We investigated the relationship between electrical alternans and cardiac electrical stability in a series of 20 dog experiments and in a pilot clinical study. Electrical alternans was detected in both the QRS complex and the ST-T wave by use of a novel multidimensional spectral technique. The magnitude of the alteration was expressed as the alternating electrocardiographic morphology index (AEMI), expressed as parts per million of waveform energy. Electrical stability in the dog preparations was assessed via the ventricular fibrillation threshold measurement, and in the clinical studies via programmed stimulation. In 10 dog experiments, systemic hypothermia resulted in a 60% decrease in ventricular fibrillation threshold (VFT) (p less than .0001) and a significant increase in both AEMI(QRS) form 3.7 +/- 3.0 to 1448 +/- 548 (p less than .0001) and AEMI(ST-T) from 43.9 +/- 18.4 to 19,178 +/- 5579 (p less than .0001). In 10 dog experiments, transient coronary artery ligation also resulted in a 60% decrease in VFT (p less than .0001), an increase from 76.3 +/- 46.5 to 245 +/- 11 in AEMI(QRS) (p less than .05), and an increase from 842 +/- 505 to 1365 +/- 392 in AEMI(ST-T) (p less than .002). In 119 observations in 20 animal experiments, the rank correlation between VFT and AEMI(QRS) was -.30 (p less than .001), with that between VFT and AEMI(ST-T) being -.55 (p less than .0001). In a double-blind pilot clinical trial consisting of 23 studies in 19 patients, the result of electrophysiologic testing was used as an independent measure of cardiac electrical stability. Alternation in waveform morphology identified the inducible patient population with a sensitivity of 92%, a positive predictivity of 70%, and a specificity of 50% (p less than .05). We conclude that analysis of subtle beat-to-beat variability in electrocardiographic morphology may provide a noninvasive measure of cardiac electrical stability.

Circulation, Vol. 77, No. 1, pp. 110-121, 1988