Choice of Renal Function Estimator Influences Adverse Outcomes with Dabigatran Etexilate in Patients with Atrial Fibrillation

Background Clinical significance of dosing dabigatran with different estimates of renal function for treatment of atrial fibrillation (AF) is unknown. Renal function is routinely estimated by the chronic kidney disease epidemiology initiative equation (CKD-EPI) and used to guide dosing. The aim of this study was to investigate the risk of adverse outcomes for patients with AF when different estimators of renal function are used. Material and Methods AF patient data were extracted from national administrative databases. Renal function was estimated using Cockcroft–Gault, CKD-EPI, and CKD-EPI adjusted for body surface area (CKD-EPI-BSA). Outcomes of cerebrovascular accident (CVA), systemic embolism (SE), and hemorrhage were extracted. Results In total, 2,425 patients were identified, of which there were hospitalizations for 138 (5.7%) hemorrhagic events, 45 (1.9%) CVA/SE, and 33 (1.4%) unspecified CVA. The level of agreement between Cockcroft–Gault with CKD-EPI and CKD-EPI-BSA yielded a weighted kappa statistic of 0.47 and 0.71, respectively. CKD-EPI and CKD-EPI-BSA significantly overestimated renal function in elderly patients resulting in higher recommended doses compared with Cockcroft–Gault. The hazard ratio for a hemorrhagic event was 2.32 (95% confidence interval, 1.22–4.42; p = 0.01) when a high dose was given compared with normal dose, based on Cockcroft–Gault. Conclusion Both CKD-EPI and CKD-EPI-BSA equations significantly overestimated renal function in the elderly population compared with the Cockcroft–Gault equation. This may lead to dose selection errors for dabigatran, particularly for those with severe impairment, increasing the risk of adverse outcome. Hence, CKD-EPI and CKD-EPI-BSA equations should not be substituted for the Cockcroft–Gault equation in the elderly for the purpose of renal dosage adjustments.