American Journal of Epidemiology Advance Access originally published online on December 21, 2007
American Journal of Epidemiology 2008 167(4):500-501; doi:10.1093/aje/kwm372
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
LETTERS TO THE EDITOR |
RE: LETTER TO THE EDITOR ON "BIAS IN CLINICAL INTERVENTION RESEARCH"
1 Department of Environmental Medicine, New York University School of Medicine, New York, NY 10016-3240
2 Department of Medicine, New York University School of Medicine, New York, NY 10016-3240
3 Department of Obstetrics and Gynecology, New York University School of Medicine, New York, NY 10016-3240
(e-mail: michael.marmor{at}med.nyu.edu)
Hemilä's letter (1) regarding Gluud's review article (2) asserted that "... there is no evidence of large and universal placebo effects" (1, p. 1219). Because Gluud's review focused on biases in intervention research generally, Hemilä's letter implies that placebos and the maintenance of blinding may not be critical to intervention research. Yet, Hróbjartsson and Gøtzsche's article (3), which Hemilä uses as the basis for his assertion, had the goal of evaluating whether placebos might be of value as solo treatments for certain conditions, and not whether placebo control groups are important in the design of clinical trials. Hróbjartsson and Gøtzsche cautioned that their findings "do not imply that control groups that receive no treatment can be substituted for control groups that receive placebo without creating a risk of bias" (3, p. 1599). These authors further cited a meta-analysis showing that clinical trials that were not double blinded yielded odds ratios that were exaggerated by an average of 17 percent (4).
Various studies provide evidence conflicting with Hemilä's dismissal of "large" placebo effects. In the treatment of hypertension, for example, response rates for placebo have on occasion equaled or exceeded those associated with effective therapies (5). Objective outcomes, such as the C-reactive protein concentration, facial swelling, and trismus, have been shown to respond to placebo treatments (6). Meissner et al. (7) furthermore identified a subgroup of "physical outcomes" that have responded significantly to placebo treatments in 50 percent of the clinical trials they analyzed; in a reanalysis of the database of Hróbjartsson and Gøtzsche, they found significant improvements in these physical outcomes in placebo compared with no-treatment study arms.
We recognize that placebo effects are not universal, but we question the relevance of this assertion by Hemilä to Gluud's review. Placebo effects can be overestimated because of regression to the mean, spontaneous improvement, additional treatment, and other factors (7). Control groups when feasible and ethical provide benchmarks for efficacy and adverse events (5) and statistical control for regression to the mean and potential confounding (7, 8). The use of placebos within control groups remains necessary to conceal the assignment of subjects to the different study arms and to reduce biases.
We also disagree with Hemilä's claim, previously asserted elsewhere (9, 10), that the clinical trial of ascorbic acid and the common cold by Karlowski et al. (11) and a review article by Chalmers (12) have been "shown to be erroneous" (1, p. 1219). Hemilä's analyses have attempted to estimate the number of subjects unblinded in the trial by Karlowski et al. and to demonstrate that the magnitude of potential bias associated with his estimated number of unblinded subjects is not great enough to explain the trial's results. The number of subjects who were unblinded in the trial by Karlowski et al. was not specified at the time of the original publication and remains today unknown. Bias associated with unblinding thus cannot be ruled out as an explanation for their findings. We agree with another group of authors who were criticized by Hemilä that, "while the interpretation [by Karlowski et al.] of a possible placebo effect has been challenged [by Hemilä], it has not necessarily been disproven" (13, p. 166).
Finally, Hemilä asserted that the trial by Karlowski et al. was used as an example of the "placebo effect in action" (1, p. 1219) in the original CONSORT statement (14). We fail to find such a statement in this reference. More importantly, the revised CONSORT statement (15) includes references to studies in which placebo controls were used (e.g., references 63, 79, 80, 91, 115, and 116 in the revised CONSORT statement), some of which showed nonzero placebo effects, although their significance was not specifically assessed.
 |
ACKNOWLEDGMENTS |
|---|
This work was supported in part by Center for AIDS Research grant P30 AI27742 (M. M.) from the National Institute on Allergy and Infectious Diseases, National Institutes of Health, and by Cancer Center grant 2P30 CA016087 (I. B-L. and A. A. A.) from the National Cancer Institute, National Institutes of Health.
Conflict of interest: none declared.
|
References |
|---|
 TOP References |
|---|
- Hemilä H. Re: "Bias in clinical intervention research." (Letter). Am J Epidemiol (2007) 165:1219.
[Free Full Text] - Gluud LL. Bias in clinical intervention research. Am J Epidemiol (2006) 163:493–501.
[Abstract/Free Full Text] - Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? An analysis of clinical trials comparing placebo with no treatment. N Engl J Med (2001) 344:1594–602.
[Abstract/Free Full Text] - Schulz KF, Chalmers I, Hayes RJ, et al. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA (1995) 273:408–12.
[Abstract/Free Full Text] - Preston RA, Materson BJ, Reda DJ, et al. Placebo-associated blood pressure response and adverse effects in the treatment of hypertension: observations from a Department of Veterans Affairs Cooperative Study. Arch Intern Med (2000) 160:1449–54.
[Abstract/Free Full Text] - Hashish I, Harvey W, Harris M. Anti-inflammatory effects of ultrasound therapy: evidence for a major placebo effect. Br J Rheumatol (1986) 25:77–81.
[Abstract/Free Full Text] - Meissner K, Distel H, Mitzdorf U. Evidence for placebo effects on physical but not on biochemical outcome parameters: a review of clinical trials. BMC Med (2007) 5:3. (http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1847831).[CrossRef][Medline]
- Ernst E, Resch KL. Concept of true and perceived placebo effects. BMJ (1995) 311:551–3.
[Free Full Text] - Hemilä H. Allocation concealment and blinding: when ignorance is bliss. (Letter). Med J Aust (2005) 183:165–6.[Medline]
- Hemilä H. Echinacea, vitamin C, the common cold, and blinding. Clin Infect Dis (2005) 41:762–3.[CrossRef][Web of Science][Medline]
- Karlowski TR, Chalmers TC, Frenkel LD, et al. Ascorbic acid for the common cold. A prophylactic and therapeutic trial. JAMA (1975) 231:1038–42.
[Abstract/Free Full Text] - Chalmers TC. Effects of ascorbic acid on the common cold. An evaluation of the evidence. Am J Med (1975) 58:532–6.[CrossRef][Web of Science][Medline]
- Forder PM, Gebski VJ, Keech AC. Allocation concealment and blinding: when ignorance is bliss—in reply. (Letter). Med J Aust (2005) 183:166.[Web of Science][Medline]
- Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA (1996) 276:637–9.
[Abstract/Free Full Text] - Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med (2001) 134:663–94.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||