Skip navigation
News

International group calls for end to selective reporting of clinical trials

Free access to all data will provide the best care for patients, says Cochrane Collaboration.

BY HANNAH HOAG | DEC 05 2011

People don’t like to reveal their failures. But when it comes to clinical trials, researchers should be compelled to make even their negative results public, according to a statement issued by an international group that reviews medical research studies. The release of all information and data from randomized clinical trials would allow physicians to provide the best possible care for patients, says the Cochrane Collaboration.

In 2004, the pharmaceutical company Merck withdrew its blockbuster arthritis drug rofecoxib (Vioxx) from pharmacies around the world because of an increased risk of heart attacks following long-term use. “The widespread use of rofecoxib has likely caused about 100,000 unnecessary heart attacks in the U.S. alone, corresponding to about 10,000 deaths, which could have been avoided by using other, equally effective drugs causing less harm,” Peter Gøtzsche, director of the Nordic Cochrane Centre in Demark, wrote in an editorial. The crux of the issue was that the clinical trial results had been selectively reported, dampening the potential risks.

“There’s been evidence of this publication bias for a long time,” says Jeremy Grimshaw, a professor of medicine at the University of Ottawa and co-chair of the steering group of the Cochrane Collaboration. “We’re not getting an unbiased view of the information. The selective reporting of trial results can overestimate the benefits of a drug and underestimate the harms,” says Dr. Grimshaw, who is also a senior scientist in the clinical epidemiology program at the Ottawa Hospital Research Institute.

In early October, the Cochrane Collaboration recommended changes to the way clinical trials are managed and reported so that patients can receive the best treatment. The group called on those who conduct trials to register all randomized clinical trials prior to patient recruitment and to make publicly available, free of charge, all data from all randomized clinical trials as well as the corresponding protocols. It also asked governments to introduce legislation to require data sharing within 12 months following the end of the randomized phase of the trial and to consider punitive measures for non-compliance.

Some headway had already been made on the first point. The member journals of the International Committee of Medical Journal Editors, which includes the Canadian Medical Association Journal, require authors to pre-register the clinical trial in a public trials registry, such as www.clinicaltrials.gov, a database managed by the U.S. National Institutes of Health, if they aim to publish their results in one of these top journals. The committee also encourages editors to publish important studies even if the results are not statistically significant. The Canadian Institutes of Health Research requires registration of all CIHR-funded randomized controlled trials, the disclosure of adverse events and submission of a final report within 12 months of the end of the trial.

Those measures, though important, aren’t sufficient to reduce publication bias, says Matthew Herder, a lawyer and assistant professor in the faculty of medicine at Dalhousie University, who recently argued in a CMAJ article that Health Canada should disclose the design and results of clinical trials. “If you’ve run a trial that doesn’t show a difference between those that received the treatment and those that didn’t, there is no requirement to disclose that information,” he says. “We need stronger enforcement measures to require information sharing.”

Monika Krzyzanowska, an oncologist at Princess Margaret Hospital, part of the University Health Network in Toronto, studies the effect of publication bias on clinical practice. “Registering the clinical trial is only part of the solution. Even if a trial is small or negative, it can be informative,” she says. “If you are selectively publishing only positive data, someone searching Medline or Pubmed will only find positive results, but if there were four other trials that were negative, then the lay of the land might not be in favour of this experimental drug.”

Professor Herder, who supports all the points the Cochrane Collaboration makes in its statement, including punitive measures for non-compliers, says university researchers may find themselves between a rock and a hard place. “There’s been a growing emphasis at universities to commercialize research and that might be operating in tension with this call for more sharing,” he says.

COMMENTS
Post a comment
University Affairs moderates all comments according to the following guidelines. If approved, comments generally appear within one business day. We may republish particularly insightful remarks in our print edition or elsewhere.

Your email address will not be published.