Predicting atrial fibrillation after cardiac surgery using a simplified risk index

Abstract

BackgroundPostoperative atrial fibrillation (POAF) is a common complication after cardiac surgery and can lead to increased risk of postoperative adverse events. However, atrial fibrillation and postoperative adverse events are preventable. In this study, a risk index was developed to predict atrial fibrillation after cardiac surgery. MethodsA prospective cohort study of 405 patients who had undergone adult cardiac surgery from 2015 September to 2016 August at Heart Institute of HCMC and Cho Ray Hospital were obtained. In order to predict POAF, a logistic regression model was developed, and a risk score was derived and validated by bootstrap. ResultsIn our study, 98 patients developed POAF (24.2%). The risk score included three significant risk factors (age ≥ 60, left atrial diameter > 41 mm, Coronary Artery Bypass Graft with concomitant mitral valve replacement or repair) that were consistent with other reports. Each of these risk factors was assigned one point. The total risk score ranges from 0 to 3 (AUC = 0.69, 95% CI: 0.63–0.75) with the best cutoff point at 1. According to this scoring system, the incidences of POAF in patients associated with each score of 0, 1, 2, and 3 were 8.6%, 30.1%, 40.8%, and 58.3% respectively. Bootstrapping with 5000 samples confirmed the final model provided was consistent with predictions. ConclusionsWe developed and validated a simple risk score based on clinical variables that can be obtained before surgery in order to accurately predict the risk of POAF in patients undergoing cardiac surgery.