Letter by Spiegel et al regarding article, "net clinical benefit of warfarin in patients with atrial fibrillation: a report from the Swedish atrial fibrillation cohort study".

BACKGROUND Known risk factors for bleeding during anticoagulant treatment are largely the same as those predicting thromboembolic events in patients with atrial fibrillation (AF). Our objective was to investigate how to maximize the likelihood of avoiding both stroke and bleeding. METHODS AND RESULTS All 182 678 subjects with atrial fibrillation in the Swedish Hospital Discharge Register were studied for an average of 1.5 years (260 000 patient-years at risk). Patients were stratified according to risk scores with the use of historic International Classification of Disease diagnostic codes in the register. Information about medication was obtained from the Swedish Drug Registry. Our primary end point was net benefit defined as number of avoided ischemic strokes with anticoagulation minus the number of excess intracranial bleedings with a weight of 1.5 to compensate for the generally more severe outcome with intracranial bleedings. The adjusted net clinical benefit favored anticoagulation for almost all atrial fibrillation patients. The exceptions were patients at very low risk of ischemic stroke with a CHA(2)DS(2)-VASc score of 0 and moderately elevated bleeding risk (-1.7%/y). The results were broadly similar with CHADS(2), except for patients with very low embolic risk; the CHA(2)DS(2)-VASc was able to identify those patients (n=6205, 3.9% of all patients) who had no net clinical benefit or even some disadvantage from anticoagulant treatment. CONCLUSIONS In almost all patients with atrial fibrillation, the risk of ischemic stroke without anticoagulant treatment is higher than the risk of intracranial bleeding with anticoagulant treatment. Analysis of the net benefit indicates that more patients may benefit from anticoagulant treatment.