Expanded Application of the DOAC Score for Predicting Bleeding Events in All Anticoagulated Patients With Atrial Fibrillation: An Integrated Analysis of 2 Prospective Japanese Registries-RAFFINE and SAKURA-AF.

Background: Bleeding risk is a concern in patients with atrial fibrillation (AF) on oral anticoagulants (OACs). The DOAC score, a novel bleeding risk score, was developed for patients on direct OACs (DOACs), but its applicability to all anticoagulated patients remains unclear. The aim of this study was to validate the DOAC score in patients with AF who received any OACs. Methods and Results: We analyzed data for 6,505 patients with nonvalvular AF receiving OACs from 2 Japanese multicenter registries (RAFFINE, SAKURA-AF; mean age 72±9 years; 28.1% female). Major bleeding occurred in 76 patients within 1 year. The DOAC score was compared to the HAS-BLED score in discrimination ability. The DOAC score showed higher area under the curve (AUC), although the difference was not statistically significant (AUC 0.653 vs. 0.601, P=0.23). The DOAC score also showed a trend towards improved net reclassification over the HAS-BLED score (net reclassification improvement (NRI): 25.4%, P=0.073). In patients without a prior history of major bleeding, the DOAC score offered a significant improvement in the NRI (30.9%, P=0.030). Conclusions: The DOAC score demonstrated moderate performance for predicting major bleeding in anticoagulated patients with AF. Although it did not significantly outperform the HAS-BLED score in this study, its applicability to both DOAC and warfarin users may offer a practical, unified approach.