Development and validation of a diagnostic model based on left atrial diameter to predict postoperative atrial fibrillation after off-pump coronary artery bypass grafting

• Published in: Journal of thoracic disease Vol. 15; no. 7; pp. 3708 - 3725
• Main Authors: Zhang, Hao, Qiao, Huanyu, Yang, Bo, Lu, Yifan, Bai, Tao, Xue, Jinrong, Liu, Yongmin
• Format: Journal Article
• Language: English
• Published: China AME Publishing Company 31.07.2023
• Subjects: Original; postoperative atrial fibrillation left atrial diameter (POAF LAD); Coronary artery bypass grafting (CABG); predictive model
• ISSN: 2072-1439; 2077-6624
• DOI: 10.21037/jtd-22-1706

This cohort study collected the clinical data of patients who underwent off-pump coronary artery bypass grafting (OPCABG) during hospitalization to observe the occurrence of postoperative atrial fibrillation (POAF), construct a POAF prediction model for CABG patients based on the left atrial diameter (LAD), and assist clinicians in making better medical decisions. In this study, all patients who had no prior history of arrhythmia and who had received isolated OPCABG between May 1, 2021, and February 1, 2022, at Beijing Anzhen Hospital Affiliated to Capital Medical University (n=749) were reviewed. Depending on an optimal cutoff obtained from receiver operating characteristic (ROC) curve analysis, patients were separated into two groups: a group with POAF (n=188) and a group without POAF (n=561). The incidence of POAF was then compared. Prediction models were built, and nomograms were plotted was plotted. Model evaluation, including calibration curve and decision curve analysis, was performed. In all, 188 out of 749 (25.1%) patients who underwent cardiac surgery experienced POAF. Multifactorial logistic regression analysis showed that age ≥66 years, LAD ≥39 mm, and post-OPCABG atrial fibrillation (AF) were independently associated. The prognostic nomogram model showed good concordance index (C-index) scores. Decision curve analysis suggested the clinical benefit of the prediction models. In this study, a prediction model for patients with POAF after OPCABG was assessed, which was shown to make more accurate predictions compared with the original risk prediction system. It may assist doctors to optimize management of patients with POAF.