OBJECTIVES:This study was conducted to examine the efficacy of dobutamine stress two-dimensional echocardiography and perfusion scintigraphy for the detection of coronary artery disease in routine practice, to establish the causes of erroneous results and to derive appropriate criteria for the selection of either or both tests. BACKGROUND:Dobutamine stress combined with echocardiography or perfusion scintigraphy may be used to detect coronary artery disease. Although both imaging approaches have demonstrated similar levels of accuracy, it is not known whether there may be particular indications for the use of one or the other technique or a rationale for their combination. METHODS:Two hundred seventeen patients without previous infarction were studied prospectively with dobutamine stress echocardiography and technetium-99m methoxy isobutyl nitrile (sestamibi) single-photon emission computed tomography at the time of diagnostic coronary angiography. The presence of coronary stenoses of > or = 50% diameter was compared with the presence of rest or stress-induced abnormalities of perfusion and regional function. The extent of these abnormalities was correlated with an equivalent score of extent of angiographic disease. RESULTS:Significant coronary artery disease was found in 142 patients; 102 (72%) were identified by dobutamine echocardiography and 108 (76%, p = NS) by perfusion imaging. In 75 patients without significant disease, the specificity of dobutamine echocardiography was 83% compared with 67% for scintigraphy (p = 0.05). Echocardiographic sensitivity was lower in patients unable to complete the test because of side effects (n = 64) than in the remainder (59% vs. 77%, p = 0.02); this influence was less apparent with scintigraphy (71% vs. 78%, p = NS). Selective use of scintigraphy in the 31 patients with a negative submaximal stress echocardiogram led to a sensitivity of 80% for this combination. Patients with left ventricular hypertrophy accounted for most of the difference in specificity between echocardiography and scintigraphy (94% vs. 59%, p = 0.02). Their respective accuracies were 76% and 73%. CONCLUSIONS:Dobutamine stress echocardiography and perfusion scintigraphy have equivalent accuracy. In patients with left ventricular hypertrophy, echocardiography appears to be the test of choice. Selective use of sestamibi scintigraphy in patients with a negative submaximal echocardiogram enhances the accuracy of stress echocardiography alone.