All-cause mortality after major gastrointestinal bleeding among patients receiving direct oral anticoagulants: a systematic review and meta-analysis

Background Although gastrointestinal (GI) bleeding represents the single most frequent site of anticoagulant-related major bleeding, outcomes after major GI bleeding including mortality are not well characterized and severity may be underappreciated. We aimed to determine 30-day all-cause mortality in adults with major GI bleeding on a direct oral anticoagulant (DOAC). Methods We searched MEDLINE, EMBASE, and Cochrane CENTRAL from inception to May 9, 2024 for randomized controlled trials and cohort studies that reported 30-day all-cause mortality after major GI bleeding in adults treated with a DOAC for venous thromboembolism or atrial fibrillation. Risk of bias was assessed using a modified QUIPS tool for prognostic studies. 30-day all-cause mortality was calculated using the random-effects inverse-variance method. Results We included 20 studies comprising 3987 DOAC-treated patients with major GI bleeds. The pooled estimate of 30-day all-cause mortality was 8.4 % (95 % confidence interval [CI], 4.9–12.5; I 2 = 83 %). In subgroup analyses, 30-day all-cause mortality was 10.3 % (95 % CI, 6.5–14.7; I 2 = 24 %) in prospective studies (9 studies, 675 major GI bleeds), 7.3 % (95 % CI, 2.2–14.4; I 2 = 90 %) in retrospective studies (11 studies, 3312 major GI bleeds), 12.9 % (95 % CI, 6.3–21.1; I 2 = 44 %) in considered at high risk of bias (9 studies, 387 major GI bleeds), and 6.1 % (95 % CI, 2.9–10.1; I 2 = 89 %) in those at low risk of bias (10 studies, 3562 major GI bleeds). Conclusion DOAC-related major GI bleeding appears to be associated with significant 30-day all-cause mortality. Further research is needed regarding the causes and contributors to mortality in this population to identify patients at high risk and develop mitigation strategies.