Abstract
OBJECTIVES: To identify predictors of intraoperative electrical cardioversion and develop a predictive model for patients undergoing radiofrequency ablation for atrial fibrillation (AF). METHODS: We retrospectively analyzed data from 1,348 patients with AF who underwent radiofrequency catheter ablation at Tongji Hospital between January 2018 and December 2023. Clinical, echocardiographic, and CT imaging data were collected. The Boruta algorithm and multivariable logistic regression were used to identify predictors and construct a nomogram. Model performance was assessed using the area under the ROC curve (AUC), calibration plots, and decision curve analysis (DCA). External validation was performed on 121 patients treated at Hubei No. 3 People's Hospital of Jianghan University from June 2023 to February 2025. RESULTS: Patients were divided into training and validation sets (7:3 ratio). Five independent predictors were identified: AF type (OR = 13.63), valvular regurgitation (OR = 3.25), BMI (OR = 1.06), left atrial diameter (OR = 1.74), and systolic blood pressure (OR = 0.96). The nomogram showed excellent discriminative ability with AUCs of 0.881 (training), 0.879 (internal validation), and 0.866 (external validation). Calibration curves demonstrated good agreement between predicted and actual outcomes. DCA confirmed the model's clinical utility. CONCLUSIONS: The proposed nomogram accurately predicts the need for intraoperative electrical cardioversion during AF ablation and may aid in individualized procedural planning.