Bleeding risk in East Asian patients with atrial fibrillation: Validation of the DOAC score in a Japanese multicenter registry

Le résumé fourni par la source

Background: Bleeding events remain relatively frequent in patients with atrial fibrillation (AF), posing significant clinical challenges in routine patient management. The recently developed direct oral anticoagulant (DOAC) score shows promise in predicting bleeding risk; however, its performance in East Asian patients is not well characterized. Objective: This study aimed to assess the performance of the DOAC score in a contemporary outpatient-based registry of Japanese patients with AF and evaluate potential modifications to enhance its predictive accuracy. Methods: The Keio Interhospital Cardiovascular Studies-Atrial Fibrillation registry is a multicenter prospective database designed according to international standards for newly diagnosed AF. Adjudicated bleeding events were prospectively collected, and we assessed each bleeding risk score's predictive capacity using the C-index and calibration analysis. Performance was evaluated at 1- and 5-year intervals. Results: Among 2101 patients (median age of 70 years; 67.9% men), the overall 1- and 5-year bleeding incidence was 2.1% and 4.7%, respectively. The DOAC score demonstrated acceptable discrimination at 1 year (C-index 0.704, 95% confidence interval [CI] 0.634-0.774), which was maintained over the extended 5-year follow-up (C-index 0.690, 95% CI 0.640-0.741). Calibration analysis showed acceptable performance across validation datasets. Notably, modifying antiplatelet scoring by distinguishing aspirin (2 points) from other antiplatelets (1 point) improved the performance (C-index 0.707, 95% CI 0.637-0.777). Conclusion: The DOAC score proved valuable for long-term bleeding risk assessment in Japanese patients with AF, with enhanced performance through antiplatelet stratification. These findings support its implementation in East Asian populations for individualized clinical decision making.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.