Outcomes associated with the use of renal replacement therapy in children receiving extracorporeal membrane oxygenation after heart surgery: a multi-institutional analysis.

Abstract

The primary objective of this investigation was to study the association between renal replacement therapy (RRT) and outcomes in children receiving extracorporeal membrane oxygenation (ECMO). Patients aged ≤18 years receiving ECMO before or after a pediatric heart operation at a Pediatric Health Information System (PHIS)-participating hospital (2004-2013) were included. The associations between RRT and study outcomes were computed using multivariate logistic regression analysis. A total of 3,502 patients from 43 hospitals qualified for inclusion. Of these, 484 (14 %) patients received RRT at some point during their hospital stay. After adjusting for patient and center characteristics, the odds of mortality were significantly higher in the RRT group (OR: 1.86, 95 % CI: 1.46- 2.37, p < 0.0001). However, there were considerable reductions in adjusted odds of mortality, compared to unadjusted odds of mortality. In adjusted models, length of ECMO was longer by 0.81 days (95 % CI: 0.13- 1.49, p = 0.02) in patients receiving RRT. We demonstrated worsening outcomes in children receiving ECMO with RRT compared to children receiving ECMO without RRT. Although the results could reflect confounding by severity of illness, they provide a rationale for prospective testing of use of RRT in critically ill children receiving ECMO with heart surgery.