Trends in utilization of antithrombotic therapy in patients with atrial fibrillation before stroke onset in a community-based study, from 1985 through 1997. From scientific evidence to practice.

Background: Despite the significant beneficial effects of antithrombotic therapy in primary prevention of stroke in patients with chronic nonvalvular atrial arrhythmia, this prevention therapy is underutilized. We conducted this population-based study to determine the rates and the trends of utilization of antithrombotic therapy for stroke patients with atrial fibrillation before stroke onset, and to evaluate indirectly the impact of medical recommendations on physician practice. Our aim was not to evaluate the efficacy of such prevention therapy.

Methods: From 2,330 men and women of any age registered for a first-ever stroke from 1985 to 1997 in a community-based study, we selected 599 patients admitted for ischemic stroke or TIA, associated with prior atrial fibrillation. Previous antithrombotic treatment before stroke onset was recorded and we evaluated the ratio of stroke patients who had received antithrombotic treatment for atrial fibrillation, from 1985 through 1997.

Results: Our study was performed to evaluate the practice of physicians in the prevention of stroke, and not to evaluate the efficacy of the anticoagulants in the prevention of stroke. Atrial fibrillation before stroke onset was identified in 599 patients. Of these, 222 (37%) received no antithrombotic therapy, 65 (10.8%) received an anticoagulant alone, 147 (24.5%) received an antiplatelet agent alone and 10 (1.7%) received both anticoagulation and antiplatelet treatment. From 1985 to 1988, the proportion of treated atrial fibrillation before stroke was small (14.6%). This increased to 21.5% within the period 1989-1991, to 40.3% within the period 1992-1994 and then to 47.6% within the period 1995-1997. It appears that the most significant change occurred within the period 1992-1994 (14.6% of treated atrial fibrillation within the period 1985-1987 constituted to 40.3% within the period 1992-1994) (P < 0.05), with a current rate of utilization of antithrombotic therapy close to 50%. The logistic regression analysis concerning anticoagulant therapy before stroke onset as a dependent variable, found that the factors independently associated with the use of anticoagulants before stroke were the lack of arterial hypertension and a history of smoking. The factors independently associated with the use of aspirin before stroke were arterial hypertension and lower limb peripheral vascular disease.

Conclusion: For primary prevention of stroke onset in patients with atrial fibrillation, therapeutic trials have changed medical practices although not to ideal levels because close to 50% of patients with atrial fibrillation experiencing an acute stroke or TIA received antithrombotic treatment. Therefore, clinical practice is inconsistent with the guidelines resulting from therapeutic trials. It is necessary to know the reasons for this inconsistency and to improve medical information about the cerebrovascular risk of atrial fibrillation and the efficacy of anticoagulants in stroke prevention in this condition.