Response by Overvad et al to Letter Regarding Article, "Female Sex Is a Risk Modifier Rather Than a Risk Factor for Stroke in Atrial Fibrillation: Should We Use a CHA ₂DS ₂-VA Score Rather Than CHA ₂DS ₂-VASc?"

BACKGROUND: Stroke risk in atrial fibrillation is assessed by using the CHA2DS2-VASc score. Sex category (Sc, ie, female sex) confers 1 point on CHA2DS2-VASc. We hypothesized that female sex is a stroke risk modifier, rather than an overall risk factor, when added to a CHA2DS2-VA (sex-independent thromboembolism risk) score scale.

METHODS: Using 3 nationwide registries, we identified patients with incident nonvalvular atrial fibrillation from January 1, 1997, through December 31, 2015. Patients receiving oral anticoagulant treatment at baseline were excluded, and person-time was censored at the time of treatment initiation (if any). CHA2DS2-VA scores were calculated for men and women, and were followed for up to 1 year in the Danish National Patient Registry. The primary outcome was a primary hospital code for ischemic stroke or systemic embolism (thromboembolism). We calculated crude event rates for risk strata as events per 100 person-years. For quantifying absolute risk of stroke, we calculated risks based on the pseudovalue method. Female sex as a prognostic factor was investigated by inclusion as an interaction term on the CHA2DS2-VA score to calculate the thromboembolic risk ratio for different score points.

RESULTS: A total of 239 671 patients with atrial fibrillation (48.7% women) contributed to the analyses. The mean ages for women and men were 76.6 years and 70.3 years, respectively; the mean CHA2DS2-VA scores were 2.7 for women and 2.3 for men. The overall 1-year thromboembolic rate per 100 person-years for women was 7.3 and 5.7 for men. The 1-year absolute risk of thromboembolism was 0.5% among men and women with a CHA2DS2-VA score of 0 and increased up to >7% among very comorbid patients (score >5). The risk ratio (male as reference) across points >1 indicated that women exhibit a higher stroke risk. The interaction was statistically significant (P<0.001).

CONCLUSIONS: Female sex is a risk modifier for stroke in patients with atrial fibrillation. Initial decisions on oral anticoagulant treatment could be guided by a CHA2DS2-VA score (ie, excluding the sex category criterion), but the Sc risk component modifies and accentuates stroke risk in women who would have been eligible for oral anticoagulant treatment on the basis of ≥2 additional stroke risk factors.