Computer recognition of cardiac arrhythmias and statistical approaches to arrhythmia analysis.

The recognition of marked spontaneous variability in the frequency of ventricular ectopy has required the analysis of various statistical approaches to define appropriate guidelines for determining that a therapeutic agent, rather than spontaneous variability alone, is the cause of an observed reduction in VPC frequency. Analysis of variance or linear regression techniques have been applied to ambulatory ECG data obtained on different patient population groups. When similar patient groups are studied with comparable assumptions, then either of these statistical techniques provides similar guidelines to detect spontaneous variation as the cause of ventricular ectopic reduction. A 75% minimal reduction in VPC frequency should be required when comparing one 24-hour ambulatory monitoring period with another. The impact of spontaneous variability of ventricular ectopy on clinical trial designs is discussed, as is the categorization of various ambulatory ECG monitoring recorders and analysis systems.