The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation.

BACKGROUND Nonrheumatic atrial fibrillation increases the risk of stroke, presumably from atrial thromboemboli. There is uncertainty about the efficacy and risks of long-term warfarin therapy to prevent stroke. METHODS We conducted an unblinded, randomized, controlled trial of long-term, low-dose warfarin therapy (target prothrombin-time ratio, 1.2 to 1.5) in patients with nonrheumatic atrial fibrillation. The control group was not given warfarin but could choose to take aspirin. RESULTS A total of 420 patients entered the trial (212 in the warfarin group and 208 in the control group) and were followed for an average of 2.2 years. Prothrombin times in the warfarin group were in the target range 83 percent of the time. Only 10 percent of the patients assigned to receive warfarin discontinued the drug permanently. There were 2 strokes in the warfarin group (incidence, 0.41 percent per year) as compared with 13 strokes in the control group (incidence, 2.98 percent per year), for a reduction of 86 percent in the risk of stroke (warfarin:control incidence ratio = 0.14; 95 percent confidence interval, 0.04 to 0.49; P = 0.0022). There were 37 deaths altogether. The death rate was markedly lower in the warfarin group than in the control group: 2.25 percent as compared with 5.97 percent per year, for an incidence ratio of 0.38 (95 percent confidence interval, 0.17 to 0.82; P = 0.005). There was one fatal hemorrhage in each group. The frequency of bleeding events that led to hospitalization or transfusion was essentially the same in both groups. The warfarin group had a higher rate of minor hemorrhage than the control group (38 vs. 21 patients). CONCLUSIONS Long-term low-dose warfarin therapy is highly effective in preventing stroke in patients with non-rheumatic atrial fibrillation, and can be quite safe with careful monitoring.