHODS: Serum total bilirubin measurements were initially made at the 6th hour of life and repeated daily for the next 4 days, and a last measurement was performed on the 7th day (150th hour) in 219 term newborns (term group) and 146 near-term newborns (near-term group). Newborns with serum total bilirubin levels of > or =8 and > or =12 mg/dL on day 2, > or =12 and > or 15 mg/dL on day 3, and > or =14 and > or =17 mg/dL on days 4, 5, and 7 for birth weights 2000 to 2500 g and >2500 g, respectively, were defined to have significant hyperbilirubinemia, and phototherapy treatment was started. The predictive ability of the 6th-hour serum total bilirubin value in determining the development of significant hyperbilirubinemia in the near-term group was assessed on the basis of the placement of any of the first week's serum bilirubin measurements in the > or =95th percentile of the study population. A Gaussian distribution curve, the 5th, 30th, 60th, and 95th percentiles, and 4 percentile tracks were obtained from mean serum total bilirubin values. On the basis of the percentile tracks with various sensitivity, specificity, and negative and positive predictive values, a nomogram demonstrating the 4 percentile tracks as risk-zone demarcators with divided risk zones was produced.
RESULTS: Twenty-three newborns (10.5%) in the term group and 37 newborns (25.3%) in the near-term group had significant hyperbilirubinemia and required phototherapy. When the daily mean serum bilirubin levels of the 2 groups were compared, the first 4 days' values did not significantly differ between the 2 groups, whereas the 5th and 7th days' values were significantly higher in the near-term group. There were significant differences between the 2 groups with respect to the incidence of significant hyperbilirubinemia, hematocrit, Apgar score, and mode of delivery. On the age-specific nomogram, the zone >95th percentile was labeled as high risk, and that <5th percentile was labeled as low risk. Serum total bilirubin values between the 5th and 30th, 30th and 60th, and 60th and 95th percentiles were designated as being in the low-intermediate, intermediate, and high-intermediate risk zones, respectively. The 5th and 95th percentiles on the nomogram had the highest sensitivity (100%) and specificity (98.2%), respectively, in predicting the subsequent development of significant hyperbilirubinemia.
CONCLUSIONS: Near-term newborns should not be treated as term newborns in the approach to management of hyperbilirubinemia, because infants of 35 to 37 weeks' gestation had significantly lower birth weights, significantly higher serum total bilirubin levels on days 5 and 7, and were 2.4 times more likely to develop significant hyperbilirubinemia than those of 38 to 42 weeks' gestation in the present study. In near-term newborns of 35 to 37 weeks' (245 to 265 days') gestation, the decision to diagnose and treat significant hyperbilirubinemia should be made on the basis of risk status (percentile distribution of the serum bilirubin values on postnatal age) rather than using birth-weight-based thresholds. A nomogram constructed from daily serum bilirubin values of each population, as we present herein, can be used in assessing the age (hour)-specific jaundice risk (high, intermediate, or low) of each near-term newborn.