Professor University of Cincinnati Mason, Ohio, United States
Background: Clinical trials do not reproduce the real-world studies due to specific protocols that they follow to achieve their endpoints.
Objectives: We aimed to review the effectiveness and safety outcomes of edoxaban versus warfarin in the real-world practice of individuals with Nonvalvular Atrial Fibrillation (NVAF).
Methods: The PubMed and EMBASE databases were searched for observational studies that compared edoxaban with warfarin of NVAF patients from January 1st, 2009 to September 21st, 2020. The title and abstract in both databases were reviewed. The focus was only on observational retrospective studies written with English language in human subjects using specific search items. The evaluated outcomes were effectiveness which includes stroke/systemic embolism (SE) and safety which includes major bleeding. The Hazard Ratios (HRs) and 95% confidence intervals (CIs) for effectiveness and safety outcomes were calculated. Because of anticipated different heterogeneity in observational studies, we decided to apply the fixed effect and random effect models.
Results: There was a total of six studies with a total of 87,236 patients on warfarin and 40,933 patients on edoxaban that met our criteria. The fixed effect model shows that patients receiving edoxaban had a significant lower risk of stroke/SE compared to warfarin HR= 0.66 (95% CI, 0.59, 0.73; P <.0001; I2 = 0%). In addition, the random effect represents a similar result of stroke/SE HR= 0.66 (95% CI, 0.61, 0.71; P <.0001; I2 = 0%). The fixed effect model shows that patients taking edoxaban had a significant lower risk of major bleeding compared to warfarin HR= 0.62 (95% CI, 0.55, 0.70; P <.0001; I2 = 47.60%). Also, the random effect represents a similar result of major bleeding HR= 0.61 (95% CI, 0.47, 0.79; P =0.0040; I2 = 45.10%).
Conclusions: Our findings indicated that edoxaban is significantly better than warfarin in minimizing the risk of stroke/SE and major bleeding in NVAF patients.
