Independent value of signal-averaged electrocardiography and left ventricular function in identifying patients with sustained ventricular tachycardia with coronary artery disease.

To determine if the signal-averaged electrocardiographic detection of late potentials is an independent marker of sustained ventricular tachycardia (VT) in patients with documented chronic coronary artery disease (CAD), 57 patients underwent signal-averaged electrocardiography. Mean ejection fraction was 47 +/- 13% in the 14 patients with sustained VT and 56 +/- 19% in the 43 patients without VT (difference not significant). The sensitivity, specificity and accuracy of late potentials for detecting patients with VT were 64% (9 of 14), 79% (34 of 43), and 75% (43 of 57), respectively. Univariate analysis and stepwise logistic regression of angiographic and electrocardiographic variables identified late potentials as an independent marker of the patient with sustained VT. The odds ratio for late potentials to detect patients with prior sustained VT was 2.6. Six-month follow-up revealed a cardiac mortality rate of 11% and an arrhythmia event rate of 22% in patients with late potentials vs a cardiac mortality rate of 3% and an arrhythmia event rate of 13% in patients without late potentials. Thus, signal-averaged electrocardiographic detection of late potentials is useful in identifying patients with prior sustained VT independent of left ventricular function.