Objective: To identify infants at risk of death and abnormal neurodevelopmental outcome following extracorporeal membrane oxygenation (ECMO) in the neonatal period.
Methods: The medical records of 82 neonates treated with ECMO were reviewed to evaluate risk of death. All survivors were followed by neurologic examinations and tested using the Bayley Scales of Infant Development or McCarthy Scale of Children's Abilities, and risk for abnormal neurodevelopmental outcome was assessed.
Results: The overall survival was 91% (75 of 82). The mean gestational age and birth weight of nonsurvivors were lower than those of survivors (37 +/- 1 weeks vs 40 +/- 0 weeks; 2734 +/- 230 vs 3325 +/- 69 gm, p < 0.05). Infants who were lost to follow-up (16%) did not differ from those with follow-up in demographic variables or clinical indicators of illness severity. Thirty-five of 63 infants (56%) with follow-up had normal neurodevelopmental outcome. Risk of abnormal outcome was higher in infants requiring assisted ventilation for > or = 15 days (relative risk [RR] 5.5; 95% confidence interval [CI] 2.0 to 14.8), supplemental oxygenation for > or = 22 days (RR 3.1; 95% CI 1.3 to 7.6), and black race (RR 8.9; 95% CI 1.3 to 62.9). None of the neuroimaging studies accurately predicted the neurodevelopmental outcome of these infants.
Conclusion: We conclude that ECMO in critically ill infants is associated with good survival. The need for prolonged respiratory support may help in identifying infants at risk for abnormal neurodevelopmental outcome.