B.C. Williams, MD, MPH
To the Editor: Biomedical research is becoming sophisticated in applying the tools of epidemiology, but slips occasionally still occur. This was the case in the recent American College of Physicians position paper (1) on parenteral nutrition in patients receiving cancer chemotherapy. The effectiveness of parenteral nutrition was reported in terms of "odds ratios," defined as "the probability that patients receiving nutritional support will experience the outcome divided by the probability that control patients will experience the same outcome." This ratio of two probabilities is a relative risk (2), not an odds ratio. An odds ratio is exactly that—a ratio of
Williams B. Odds Ratio and Relative Risk. Ann Intern Med. ;111:444–445. doi: 10.7326/0003-4819-111-5-444
Download citation file:
Published: Ann Intern Med. 1989;111(5):444-445.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use