Accuracy of exercise 201Tl myocardial scintigraphy in asymptomatic young men.

BACKGROUND Little is known about the diagnostic usefulness of 201Tl scintigraphy for detecting asymptomatic coronary artery disease in apparently healthy men. We thus evaluated planar 201Tl exercise myocardial scintigraphy in 845 asymptomatic male military aircrew undergoing coronary arteriography because of abnormal noninvasive tests suggesting possible myocardial ischemia. METHODS AND RESULTS Patients were stratified by prior disease risk into six subgroups using age (< 45 and > or = 45 years) and ratio of total to high density lipoprotein cholesterol (< 4.5, 4.5-6.0, and > 6.0). Significant coronary artery disease (> or = 50% diameter stenosis in any major coronary artery) was present in 143 (16.9% prevalence). Overall sensitivity and specificity of 201Tl scintigraphy adjusted for verification bias were estimated to be 45 +/- 4% and 78 +/- 1%, respectively. These values are lower than corresponding values accepted for clinical populations. Positive and negative predictive values varied across subgroups. A normal thallium scan indicated low risk of disease, but an abnormal test was likely to be a false-positive result. A logistic equation was retrospectively fit to the data for estimating the probability of disease given age, cholesterol ratio, and thallium results. Within each quintile of estimated risk, the average risk did not differ significantly from the observed disease prevalence. CONCLUSIONS Exercise 201Tl scintigraphy is limited by the frequent occurrence of false-positive tests in detecting asymptomatic, anatomic coronary artery disease in young men in accordance with Bayesian probability theory.