Fragmented QRS may predict postoperative atrial fibrillation in patients undergoing isolated coronary artery bypass graft surgery

Objective: Fragmented QRS complexes (fQRS) are defined as various RSR′ patterns in 2 contiguous leads corresponding to a major coronary artery territory. Although the reason of association between fQRS and cardiac events was documented as cardiac fibrosis, the predictive role of fQRS was not studied for postoperative atrial fibrillation (POAF) which is a frequent and serious complication in patients undergoing isolated coronary artery bypass graft (CABG) surgery. Therefore, this issue was investigated in the present study. Methods: The current study has a prospective observational design. Two hundred and seventy two eligible patients who underwent isolated CABG surgery were enrolled consecutively. The patients were divided in two groups with post-op atrial fibrillation (AF) and non-AF. The occurrence of new-onset AF following CABG and the relationship with fQRS were searched. The logistic regression analysis was used to determine independent predictors for POAF. The sensitivity and specificity of study variables in predicting POAF were calculated using a receiver-operating characteristic curve (ROC). Results: POAF occurred in 62 of 272 patients (22.8%). Patients with POAF were generally older (p 0.001) and female (p=0.006), with preexisting hypertension (p=0.008), lower hemoglobin levels (p=0.011), chronic obstructive lung disease (p=0.003), prolonged QRS time (p=0.004), and higher EUROSCORE (p 0.001) compared to non-AF patients. Patients with POAF had lower left ventricular ejection fraction (p 0.001) and high fQRS rate (p 0.001), but similar left atrial size (LA, p=NS). Interestingly, LA size was significantly enlarged in patients with fQRS (3.8±0.6 vs. 4.1±0.5 cm, p=0.002). In addition, the patients with POAF had prolonged stay in the cardiac surgery intensive-care unit (p 0.001) and extended in-hospital care (p=0.001). New-onset POAF was significantly related to the presence and number of fQRS. In the logistic regression analysis, only age (OR: 1.044, 95% CI: 1.008-1.082, p=0.016), female gender (OR: 2.347, 95% CI: 1.079-5.106, p=0.031), the presence and number of fQRS (OR: 6.020, 95% CI: 3.152-11.5 and OR: 1.522, 95% CI: 1.282-1.807, both of, p 0.001) were independent predictors of POAF. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and the diagnostic accuracy (DA) of presence of fQRS on pre-op electrocardiogram to predict POAF were 66%, 76%, 45%, 88% and 74%, respectively. The area under ROC was found as 0.733 (p 0.001, 95% CI: 0.657-0.810). Conclusion: In our study, we found that new-onset POAF was independently related to the presence and number of fQRS in patients undergoing CABG surgery. In addition, fQRS on pre-op surface ECG had high predictive values for new-onset POAF.