Research contracts violate guidelines

DURHAM, N.C., Oct. 23 (UPI) -- A review of more than 100 contracts between academic research institutions and corporate sponsors found they repeatedly failed to adhere to international guidelines designed to protect the integrity of the research, a new study released Wednesday reveals.

Such a failure to follow these guidelines, which include items such as design of clinical trials, who has access to the clinical trial data and publication rights, could threaten research investigators and even patients, the study concludes.


Led by researchers at Duke University Medical Center and the Duke University School of Law in Durham, N.C., the study looked at 108 institutional members of the Association of American Medical Colleges. The study, which lasted from November 2001 to January 2002, examined the extent to which contracts between these institutions and industry sponsors complied with three key provisions put out by the International Committee of Medical Journal Editors.


The ICMJE guidelines are used by more than 500 medical journals as a standard for scientists submitting their research for publication. The three provisions examined were:

-- Whether research authors of multiple center clinical trials had access to all trial data;

-- Whether research authors or an independent committee controlled the editorial and publication decisions; and

-- Whether research investigators were involved in the design and protocol of the studies.

The investigators found academic institution contracts with corporate sponsors rarely included these provisions. Their analyses showed, on average, only 1 percent of these contracts required research authors in multiple center clinical trials to have independent access to data. However, academic institutions coordinating multiple-center studies had a higher level of access in half of the contracts reviewed, researchers report in the October 24 issue of The New England Journal of Medicine.

Lead study author Dr. Kevin A. Schulman, a professor of medicine, said the low compliance rate was a bit of a shock.

"How low the level was kind of surprising and also what was surprising was how powerless the institutions felt in negotiating," Schulman told United Press International. "There's a failure of protections."

Academic centers might feel as though they have minimal power in negotiating better contracts that guarantee freedom from corporate influence and research integrity, Schulman explained. "With a mature industry, this shouldn't be a threat at all," Schulman said.


However, a highly competitive environment to find effective drugs tends to make some sponsors unwilling to share control over a project. "I think we will see the academic centers be much more aggressive," he said. "After our paper comes out, they'll have a little more ammunition."

Dr. David Korn, senior vice president for biomedical and health sciences research for the Association of American Medical Colleges in Washington, D.C., said academic centers vary greatly in their size, staff and budget and those factors can affect how they bargain with sponsors.

The finding that compliance was "not 100 percent (is) very troubling," Korn told UPI. However, he said, there were some flaws in the research approach. "I think the way the study was conducted gives somewhat of a false alarm."

He added it is standard practice that not all researchers in multiple-center clinical trials have access to the data. "That would be a logistically difficult thing to accomplish on these multi-center trials," he said, adding it also is standard practice to give access to the data to the coordinating center and lead authors who will have their names appear on the published study. However, not every research team at every single site in these multiple-center studies typically gets to share in that access.


In an accompanying editorial, Dr. Jeffrey M. Drazen, professor of medicine at Harvard School of Public Health in Boston, wrote, "The study by Schulman et al. is imperfect." Because contract officers were surveyed, Drazen said, considering human nature's imperfect memory, they are subject to not recalling details correctly. Another flaw, Drazen explained, is research investigators, "who may be more aware of the terms of their contracts than contracting officers," were not surveyed.

Although there are disagreements about the study's interpretations, its findings illuminate the delicate power balance between sponsors and academic centers. A report in the same NEJM issue highlights an egregious example between a researcher and a sponsor. The 1998 case involved University of Toronto researcher Nancy Olivieri and Apotex, Inc., a Canadian pharmaceutical firm that funded Olivieri's research on a drug called deferiprone. When Olivieri found deferiprone ineffective and wanted to publish the results, Apotex halted the clinical trials and threatened legal action.

A similar case occurred around the same time, this time between the University of California, San Francisco and The Immune Response Corp. of Carlsbad, Calif. The dispute was over the effectiveness of an AIDS vaccine.

"(Immune Response) objected not to the data, but to how we analyzed the data," Dr. James Kahn, a professor of medicine at the University of California, San Francisco and principal investigator of that study, told UPI. When UCSF said they did not want to interpret the data differently to make vaccine results more favorable, the company threatened to withhold data and then also threatened legal action. The suit eventually was dropped, Kahn said. The results, which stated the vaccine's ineffectiveness, were published in the November 1, 2000 issue of Journal of the American Medical Association.


Maximizing results is not a violation of guidelines, however. For complex clinical trials, scientists can publish selected portions of the study, but must note in the publication whether portions of their same research had appeared elsewhere, explained Dr. Steven Lagakos, professor at Harvard's School of Public Health and an investigator in the AIDS vaccine study funded by Immune Response Corp.

"It's called salami science," Schulman said, "when you cut the study up into a lot of little pieces"

The problem is one of misrepresenting the data. Data interpretation is key and how the data are interpreted can skew the results of studies called meta-analyses, where researchers combine the results of multiple studies on a single topic. Skewed study results could impact patients.

"Right now we have these mass marketing drugs," Schulman said, and "the truth is there's probably a group of patients who probably don't benefit from some of these classes of drugs."

Dr. Drummond Rennie, a deputy editor at JAMA and a professor of medicine at the Institute of Health Policy Studies at USCF addressed this issue in an article published in the November 10, 1999 issue of JAMA. Rennie sited an example published in a 1997 issue of British Medical Journal. In that scenario, researchers look at 84 trials involving 11,980 patients to investigate the effective of a drug called ondansetron. Rennie said the researchers found that data from nine clinical trials had been published in 14 reports and that 17 percent of the data was duplicate.


It also turned out data on 28 percent of the total patient population evaluated in the meta-analyses had been reported twice. The biggest point, Rennie noted, is the duplicate data resulted in a 23 percent over-estimation of the efficacy of ondansetron.

"Any study that has strong pharmaceutical backing is highly likely to be defective," Rennie told UPI. "You get the data the drug company wants you to have."

Rennie said Schulman's report began just went ICMJE guidelines had been implemented. "I'm not surprised (by the findings) because this shows what an appalling state everyone was in. It takes years to changes things like this."

He added, "If the same study was conducted in five years time, I would hope there would be a huge difference."

(Reported by Katrina Woznicki, UPI Science News, in Washington)

Latest Headlines