The trusted source for
healthcare information and
Study finds conflicts of interest pollute science
Widespread financial ties permeate field
With nearly two-thirds of biomedical research funded by industry, it is perhaps not surprising that there are financial conflicts of interest between companies, investigators, and universities. A study published in the Journal of the American Medical Association in January quantifies it and discusses its impact.1 The results: There is more than we would like, with more influence than is good.
In some respects, the influx of private money into biomedical research has been great, note the authors — there have been medical advances and impetus for further research. However, many studies, they say, have indicated that close ties between scientists and industry may compromise the integrity of research.
The article included 37 studies that looked at the extent and impact of financial relationships. Eleven of them determined that industry-sponsored research tends to yield pro-industry conclusions.
The lead author, Justin Bekelman, MD, an intern at Johns Hopkins Hospital in Baltimore, says that there could be many reasons for this. "Some suggest that publication bias in the literature — positive studies are published more often — contributes to this finding."
Part of the publication issue is that pharmaceutical companies that experience negative results don’t send in those studies for publication, says the former editor of the New England Journal of Medicine, Jerome Kassirer, MD, a distinguished professor at the Tufts University School of Medicine and an adjunct professor at Yale’s medical school. But the critical question with any study sent for publication isn’t whether it’s positive or negative, but whether it will make a difference.
"If a study says that the treatment with a new and unused drug doesn’t work, why publish it?" he asks. "What you care about is whether a study changes practice." Recent studies on hormone replacement therapy can certainly be construed as negative, yet because they changed the practice of medicine, Kassirer says, there was no question that they would be published.
But positive results can’t only be the result of publication bias. "We have found evidence of a funding effect," says Sheldon Krimsky, PhD, a professor in environmental and urban planning and policy at Tufts University who has written extensively on financial conflicts of interest in science. "That means that on a population scale, not on a one-to-one determinative scale, we see that privately funded research tends toward the values and interests of the funder."
Take any group of studies on the same topic, and as a group, those that are funded privately would be more positive toward the people providing the money than those that don’t rely on that money.
Bekelman also notes the impact of industry on trial design, and four of the studies analyzed in his article look at that topic. "We found that industry tends to sponsor trial designs that favor pro-industry results," he says. For example, in one analysis of multiple myeloma trials, industry-sponsored studies were substantially more likely to use placebo controls than were nonindustry-sponsored studies. "The authors of this study also found that the use of placebo controls increased the likelihood of positive study results. One of the fundamental ethical principles of medical research is that there needs to be uncertainty about which treatment is better. Otherwise, it’s not fair to ask patients to volunteer for studies if you already know that one treatment is inferior."
Another issue raised in Bekelman’s article involves financial conflicts of interest between researchers and sponsoring companies. One of the studies looked at how institutions dealt with financial conflicts of interest and found that among 250 institutions, management of conflicts and penalties for nondisclosure were almost universally discretionary. Only one of 10 research-oriented medical schools prohibited investigators from having equity, consulting agreements, or decision-making positions in a company sponsoring their research.
In 2001, a series of articles in the Seattle Times about the Fred Hutchinson Cancer Research Center (the Hutch) in Seattle pointed to some serious conflicts of interest in several patient protocols that may or may not have clouded the judgment of researchers and harmed patients. But the Hutch was hardly alone.
"It would be very rare now for a researcher, particularly the most senior of them, not to have their foot in the corporate door," says John Pesando, MD, PhD, a medical consultant and former researcher at the Hutch who was one of the whistle-blowers in the case. "It is a very large and weakly controlled issue."
In part due to the investigational reports in the media, Fred Hutchinson executives made some changes to the center’s financial disclosure policies. It now prohibits ownership of stock or receipt of royalty payments on patents by a researcher directly and significantly related to a clinical trial in which a researcher is involved. Scientists involved in the conduct of human subjects’ trials are also now required to disclose to patients and in scientific publications any financial interest that they have in the for-profit company sponsoring the trial in which they are involved.
Purity of science is paramount
Medical research is a matter of life and death, says Bekelman. "The guidance and medicine that patients receive from their doctors relies upon valid scientific evidence," he notes. "While industry sponsorship of medical research has led to considerable scientific progress over the past few decades, it is imperative to ensure that the integrity of the scientific process and the safety of research participants is never jeopardized."
The funny thing, Bekelman says, is that pharmaceutical companies and medical device manufacturers have a greater interest in valid scientific research than positive findings.
Kassirer says he would like to see clinical trials staff simply be more aware of the financial conflicts that are there. "Who is a consultant for the company? Who is on their speakers’ bureau? Who has stock? It’s not that it has to impact whether the science is good, but we have no idea."
In an ideal world, no one with a financial interest in the sponsoring company would do any medical research, says Kassirer, who just finished writing a book on the subject of industry’s infiltration into medicine. The potential for a bias is simply too great. "But as a start, I would like to see more disclosures of the relationship between researchers and companies — the amount and type of involvement."
Academic and industry collaboration has been fundamental to many advances in medical care, says Bekelman. "Our findings demonstrate that such collaboration is extensive and may have an impact on trial outcome. The question is not whether to prohibit collaboration but instead how to manage collaboration appropriately. Researchers need to be cognizant of the fact that industry sponsorship is associated with study bias, publication delay, and data withholding. Then they should design their collaborations in such ways as to minimize these unintended consequences."
A lot of people have interesting ideas on how to do that, he continues, citing Harvard researchers as being "pretty close to getting it."
Bekelman thinks part of the answer is that guidelines must be adopted and practiced by medical journal editors and medial professional societies. "We also suggest a comprehensive clinical trials registry so that the results of all trials will be available with any financial interests of the investigators."
Something has to change, says Krimsky. "Imagine if you are in a courtroom and a judge, before meting out a sentence on a felon, stands up and says, Before I issue this sentence, I want to disclose that I will be sending this person for the rest of his life to a for-profit prison in which I have a financial interest.’ Most people would cringe at the idea, that the judge, who has disclosed his interest, has this interest in the first place. But we have become acclimated to the fact that scientists make these disclosures all the time. That’s just the way things are."
Some value systems — such as that of universities and industry — just shouldn’t be blended, Krimsky says. Perhaps, over time, they will become less enmeshed as the policies Bekelman mentioned at journals and in institutions are created.
But a policy or rule won’t be enough, says Pesando. "We have some great rules out there that are designed to protect patients. But you have to have enforcement. Otherwise, it’s like having speed limits but no police to enforce them. You won’t control anyone that way."
1. Bekelman JE, Li Y, Gross CP. Scope and impact of financial conflicts of interest in biomedical research: A systematic review. JAMA 2003; 289:454-465.