Background: Atrial fibrillation (AF) is a common medical problem with increasing prevalence among the elderly. Warfarin is effective in the prevention of AF-related-stroke but is often underutilized, especially in high-risk populations, like the elderly.
Objectives: To determine, in a group of elderly patients with AF, if those treated in-line with the clinical recommendations differ from patients who were not, regarding morbidity and mortality and also to determine independent predictors of mortality. A second objective was to verify if the CHADS2 score is a good predictor of thromboembolic risk in the elderly.
Population and methods: A total of 161 consecutive elderly patients with AF admitted in a single centre were evaluated. Clinical follow-up was available for 88.4%, with a mean duration of 9 months.
Results: Mean age was 80.9 ± 6.6 years; 96.3% of the patients had permanent AF, with controlled ventricular rate in 56.4%. Previous stroke was verified in 30.4%. Only 37.3% had oral anticoagulation at hospital discharge, despite 87.6% had guideline recommendation. Cumulative mortality rate in follow-up was 48.4% and the thromboembolism rate was 8.1%. We verified that CHADS2 score was a good predictor of thromboembolic risk in this population (c-statistic=0.742). Clinical follow-up showed that patients treated according with the clinical recommendations were more likely to survive (33.33% vs 53.93%; p=0.048). Multivariate analysis showed that age >80 years, renal disease, neoplasm and neuropsychic disease as independent predictors of mortality (c-statistic=0.83).
Conclusion: A gap of 50% existed between the guideline recommendations and actual practice. The use of risk stratification scores can help guide the decision to use anticoagulation in older patients with AF. Elderly patients treated according with the clinical recommendations had a better prognosis.