DOAC Score Versus HAS‐BLED and ORBIT for Predicting Bleeding Events in Atrial Fibrillation on Direct Oral Anticoagulants

BACKGROUND: The comparative performance of the DOAC score versus established bleeding risk scores in patients with atrial fibrillation (AF) receiving direct oral anticoagulants (DOACs) remains uncertain. This meta-analysis evaluated the predictive ability of the DOAC score compared with HAS-BLED and ORBIT. METHODS: PubMed and Embase were systematically searched to identify studies assessing the predictive performance of the DOAC score in AF patients treated with DOACs. Pooled C-indices were calculated to compare discrimination. Reclassification metrics (net reclassification improvement NRI, integrated discrimination improvement IDI), calibration analyses, and decision curve analyses (DCA) were synthesized qualitatively. RESULTS: Nine studies comprising 12 cohorts were included (n = 89 688). The DOAC score demonstrated significantly superior discrimination for major bleeding compared with HAS-BLED (C-index 0.68 vs. 0.63). No significant differences were observed for intracranial hemorrhage, gastrointestinal bleeding, or clinically relevant non-major bleeding, nor in comparisons with ORBIT. Reclassification analyses showed heterogeneous findings, with several studies reporting no incremental benefit of the DOAC score, although one large cohort demonstrated improved NRI and IDI over HAS-BLED. Calibration analyses revealed good performance across scores, though both HAS-BLED and DOAC tended to overestimate bleeding risk in high-risk groups. DCA suggested variable but occasionally greater net benefit of the DOAC score at clinically relevant risk thresholds. CONCLUSIONS: The DOAC score provides modest but statistically significant improvement in predicting major bleeding compared with HAS-BLED, with comparable performance to ORBIT. However, reclassification, calibration, and clinical utility vary across settings, underscoring the need for further prospective validation.