[Preventive treatment of systemic embolic complications of atrial fibrillation]

Non-valvular atrial fibrillation multiplies the risk of presumed embolic events by a factor of four. The hemorrhagic risk of anticoagulant treatment varies considerably and its prophylactic efficacy was not tested in any randomised trial before the end of 1989. The recommendations of experts at that time recognised that data were inadequate. The publication since of four randomised trials involving 3,049 patients has provided a more objective base for management decisions, highly in favour of the anticoagulation of cases of non-isolated atrial fibrillation in the absence of contraindications.