Anticoagulant activity of apixaban can be estimated by multiple regression analysis

BACKGROUND: Information on apixaban anticoagulant activity is required to prevent major bleeding or thrombosis during its use. METHODS: We enrolled 194 patients with nonvalvular atrial fibrillation (NVAF) in whom warfarin was replaced with apixaban: 105 (54.1%) received the standard dose of apixaban (5 mg twice daily [BID]; 5 mg group) and 89 (45.9%) received a reduced dose (2.5 mg BID; 2.5 mg group). Multiple regression analysis was performed to predict the prothrombin time of apixaban (PTa) based on factors including age, body weight (BW), serum creatinine, and CHA(2)DS(2)‐VASc score. RESULTS: PTa and PT of warfarin (PTw) were significantly correlated in both groups (correlation coefficient R = 0.239 [P = .014] in the 5 mg group; R = 0.248 [P = .019] in the 2.5 mg group). PTa in the 5 mg group was predicted as follows: 16.952‐0.036 × BW +0.299 × CHA(2)DS(2)‐VASc score (P < .0004; R = 0.378). However, in the 2.5 mg group, PTa could not be predicted. The mean of the predicted and measured PTa values in the 5 mg group was 15.6 s, which was similar to the mean measured PTa of 15.5 s in the 2.5 mg group. CONCLUSIONS: PT can be predicted by a formula including simple clinical parameters in patients receiving the standard dose of apixaban. This simple predictive formula may help to stratify bleeding and thrombosis risks in patients treated with apixaban.