Anticoagulation versus no anticoagulation after intracranial hemorrhage in patients with atrial fibrillation: an updated systematic review and meta-analysis.

BACKGROUND: Patients with atrial fibrillation (AF) and prior intracranial hemorrhage (ICH) present a therapeutic dilemma, where stroke prevention benefits of anticoagulation must be weighed against recurrent hemorrhage risks. Existing evidence remains inconclusive due to limited sample sizes. This updated meta-analysis synthesizes data from recent randomized controlled trials (RCTs) to determine whether anticoagulation provides net clinical benefit in this population. METHODS: We searched PubMed, Embase, and ClinicalTrials.gov until May 2025 for RCTs comparing anticoagulation versus no anticoagulation in AF patients with prior ICH. The primary efficacy and safety outcomes were ischemic stroke and recurrent ICH, respectively, with secondary outcomes of major adverse ischemic events (composite of ischemic stroke, systemic embolism, pulmonary embolism, and myocardial infarction), all-cause mortality, and major bleeding. Pooled effect estimates (hazard ratios or risk ratios) with confidence intervals. Number needed to treat (NNT) and number needed to harm (NNH) were also estimated to provide clinical context. RESULTS: Five RCTs with 733 patients (371 in the anticoagulation group and 362 in the no anticoagulation group) were included. Anticoagulation therapy demonstrated an 80% reduction in ischemic stroke risk (RR 0.20, 95% CI 0.06-0.71; NNT = 9) and a 64% reduction in major adverse ischemic events (HR 0.36, 95% CI 0.20-0.64, NNT = 11) while showing a nearly four-fold increased risk of intracranial hemorrhage (HR 3.69, 95% CI 1.54-8.84; NNH = 24) and a significantly increased risk of major bleeding (HR 2.40, 95% CI 1.20-4.81; NNH = 23). No significant difference was observed for all-cause mortality between groups (HR 1.17, 95% CI 0.80-1.72; NNH = 177). CONCLUSION: Anticoagulation in AF patients with prior ICH shows clinically meaningful reductions in thromboembolic events but significantly increases bleeding risks in this population. The balance between efficacy and safety underscores the need for individualized patient selection.