Outcomes of Non–Vitamin K Antagonist Oral Anticoagulants Versus Warfarin in Patients with Atrial Fibrillation and Diabetes Mellitus: A Systematic Review of Randomized Controlled Trials and Observational Studies

OBJECTIVES: Concomitant diabetes mellitus (DM) increases risk of unfavorable outcomes among patients with atrial fibrillation (AF). The beneficial effect of non–vitamin K antagonist oral anticoagulants (NOACs) vs warfarin on both stroke and bleeding prevention within this population was established in meta-analyses. Comparisons between NOACs via meta-analyses may not be appropriate, considering the heterogeneity between studies. This study aimed to identify differences between original studies of NOACs vs warfarin to support interpretation of results across studies.

METHODS: A systematic search of PubMed was performed to identify relevant outcome studies.

RESULTS: Four randomized controlled trials (RCTs) and 5 observational studies comparing NOAC(s) vs warfarin in patients with AF and DM were identified. Among the 4 pivotal RCTs for each NOAC, patients from the edoxaban trial had the highest baseline stroke risk score; a baseline bleeding risk score or a common proxy was not available across studies. Dabigatran and edoxaban were the only NOACs to significantly reduce stroke (hazard ratio [HR]: 0.61, 95% confidence interval [CI]: 0.41–0.91) and major bleeding risk (HR: 0.79, 95% CI: 0.65–0.96) compared to warfarin. Among the 5 observational studies, either matching/weighting using propensity score or regression analyses were applied to adjust for covariates and balance between groups. However, baseline bleeding risk and diabetic-related factors were not assessed in every study. Definition of effectiveness and safety endpoints also differed across studies. Overall, NOACs were better than or comparable with warfarin with regards to stroke and bleeding prevention.

CONCLUSIONS: NOACs are more effective and safer than warfarin in patients with AF and DM in both clinical-trial and real-world settings. Appropriate covariate adjustment and standardized outcome measures will be required in future studies to make a comparison between NOACs across different studies possible.