Analysis of the spontaneous variability of ventricular arrhythmias: consecutive ambulatory electrocardiographic recordings of ventricular tachycardia.

Results are reported of analysis of the variability of complex ventricular arrhythmias in a cohort of 110 patients selected for the presence of ventricular tachycardia (VT). All patients were enrolled in investigational antiarrhythmic drug trials and had an average of 4 consecutive days of placebo ambulatory electrocardiographic recording to serve as the database for this study. Using a statistical approach incorporating analysis of variance, the minimum percent reductions of ventricular premature complexes, couplets and VT were calculated to establish "drug effect" rather than variability at a significance level of 0.05. The relative variability of ventricular arrhythmias in prognostically important groups was also analyzed: (1) coronary artery disease (CAD) (n = 57) vs no CAD (n = 53); (2) patients with a left ventricular ejection fraction of 40% or less (n = 52) vs those with an ejection fraction greater than 40% (n = 58); and (3) patients with frequent runs of VT (10 or more runs/day, n = 63) vs infrequent VT (n = 47). Multiple regression analysis revealed that patients with CAD have significantly greater premature ventricular complex variability than patients without CAD (p less than 0.01). Also, patients with frequent VT runs have greater VT variability than that previously reported in smaller studies, thus requiring greater VT reductions to establish drug effect. Whether the variability of ventricular arrhythmia is itself an independent risk factor for sudden cardiac death is unknown.