Using quantitative coronary arteriography to redefine SPECT sensitivity and specificity.

Previous studies looking at the sensitivity, specificity, and predictive accuracy of single photon emission computed tomography (SPECT) have been based upon the results obtained by visual interpretation of coronary arteriograms. Since the results of visual and quantitative determination of percent diameter stenosis have been shown to be statistically different, the results obtained from SPECT imaging when compared to quantitative methods for assessing coronary artery disease would be expected to provide a more correct assessment of sensitivity, specificity, and predictive accuracy. To determine the "true" sensitivity, specificity, and predictive accuracy of SPECT in diagnosing coronary artery disease, this study compared the results obtained in 44 SPECT images (20 thallium and 24 teboroxime) with the results obtained when quantitative coronary arteriography was used to analyze the coronary arteriograms. These 44 cases were then compared against 8 different definitions of significant coronary artery disease, varying from 30 to 80%, to yield 352 comparisons. The maximum specificity and predictive accuracy was found when 45% diameter stenosis was used to define the presence or absence of significant disease. At 45% diameter stenosis, SPECT imaging demonstrated an 86% sensitivity, 78% specificity, and 94% predictive accuracy with only 6% false positives. In 100% of the cases where 45% diameter stenosis was used to define the presence of disease and exercise failed to demonstrate ST segment changes or angina, when SPECT imaging demonstrated a perfusion defect(s), quantitative coronary arteriography agreed with SPECT imaging results.(ABSTRACT TRUNCATED AT 250 WORDS)