Neonatal jaundice in ABO incompatible infants. Computer-assisted evaluation of risk of hyperbilirubinaemia and analysis of differences between sexes

A set of maternal and neonatal variables have been analyzed in a sample of ABO incompatible infants in order: (i) to attempt an evaluation, at birth, of the risk of hyperbilirubinaemia during the first few days of life and (ii) to elucidate the role of sex regarding the hyperbilirubinaemia in ABO incompatible infants. The following variables are included: Coombs test, gestational length, birth weight, birth order, ABO maternal phenotype, type of feto-maternal ABO incompatibility (A or B) and two factors of placental alkaline phosphatase (PAP) polymorphic system (Plf1 and Pl11). In males only the Coombs test showed a statistically significant contribution to the separation between infants with serum bilirubin level < 10 mg/dl and those with a level ≥10 mg/dl. In females the contribution of Coombs test and gestational length were highly significant and that of birth weight was very near the level of significance. Among female infants a group with a very slight risk of hyperbilirubinaemia (1%) can be separated at birth. This group comprises more than 50% of infants. In males the separation is much less useful. The present data indicate that although the overall frequency of hyperbilirubinaemia is very similar in ABO incompatible males and females, the set of variables considered are rather good predictors of serum bilirubin levels in females but not in males. Multivariate analysis also suggests that the two sexes may be very unlike concerning the pattern of relations among variables. In fact PAP phenotype appears a predictor of jaundice much more important in males than in females; on the contrary gestational length is very important in females but not in males.