The risk of stroke is more than five times greater in patients with atrial fibrillation than in those who have normal sinus rhythm. The annual risk in patients younger than age 60 with lone atrial fibrillation is 0.5 percent, which is similar to that of a control population matched for age and gender. In elderly patients the risk is much higher, often more than 10 percent annually. Ezekowitz and Levine conducted a MEDLINE search for studies of patients with atrial fibrillation that evaluated the effectiveness of warfarin, aspirin or combination therapy in preventing stroke.

Pooled data from five primary prevention trials showed that warfarin was associated with a 68 percent reduction in stroke in patients with atrial fibrillation, with no significant increase in major bleeding complications. Follow-up was maintained for 1.2 to 2.3 years. During that time, four of the five trials were discontinued early because of the significant overall benefit occurring in patients taking warfarin. Independent risk factors for stroke include increasing age, history of hypertension, diabetes and previous transient ischemic attack. Patients younger than 65 years of age with atrial fibrillation but no risk factors do not benefit from warfarin treatment. However, all other patients benefited, particularly women, who experienced an 84 percent reduction in risk.

The Stroke Prevention in Atrial Fibrillation (SPAF III) study showed that low-risk patients between 65 and 75 years of age with atrial fibrillation benefited from receiving aspirin daily. Low-risk patients were characterized as those who had no recent history of congestive heart failure, previous thromboembolism or systolic blood pressure higher than 160 mm Hg, and were not women older than 75 years of age. A previous stroke prevention trial had shown a 42 percent reduction in stroke in low-risk patients taking aspirin alone. While a specific dosage was not recommended, most studies have tested 325 mg per day. SPAF III also compared combination therapy (consisting of low-intensity fixed-dose warfarin and aspirin) with adjusted-dose warfarin alone. The fixed dose met an International Normalization Ratio (INR) of 1.2 to 1.5, whereas the INR of the adjusted dose was 2.0 to 3.0. Patients taking combination therapy had a much higher annual rate of stroke (7.9 percent) compared with those taking warfarin alone (1.9 percent). Overall, the primary prevention studies indicated that use of therapeutic doses of warfarin alone were superior to aspirin alone and combination therapy.

The one study addressing secondary prevention randomized patients to receive anticoagulation, aspirin or placebo. These data were evaluated separately from those in the primary prevention trials. Patients who were unable to take warfarin were randomized to receive either aspirin or placebo. Follow-up was maintained for approximately 2.3 years. Warfarin reduced the risk of stroke by 66 percent. Patients taking aspirin or placebo did not experience such dramatic reductions (15 and 19 percent, respectively). Recurrence of stroke was about three times higher in the placebo group in this study than in the placebo groups in the primary prevention trials.

The authors conclude that warfarin is effective in preventing stroke in patients with atrial fibrillation, although aspirin is adequate in low-risk patients.