Electrocardiography is Useful to Predict Postoperative Ventricular Arrhythmia in Patients Undergoing Cardiac Surgery: A Retrospective Study.

Preoperative detection of high-/low-risk postoperative ventricular arrhythmia (POVA) patients using a noninvasive method is an important issue in the clinical setting. This study mainly aimed to determine the usefulness of several preoperative electrocardiographic (ECG) markers in the risk assessment of POVA with cardiac surgery. We enrolled 1024 consecutive patients undergoing cardiac surgery, and a total of 823 patients were included in the study. Logistic regression analysis determined preoperative ECG markers. A new risk predicting model were developed to predict occurrence of POVA, and the receiver operating characteristic curve (ROC) was used to validate this model. Of these, 337 patients experienced POVA, and 485 patients did not experience POVA in this retrospective study. Among 15 ECG markers, a univariate analysis found a strong association between POVA and preoperative VA, the R-wave in lead aVR, the QRS wave, index of cardiac electrophysiological balance (iCEB), QT interval corrected (QTc), Tpeak-Tend interval (Tpe) in lead V, the J wave in the inferolateral leads, pathological Q wave, and S+R>35 mm. Multivariate analysis showed that a preoperative J wave [adjusted odds ratio (AOR): 3.80; 95% CI: 1.88-7.66; < 0.001], Tpe >112.5-ms (AOR: 2.80; 95% CI: 1.57-4.99; < 0.001), and S+R >35 mm (AOR: 2.92; 95% CI: 1.29-6.60; = 0.01) were independently associated with POVA. A new risk predicting model were developed in predicting POVA. The ECG biomarkers including J wave, Tpe >112.5 ms, and S+R >35 mm were significantly predicted POVAs. A risk predicting model developed with electrocardiographic risk markers preoperatively predicted POVAs.