This systematic review and meta-analysis aimed to evaluate clinical outcomes associated with different systemic heparin protocols (conservative v standard intensity) during extracorporeal membrane oxygenation (ECMO) support. Included studies comprised both venovenous and venoarterial ECMO modalities.

A systematic review and meta-analysis were conducted per Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. The PubMed, Cochrane Library, and Scopus databases were searched through November 15, 2024, for studies comparing conservative anticoagulation (no or low-target activated partial thromboplastin time [aPTT] of 40-60 seconds, anticoagulation monitoring) with standard anticoagulation (standard heparin target aPTT of 60-80 seconds).

Studies were conducted in intensive care units across multiple international centers.

Eight studies (601 patients; 280 in the conservative group and 321 in the standard group) were included in the meta-analysis.

Conservative (no or low-target aPTT of 40-60 seconds, anticoagulation monitoring) versus standard anticoagulation (standard heparin target aPTT of 60-80 seconds).

Primary outcomes included mortality, major bleeding, and thrombosis; secondary outcomes were stroke, limb ischemia, and circuit thrombosis. Random-effects models assessed outcomes, with heterogeneity evaluated using the Q test and I² statistic. A total of 698 studies were identified (PubMed, 461; Cochrane Library, 94; Scopus, 143). After removal of duplicates, 358 records were screened, and 245 studies were excluded. Twenty-seven studies underwent full-text review, with 13 meeting the inclusion criteria. Eight studies (601 patients; 280 in the conservative group and 321 in the standard group) were included in the meta-analysis. Mortality was similar between the conservative group (131 of 280 patients [46.8%]) and the standard group (159 of 321 [49.5%]) (odds ratio [OR], 1.04 [95% confidence interval (CI), 0.53-2.05]; I² = 64%). Thrombotic events were assessed in 4 studies (26 of 123 in the conservative group [21.1%] v 47 of 178 in the standard group [26.4%]; OR, 0.60 [95% CI, 0.33-1.09]; I² = 0%). Stroke (5 of 127 v 5 of 178), limb ischemia (10 of 87 v 20 of 147), and circuit thrombosis (15 of 141 v 15 of 191) did not differ between groups. However, major bleeding significantly favored conservative strategies (7 studies; 74 of 251 [29.5%] v 157 of 297 [52.9%]; OR, 0.40 [95% CI, 0.18-0.91]; p = 0.03; I² = 72%).

The findings of this systematic review and meta-analysis indicate a potential association between conservative anticoagulation strategies during ECMO and reduced bleeding events without increased mortality or thrombotic complications, including stroke, limb ischemia, and circuit thrombosis. Given the low event rates, wide CIs, and absence of a formal power calculation, these null findings should not be interpreted as evidence of equivalence.