Background: Infants in many areas of the world are being discharged from the hospital before the age at which significant hyperbilirubinemia typically develops.

As a result, phototherapy is one of the most commonly reported reasons for neonatal readmission to the hospital. The American Academy of Pediatrics recommends plotting bilirubin measurements on an hour-specific nomogram to help identify infants at risk of hyperbilirubinemia. Dalal and colleagues postulated that a single value could better predict subsequent hyperbilirubinemia than determining the change between two bilirubin measurements taken within the first 48 hours of life.

The Study: This prospective cohort study was conducted between October 2006 and April 2007 at a hospital in New Delhi, India. Infants who were born at 35 weeks of gestation or later, lived within 100 km (62.5 miles) of the hospital, and able to return for follow-up were eligible for inclusion. Neonates who received phototherapy before 48 hours of age, had Rh isoimmunization, and required neonatal intensive care unit admission for longer than 24 hours, or had major congenital malformations were excluded. All mothers were encouraged by hospital staff to breastfeed exclusively.

Each infant enrolled in the study had two transcutaneous bilirubin (TcB) levels measured, one at 24 ± 6 hours and another at least 12 hours later, but within the first 48 hours of age. Each value was plotted on the hour-specific nomogram. The difference between the two values was calculated for each participant and percentiles of change were determined. The primary outcome was hyperbilirubinemia requiring phototherapy. Neonates were assessed clinically for jaundice twice a day by a team blinded to the TcB findings, and were subsequently discharged according to hospital policy. Parents of enrolled newborns were instructed to return on the fifth day of life for follow-up or earlier if they recognized worsening jaundice. Follow-up included clinical assessment for jaundice, with serum bilirubin testing as indicated.

Results: Of the 972 term and near-term infants born during the study, 358 met inclusion criteria. Ninety percent of participants finished the study, and 15 percent required phototherapy. For measurements taken 24 hours or less apart, the magnitude of change between measurements was not affected by the time interval between them (0.14 ± 0.08 mg per dL [2.39 ± 1.37 μmol per L] per hour at 12 to 18 hours between the two TcB measurements, and 0.14 ± 0.07 mg per dL [2.39 ± 1.20 μmol per L] per hour at 18 to 24 hours between them). However, if more than 24 hours elapsed between the first and second readings, the change between them was smaller (0.10 ± 0.5 mg per dL [1.71 ± 8.55 μmol per L] per hour). In general, TcB readings increased at an average rate of 0.14 mg per dL per hour in the first 48 hours of life, with the exceptions of a greater change between the two readings in neonates who subsequently required photo-therapy than in neonates who did not, and late-preterm neonates versus full-term neonates. Single TcB readings at 30 to 48 hours of life were more predictive of future hyperbilirubinemia than the readings taken between 18 and 30 hours of age. The positive likelihood ratios were equivalent for the second TcB reading and the change in readings, and were higher than the positive likelihood ratio for the first TcB measurement.

Conclusion: The authors conclude that a single TcB reading at 30 to 48 hours of age can reasonably predict subsequent hyperbilirubinemia in the neonate and is as accurate as calculating the change between two readings.