Leonard Leibovici, MD
To the editors: Two points in the article by Bates and colleagues (1) merit reconsideration. First, rapidly fatal disease was included in the final model as a significant and independent predictor of true bacteremia. However, "true bacteremia" was defined by reviewers who were not blinded to the outcome of the episode (whether or not the patient "had done well"). If the original assessment of a rapidly fatal disease was accurate, than "true bacteremia" was defined with foreknowledge of a factor (outcome of the patient) closely related to one of the independent variables. Thus, a strong bias may have been introduced
Leibovici L. Predicting Bacteremia. Ann Intern Med. 1991;114:703. doi: 10.7326/0003-4819-114-8-703_1
Download citation file:
Published: Ann Intern Med. 1991;114(8):703.
Multi-Organ Failure and Sepsis, Pulmonary/Critical Care.
Results provided by:
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use