AMA Council Suggests Way To Reduce Research Bias
No one will be surprised to learn one of the conclusions in a report on scientific publication bias by the AMA's Council on Scientific Affairs: money matters in research.
The report, issued at last month's House of Delegates meeting in Chicago, states that “studies with positive findings are more likely to be published than studies with negative or null results, and an association exists between pharmaceutical industry sponsorship of clinical research and publication of results favoring the sponsor's products.”
Additionally, the publication of negative results may be delayed compared with the time to publication of studies with positive results. This relative time lag is not limited to industry-sponsored trials.
The council conducted a literature search of articles published between 1985 and 2003 and identified 12 systematic reviews on the relationship between pharmaceutical-industry sponsorship and research outcome, quality, or publication bias. Three of these reviews evaluated previously published information on the relationship between industry funding and outcome or quality; more than 2,000 original studies were included in the original studies covered by these reviews.
Four sources of bias were identified: investigators and authors, journal editors and reviewers, clinical-trial agreements, and outcome bias.
“When control lies with the commercial rather than academic or public sector, bias can also envelop the process through the trial design,” the report stated. “Outcome bias can result from the use of unreliable methods or instruments, as well as inadequate sample size or comparison groups.”
On the basis of the findings, the Council on Scientific Affairs recommended the following:
That all medical journal editors and authors adhere to the revised“ Consolidated Standards for Reporting of Trials (CONSORT) Group Statement” (posted online at<www.consort-statement.org>) and the International Committee of Medical Journal Editors' “Uniform Requirements for Manuscripts Submitted to Biomedical Journals” (posted online at<www.icmje.org>). | |||||
That the Department of Health and Human Services establish a comprehensive registry for all clinical trials conducted in the United States; that every clinical trial should have a unique identifier; and that all results from registered clinical trials be made publicly available through either publication or an electronic data repository. | |||||
That the AMA urge institutional review boards to consider registration of clinical trials in an existing registry as condition of approval. | |||||
The council's report, “Influence of Funding Source on Outcome, Validity, and Reliability of Pharmaceutical Research,” is posted online at<www.ama-assn.org/ama/pub/article/print/2036-8608.html>.▪
