| Abstract: | Vitamin K-inhibiting anticoagulants, especially warfarin, have become standard treatment for reducing the incidence of strokes and systemic emboli in patients with nonvalvular atrial fibrillation (NVAF). The randomized controlled trials that form the scientific basis for the efficacy of anticoagulants in prophylaxis of embolic events show small but statistically significant benefit with warfarin and other vitamin K inhibitors. The generalizability of these randomized trials to clinical practice is highly questionable because of the low percentage of NVAF patients from the participating institutions that entered the trials, the relatively young age of the patients, and the superior anticoagulation monitoring compared with that in general practice. Indirect comparisons of warfarin with aspirin by looking at separate meta-analyses of placebo-controlled randomized trials give potentially biased results. The meta-analyses of trials directly comparing warfarin with aspirin have diametrically opposing conclusions. In contrast to observational studies of general medical practice, randomized trials significantly underestimate the bleeding risks of warfarin. Anticoagulants for stroke prophylaxis for NVAF cause about 17,000 major bleeds in the United States per year, of which about 4000 are fatal. In 5 randomized trials with follow-up periods of 1.3-2.3 years, 10% to 38% of patients permanently discontinued anticoagulants. The 2-year average follow up of patients in the randomized trials is too short to predict the long-term impact of anticoagulation on the natural history of NVAF. Aspirin should be preferred over anticoagulants in prophylaxis against cardiogenic embolism in NVAF patients. |
