Industry funding of both internal and academic medical research skyrocketed between 1980 and 2000, from about $1.5 billion to $22.4 billion. Even as the new funding led to important new medical products and technologies, it provoked questions about the financial relationships between industry and the scientists conducting that research.

"It seems like we can't go a week these days without reading something in the paper about some form of conflict of interest in medicine or research," says Eric Campbell, an associate professor of medicine health policy at Harvard Medical School and Massachusetts General Hospital, both in Boston.

There is no inherent conflict of interest in working with industry, says Harry Greenberg, a professor of medicine at Stanford School of Medicine in Palo Alto, California. In fact, he believes, industry funding is vital for translating basic discoveries into clinical therapies. "Every academic institution that I am aware of encourages physician-scientists and other scientists to do industry-sponsored research," Greenberg says. "It's only when there is some outside financial arrangement, such as giving talks for the company or consulting, when questions arise. Some of those are manageable, some, perhaps, are not."


Courtesy, Eric Campbell
Eric Campbell

Conflict-of-interest guidelines generally apply to any researcher engaged in federal or academic research that's somehow tied to industry. But the issues are more acute for physician-scientists, for several reasons. For one thing, clinical trial data are open to interpretation and, potentially, individual discretion, so a lack of objectivity in analyzing those trials can have more dramatic scientific consequences. For another, conflict of interest is as much a product of public perception as it is inappropriate influence, and human health–related research can make physician-scientists more visible to the public than other scientists.

Under growing pressure, many medical schools and research institutions are reexamining their conflict-of-interest policies, fine-tuning disclosure procedures, and ensuring they're in line with the policies of federal funding agencies. How successful these policies will be in weeding out undue influence is unclear, but the recent changes almost certainly will put more pressure on researchers to both report their contacts with industry and justify any payments they receive.

In the public eye

Ongoing U.S. Senate investigations have put the spotlight on universities and scientists who have failed to disclose compensation from drug companies. For example, late this summer, the U.S. National Institutes of Health (NIH) suspended a $9 million grant for a depression study at Emory University in Atlanta, Georgia. The suspension was an apparent response to a Senate committee's finding that Charles Nemeroff, a professor of psychiatry and behavioral sciences at Emory, failed to report at least $1.2 million of income from drug and device companies (Science, 24 October, p. 513). Nemeroff stepped down as chair of the psychiatry department in early October and voluntarily removed himself from all NIH-funded projects. The same day the NIH grant suspension was announced, Emory released a statement announcing the creation of a new central office "to oversee administration and enforcement of conflict of interest policies."

Few early-career scientists and physician-scientists have the opportunity to refuse or accept millions in personal payments from industry. Yet, even younger scientists may face subtle potential conflicts. A report released this June by the Association of American Medical Colleges (AAMC) in Washington, D.C., which represents 129 medical schools in the United States, gives one example of such a potential conflict. "Our department chair encouraged us to attend an industry-sponsored retreat in a beach location where he was speaking," a junior faculty member told the AAMC working group that produced the report. "There was a competition to see who could go. It seemed prestigious for those who went." Polls show that scientists believe such interventions don't influence their decision-making. But even if this is true, industry-sponsored trips to exotic locations create at least a perception of conflict.

A perk as small as a nice meal can set a dangerous precedent, especially when it happens on a regular basis. "Our institution does not allow industry representatives on campus during our noon educational lectures and has a very limited fare of sandwiches. One day a week, however, the industry representatives would have a catered lunch at a hotel across the street from the hospital. The residents would take us over there for a better lunch," said another student interviewed for the report.

Even when junior scientists don't benefit directly from industry gifts, a department chair with a financial stake in a particular company can influence research within the department. "You're the junior faculty member who's doing the research that will determine whether the chairman is going to make a lot of money. And by the way, he or she is going to vote on whether you're going to be promoted. This probably happens in every single institution in America," Campbell says. "That's why [the research community has] developed, on the individual basis, a very elaborate system of disclosures of relationships and subsequent management of [problematic] relationships." He adds that the best thing junior researchers can do is make sure they are in compliance with national and institutional conflict-of-interest policies.

The AAMC task force report, which specifically looked at industry funding within the context of medical education, called for medical schools and teaching hospitals to adopt policies by next July that prohibit researchers from taking gifts, including meals, from industry and that prohibit company representatives from ghostwriting papers, reports, or presentations. The report did not, however, discourage physicians from attending company-sponsored educational forums or serving on industry advisory boards.

Pushing universities to adopt stronger policies


Courtesy, Gabriel Silverman
Gabriel Silverman

For their part, universities and medical schools are adopting policies that offer guidance to researchers and promote disclosure of real and apparent conflicts. "A year ago, there were only a handful of schools that had comprehensive conflict-of-interest policies, and now over 20 schools have very strong policies and about 30 more are developing them. That's a third of the 150 medical schools," says Gabriel Silverman, an M.D./Ph.D. student at the University of Pittsburgh School of Medicine in Pennsylvania.

Silverman headed this year's edition of the American Medical Student Association's annual assessment of conflict-of-interest policies, called the PharmFree Scorecard 2008. The group found that the majority of schools have either inadequate or nonexistent policies. "What we took away from this year's scorecard is that most U.S. medical schools are failing to address aggressive industry marketing in medicine at this moment in time," Silverman says.

The students examined the policies of 151 medical schools and schools of osteopathy across a broad spectrum of potential sources of conflict, including acceptance of gifts and meals, consulting and speaking relationships, pharmaceutical samples, the participation of individuals with financial conflicts in purchasing decisions, industry funding of educational events, support for scholarships and training, and the presence of sales reps on campus or in hospitals. They also rated the schools' policies in terms of disclosure requirements, oversight, and sanctions.

What's needed is a good balance of policies that enables scientists to work with industry while at the same time safeguarding against conflicts, Silverman says. "We recognize that industry relies on advice from physicians to improve their products and that medicine relies on industry to bring treatment innovations to market," he says.

Advice on (potential) conflicts of interest

Early-career researchers working with industry need to acknowledge the potential for bias. "What researchers need to be thinking about is, 'how can bias be introduced into the way I am evaluating my research?' " Greenberg says. He adds: When in doubt, ask.

Be aware of existing disclosure policies and compensation limits at your institution, funding agency, and journals that you regularly submit papers to. "Know first of all what is required to be disclosed and obey those rules even if no one at your institution necessarily cares or checks up on you," Campbell says.

It's also important to learn of any industry ties that primary investigators have before going to work for them. Ask questions about those relationships and find out whether any outside compensation influences the work they do. "If a senior faculty member has equity relationships with a company, will those companies be funding research in the lab?" asks Campbell. "How will they separate the products of the company from products of other companies? How will they divide up who works for what publications? How will they resolve conflicts? All of those are key questions about how a lab functions."

Images, top to bottom: Comstock Business Impacts. Courtesy, Eric Campbell. Courtesy, Gabriel Silverman

Joel B. Finkelstein writes from Washington, D.C.

Joel B. Finkelstein writes from Washington, D.C.
10.1126/science.caredit.a0800165