Abstract
BACKGROUND: Extracorporeal membrane oxygenation (ECMO) may provide benefit in certain populations of adults, including those with severe cardiac failure, severe respiratory failure, and cardiac arrest. However, it is also associated with serious short- and long-term complications, and there remains a lack of high-quality evidence to guide practice. Recently several large randomized controlled trials (RCTs) have been published, therefore, we undertook an update of our previous systematic review published in 2014. OBJECTIVES: To evaluate whether venovenous (VV), venoarterial (VA), or ECMO cardiopulmonary resuscitation (ECPR) improve mortality compared to conventional cardiopulmonary support in critically ill adults. SEARCH METHODS: We used standard, extensive Cochrane search methods. The latest search date was March 2022. The search was limited to English language only. SELECTION CRITERIA: We included RCTs, quasi-RCTs, and cluster-RCTs that compared VV ECMO, VA ECMO or ECPR to conventional support in critically ill adults. DATA COLLECTION AND ANALYSIS: We used standard Cochrane methods. Our primary outcome was 1. all-cause mortality at day 90 to one year. Our secondary outcomes were 2. length of hospital stay, 3. survival to discharge, 4. disability, 5. adverse outcomes/safety events, 6. health-related quality of life, 7. longer-term health status, and 8. cost-effectiveness. We used GRADE to assess certainty of evidence. MAIN RESULTS: Five RCTs met our inclusion criteria, with four new studies being added to the original review (total 757 participants). Two studies were of VV ECMO (429 participants), one VA ECMO (41 participants), and two ECPR (285 participants). Four RCTs had a low risk of bias and one was unclear, and the overall certainty of the results (GRADE score) was moderate, reduced primarily due to indirectness of the study populations and interventions. ECMO was associated with a reduction in 90-day to one-year mortality compared to conventional treatment (risk ratio [RR] 0.80, 95% confidence interval [CI] 0.70 to 0.92; P = 0.002, I(2) = 11%). This finding remained stable after performing a sensitivity analysis by removing the single trial with an uncertain risk of bias. Subgroup analyses did not reveal a significant subgroup effect across VV, VA, or ECPR modes (P = 0.73). Four studies reported an increased risk of major hemorrhage with ECMO (RR 3.32, 95% CI 1.90 to 5.82; P < 0.001), while two studies reported no difference in favorable neurologic outcome (RR 2.83, 95% CI 0.36 to 22.42; P = 0.32). Other secondary outcomes were not consistently reported across the studies. AUTHORS' CONCLUSIONS: In this updated systematic review, which included four additional RCTs, we found that ECMO was associated with a reduction in day-90 to one-year all-cause mortality, as well as three times increased risk of bleeding. However, the certainty of this result was only low to moderate, limited by a low number of small trials, clinical heterogeneity, and indirectness across studies.