Background: Clinical characteristics of patients, angiographic referral bias, and several technical factors may all affect the reported diagnostic accuracy of tests. The aim of this study was to assess their influence on the diagnostic accuracy of dobutamine stress echocardiography (DSE).
Methods: The medical literature from 1991 to 2006 was searched for diagnostic studies using DSE and meta-analysis was applied to the 62 studies thus retrieved, including 6881 patients. These studies were analyzed for patient characteristics, angiographic referral bias, and several technical factors.
Results: The sensitivity of DSE was significantly related to the inclusion of patients with prior myocardial infarctions (0.834 vs 0.740, P < .01) and defining the results of DSE as already positive in case of resting wall motion abnormalities rather than obligatory myocardial ischemia (0.786 vs 0.864, P < .01). Specificity tended to be lower when patients with resting wall motion abnormalities were included in a study (0.812 vs 0.877, P < .10). The presence of referral bias adversely affected the specificity of DSE (0.771 vs 0.842, P < .01).
Conclusion: This analysis suggests that the reported sensitivity of DSE is likely higher and the specificity lower than expected in routine clinical practice because of the inappropriate inclusion of patients with prior myocardial infarctions, the definition of positive results on DSE, and the negative influence of referral bias. However, in the patient subset that will be sent to coronary angiography, the opposite results can be expected.