Background Nonvitamin K oral anticoagulants require dose adjustment based on kidney function.The most common estimate of kidney function employed in clinical practice is estimated glomerular filtration rate (eGFR); however, product monographs recommend the use of the Cockcroft-Gault estimated creatinine clearance (eCrCl) for dose adjustment. Methods and Results The authors included patients enrolled in the ORBIT-AF II (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation AF II) trial. Dosing was considered inappropriate when use of eGFR resulted in a lower (undertreatment) or higher (overtreatment) dose than that recommended by the eCrCl. The primary outcome of major adverse cardiovascular and neurological events was a composite of cardiovascular death, stroke or systemic embolism, new-onset heart failure, and myocardial infarction. Among 8727 in the overall cohort, agreement between eCrCl and eGFR was observed in 93.5% to 93.8% of patients. Among 2184 patients with chronic kidney disease (CKD), the agreement between eCrCl and eGFR was 79.9% to 80.7%. Dosing misclassification was more frequent in the CKD population (41.9% of rivaroxaban users, 5.7% of dabigatran users, and 4.6% apixaban users). At 1 year, undertreated patients in the CKD group had significantly greater major adverse cardiovascular and neurological events (adjusted hazard ratio, 2.93 [95% CI, 1.08-7.92]) compared with the group with appropriate nonvitamin K oral anticoagulants dosing (P=0.03). Conclusions The prevalence of misclassification of nonvitamin K oral anticoagulants dosing was high when using eGFR, particularly among patients with CKD. Among patients with CKD, potential undertreatment due to inappropriate and off-label renal formulae may result in worse clinical outcomes. These findings highlight the importance of using eCrCl, and not eGFR, for dose adjustment in all patients with AF receiving nonvitamin K oral anticoagulants.
Keywords: ORBIT‐AF II; atrial fibrillation; nonvitamin K oral anticoagulant; renal dose adjustment.