부산대학교 도서관

Abstract Objectives This study aimed to identify the risk factors for postoperative atrial fibrillation in patients with valvular atrial fibrillation, and establish predictive models of atrial fibrillation recurrence. Methods Overall, 224 patients who underwent radiofrequency ablation of atrial fibrillation from November 2014 to November 2020 were included. The statistical package for social sciences, X-tile, and R-studio were used for statistical analysis. Results Patients were divided into training and validation sets according to a ratio of 3:1. The training set was analysed using univariate and multivariate Cox regression analysis and showed that preoperative uric acid > 401 μmol/L (P = 0.006), B-type natriuretic peptide > 202 ng/L (P = 0.042), hypersensitivity C-reactive protein > 6.1 mg/L (P = 0.026), erythrocyte sedimentation rate > 7.0 mm/h (P = 0.016), preoperative left atrial diameter > 48 mm (P = 0.031) were significantly correlated with the recurrence of atrial fibrillation after radiofrequency ablation in patients with valvular atrial fibrillation. In the training set, a Cox regression model of the five related factors was established using the R language. The C-index of the model was 0.82, and the area under the receiver operating characteristic curve was 0.831 (P