Improved selection of patients for programmed ventricular stimulation by frequency analysis of signal-averaged electrocardiograms.

Improved selection of patients with suspected sustained ventricular tachycardia (VT) for programmed ventricular stimulation is needed. To determine if frequency analysis detects patients in whom sustained VT might be induced, we first obtained fast-Fourier transforms (FFT) of signal-averaged electrocardiograms (ECGs) from 20 patients with spontaneous sustained VT (group 1) and compared them with the results of programmed ventricular stimulation with single and double extrastimuli during two cycle lengths and burst pacing from two right ventricular sites. The FFT data were expressed as an area ratio that quantified the relative contributions of 20 to 50 Hz frequencies in the terminal QRS and ST segment. A logistic regression with inducibility as the dependent variable was used to help define area ratio values greater than 20 as abnormal. Sustained monomorphic VT was induced in 18 patients, each with an area ratio value greater than 20. Sustained VT was not induced in two patients, each with an area ratio value less than 20. FFT data were then compared prospectively with the results of programmed stimulation in 38 patients (group II) with nonsustained VT (12 patients) or syncope (26 patients) referred for electrophysiologic study. In none of the 26 patients in group II with normal FFT values was VT inducible. Sustained monomorphic VT was induced in five of 12 patients with abnormal FFT values. Thus, the results of FFT analysis correctly predicted the results of programmed ventricular stimulation in 88% of patients studied and in 82% of patients in group II with syncope or nonsustained VT. Moreover, all five patients in group II in whom sustained VT was induced were identified correctly.(ABSTRACT TRUNCATED AT 250 WORDS)