LETTER
Bias in Observational Studies of Treatment
Dennis H. Osmond, MD;
Edwin Charlebois, MD; and
Andrew R. Moss, MD
1 December 1996 | Volume 125 Issue 11 | Page 941
TO THE EDITOR:
In their recent article, Glesby and Hoover [1] cite our paper on changes in survival among patients with the acquired immunodeficiency syndrome (AIDS) [2] as an analysis subject to selection bias. We agree that this is a potentially important bias, but we were surprised that they ignored our explicit concern about this effect. We noted that "because subjects with rapid CD4 lymphocyte loss were removed from the cohorts by death at a disproportionately high rate during the early years of followup ... subjects dying early in follow-up or progressing rapidly had less chance to receive therapy" [2].
We do not believe our results are attributable to selection bias. Glesby and Hoover suggest that our conclusions are based on a proportional hazards model. We concluded that treatment and prophylaxis for Pneumocystis carinii pneumonia had a larger effect on improved survival than did antiretroviral therapy because we saw improved survival over the decade only in men with a diagnosis of P. carinii pneumonia, an outcome that is hard to explain if antiretroviral therapy were responsible. We suggested, with qualification by the recognition of possible bias, that the proportional hazards model showed a lack of survival benefit from initiation of zidovudine therapy at higher CD4 lymphocyte counts. This was subsequently found in clinical trials of early compared with late initiation of zidovudine therapy.
Glesby and Hoover hypothesized that "those who received little or no benefit from antiretroviral therapy would have died sooner and, by default, would never have used the prophylaxis." We observed almost no deaths among human immunodeficiency virus positive-patients with CD4 counts greater than 200 cells/mm3; by 1991, nearly all cohort members with CD4 counts less than 200 cells/mm3 reported some use of antiretroviral and prophylactic therapies (90% and 92%, respectively) [3]. We believe that the effect of zidovudine use on survival has been modest, but that, in the early years of the epidemic, survival was improved with treating and preventing P. carinii pneumonia.
|
Author and Article Information
|
|---|
University of California, San Francisco, San Francisco, CA 94341.
1. Glesby MJ, Hoover DR. Survivor treatment selection bias in observational studies: examples from the AIDS literature. Ann Intern Med. 1996; 124:999-1005.
2. Osmond DH, Charlebois E, Lang W, Shiboski S, Moss A. Changes in AIDS survival time in two San Francisco cohorts of homosexual men, 1983 to 1993. JAMA. 1994; 271:1083-7.
3. Lang W, Osmond D, Page-Bodkin K, Moss A, Winkelstein W. Population-based estimates of antiretroviral therapy and anti-pneumocystis prophylaxis in San Francisco: 1991. J Acquir Immune Defic Syndr. 1993; 6:191-3.
About Letters
The Editors welcome submissions for possible publication in the Letters section. Authors of letters should:
•Include no more than 300 words of text, three authors, and five references
•Type with double-spacing
•Send three copies of the letter, an authors' form signed by all authors, and a cover letter describing any conflicts of interest related to the contents of the letter.
Letters commenting on an Annals article will be considered if they are received within 6 weeks of the time the article was published. Only some of the letters received can be published. Published letters are edited and may be shortened; tables and figures are included only selectively. Authors will be notified that the letter has been received. If the letter is selected for publication, the author will be notified about 3 weeks before the publication date. Unpublished letters cannot be returned.
Annals welcomes electronically submitted letters.
Top
Author & Article Info
References
Article
