Prognostic significance of the signal averaged electrocardiogram in patients with chronic stable coronary artery disease. Analysis in the time domain and by spectral temporal mapping

BACKGROUND: Ventricular late potentials detected by the signal averaged electrocardiogram (SAECG) have been used to predict cardiac death in patients after recent myocardial infarction. The goal of this study was to investigate the prognostic significance of the SAECG in a population of chronic coronary artery disease, with and without previous myocardial infarction. METHODS: SAECG was recorded in 698 patients with angiographically proven coronary artery disease and analyzed by time domain analysis (TDA) and by spectral temporal mapping (STM). Cardiac death or ventricular fibrillation (= cardiac event) were used as the primary endpoint for follow-up (25 to 33 months). RESULTS: A cardiac event occurred in 46 out of 698 patients (6.6%). An abnormal SAECG using TDA was found in 43% of patients with a cardiac event, as compared to 21.7% in those without (p<0.0005). The probability of a cardiac event during follow-up was 4.4% when TDA and STM were both normal, 9.5% and 10.2% when either STM or TDA were abnormal and 28.5% when both were abnormal. A duration of the averaged QRS complex of more than 120 ms and a left ventricular ejection fraction of less than 45% were the only independent predictors of a cardiac event. Logistic regression analysis could predict a cardiac event with a sensitivity of 54% and a specificity of 88%. CONCLUSIONS: In patients with chronic coronary artery disease the duration of the signal averaged QRS complex and left ventricular ejection fraction are independent predictors of cardiac death or ventricular fibrillation.