ABSTRACT
In patients with malignant ventricular arrhythmia, the utility of exercise testing for exposing arrhythmias and guiding antiarrhythmic drug selection and dosage is not yet generally accepted. A major reservation relates to the issue of reproducibility of arrhythmia provocation. This prospective study comprised 28 patients referred for evaluation of ventricular arrhythmia. In half of these subjects the clinical arrhythmia was sustained ventricular tachycardia or ventricular fibrillation. Maximal exercise testing adhering to a Bruce protocol was performed on 2 separate days. No antiarrhythmic drugs were administered at the time of testing. Of the 28 patients, 27 had an increase in arrhythmia with exercise. The prevalence rates of arrhythmia were greater than 80% and did not significantly differ between test 1 and test 2. Excluding infrequent single ventricular premature complexes, the reproducibility of a test with positive outcome was 76%. Similarly, test-retest agreement, a bidirectional measure of reproducibility, was greater than 74%. Kappa coefficients ranged from 0.26 to 0.36 for all grades of arrhythmia, indicating that reproducibility was consistently greater than by chance alone. Linear regression analysis of arrhythmia frequency in the 2 tests yielded R values of 0.76 to 0.96 and slopes of 0.82 to 1.2. No patient had enough spontaneous variability in repetitive forms to suggest spurious efficacy or aggravation. For ventricular premature complexes there was an 8 to 14% incidence of spurious efficacy and a 4 to 7% incidence of spurious aggravation. These results suggest that exercise-induced ventricular arrhythmia is sufficiently reproducible to serve as an adjunct method in the evaluation and management of patients with potentially life-threatening ventricular arrhythmias.