Short-term variability of ventricular arrhythmia and rapid assessment of drug efficacy.
 Statistical criteria for suppression and aggravation of ventricular arrhythmia were defined by means of 50 short-term drug tests performed in 24 patients.
 Each patient's spontaneous variability (SV) was evaluated by linear regression analysis of hour-to-hour changes in ectopy during 24- to 48-hour Holter monitoring.
 The response to a single oral dose of disopyramide, 300 mg, flecainide, 200 mg, and propafenone, 450 mg, was measured during a trial lasting 4 hours.
 Lidocaine was administered intravenously in incremental doses of up to 4 mg/min and was evaluated over 3 hours.
 Threshold values of ventricular arrhythmia corresponding to 95% confidence limits were calculated from baseline recordings and were used to ascertain the likelihood of a true drug effect.
 The minimum decrease in hourly ectopy indicating arrhythmia suppression averaged 90.9%, while an increase of at least 947% was required for a proarrhythmic effect.
 When these efficacy criteria were applied, 16 of 50 short-term tests revealed no drug effect.
 In contrast, when a 70% threshold derived from studies of daily variability was employed, only 7 of 50 trials were negative.
 Thus individual determination of hourly arrhythmia variability yields more stringent criteria than extrapolation from day-to-day spontaneous variation.
