Discriminative Ability of CHA2DS2-VASc and HAS-BLED Score in Whites and Nonwhites.

ABSTRACT

The CHA2DS2-VASc and HAS-BLED scoring systems are used in patients with atrial fibrillation (AF) to estimate risk of stroke and bleeding, respectively. Both were developed in minimally diverse European populations and these scores have not yet been extensively studied in US whites and nonwhites. In a retrospective cohort study, we included patients with AF who received inpatient or outpatient care in a large integrated academic health system from 2011 to 2017. Cox proportional hazards were used to analyze associations between stroke and CHA2DS2-VASc score in AF patients not prescribed anticoagulation and between incident bleeding and HAS-BLED score in anticoagulated patients. After exclusions for previous stroke, the cohort included 21,648 patients with a mean age of 66.8 ± 15.8. Anticoagulation was prescribed in 52% of whites and 46% of nonwhites (p < 0.001) with a CHA2DS2-VASc score of ≥2. Mean CHA2DS2-VASc scores were 2.4 ± 1.6 in whites and 2.2 ± 1.6 in nonwhites and mean HAS-BLED scores was 1.5 ± 1.1 in whites and 1.3 ± 1.0 in nonwhites. After adjusting for baseline differences, the discriminative ability of CHA2DS2-VASc and HAS-BLED was similar in whites and nonwhites (p = 0.52, 0.33, respectively). The discriminative ability of HAS-BLED was similar in patients on vitamin K antagonists and direct oral anticoagulants. In conclusion, oral anticoagulation was prescribed less frequently in nonwhites. However, the discriminative ability of CHA2DS2-VASc and HAS-BLED were similar in whites and nonwhites.