Predictive abilities of the HAS-BLED and ORBIT bleeding risk scores in non-warfarin anticoagulated atrial fibrillation patients: An ancillary analysis from the AMADEUS trial

Background

Simple bleeding risk scores have been proposed to predict bleeding events, in patients anticoagulated using non-warfarin anticoagulants. We compared the relative predictive values of two bleeding risk scores, HAS-BLED and ORBIT, in non-warfarin anticoagulated patients with atrial fibrillation (AF).

Methods and results

In a post-hoc ancillary analysis of ‘clinically relevant bleeding’ events amongst 2283 patients in the idraparinux arm in the AMADEUS trial. The two scores performed modestly in predicting both bleeding outcomes, although there was a trend for better HAS-BLED score performance in predicting any clinically relevant bleeding [c-indexes in HAS-BLED vs. ORBIT; 0.61 (95% CI; 0.58–0.64) vs. 0.58 (95% CI; 0.55–0.61); c-index difference = 0.03, z-score = 1.84, p = 0.06)]. Using the HAS-BLED score compared with the ORBIT score correctly and significantly reclassified 15.6% of the population (95% CI: 4.3 to 27.0; p = 0.007). Decision curve analyses confirmed the increasing ability to correctly identify patients who would bleed using the HAS-BLED score versus the ORBIT score, over a wide range of thresholds for any clinically relevant bleeding risk predictions.

Conclusion

In this comparison of the HAS-BLED and ORBIT scores in a cohort of non-warfarin anticoagulated patients with AF, we show that the HAS-BLED score more accurately predicted any clinically relevant bleeding amongst patients with AF who were anticoagulated with a non-warfarin anticoagulant, when compared with the ORBIT score.