Peripheral versus central extracorporeal membrane oxygenation for postcardiotomy shock : Multicenter registry, systematic review, and meta-analysis

Background: We hypothesized that cannulation strategy in venoarterial extracorporeal membrane oxygenation (VA-ECMO) could play a crucial role in the perioperative survival of patients affected by postcardiotomy shock. Methods: Between January 2010 and March 2018, 781 adult patients receiving VA-ECMO for postcardiotomy shock at 19 cardiac surgical centers were retrieved from the Postcardiotomy Veno-arterial Extracorporeal Membrane Oxygenation study registry. A parallel systematic review and meta-analysis (PubMed/MEDLINE, Embase, and Cochrane Library) through December 2018 was also accomplished. Results: Central and peripheral VA-ECMO cannulation were performed in 245 (31.4%) and 536 (68.6%) patients, respectively. Main indications for the institution VA-ECMO were failure to wean from cardiopulmonary bypass (38%) and heart failure following cardiopulmonary bypass weaning (48%). The doubly robust analysis after inverse probability treatment weighting by propensity score demonstrated that central VA-ECMO was associated with greater hospital mortality (odds ratio 1.54; 95% confidence interval, 1.09-2.18), reoperation for bleeding/tamponade (odds ratio, 1.96; 95% confidence interval, 1.37-2.81), and transfusion of more than 9 RBC units (odds ratio, 2.42; 95% confidence interval, 1.59-3.67). The systematic review provided a total of 2491 individuals with postcardiotomy shock treated with VA-ECMO. Pooled prevalence of in-hospital/30-day mortality in overall patient population was 66.6% (95% confidence interval, 64.7-68.4%), and pooled unadjusted risk ratio analysis confirmed that patients undergoing peripheral VA-ECMO had a lower in-hospital/30-day mortality than patients undergoing central cannulation (risk ratio, 0.92; 95% confidence interval, 0.87-0.98). Adjustments for important confounders did not alter our results. Conclusions: In patients with postcardiotomy shock treated with VA-ECMO, central cannulation was associated with greater in-hospital mortality than peripheral cannulation