Prospective evaluation of radionuclide angiocardiography for the diagnosis of coronary artery disease.

This investigation assesses prospectively the accuracy of rest and exercise radionuclide angiocardiography (RNA) in detecting coronary artery disease (CAD). By retrospective analysis of 496 patients, optimal RNA criteria were determined for the presence or absence of CAD. Multivariate analysis of patients with normal coronary arteries on catheterization provided a formula to predict normal exercise ejection fraction (EF) in a given patient. The presence of CAD was indicated by 1 or more of the following RNA measurements: (1) rest EF less than 0.50, (2) exercise EF at least 0.06 less than the predicted value, (3) exercise increase in end-systolic volume greater than 20 ml, (4) exercise-induced wall motion abnormality. The absence of CAD was indicated by the absence of all 4 criteria. After applying these criteria to 221 consecutive patients, the RNA determinations were compared with the catheterization, determinations. Significant CAD was present in 71% of the patients studied. The sensitivity of the test was 0.87 and the specificity 0.54. Thus, because of its high sensitivity, RNA is of value in screening patients under consideration for cardiac catheterization. No patient with significant left main coronary narrowing and only 4 of 65 patients with 3-vessel disease were misdiagnosed. The poor specificity of the test, however, limits its overall accuracy.