Detecting arrhythmia by exercise electrocardiography in pediatric patients: assessment of sensitivity and influence on clinical management.

We studied 49 patients with suspected arrhythmias (group 1) and 92 patients with previously documented arrhythmias (group 2) who underwent exercise electrocardiography (EECG) at our institution between 1979 and 1987. Among group 1 patients, all of whom had sinus rhythm before exercise, 10 (20%) had abnormal findings on EECG, and treatment was modified in 4 of these 10 (8% of group 1). Further testing (24-hour ambulatory or transtelephonic electrocardiographic monitoring or electrophysiologic study) of the 39 patients with normal EECG findings revealed 8 additional patients with arrhythmias. In group 1, the sensitivity of EECG was 56%, and its negative predictive value was 79%. Group 2 consisted of 38 patients with atrial arrhythmias, 31 with ventricular arrhythmias, and 23 with atrioventricular conduction abnormalities before EECG. Of these 92 patients, 68 (74%) had abnormal EECG findings. All but 1 of the 24 patients with normal findings underwent further testing, and rhythm abnormalities were induced in 16. Patients with atrial arrhythmias were more likely to have normal EECG results (42%) than were those with ventricular arrhythmias (23%) or an atrioventricular conduction abnormality (4%). Of the 35 patients who had been referred for suppression of an arrhythmia, 25 (71%) had abnormal rhythm suppressed during exercise. Clinical management was modified in 27% of group 2 patients on the basis of EECG findings.