A Prospective Study of Atrial Fibrillation after Cardiac Surgery: Multivariate Risk Analysis Using P Wave Signal-Averaged ECG and Clinical Variables

Background: Atrial fibrillation (AF) is a commonly encountered arrhythmia following cardiac surgery and when sustained, may be associated with significant morbidity. Methods: This large prospective investigation examined a variety of clinical and P wave signal-averaged electrocardiogram (SAECG) parameters to identify independent predictors of AF following cardiac surgery. A total of 272 patients underwent P wave SAECG recording and analysis prior to surgery. Information on their clinical, surgical, and hemodynamic characteristics as well as hospital course was collected. Patients were followed during their postoperative course with telemetry and ECGs. Results: During an observation period of up to 14 days, 79 patients (29%) developed AF 2.5 ± 1.9 days after surgery. Patients who developed AF following cardiac surgery were more likely to be older, undergo valve surgery, to have ejection fraction (EF) <40%, to have P wave duration on SAECG >140 ms (all P < 0.01), and to take digoxin preoperatively (P < 0.05). A multivariate analysis found that only P wave duration on SAECG >140 ms and EF <40% were independent predictors of AF following cardiac surgery. The odds ratio of P wave duration on SAECG >140 ms and EF <40% for the development of AF following cardiac surgery was 3.1 and 2.8, respectively, and 8.7 when combined. Conclusions: Thus, the presence of preexisting abnormal atrial substrate as detected by P wave prolongation on SAECG, and implicated by EF <40%, clearly predicted a higher risk of AF following cardiac surgery and may provide clinicians with an effective means of identifying those at greatest risk.