JEFFREY GLASSROTH, M.D.
To the editor: The recent article by Taylor and coauthors (1) shows that in decision analysis the decision depends on the data analyzed. Although existing data on isoniazid are imperfect, in the report of the study by Taylor and colleagues the information introduced into the decision analysis system was not the best available. Comstock (2) has pointed out in his accompanying editorial many of the study's deficiencies. In addition, I think the following three points deserve comment.
First, in defining tuberculosis incidence Taylor and associates use, in part, data generated in several surveys of Southeastern U.S. populations (3). Nontuberculous mycobacterial
GLASSROTH J. Tuberculosis Prophylaxis with Isoniazid. Ann Intern Med. 1981;95:517–518. doi: 10.7326/0003-4819-95-4-517
Download citation file:
Published: Ann Intern Med. 1981;95(4):517-518.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use