[Skip to Navigation]
Sign In
Editor's Correspondence
November 26, 2001

Electron-Beam Computed Tomography as a Population Screening Tool

Arch Intern Med. 2001;161(21):2624-2625. doi:

In their article, Nallamothu et al1 conclude that, "The performance of EBCT as a diagnostic test for obstructive CAD is reasonable based on sensitivity and specificity rates." Unfortunately, the most common indication for EBCT currently in our market is the screening of asymptomatic adults, often driven by newspaper and billboard advertisements. It would have been helpful for the authors to discuss the implications of their data for such widespread screening of low-risk populations with EBCT. To be specific, if the diagnostic threshold is set to define a sensitivity of 92% and a specificity of 51%, consistent with the studies they summarize, and if the test is applied to a cohort of 1000 asymptomatic adults with a true CAD prevalence of 5% (close to the overall prevalence at age 50 years), the positive predictive power of EBCT will then be only 8.7%. If the sensitivity is reduced to 80%, and the specificity increased to 71%, a compromise that might be suggested for low-risk populations, the positive predictive power increases to only 12.7%. The true values may be lower since the specificity itself can decrease when the test is applied to other populations, eg, older patients with more incidental calcification.

Add or change institution