Share

Bias in industry studies

Drug reviews that are supported by the pharmaceutical industry tend to be more biased and rate the drugs more favourably than independent reviews, a new study contends.

As a result, industry-supported reviews of existing drug research need to be taken with a grain of salt, said the authors of the study in the October 7 issue of the British Medical Journal.

"It comes as no surprise that certain methodological limitations and selection biases would emerge when a purely academic meta-analysis approach - or one driven by a pharmaceutical sponsor - was taken," said Dr Mark Fendrick, a professor of health management and policy at the University of Michigan School of Public Health, and co-editor-in-chief of the American Journal of Managed Care.

The new study corroborates a number of other recent studies that have shown that clinical trials of drugs funded by pharmaceutical companies and other for-profit entities were more likely to report positive findings than similar trials funded by nonprofit groups.

Industry funding more studies
At the same time, the drug industry is paying for more and more medical research, with a full half of studies now funded solely by the private sector, another study found.

Independent reviews of drugs - such as those conducted by the Cochrane Collaboration, an independent body - can have a much-needed counterbalancing effect, said the authors of the new study, who are based at a Cochrane centre in Denmark. Cochrane reviews collect and analyse existing data using a rigorous, standardised system, the researchers said.

For the new study, the researchers compared 24 Cochrane reviews with other so-called meta-analyses, or studies of studies. Eight of the 24 reviews were supported by the pharmaceutical industry, nine had undeclared support, and seven had no support or were supported by non-industry sources.

On a scale of 0 to 7, the median quality score for a Cochrane review was 7, but only 3 for the other reviews, the authors of the new study said.

Cochrane reviews more often considered the potential for bias within the review, compared with industry-supported reviews and reviews with undeclared support. And, in general, Cochrane reviews had fewer methodological problems, although they were not without problems altogether, the study authors said.

Industry recommends more freely
All seven of the industry-supported reviews recommended the drug in question without hesitation, while none of the Cochrane reviews did. In both cases, however, the estimated effect of the drug was similar.

"This suggests that the main problem with industry-supported reviews lies in how conclusions are formulated," the authors stated.

As one example, a meta-analysis funded by Merck & Co. concluded in 2001 that there was no increased risk for arterial thrombosis (blood clots within an artery) with the company's now-withdrawn arthritis drug Vioxx. But another, non-industry-supported meta-analysis did show an increased risk for clots. That increased risk was evident in publications that were available to the authors of the first report, the authors of the new review said.

In September 2004, Merck withdrew Vioxx from the market after studies found the painkiller increased the risk of heart attack and stroke.

In June, the New England Journal of Medicine issued a rare correction on an influential 2005 study on Vioxx. The correction involved a crucial point: That cardiovascular risks linked to the controversial cox-2 inhibitor kicked in at four months to six months of use, not after 18 months as had been reported in the article first published in 2005. The original article outlined the results of the so-called APPROVe study, which was funded by Merck.

The authors of this week's BMJ study found that, generally, drugs reviews with undeclared support, or not-for-profit support, or no support drew conclusions similar in cautiousness to the Cochrane reviews.

Call for transparency
As a result of the new findings, the authors of the new study are calling for more transparency in industry-supported drug trials, including more information on the methodology used and the estimated effects of the drugs.

"They're not saying industry-supported meta-analyses should not be done or should not be published but, once again, the issue of sponsorship should be highlighted," Fendrick said. "Sponsorship of scientific research, whether a single study, a meta-analysis or anything else, should be emphasised to be sure that the potential for these biases are made immediately apparent."

Asked to comment on the new study, Jeff Trewhitt, a spokesman for the Pharmaceutical Research and Manufacturers of America (PhRMA), a drug industry group, said, "We still need to study the report. But it is important to note that the vast majority of clinical studies that companies conduct are rigorously reviewed by regulators as part of the drug approval process. It's in our interests to make sure that the studies are of the highest quality.

"What's more, PhRMA and its member companies are committed to accuracy and transparency in clinical trials. PhRMA, two years ago, launched www.ClinicalStudyResults.org and, so far, member companies have submitted data summaries covering nearly 300 medicines. The summaries cover positive and negative data. PhRMA also earlier this year urged companies to register clinical trials for all medicines and not just drugs for life-threatening diseases on www.clinicaltrials.gov at the National Institutes of Health," he added. – (HealthDayNews)

Read more:
Research Centre
One in three breakthroughs refuted

October 2006

We live in a world where facts and fiction get blurred
Who we choose to trust can have a profound impact on our lives. Join thousands of devoted South Africans who look to News24 to bring them news they can trust every day. As we celebrate 25 years, become a News24 subscriber as we strive to keep you informed, inspired and empowered.
Join News24 today
heading
description
username
Show Comments ()
Editorial feedback and complaints

Contact the public editor with feedback for our journalists, complaints, queries or suggestions about articles on News24.

LEARN MORE