Background Risk stratification of atrial fibrillation patients with a congestive heart failure (C), hypertension (H), age ≥ 75 (A), diabetes (D), stroke or transient ischemic attack (TIA) (S2) (CHADS2) score of <2 remains imprecise, particularly in women. Our objectives were to validate the CHADS2 and congestive heart failure (C), hypertension (H), age ≥ 75 (A2), diabetes (D), stroke, TIA or prior thromboembolic disease (S2)- vascular disease (V), age 65-74 (A), female gender (S) (CHA2DS2-VASc) stroke risk scores in a healthy cohort of American women with atrial fibrillation and to determine whether CHA2DS2-VASc further risk-stratifies individuals with a CHADS2 score of <2. Methods We identified a cohort of 5981 women with atrial fibrillation not on warfarin at baseline (mean age 65.9 ± 7.2 years) enrolled in the Women's Health Initiative and followed for a median of 11.8 years. Univariate and multivariate proportional hazards analyses were used to examine these 2 risk scores, with main outcome measures being annualized event rates of ischemic stroke or transient ischemic attack stratified by risk score. Results Annualized stroke/transient ischemic attack rates ranged from 0.36% to 2.43% with increasing CHADS2 score (0-4+) (hazard ratio [HR] 1.57; 95% confidence interval [CI], 1.45-1.71 for each 1-point increase) and 0.20%-2.02% with increasing CHA2DS 2-VASc score (1-6+) (HR 1.50; 95% CI, 1.41-1.60 for each 1-point increase). CHA2DS2-VASc had a higher c statistic than CHADS2: 0.67 (95% CI, 0.65-0.69) versus 0.65 (95% CI, 0.62-0.67), P <.01. For CHADS2 scores <2, stroke risk almost doubled with every additional CHA2DS2-VASc point. Conclusions Although both CHADS2, and CHA2DS2-VASc are predictive of stroke risk in postmenopausal women with atrial fibrillation, CHA 2DS2-VASc further risk-stratifies patients with a CHADS2 score <2.
- Atrial fibrillation