OBJECTIVETo investigate the predictive value of umbilical cord serum (UCS) bilirubin for subsequent jaundice in healthy term newborns.

METHODSFive hundred and twenty-three healthy term newborns (275 boys, 248 girls) were selected. The cord blood total serum bilirubin concentration and the serum albumin concentration were determined. All the infants were assessed for jaundice daily by measurement of transcutaneous bilirubin (TCB). When the infant's TCB was >or= 18 within the first 24 h after birth, >or= 21 at 48 h, >or= 25 at or after 72 h, the venous total serum bilirubin (TSB) was determined and treatment against jaundice was applied as needed. The infants were aligned into four groups according to their UCS bilirubin levels, starting from < 30 micromol/L(group 1); >or= 30 micromol/L(group 2); >or= 36 micromol/L(group 3); >or= 42 micromol/L(group 4). The frequency of hyperbilirubinemia and phototherapy (PT) were compared among the four groups. An analysis of UCS bilirubin as a predictor of later development of jaundice was performed. The characteristics of the infants who became jaundiced (jaundiced group) were compared with the normal infants (non-jaundiced group).

RESULTSA clear correlation between UCS bilirubin level and the development of hyperbilirubinemia was found in all populations of the four groups. Only eight of the 194 infants in group 1 showed a TCB index >or= 25. TSB values > 205 micromol/L but < 257 micromol/L were observed in 2 newborns. None of the infants in this group showed TSB > 257 micromol/L or needed PT. Thirty-two infants in group 2 showed TCB >or= 25, 12 infants had TSB > 205 micromol/L but < 257 micromol/L, 2 infants had TSB > 205 micromol/L and received PT. In group 3, one infant developed hyperbilirubinemia at 48 h after birth and received PT. Thirty-nine infants showed TCB >or= 25, 16 infants TSB > 205 micromol/L but < 257 micromol/L, 2 infants had TSB > 205 micromol/L and also received PT. In group 4, 4 infants showed a range of TSB from 200 to 215 micromol/L at 48 h and received PT. Twenty-two infants showed TCB >or= 25, 17 of them showed TSB > 205 micromol/L but < 257 micromol/L, and 5 of them had TSB > 205 micromol/L and received PT. The frequency of TSB > 205 micromol/L increased from 1.03% in group 1, 5.77% in group 2, 19.75% in group 3 and to 42.5% in group 4. None of the 194 newborns in group 1 needed phototherapy, whereas 0.96%, 3.70% and 22.5% of the newborns in groups 2 - 4, needed PT. The frequency of patients with hyperbilirubinemia or phototherapy increased with increasing UCS bilirubin levels. For the prediction of TCB >or= 25 using a UCS bilirubin cut-off level, such as >or= 35 micromol/L, we found a positive predictive value of 45.68% and sensitivity of 68.27%. It is significant to predict neonatal jaundice by UCS bilirubin levels (P < 0.001). In the jaundiced group (TCB >or= 25) UCS bilirubin levels were significantly higher than those in the non-jaundiced group (t = 10.96, P < 0.001). No significant differences were found in the cord blood serum albumin concentration (t = 2.38, P > 0.05), the gestational age (t = -0.90, P > 0.05), and birthweight (t = 0.10, P > 0.05) between the jaundiced and non-jaundiced groups.

CONCLUSIONSUCS bilirubin level is useful in predicting the subsequent jaundice in healthy term infants. The use of UCS bilirubin values may help detect infants at low or high risk for hyperbilirubinemia and minimize an unnecessary prolongation of hospitalization.