Survey finds conflicts of interest in clinical drug trials

5/26/2005

Many U.S. medical schools are willing to give companies that sponsor studies of new drugs and treatments considerable control over the results, according to survey results that some doctors found troubling.

Half the schools said they would let pharmaceutical companies and makers of medical devices draft articles that appear in medical journals, and a quarter would allow them to supply the actual results. But academics draw the line at gag orders that keep researchers from publishing negative findings.

This is totally beyond reasonable practice. What you re seeing here is a willingness by some institutions to give more leeway than they should, said Dr. Harlan Krumholz, a Yale University cardiologist and epidemiologist who was not involved in the survey.

Private industry funds more than two-thirds of medical research at U.S. universities, a situation that has led increasingly to conflict-of-interest suspicions. Two decades ago, the federal government was the main benefactor.

The study, led by Michelle Mello of the Harvard School of Public Health, appears in today s New England Journal of Medicine.

Harvard researchers sent surveys to the nation s 122 accredited medical schools to gauge what kinds of standards exist between researchers and sponsors. All but 15 responded.

Dr. Doug Wilkerson, associate vice president of research at the Medical University of Ohio in Toledo, thinks MUO took part in the survey, though he wasn t sure because his institution has taken part in several similar studies.

Dr. Wilkerson said he s surprised that many medical schools apparently allow so much control by drug companies, saying the types of behaviors described by the researchers are not allowed at MUO.

We certainly try to do what s right, he said.

Dr. Wilkerson added that though drug companies have a large role in research funding nationwide, less than 10 percent of research at MUO is funded by drug companies.

The researchers did not directly establish exactly how much control universities actually give to companies. But the medical schools overwhelmingly agreed that they would not enter into contracts that would allow companies to edit research articles or suppress negative results. The schools were split on other issues. Fifty percent would allow companies to draft research papers, while nearly 25 percent would let them provide the data.

These results are really bothersome, said Dr. Jerome Kassirer, former editor-in-chief of the journal and author of a recent book about conflicts of interest in research. Some investigators may be willing to accept constraints just to maintain good relations with the company, said Dr. Kassirer, who had no role in the survey.

Dr. Wilkerson said dealing with drug company funding can be complicated.

Contracting for clinical trials is a dance, he said. We always start with an agreement provided by the company, that s just the name of the game, and it s always very one-sided, so we dance back to the middle. One issue not mentioned [in the journal] is who pays for treatment of any subjects injured. We always insist the company does.

The Pharmaceutical Research and Manufacturers of America, a trade group, insists that corporate sponsors do not interfere with researchers independence. The group publishes voluntary research guidelines stating that companies will sometimes help analyze and interpret results and have the right to review articles before publication. The guidelines also note that sponsors own the data and have sole discretion over who has access to the information.

Recent controversies involving companies accused of suppressing unfavorable results have led to demands for more public disclosure of industry-sponsored research. Drug manufacturers GlaxoSmithKline and Merck were recently accused of hiding information about the antidepressant Paxil and the painkiller Vioxx, respectively.

Dr. Wilkerson said MUO would never agree to a gag order on research.

For a company to say, If you find some negative findings at MUO, you re not allowed to publish them, we just don t go there, he said.