Adverse events associated with neonatal exchange transfusion in the 1990s

J Pediatr. 2004 May;144(5):626-31. doi: 10.1016/j.jpeds.2004.01.054.


Objective: To determine the rates of adverse events associated with neonatal exchange transfusions performed for hyperbilirubinemia.

Study design: Retrospective chart review of 55 neonates who underwent 66 exchange transfusions at two perinatal centers in Cleveland between 1992 and 2002. Demographic data, causes of jaundice, details of exchange method, and adverse events occurring within one week of exchange were recorded. At the time of exchange, 62% of infants had other neonatal morbidities. Outcomes were stratified according to gestational ages < or =32 weeks, 33 to 36 weeks, and > or =37 weeks.

Results: Overall, 74% of exchanges were associated with an adverse event. The most common events were thrombocytopenia (44%), hypocalcemia (29%), and metabolic acidosis (24%), of which 69%, 74%, and 44%, respectively, required treatment. There were two serious adverse events, both in infants with other serious neonatal morbidities: seizures in one infant and the death of a critically ill preterm infant (body weight 731 g, gestational age 25 weeks). There were no cases of sepsis, necrotizing enterocolitis, or cardiac arrest. Adverse events were more frequent in exchanges done on preterm infants: < or =32 weeks (87%), 33 to 36 weeks (78%), and > or =37 weeks (67%), and in infants with other neonatal morbidity (79% vs 57%; P=.08). Controlling for neonatal morbidity, we found the odds of an adverse event were significantly higher when both umbilical venous and arterial catheters were used compared with other methods of exchange (88% vs 58%; OR, 5.17; 95% CI, 1.1, 34.2; P=.03).

Conclusions: The majority of adverse events associated with exchange transfusion are laboratory abnormalities and are asymptomatic and treatable.

MeSH terms

  • Exchange Transfusion, Whole Blood / adverse effects*
  • Female
  • Gestational Age
  • Humans
  • Infant, Newborn
  • Jaundice, Neonatal / etiology
  • Jaundice, Neonatal / therapy*
  • Logistic Models
  • Male
  • Morbidity
  • Ohio / epidemiology
  • Retrospective Studies
  • Statistics, Nonparametric