Study: Industry Drug Trials Show More Favorable Results

A study published in today’s issue of the Annals of Internal Medicine reviews drug trials appearing on the Web site ClinicalTrials.gov, and finds those trials funded by industry sources are more likely to report data favorable to the drug being tested.

The investigators — Florence Bourgeois and Kenneth Mandl of Children’s Hospital Boston in Massachusetts and Srinivas Murthy of the Hospital for Sick Children in Toronto, Canada — reviewed 546 drug trials conducted between 2000 and 2006, listed with ClinicalTrials.gov, a federal registry of clinical trials. The study focused on five classes of drugs: cholesterol-lowering drugs, antidepressants, antipsychotics, proton-pump inhibitors that reduce the production of stomach acid, and vasodilators that dilate blood vessels to treat hypertension. The researchers also reviewed medical literature for publications associated with each trial, checking four separate databases, and contacted trial investigators if necessary.

The authors then abstracted the data from the trial records and publications, including timing of registration, elements of the study design, funding source — industry, government agencies, and non-profit or non-federal organizations — publication date, and study outcomes.

The results show industry funded nearly two-thirds (63%) of the trials on ClinicalTrials.gov, with non-profit/non-federal sources supporting about a quarter (23%), and government covering the remaining 14 percent. About two-thirds (66%) of the trials had published results within three years.

The vast majority (85%) of industry-funded trials reported positive results, as did a somewhat smaller percentage (72%) of trials supported by non-profit/non-federal sources. In contrast, half (50%) of the federal government trials reported positive outcomes. Among the non-profit/non-federal funded trials, 85 percent of those that had some industry support reported positive results, compared to 61 percent of those without industry contributions.

Other findings: Nearly all (89%) of the industry entries reported on later-stage (Phase 3 or 4) trials. Industry trials were also more likely to have larger numbers of participants — median sample size of 306 cases.

In a press release, the authors highlight initiatives to register all clinical trials, thus giving a more complete record of outcomes, and reducing publication bias.  In 2005, the International Committee of Medical Journal Editors began requiring that a trial be registered before enrolling patients in order to be considered for publication. In 2007, the U.S. Food and Drug Administration — a partner with National Institutes of Health in developing ClinicalTrials.gov — expanded the scope of the site, requiring the sponsors of all drug, biologic, and device trials after Phase 1 to register their studies.

Photo: Jorge Franganillo/Flickr