RANDALL D. CEBUL, M.D.
To the editor: Those of us who use probabilistic reasoning in our approaches to patient care welcome the series of articles introduced by Dr. Sox in the January issue (1). Although far more is known about these quantitative approaches than could be introduced in Dr. Sox's brief review, I want to draw attention to a somewhat misleading definition of the "likelihood ratio"for a diagnostic test result—sensitivity/(1—specificity)—that appeared in the glossary.
The likelihood ratio expresses the odds that a given level of test abnormality would be expected in patients with the disease under consideration. Specifically, the likelihood ratio contrasts the proportion
CEBUL RD. The Likelihood Ratio and Diagnostic Tests. Ann Intern Med. ;104:888–889. doi: 10.7326/0003-4819-104-6-888_2
Download citation file:
Published: Ann Intern Med. 1986;104(6):888-889.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use