Survival predictors in ventricular assist device patients with prior extracorporeal life support: selecting appropriate candidates.

Several centers turn patients down for long-term ventricular assist devices (VADs) once they have received extracorporeal life support (ECLS) due to the expected poor outcome in these patients. The aim of this study was to identify survival predictors in this cohort of patients. Data of patients undergoing VAD support between January 2010 and November 2013 were retrospectively reviewed. Patients on ECLS support before implantation were considered eligible for inclusion. Outcome in survivors following long-term VAD support was compared with outcomes in nonsurvivors. Student's t-test and χ(2)-test were used as applicable. A total of 65 long-term VADs were implanted. The inclusion criteria were met by 24 patients. Eight patients did not survive the first 30 days. All preoperative characteristics were comparable between the two groups except for statistically higher Model for End-stage Liver Disease (MELD) score, bilirubin, white blood cell count, and blood urea nitrogen in nonsurvivors (P = 0.002, 0.01, 0.01, and 0.003, respectively). Stepwise discriminant analysis revealed MELD score as the most important survival predictor. Based on this analysis, an outcome predictor formula was generated. The 30-day and 1-year survival rates were 67% and 54%, respectively. In this study, we were able to determine survival predictors in VAD patients with prior ECLS support. The outcome in these patients is limited and associated with higher postoperative complications, particularly right ventricular and respiratory failure. The pre-VAD MELD score is an important predictor of poor outcome.