Prognostic Significance and Clinical Utility of Intraventricular Conduction Delays on the Preoperative Electrocardiogram

Abstract

The prognostic significance of the preoperative electrocardiogram (ECG), particularly intraventricular conduction delays (IVCDs), on postoperative outcomes among patients undergoing noncardiac surgery is uncertain. In a retrospective cohort, we evaluated the risk associated with preoperative IVCDs on in-hospital death and postoperative myocardial infarction (POMI). The 152,479 patients who underwent noncardiac surgery were categorized by preoperative electrocardiographic findings: normal (36.1%), left bundle branch block (LBBB, 1.2%), right bundle branch block (2.9%), nonspecific IVCD (3.3%), and any other ECG abnormality (56.5%). The primary and secondary outcomes were postoperative in-hospital mortality and POMI, respectively. In multivariable-adjusted models, compared with normal ECGs, each electrocardiographic abnormality category was associated with increased risk of postoperative death: LBBB odds ratio (OR) 1.89 (95% confidence interval [CI] 1.35 to 2.65), right bundle branch block OR 1.73 (95% CI 1.33 to 2.24), nonspecific IVCD OR 1.95 (95% CI 1.53 to 2.48), and other abnormal ECG OR 1.94 (95% CI 1.68 to 2.25). ECGs with conduction delays did not confer increased risk of postoperative death compared with other ECG abnormalities. Moreover, receiver operating characteristic analysis of models incorporating demographic and co-morbidity data demonstrated marginal additive benefit of any electrocardiographic data. Risk of POMI was not significantly increased among ECGs with conduction delays compared with both normal and other abnormal ECGs. In conclusion, patients with intraventricular conduction disease, including LBBB, on preoperative ECG are not at greater risk of postoperative in-hospital death or POMI compared with patients with other ECG abnormalities. Furthermore, any preoperative electrocardiographic abnormalities, including intraventricular delays, provide marginal clinical utility beyond demographic and clinical history for predicting postoperative in-hospital death or POMI.