Levosimendan is often given during venoarterial extracorporeal membrane oxygenation (VA-ECMO) support with the intention of facilitating weaning. The available evidence, however, has been derived largely from nonrandomized studies and remains vulnerable to treatment-timing bias. An updated systematic review and meta-analysis was undertaken incorporating the LEVOECMO randomized trial and newer comparative observational studies.

PubMed/MEDLINE, Embase, the Cochrane Library, Web of Science, China National Knowledge Infrastructure, SinoMed, Wanfang Data, and the VIP Database were searched from the earliest date available in each database to May 21, 2026. Eligible studies enrolled adults with cardiogenic shock supported by VA-ECMO and compared levosimendan with no levosimendan, placebo, or standard care. The primary outcome was successful VA-ECMO weaning. The key secondary outcome was 28- and/or 30-day all-cause mortality. VA-ECMO duration and ventricular arrhythmias were assessed as exploratory outcomes. Random-effects models were prespecified as the primary analytical approach.

Nine studies were included, comprising one randomized trial and eight observational comparative studies. Six studies contributed to the primary analysis. Levosimendan was associated with a higher successful weaning rate overall (odds ratio [OR] 1.60, 95% confidence interval [CI] 1.17-2.17; p = 0.003; I² = 0%), whereas no benefit was seen in the randomized trial alone (OR 1.00, 95% CI 0.56-1.81). In sensitivity analysis restricted to the randomized trial and the more methodologically rigorous propensity score-based comparative studies, the association was attenuated and no longer significant (OR 1.53, 95% CI 0.82-2.86; p = 0.18). Levosimendan was not associated with lower 28- or 30-day mortality (OR 1.15, 95% CI 0.80-1.66; p = 0.45; I² = 0%). No significant difference was observed for VA-ECMO duration (mean difference 0.28 days, 95% CI −1.31 to 1.87; p = 0.73; I² = 83%). Ventricular arrhythmias were numerically more frequent with levosimendan, but the estimate was imprecise (OR 1.65, 95% CI 0.80-3.42; p = 0.18).

The apparent weaning benefit of levosimendan was driven mainly by nonrandomized evidence and was not confirmed by the LEVOECMO randomized trial. No reduction in 28- and/or 30-day mortality was observed, and current evidence does not support routine levosimendan use during VA-ECMO.