Arrhythmias & electrophysiology
Development and external validation of the GLORIA-AF Bleeding Weighted Risk Score for patients with atrial fibrillation
Authors' conclusion
The GLORIA-AF Bleeding Score showed modestly higher discrimination than HAS-BLED for 1-year major bleeding prediction while retaining clinical interpretability and usability. External validation in independent European and Asia-Pacific registries further supports its transportability beyond the derivation population.
Study context
- Population
- Patients with atrial fibrillation (AF) (GLORIA-AF Phase II/III derivation; EORP-AF and APHRS-AF external validation)derivation cohort
- Participants
- 20,250 patients
- Comparison
- GLORIA-AF Bleeding Weighted Risk Score vs HAS-BLED, unweighted GLORIA-AF score and full multivariable Cox model
- Outcome
- Major bleeding prediction (discrimination by C-index)
- Follow-up
- 1 year
Methodological detailsprospective, multinational, LASSO-penalized regression with stability selection, external validation in EORP-AF and APHRS-AF registries
- Prospective
- Multinational
- LASSO-penalized regression with stability selection
- External validation in EORP-AF and APHRS-AF registries
Performance
As reported in the abstract
Major bleeding, discrimination (derivation cohort)
| Test or model | C-index |
|---|---|
| GLORIA-AF Bleeding Weighted Risk Score | 0.688 (95% CI: 0.660-0.717) |
| HAS-BLED | 0.631 (95% CI: 0.604-0.658) |
| Full multivariable regression | 0.690 (95%CI: 0.661-0.719) |
C-index
GLORIA-AF Bleeding Weighted Risk Score0.688 (95% CI: 0.660-0.717)
HAS-BLED0.631 (95% CI: 0.604-0.658)
Full multivariable regression0.690 (95%CI: 0.661-0.719)
- GLORIA-AF Bleeding Weighted Risk Score vs HAS-BLED: P < 0.001
- GLORIA-AF Bleeding Weighted Risk Score vs Full multivariable regression: P = 0.084 — comparable discrimination
- Major bleeding during 1-year follow-up
- 317 (1.6%)
- Calibration (O:E ratio)
- 0.996
Further findings
Discrimination for 1-year major bleeding, external validation (EORP-AF and APHRS-AF)
- Derived score C-index in external validation
- 0.674 (95% CI 0.643-0.705; P < 0.001)
- Better discrimination, calibration and clinical benefit than HAS-BLED
- n = 9,752 external validation patients; 148 (1.5%) major bleeding
Original publication
Development and external validation of the GLORIA-AF Bleeding Weighted Risk Score for patients with atrial fibrillation. Eur J Prev Cardiol. Published 29 August 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →