Background: Atrial fibrillation (AF) is the most common arrhythmia, especially in elderly subjects (more than 10% of those older than 80). The main complication is the occurrence of ischemic stroke. Oral anticoagulants reduce by 2/3 the thromboembolic risk but increase the risk of bleeding. The question for the clinician is to identify patients at highest thromboembolic risk who would the most benefit from anticoagulant therapy: in this way, the dosage of D-dimer, a marker of activation of coagulation may be of interest.

Primary aim: To assess the predictive value of D-dimer levels on the occurrence of thromboembolic event or death.

Methods: Patients with AF were included whatever the treatment they received, when they were in a stable clinical condition and when the antithrombotic treatment remained unchanged during the last month. It is a prospective multicentric study performed in 25 departments of AP-HP. Patients had visits and biologic dosages including D-dimer every 4 months. At every visit, patients were questioned about ischemic events (stroke, transient attack, systemic embolism, myocardial infarction), venous thromboembolic events, the occurrence of disease and bleeding events since the last visit. D-dimer dosage was centralised in the Laboratory of Pr Drouet (Lariboisière hospital). The usual management of the patient was in charge of the practitioner and not modified by the study. The planned duration of follow-up was 1 year to 2 years.

Results: 434 patients were included (mean age 80 years) from 02/01/01 to 04/06/30. Antithrombotic treatments at inclusion were: oral anticoagulant (n=240), antiplatelet agents (n=57), no antithrombotic therapy (n=45). Follow-up has ended on 05/06/30.

Discussion: Clinical events have been recorded. D-dimer measurements are ongoing. Results will be provided at the meeting (including D-dimer levels and clinical events).

Author notes

Corresponding author