Stumble Upon

del.icio.us

Digg

Facebook

Print

Email

Speak Up

TORONTO - Researchers who scrutinized news releases put out by top-selling pharmaceutical companies say they've found some shortcomings.

Dr. Chaim Bell, the senior author of the study, said he began the project after noticing press releases and news articles that dealt with "really preliminary information" but didn't itemize caveats or limitations of the studies.

The study, published in the journal Public Library of Science One, was conducted by Bell and other researchers at St. Michael's Hospital and the University of Toronto.

"The classic is that there's always some new cure for cancer that comes out," explained Bell, a clinician-scientist and assistant professor of medicine and health policy, management and evaluation.

"And unfortunately people become disappointed when it doesn't come to fruition, but part of that is because it was preliminary findings, and then when there are more mature findings, or when it's analyzed a different way, with ... a larger sample size, it really didn't come out to those same findings."

They decided to look at the electronic press releases issued in a one-year period by the 10 top-selling pharmaceutical companies, but had to exclude one company's releases because they weren't publicly available for the chosen year, 2005.

Altogether, they counted 1,028 press releases that were issued. Of these, 235 concerned an announcement about original research. When these were studied, it was found that more than half were concerning preliminary research of the sort that might be presented at a conference.

"I was surprised that of the scientific communications that companies give out, that about 60 per cent of them are based on findings from abstracts," Bell said in an interview Wednesday, referring to preliminary data presented at scientific conferences.

"So the majority of the findings of the press releases are related to preliminary findings that aren't fully peer-reviewed."

He noted that such abstracts are selected by a committee for conferences, but the information isn't normally subjected to the same rigorous examination that it would get prior to being published in a medical journal.

"About a fifth of those, they didn't really say where they got their data. They just said 'findings of a study.' It was a real Sherlock Holmes type of enterprise to try and track down where this study came from or was this original data, was this abstract data or was it presented at a meeting."

In addition, about 30 per cent did not quantify the study results, Bell said. For example, a news release might say one drug was better than another rather than saying there was a 30 per cent decreased risk related to using one drug over another.

"But really, they didn't talk about the study limitations and the idea that this is preliminary studies or preliminary data. Almost universally this was not discussed."

On the plus side, 20 per cent of the news releases described studies published in peer-reviewed journals.

And Bell found it "encouraging" that the majority of news releases - 76 per cent - talked about adverse effects of the drugs.

"They often talk about adverse events or harms in this study, similar to what they would do with commercials on TV where they list a variety of adverse events or side-effects or harms," he said.

Russell Williams, president of Canada's Research-Based Pharmaceutical Companies (Rx&D), issued a statement Wednesday saying that his group hasn't had an opportunity to fully review the findings of Bell and his colleagues.

"However, we can clearly state that as the community that develops medicines and vaccines that improve and save lives, our members are committed to accuracy and transparency in all public communications," he said in an email.

"Recognizing the space limitations of a press release and the complexity of the research and development of new medicines and vaccines, other sources are readily available and accessible to journalists seeking further information including websites and contacts."

He noted that the organization's members have created a website to report ongoing clinical trials and results.

As for the limitation of his own study, Bell noted that it only looked at a one-year period, and only the top-selling pharmaceutical companies.

"It may be that other companies differ from this in terms of quality," he said of their news releases.

-

On the Net:

Drug industry clinical trials:

http://clinicaltrials.ifpma.org/no-cache/en/myportal/index.htm

Study in PLoS One journal:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0 002828