Reproducibility of a multi-stage near maximal treadmill test for exercise tolerance in angina pectoris.

A retrospective analysis was made of the reproducibility of a treadmill exercise tolerance test performed on patients with angina pectoris to an end point of moderately severe (3+) angina. One hundred and forty tests (70 pairs) in 63 patients were found to meet criteria for inclusion in the study. In 30 patients a comparison of the first test the patient had ever had with a subsequent test revealed a significantly poorer performance on the first test; mean differences + one standard error of the mean between the first test and the subsequent tests were 0.90 + 0.20 min of walking from the start of the test to the onset of angina, 2.3 + 0.33 total min of walking, and 5.8 + 0.86 "exercise units" attained at the end point of "3+" angina. In 40 patients a comparison of two tests, neither of which was the patients' first test, revealed no significant differences in performance; mean differences between the two tests were 0.20 + 0.23 min of walking at the onset of angina, 0.11 + 0.07 min of walking and 0.18 + 0.15 exercise units attained at the end point of "3+" angina. There were no significant differences in maximum heart rates attained between any of the tests. In using treadmill testing to quantitate exercise capacity in angina pectoris, the first test is likely to yield a poorer performance than subsequent ones. Tests subsequent to the first are highly reproducible when a specifically defined symptomatic end point of moderately severe intensity is employed.