We sought to evaluate extracorporeal membrane oxygenation (ECMO) use for low cardiac output syndrome (LCOS) after congenital heart surgery. Our primary aims were (1) to compare hospital mortality and morbidity for proactive ECMO versus extracorporeal cardiopulmonary resuscitation (ECPR) and (2) to assess the impact of cardiopulmonary resuscitation (CPR) duration. Our secondary aim was to investigate pre-ECMO vasoactive inotropic scores (VIS).

This is a retrospective analysis from the Pediatric Cardiac Critical Care Consortium that includes all patients after cardiopulmonary bypass surgery cannulated to ECMO in the first 72 postoperative hours for LCOS. We performed univariable, multivariable, and propensity score analyses to quantify the effect of proactive ECMO versus ECPR on hospital mortality.

ECPR had a greater rate of hospital mortality than proactive ECMO (odds ratio, 1.82; 95% CI 1.32-2.51). Patients who received ECPR had shorter aortic clamp time (median, 84.5; range, 54-124 vs 108, 68-178). STAT category was not a significant predictor of hospital mortality. Risk-adjusted mortality for duration of CPR less than 30 minutes was not different from proactive ECMO (odds ratio, 0.81; 95% CI, 0.48-1.36). The greatest pre-ECMO VIS was lower with ECPR (median, 12.5, interquartile range, 8.0-18.0) than with proactive ECMO (18, 12.5-26.0).

For early postoperative ECMO for LCOS, ECPR has a greater risk-adjusted hospital mortality than proactive ECMO, with CPR duration of less than 30 minutes being the exception. Pre-ECMO VIS was lower for ECPR. Lower VIS, shorter aortic crossclamp times, and an unexpectedly high proportion of lower STAT categories indicate that ECPR is not uncommon in patients deemed lower risk.