New score for predicting thromboembolic events in patients with atrial fibrillation using direct oral anticoagulants

Abstract

Determinants of thrombotic events remain uncertain in patients with atrial fibrillation treated with direct oral anticoagulants (DOACs). Our aim was to identify risk factors associated with thromboembolism in patients with at atrial fibrillation on DOACs and to construct and externally validate a predictive model that would provide a validated tool for clinical assessment of thromboembolism. In the development cohort, prediction model was built by logistic regression, the area under the curve (AUC), and Nomogram. External validation and calibration of the model using AUC and Hosmer–Lemeshow test. This national multicenter retrospective study included 3263 patients with atrial fibrillation treated with DOACs. The development cohort consisted of 2390 patients from three centers and the external validation cohort consisted of 873 patients from 13 centers. Multifactorial analysis showed that heavy drinking, hypertension, prior stroke/transient ischemic attack (TIA), cerebral infarction during hospitalization were independent risk factors for thromboembolism. The Alfalfa-TE risk score was constructed using these four factors (AUC = 0.84), and in the external validation cohort, the model showed good discriminatory power (AUC = 0.74) and good calibration (Hosmer–Lemeshow test P value of 0.649). Based on four factors, we derived and externally validated a predictive model for thromboembolism with DOACs in patients with atrial fibrillation (Alfalfa-TE risk score). The model has good predictive value and may be an effective tool to help reduce the occurrence of thromboembolism in patients with DOACs.