The trusted source for
healthcare information and
Conflict of interest issues require more IRBs and institution attention in future
Duke University study highlights problem
In the best of all worlds, there wouldn’t be research conflicts of interest. Investigators would be focused on the intent of research — uncovering previously unknown information. They’d be able to publish findings, positive or negative, regardless of the funding source. Institutions wouldn’t accept money from sponsors who demand the final word on whether findings can be published, and IRBs could be well assured that every protocol arrives free and clear of undue political and financial pressure from outside sources.
However, the reality falls a little short of the ideal, which is why many IRBs have had to deal with conflict of interest issues in addition to their primary goals and duties.
"Conflicts of interest has been a problem in research since almost the beginning of time," says Robert Levine, MD, professor of medicine and co-chair of Yale University Interdisciplinary Program in Bioethics at Yale University in New Haven, CT. Levine worked in the 1970s with the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, which developed the Belmont Report. The publications of the National Commission formed the basis for the "Common Rule," the federal government’s Regulations for Protection of Human Research Subjects. "It’s a problem whenever human beings can better their position by doing some kind of work," he says.
"Researchers don’t want to distort the meaning of their work because their primary capital in the world of research is their good name, and they don’t want to be associated with anything that isn’t done as well or accurately as they possibly can do," Levine adds. "But there are a small number of people who are willing to distort information in the pursuit of some goal they think is more important."
Because the world doesn’t function perfectly, IRBs may have to become more involved in reducing conflicts of interest, says Jesse Goldner, JD, professor of law and IRB chairperson at Saint Louis (MO) University.
"The rule is that if an IRB says No’ to a study the university can’t go ahead and say, Yes, a study can be done.’ So the IRB is kind of the top dog," he says. "With that structure in mind, you have to have an IRB that would make a final decision about whether a conflict of interest exists."
Again, the better solution would be for research institutions to have separate conflict of interest committees that would send their determinations to the IRB, but that doesn’t seem to be happening as often as it should, Goldner notes.
"I talk to people around the country, and it’s clear that at some institutions there is very little management of conflict of interest in research projects," he says.
Sponsors typically contract with institution
A recent study would appear to confirm this observation. Investigators from Duke University in Durham, NC, surveyed medical schools about research and their agreements with industry sponsors of multicenter clinical trials and concluded that academic institutions routinely fail to adhere to research publication, trial design, and access to data guidelines issued by the International Committee of Medical Journal Editors (ICMJE).1
"We did the study because ICMJE guidelines really speak to the role of protecting academic freedom for investigators, and the truth is that contracts are between institutions and companies, and investigators aren’t party to contracts," explains Kevin A. Schulman, MD, lead author of the Duke study, and a professor of medicine at the Duke University School of Medicine.
"We wanted to see if sites use the standards of ICMJE," Schulman says. "The answer generally is No,’ and a lot of that is because institutions felt powerless to negotiate with a sponsor when the sponsor can go to a different site that has a different standard for research projects.
"What we need are standards that academic medical centers can all agree to," Schulman adds. "We can’t have competing standards of research ethics between private centers and academic centers."
The Association of American Medical Colleges (AAMC) of Washington, DC, issued in October 2002 its own recommendations for oversight of potential financial conflicts of interest in human subjects research. AAMC’s report states that its recommendations complement and further develop the earlier guidelines on conflicts of interest that were issued by the Association of American Universities of Washington, DC.
One example of how a conflict of interest can impact the integrity of clinical research is the case of Nancy Olivieri, a Canadian hematologist and clinical researcher at the Hospital for Sick Children in Toronto.2 Olivieri studied the drug deferiprone in a short-term, uncontrolled clinical trial of patients with thalassemia who had iron overload. The study was a collaboration with Apotex, a pharmaceutical company that produced the drug, and the trial was sponsored by a grant from Apotex to the Hospital for Sick Children.
Initial results, published in 1995, were encouraging, but results from a second prospective trial that compared deferiprone treatment with the standard drug deferoxamine, were not. Deferiprone treatment failed to reduce hepatic iron levels below baseline or increased them. The drug also appeared to increase hepatic fibrosis in some patients.
Olivieri reported these findings to the IRB, presented them at a scientific meeting, and submitted them for publication. Apotex reacted by stopping all of Olivieri’s clinical trials and threatened to sue her for violation of a confidentiality agreement she had signed prior to the first clinical trial. When she ignored the threat and submitted the study for publication, the company sued her for $20 million.
The result of the legal fight was that Olivieri was publicly and personally harassed and received no support from her hospital, which eventually referred her to a Canadian licensing board for research misconduct. The hospital had even publicized this action, despite policies to the contrary. Recently, the licensing board vindicated Olivieri on all counts.
While extreme, this case underscores a conflict of interest problem that should be brought to an IRB’s attention because this can impact information that should be included in informed consent materials, Schulman says.
Conflicts of interest impact informed consent
From an IRB’s perspective, the issue is what did the study participant consent to when participating in a research study with regard to helping to create general knowledge, Schulman asks.
"What we found in our study is that institutions are enrolling people in what they are calling research, but there’s no obligation to publish the study, so there’s no ability to fulfill part of the promise," he says.
"Human subjects research is not marketing, and we need to critically evaluate the reason we’re recruiting patients to participate in these activities, and that’s an area that’s just missing," Schulman adds.
At issue is whether people who participate in clinical trials chiefly in hopes of improving treatments for future generations would be as interested in participating if they were told that there is a possibility the research will not be published.
This is why IRBs need to ask these tough questions of investigators about publication policies and require all conflicts of interest to be clearly described on informed consent forms, Goldner says.
"People ought to be told that the university is receiving financial support from XYZ drug company to conduct this study," Goldner says. "Then the question is whether disclosure is enough because in some circumstances disclosure is not enough."
While research conflict of interest problems date back centuries, the present research environment involves such high personal stakes that the financial pressure on investigators has increased, Levine says.
"Back in the 1960s, you couldn’t become an instant millionaire by discovering something in the lab, but now if you come up with the right recombinant molecule, you could," he says.
"You go out and meet with a venture capitalist, who puts up a few million dollars, and then you take the company public and the initial public offering brings in millions of dollars, and everyone lives happily ever after — so this is an enormous problem."
Findings review requires skilled scientific eye
IRBs, conflict of interest committees, and editors of journals are vulnerable to overlooking research that has been distorted because of financial conflicts because investigators can easily alter or withhold data from their reports. It is not easy to catch such investigators, Levine says.
"It is much easier to fake reports that look excellent than it is to do the research needed to generate valid reports," he says.
While scientists and researchers traditionally have served on IRBs, some of the recent federal changes and regulatory burdens have discouraged some of the best and brightest from serving on IRBs and that, in turn, could undermine the ability of an IRB to catch research fraud and abuse, Levine says.
A partial solution would be to make certain that all IRBs include in their membership some very skilled and experienced researchers and scientists, who are better able to spot discrepancies in protocols and reports, he suggests.
"It may be possible to con the system, but it’s harder to con a scientist," Levine says.
1. Schulman KA, Seils DM, Timbie JW, et al. A national survey of provisions in clinical-trial agreements between medical schools and industry sponsors. N Engl J Med 2002; 347(17):1,335-1,341.
2. Nathan DG, Weatherall DJ. Sounding board: Academic freedom in clinical research. N Engl J Med 2002; 347(17):1,368-1,371.