Nomogram predicts atrial fibrillation after coronary artery bypass grafting

OBJECTIVE: Using the nomogram to intuitively predict atrial fibrillation after coronary artery bypass grafting. Identify high-risk patients with atrial fibrillation and provide preoperative protective therapy. METHODS: A total of 397 patients that underwent coronary artery bypass grafting were consecutively enrolled. Independent predictors of patients were analyzed by multivariate logistic regression. Two nomograms were constructed to predict postoperative atrial fibrillation. RESULTS: , Insulins, and Statins were independently associated with atrial fibrillation after isolated coronary artery bypass grafting. The nomogram of postoperative atrial fibrillation in patients was constructed using total predictor variables (AUC = 0.727, 95% CI 0.673-0.781). The model was internally validated (AUC = 0.701) by K-fold Cross-validation resampling (K = 5, Times = 400). To make an early intervention, the intraoperative information of the patients was excluded. Only 6 variables before surgery were used to establish the brief nomogram to predict postoperative atrial fibrillation (AUC = 0.707, 95% CI 0.651-0.764). The brief model was internally validated (AUC = 0.683) by resampling with K-fold Cross-validation resampling. CONCLUSIONS: These two nomograms could be used to predict patients at high risk for atrial fibrillation after isolated coronary artery bypass grafting.