[Atrial fibrillation: embolic risk and prevention].

When atrial fibrillation (AF) complicates rheumatic heart disease, the risk of stroke is 17 times that of patients in sinus rhythm and full anticoagulation is mandatory. Non-rheumatic AF carries a lower risk--5% per annum, a 5-fold increase. Four major trials (SPAF, AFASAK, BAATAF, CAFA) have lately examined thromboembolic prophylaxis in this group of patients. These randomized prospective open studies showed a 56-86% reduction in stroke and systemic embolism in patients receiving full anticoagulation compared with placebo. In older people, the BAATAF trial of low-dose warfarin (INR = 1.5-2.7) showed an 86% reduction in stroke and a significant reduction in mortality. In all 4 studies the incidence of hemorrhagic complication was very low (0.5%). In SPAF trial, aspirin, 325 mg/day was found to be effective, but this was not the case in AFASAK, which used 75 mg/day and had an older population. In a double-blind randomized trial indobufene, 100 mg bid, was found effective in the 67% reduction of stroke, systemic and pulmonary embolism in patients with various cardiac diseases in AF or in sinus rhythm. Consequently, a reasonable policy would be to give full or low-dose anticoagulation to those patients with chronic AF who have structural heart disease or are over 65 years old; to consider low-dose anticoagulation or aspirin or indobufene in younger patients with chronic lone AF; and to give indobufene or aspirin or nothing to those with episodes of paroxysmal AF lasting hours only. In borderline cases, the use of transesophageal echocardiography to exclude left atrial thrombus and spontaneous echo contrast may aid decision-making.