Criteria for exchange transfusion in jaundiced newborns.

Abstract

A widely published set of exchange transfusion criteria lowers critical bilirubin concentrations when the serum albumin falls to < 2.5 g/dL. Although acknowledging the role of bilirubin-albumin binding in bilirubin neurotoxicity, this approach may inadvertently produce two critical bilirubin concentrations. This study investigates using the bilirubin/albumin ratio instead of the single albumin concentration to eliminate this potential ambiguity in the criteria. The bilirubin/albumin ratio was defined as a reliable indicator of bilirubin-albumin binding if the frequency curves of specific unbound bilirubin concentrations are normally distributed functions of the ratio. Therefore the bilirubin/albumin ratios at which the unbound bilirubin reached 10, 15, and 20 nmol/L were determined by the peroxidase method in 35 well full-term, 10 ill full-term, and 19 ill preterm neonates. The frequency curves for each unbound bilirubin concentration plotted against the bilirubin/albumin ratio were tested for normality. The frequency of each unbound bilirubin concentration was a normally distributed function of the bilirubin/albumin ratio. Furthermore, the mean ratio at which each unbound bilirubin occurred did not differ significantly among the groups of neonates. The bilirubin/albumin ratio is a simple, nonambiguous way of incorporating the serum albumin concentration into exchange transfusion criteria.