OBJECTIVES: The purpose of this meta-analysis was to compare efficacy and safety of direct oral anticoagulants (DOACs) to warfarin for secondary stroke prevention among adult patients with atrial fibrillation and prior stroke.

METHODS: Major repositories were screened for randomized controlled trials (RCTs), RCT subgroups, and observational studies (OBSs, divided in claims and non-claims). Occurrence of ischemic stroke or transient ischemic attack, systemic embolism, all-cause mortality, intracranial hemorrhage (ICH), and major bleeding were outcomes of interest. Hazard ratios (HRs) and their confidence intervals (95%CIs) were pooled using random-effects models for each study design. Claims studies were analyzed separately from non-claims, while RCT subgroups were treated as OBSs as the randomization was broken.

RESULTS: Of 8,647 articles, 20 were included (one RCT, six RCT subgroups, nine claims, and four non-claims). Comparing DOACs to warfarin, pooled HRs (95%CI) were consistently in favor of DOACs although some did not reach statistical significance: for ischemic stroke 0.84 (0.66-1.07) in claims; 0.90 (0.77-1.06) in non-claims and RCT subgroups; for systemic embolism 0.77 (0.62-0.96) in claims; 0.86 (0.77-0.96) in non-claims and RCT subgroups; for all-cause mortality 0.57 (0.33-0.99) in claims; 0.87 (0.79-0.96) in non-claims and RCT subgroups; for ICH 0.72 (0.39-1.33) in claims; 0.51 (0.38-0.67) in non-claims and RCT subgroups; and for major bleeding 0.86 (0.71-1.03) in claims; 0.90 (0.76-1.08) for non-claims and RCT subgroups.

CONCLUSIONS: DOACs were associated with better efficacy and safety profiles than warfarin in atrial fibrillation patients with prior stroke; more specifically, a lower risk of systemic embolism, all-cause mortality, and ICH.