RÉSUMÉ
AIMS: Non-valvular atrial fibrillation (NVAF) accounts for 25% of completed strokes (CS) of a cardioembolic origin in patients over 60 years old. Our aim was to define the predictors of a good and poor prognosis after a CS secondary to an NVAF in our milieu. PATIENTS AND METHODS: We evaluated the risk factors (RF) and severity of CS in relation to death, functionality and recurrence at 5 years. 81 patients between the ages of 49 and 88 were followed up consecutively for 1 to 90 months; 38 (46.9%) of them were males. Multivariate analysis was performed with the following independent variables: age, gender, smoking, hypertension, heart disease, diabetes mellitus and characteristics of the stroke. The severity of the CS was assessed by means of the modified Rankin scale, which was dichotomised into a good prognosis (0-2) and a poor prognosis (> or = 3), both basal and at the end of the clinical control. We also evaluated the secondary preventive treatment used and its relation with recurrence, prognosis, death and complications. RESULTS: No RF was linked to a poor prognosis or recurrence; 88% had a poor prognosis. Antiplatelet drugs were used in 42% of cases and 39% received anticoagulants. A good final progression was observed in 9.5% of the patients treated with antiplatelet drugs versus 35% of those receiving anticoagulation therapy (p = 0.004). Severity of the CS on admission was worse in the aspirin group, with no differences in recurrence and mortality. A better prognosis was observed in patients from urban areas. CONCLUSIONS: Use of antiplatelet drugs, living in a rural area and a Rankin score of > or = 3 on admission are factors suggesting a poor prognosis in the clinical control at 5 years.