GEORGE A. DIAMOND, M.D.
To the editor: Detrano and Froelicher (1) recently proposed using epidemiologic follow-up as a referent standard for defining the sensitivity and specificity of screening tests for coronary artery disease. Despite the appeal of this approach, it has an important fundamental limitation.
Assume that a hypothetical test is used to define two prognostic subsets: a high-risk group of 200 patients with positive test responses, 20 of whom die during the next year (a 10% annual mortality), and a low-risk group of 800 patients with negative test responses, only 8 of whom die during the next year (a 1% annual mortality). At
DIAMOND GA. Screening for Coronary Artery Disease. Ann Intern Med. 1987;107:594–595. doi: 10.7326/0003-4819-107-4-594_2
Download citation file:
© 2018
Published: Ann Intern Med. 1987;107(4):594-595.
DOI: 10.7326/0003-4819-107-4-594_2