Abstract
BACKGROUND: Atrial fibrillation (AF) is a common arrhythmia, with risk of systemic embolism and death. It presents rheumatic etiology in up to 32% of developing countries, whose anticoagulation and evolution data are scarce. OBJECTIVES: to determine the predictors of cardiac death considering the clinical profile, thromboembolism and bleeding scores of patients with AF of a single center, with high prevalence of rheumatic heart disease. METHODS: 302 patients with AF were studied, mean age 58.1 years; 161 women; 96 pts with rheumatic etiology. Patients underwent clinical and laboratory evaluation, measurement of risk scores and the mean follow-up of 12.8 months. RESULTS: 174 were using warfarin. The averages of the HAS-BLED and ATRIA scores were 1.4 and 1.2, respectively. Percent time in therapeutic range of international normalized ratio was 45.8%. Thirty patients (9.9%) had cardiac death and 41 had some type of bleeding due to warfarin. By univariate analysis, there was statistical significance between cardiac death and permanent AF, blood pressure, systolic dysfunction, R2CHADS2, CCS, EHRA and HAS-BLED. There was no association with valvular AF. By multivariate analysis, systemic arterial and pulmonary artery pressures, classification CCS and systolic dysfunction showed statistical significance. CONCLUSIONS: There was no association between cardiac death and valvular AF. Independent predictors of cardiac death were low measures of blood pressure, higher score CCS classification and the presence of systolic ventricular dysfunction.