Early prediction of neonatal hyperbilirubinemia.

The study aim was to predict, using serum bilirubin level measured 18 to 24 hours (SB, 18-24) after birth, the occurrence of peak serum bilirubin level > 15 mg/dL (hyperbilirubinemia) or the requirement of phototherapy, any time from the second to fifth postnatal day. The study was conducted on a prospective cohort of 274 neonates born in north India. The main outcome measures were (a) hyperbilirubinemia and (b) phototherapy. Serum bilirubin level was estimated at 18-24 hours of age and then daily from second to fifth postnatal day. Exclusion criteria were Rh incompatibility, asphyxia and life threatening congenital malformations; and neonates of women with gestational diabetes or history intake of drugs affecting the fetal liver. Hyperbilirubinemia was found in 12.8%; and 19.3% neonates received phototherapy. Dichotomous SB 18-24, using a cut-off of > 3.99 mg/dL, as the "prediction test" had the following sensitivity and specificity for predicting (a) hyperbilirubinemia: 67% and 67%, respectively, and (b) the treatment with phototherapy: 64% and 68%, respectively. We concluded that by using SB 18-24 as the "prediction test", approximately two-thirds of neonates were test negative and had about one in ten chances of re-admission for treatment of hyperbilirubinemia, if discharged. After further validation, our results will be of benefit to neonates delivered in developing countries.