The association of anticoagulation, ischemic stroke, and hemorrhage in elderly adults with chronic kidney disease and atrial fibrillation

The utility of anticoagulants for ischemic stroke prophylaxis in elderly patients with chronic kidney disease (CKD) and atrial fibrillation remains uncertain. In this population-based retrospective cohort study, we determined the association of anticoagulant use with ischemic stroke or hemorrhage in elderly patients (66 years and older) with advanced chronic kidney disease (eGFR under 45 ml/min/1.73m 2 ) and atrial fibrillation. We followed 6,544 patients with CKD and new onset atrial fibrillation, of whom 1,475 filled a prescription for an anticoagulant. We used propensity-score matched Cox proportional hazards and competing risk models to determine the time to first event of ischemic stroke, hemorrhage or mortality. After matching to examine exposure to anticoagulants, 1,417 matched pairs were identified. The crude rate of ischemic stroke and hemorrhage were 41.3 and 61.3 with anticoagulants and 34.4 and 34.3 without anticoagulants per 100 person-years, respectively. The hazard ratios of ischemic stroke, hemorrhage, and mortality for receipt of an anticoagulation prescription were 1.10 (95% confidence interval, 0.78-1.56), 1.42 (1.04-1.93), and 0.74 (0.62-0.88) as compared to non-receipt of anticoagulation. After accounting for the competing risk of death, the hazard ratios for ischemic stroke and hemorrhage were 1.12 (0.90-1.39) and 1.60 (1.31-1.97), respectively. The findings were consistent in a sensitivity analysis accounting for time varying anticoagulant exposure. Thus, in older patients with CKD and atrial fibrillation, receipt of an anticoagulant was not associated with a lower risk of ischemic stroke, but a higher risk of hemorrhage and a lower risk of mortality.