Umbilical cord bilirubin as a predictor of neonatal jaundice: a retrospective cohort study

Background Hyperbilirubinaemia is a major cause of neonatal morbidity. Early identification of those infants most at risk might allow the development of targeted primary preventative therapy and follow-up. The objective of this study was to assess whether arterial umbilical cord bilirubin (aUCB) level at delivery predicts the development of neonatal jaundice in term deliveries. Methods Retrospective analysis of hospital biochemistry records identified term deliveries with recorded aUCB. Infant medical records were reviewed to identify those who developed neonatal hyperbilirubinaemia (requiring treatment according to UK NICE guidelines) with/without a positive direct antiglobulin test (DAT). Results Of 1411 term deliveries with a clearly recorded aUCB, 30 infants developed clinically-significant jaundice (2.7%), of whom 8 were DAT + ve (0.6%) mostly due to ABO incompatibility. aUCB strongly predicted the development of DAT + ve jaundice (area under the ROC curve = 0.996), as well as all-cause jaundice (area under the ROC curve = 0.74). However, this effect was critically dependent on maternal blood group. Amongst infants at risk of ABO incompatibility (maternal blood groups O + ve/O-ve, 39.7%) the predictive value of aUCB for all cause jaundice was strengthened (area under the ROC curve = 0.88). Amongst those not at risk (defined maternal blood group not O + ve/O-ve, 51.0%) it disappeared completely (area under the ROC curve = 0.46). A cutoff of 35 μmol/l for mothers with blood group O + ve/O-ve increased the pre-test probability for all-cause jaundice of 4% to a post-test probability of 30%. Conclusions For infants of mothers with blood group O, aUCB predicts development of neonatal jaundice. There was no evident utility for infants of mothers with other blood groups. Estimation of aUCB should be considered as a strategy for early identification of those at risk of neonatal haemolytic jaundice.