Model for end-stage liver disease scores in veno-arterial extracorporeal membrane oxygenation

Veno-arterial extracorporeal membrane oxygenation is a valuable therapeutic approach in patients with severe heart failure due to different etiologies. Current prognosis with veno-arterial extracorporeal membrane oxygenation is unsatisfactory, and the risk stratification is still challenging. Therefore, we aimed to evaluate the predictive value of different baseline model for end-stage liver disease scores for survival in patients with veno-arterial extracorporeal membrane oxygenation.

We conducted an observational, retrospective study of consecutive veno-arterial extracorporeal membrane oxygenation-treated patients between January 2012 and August 2018. The four types of model for end-stage liver disease scores-model for end-stage liver disease, international normalized ratio-excluded model for end-stage liver disease, modified model for end-stage liver disease, and model for end-stage liver disease with sodium-were calculated preoperatively. Veno-arterial extracorporeal membrane oxygenation was used based on the four clinical indications: primer graft failure after heart transplantation, weaning failure from cardiopulmonary bypass, acute myocardial infarction with refractory cardiogenic shock, and bridge to transplantation or bridge to candidacy. The primary endpoint of the study was overall mortality. The secondary endpoint was in-hospital mortality. We performed univariable and multivariable Cox regression analyses.

Data from 135 patients were analyzed. The median follow-up was 952 days (interquartile range = 417-1555 days). In-hospital mortality was 62.2%, and overall mortality was 71.1%. The multivariable Cox regression analysis is adjusted for indication, and the survival after veno-arterial extracorporeal membrane oxygenation score showed that the following scores were associated with overall mortality: model for end-stage liver disease (hazard ratio = 1.04; 95% confidence interval = 1.01-1.07; p = 0.016), modified model for end-stage liver disease (hazard ratio = 1.04; 95% confidence interval = 1.01-1.06; p = 0.006), and model for end-stage liver disease with sodium (hazard ratio = 1.05; 95% confidence interval = 1.02-1.08; p = 0.001).

Model for end-stage liver disease, modified model for end-stage liver disease, and model for end-stage liver disease with sodium scores could be useful in the risk stratification of veno-arterial extracorporeal membrane oxygenation treatment in varying clinical indications.