Comparison of early versus delayed strategies for repair of congenital diaphragmatic hernia on extracorporeal membrane oxygenation

Purpose For the last seven years, our institution has repaired infants with CDH that require ECMO early after cannulation. Prior to that, we attempted to decannulate before repair, but repaired on ECMO if we were unable to wean after two weeks. This study compares those strategies. Methods From 2002 to 2016, 65 infants with CDH required ECMO. 67.7% were repaired on ECMO, and 27.7% were repaired after decannulation. Data were compared between patients repaired ≤5days after cannulation ("early protocol", n=30) and >5days after cannulation or after de-cannulation ("late protocol", n=35). We used Cox regression to assess differences in outcomes between groups. Results Survival for the early and late protocol groups was 43.3% and 68.8%, respectively (p=0.0485). For patients that were successfully decannulated before repair, survival was 94.4%. Moreover, the early repair protocol was associated with prolongation of ECMO (16.8±7.4 vs. 12.6±6.8days, p =0.0216). After multivariate regression, the early repair protocol was an independent predictor of both mortality (HR=3.48, 95% CI=1.28–9.45, p =0.015) and days on ECMO (IRR=1.39, 95% CI=1.07–1.79, p =0.012). All bleeding occurred in patients repaired on ECMO (29.5%, 13/44). Conclusions Our data suggest that protocolized CDH repair early after ECMO cannulation may be associated with increased mortality and prolongation of ECMO. However, early repair is not necessarily harmful for those patients who would otherwise be unable to wean from ECMO before repair. Further work is needed to better move towards individualized patient care. Type of study Treatment Study. Level of evidence Level III.