There is a wide variation in reported accuracy ofnoninvasive stress myocardial imaging as a screening tool for coronary artery disease (CAD). This study was undertaken to determine its current accuracy in a wide spectrum of patients with chest pain syndromes using invasive coronary angiography as the gold standard.
The patient population consisted of consecutive patients undergoing coronary angiography in whom noninvasive stress imaging, either nuclear or echocardiographic, was performed within 6 months prior to the angiogram. The specificity, sensitivity, positive and negative predictive values, and diagnostic accuracy for detecting > or =1 lesions with > or =50% diameter coronary stenosis were determined for each modality.
Of the 227 eligible patients, 141 were men and 86 were women; 70% had significant CAD. The diagnostic accuracy overall was 71% and was no different for nuclear or echocardiographic testing. The positive predictive value (86% vs. 52%; P = 0.002) and diagnostic accuracy (83% vs. 51%; P = 0.002) were better in men than in women.
In this study, noninvasive stress imaging lacked the accuracy of a good screening test for significant CAD. This finding was particularly true for women, for whom it was not much better than a coin toss.