AbstractBackground and Purpose—
As the threshold of stroke risk for initiating oral anticoagulants is lowered after the introduction of the nonvitamin K antagonist oral anticoagulants, the focus of stroke prevention in patients with nonvalvular atrial fibrillation has shifted away from predicting high-risk patients toward initially identifying patients with a truly low risk of ischemic stroke, who do not need antithrombotic therapy. We tested the predictive ability of the congestive heart failure, hypertension, age ≥75, diabetes mellitus, prior stroke or transient ischemic attack (doubled; CHADS2), congestive heart failure, hypertension, age ≥75 (doubled), diabetes mellitus, prior stroke or transient ischemic attack (doubled), vascular disease, age 65 to 74, female (CHA2DS2-VASc), and Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) risk stratification schemes in oral anticoagulants naive patients with atrial fibrillation in a Korean nationwide sample cohort.Methods—
From January 2002 to December 2008, a total of 5855 oral anticoagulant naive patients with nonvalvular atrial fibrillation aged ≥20 years were enrolled from Korea National Health Insurance Service-Sample Cohort database and were followed-up until December 2013.Results—
At baseline, the proportions categorized as low risk using CHADS2, CHA2DS2-VASc, and ATRIA risk stratification schemes were 1049 (17.9%), 860 (14.7%), and 3280 (56.0%), respectively. During follow-up, the low-risk category using CHADS2, CHA2DS2-VASc, and ATRIA scores was retained in 811 (13.9%), 667 (11.4%), and 2729 (46.6%) patients, respectively. Rates of ischemic stroke (100 person-years) in the low risk categories of CHADS2, CHA2DS2-VASc, and ATRIA scores were 0.42, 0.26, and 1.43, respectively. CHA2DS2-VASc had the best sensitivity (98.8% versus 85.7% in CHADS2 and 74.8% in ATRIA) and negative predictive value (98.8% versus 95.3% for CHADS2 and 93.7% for ATRIA) for the prediction of stroke incidence and was best for the prediction of the absence of ischemic stroke during 5 years of follow-up (odds ratio, 16.4 [95% confidence interval, 8.8–30.8]).Conclusions—
The CHA2DS2-VASc score shows good performance in defining truly low-risk Asian patients with atrial fibrillation for stroke compared with CHADS2 and ATRIA scores.