Comparative sensitivity of the exercise electrocardiogram, thallium imaging and stress radionuclide angiography to detect the presence and severity of coronary heart disease.

Abstract

The relative sensitivity and specificity of individual and combined noninvasive tests to detect coronary heart disease were evaluated in 75 patients with chest pain admitted for cardiac catheterization and coronary arteriography. Of the 75 patients, 56 had coronary heart disease. Exercise-induced ST-segment abnormalities (.gtoreq. 1 mm) were found in 58%. Computer-processed exercise 201Th perfusion imaging detected 82% (P < 0.01) and assessment of regional ejection fraction determined at rest and during isometric exercise by radionuclide angiography detected 82% (P < 0.02). Pathologic Q waves were present in 20%. Of 9 patients with single-vessel disease, only 1 had exercise ST-segment abnormalities, while 4 had abnormalities in 201Th perfusion and 5 in regional ejection fraction. Of 16 patients with 2-vessel disease, 10 had ST-segment abnormalities, 14 had defects on 201Th imaging and 13 had abnormalities in regional ejection fraction. Of 31 patients with 3-vessel coronary heart disease, 23 had exercise-induced ST-segment changes, while 28 had 201Th perfusion defects and 28 had abnormalities in regional ejection fraction. Combined noninvasive testing using pathologic Q waves and exercise ST-segment abnormalities detected 71% of patients with coronary heart disease. Addition of exercise 201Th imaging resulted in 88% of patients being detected and addition of regional ejection fraction detected 96%. If an abnormality in any of the 4 tests was considered, 55 of 56 patients (98%) with coronary heart disease were detected. In 19 patients with normal coronary arteries, the specificity of the exercise ECG was 84%, exercise 201Th imaging 89% and assessment of regional ejection fraction 79%. If all noninvasive tests were considered, the specificity decreased to 58%. Either exercise 201Th imaging or assessment of regional ejection fraction is superior to exercise-induced ST-segment abnormalities. Combined testing results in a very high sensitivity, but there is a concomitant reduction in specificity.