The Balance Between Stroke Prevention and Bleeding Risk in Atrial Fibrillation

See related article, pages 1482–1486. Although atrial fibrillation (AF) is well recognized to confer a risk of stroke, this risk is not homogeneous. Oral anticoagulation (OAC) with warfarin is highly beneficial, but such therapy is inconvenient and carries a risk of bleeding. Thus, stroke risk stratification schemes have been devised to identify “high risk” AF patients for whom the absolute benefits of OAC exceed its risks.In general, present treatment guidelines recommend OAC for those classed at high risk of stroke, and aspirin for those at “low risk.” In those at “moderate risk,” either OAC or aspirin is recommended. There are many ways of classifying stroke risk, and in a recent comparison of 12 risk stratification schemes to predict stroke in patients with nonvalvular AF, the Stroke Risk in Atrial Fibrillation Working Group1 identified 7 schemes that were based directly on event-rate analyses (largely been identified from non-OAC arms of clinical trials, and occasionally from cohort studies), whereas 5 resulted from expert panel consensus. The most frequently included features were prior stroke/TIA (in 100% of schemes), patient age, hypertension and diabetes mellitus.Of the various published schema, the CHADS2 score is probably the most popular, which is well validated and easy to use, where 1 point is given for Congestive heart failure, Hypertension, Age >75 and Diabetes, whereas 2 points are given for Stroke or transient ischemic attack (TIA).2Based on published test study cohorts, the absolute stroke rates for the different stroke risk schema varied rather widely, and the proportions of patients categorized by the different schemes as low-risk or high risk can differ fairly substantially.1 Indeed, validation studies have shown that no published risk stratification schema is ideal and all can frequently underestimate stroke risk.2,3 The c -statistic (used to assess the predictive …