True sensitivity of cardiac exercise testing. A combined clinical evaluation of multiple parameters

Maximal exercise tests were performed in 308 patients with angiographically defined coronary disease, in 38 patients whose coronary vessels were not obstructed, and in 236 clinically normal volunteers. The sensitivity of parameters indicating ischaemia was evaluated singly and in combination in each group. The most sensitive individual parameters were work capacity restriction (65%), ST segment depression (56%) and test angina (55%). When the sequence of parameters were assessed successively in the coronary group, the sensitivity rose from 56% with ST depression to 70% with the addition of test angina, to 85% when work capacity restriction was added, and to 87% when inadequate heart rate or blood pressure responses or repetitive ventricular arrhythmias were included. In the group of patients whose coronary vessels were not obstructed, substantial clinical abnormalities and a bias to carry out angiography in those with abnormal exercise responses led to false ST segment changes in 34%, and, in a further 20%, to positivity in the other parameters. In clinically normal subjects a 15% "false" positive rate occurred. One parameter (R wave amplitude change) was falsely positive in 24%, and was therefore excluded. Grouped parameter sensitivity for patients who exercised on a bicycle and those who used a treadmill was similar.