ALTHOUGH THE technique of exchange transfusion is now two decades old, a lively discussion continues as to the indications for its use1-3 and the safety of the procedure.4-6 Those with extensive experience with exchange transfusion recognize that sudden and poorly understood deaths occur during and soon after this procedure which tempers the enthusiasm for its use despite its admitted effectiveness as therapy for severe erythroblastosis fetalis. Trolle5 has gone so far as to suggest that the use of exchange transfusion for idiopathic hyperbilirubinemia in full-term and premature infants would result in a net loss of life in that the mortality of the procedure would exceed that of kernicterus, using present indications for exchange transfusion. Boggs,7 on the other hand, reports no mortality related to 506 exchange transfusions performed for idiopathic hyperbilirubinemia.
The purpose of this study was to investigate the exchange transfusion-related mortality on a large