Predictors associated with successful weaning of veno-venous extracorporeal membrane oxygenation and mortality in adult patients with severe acute lung failure: Protocol of a pooled data analysis of cohort studies

Background Severe acute lung failure (ALF) often necessitates veno-venous extracorporeal membrane oxygenation (VV-ECMO), where identifying predictors of weaning success and mortality remains crucial yet challenging. The study aims to identify predictors of weaning success and mortality in adults undergoing VV-ECMO for severe ALF, a gap in current clinical knowledge. Methods and analysis PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials will be searched for cohort studies examining the predictive factors of successful weaning and mortality in adult patients on VV-ECMO due to severe ALF. Risk of bias assessment will be conducted using the Newcastle-Ottawa scale for each included study. The primary outcomes will be successful weaning from VV-ECMO and all-cause mortality. Between-study heterogeneity will be evaluated using the I 2 statistic. Sensitivity, subgroup, and meta-regression analyses will be performed to ascertain potential sources of heterogeneity and assess the robustness of our results. We will use the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) tool to recommend the level of evidence. Discussion This study seeks to provide clinically significant insights into predictors for weaning and mortality during VV-ECMO treatment for ALF, aiming to support clinical decisions and potentially influence health policy, thereby improving patient outcomes. Ethics and dissemination Given the absence of direct engagement with human subjects or access to personal medical records, ethical approval for this study is deemed unnecessary. The study findings will be shared at a scientific conference either at the global or national level. Alternatively, the results will be presented for publication in a rigorously peer-reviewed journal regarding critical care medicine.