Stroke Risk Stratification in a "Real-World" Elderly Anticoagulated Atrial Fibrillation Population

Methods: Six hundred and sixty-two consecutive AF patients (mean [SD] age 74 [7.7] years; 36.1% female) referred to the Anticoagulation Clinic of the Azienda Ospedaliera Careggi of Florence, Italy, were included and followed-up for a mean 3.6 +/- 2.7 years for the incidence of thromboembolic (TE) events. The ability of the new CHA(2)DS(2)-VASc schema to predict TE was compared with other contemporary stroke risk schema (including CHADS(2), NICE 2006, ACC/AHA/ESC 2006, and ACCP 2008), by determining the c-statistic. Results: Univariate predictors of TE events were female gender (odds ratio 1.9; 95%CI [confidence intervals] 1.01-3.70) and previous stroke/transient ischemic attack (TIA)/TE (OR 5.6; 95%CI 2.70-11.45), although after adjustment only previous stroke/TIA/TE was an independent predictor of TE (OR 5.5; 95%CI 2.68-11.31; P = 0.0001). All stroke risk schema had modest discriminating ability, with c-statistics ranging from 0.54 (atrial fibrillation investigators [AFI]) to 0.72 (CHA(2)DS(2)-VASc). The CHADS(2) and CHA(2)DS(2)-VASc schemes having the best c-statistics (0.717 and 0.724, respectively) with significant discriminating value between risk strata (both P <0.001). The proportion of patients assigned to individual risk categories varied widely across the schema, with those categorized as "moderate-risk" ranging from 5.3% (CHA(2)DS(2)-VASc) to 49.2% (CHADS(2)-classical). Conclusion: In this "real world" cohort, current published risk schemas have modest predictive ability, with the CHADS(2) and CHA(2)DS(2)-VASc schemes having the best predictive value for thromboembolism. Future trials could assess the value of alternative strategies for thromboprophylaxis in high-risk anticoagulated patients identified by these schemes. (J Cardiovasc Electrophysiol, Vol. 22, pp. 25-30, January 2011).