10/11/2011 06:43 EDT | Updated 12/11/2011 05:12 EST

Conflicts of interest remain common on expert panels struck to advise doctors

TORONTO - Expert medical panels struck to issue advice on the treatment of patients remain rife with financial conflicts of interest, a new study suggests.

The problem is more acute among panels struck by specialty medicine groups than government, and more pronounced — at least in this small series — in Canada than in the United States, the researchers said in the study, published Wednesday in the journal BMJ.

The study looked at 11 American and three Canadian expert panels convened to draft advice for doctors on the treatment of high cholesterol and diabetes, conditions which generate drug sales in the order of US$70 billion a year for pharmaceutical companies.

Some were struck by government groups such as the Canadian and U.S. task forces for preventive health care. Others were convened by specialty medical groups such as the American Diabetes Association and the Canadian Cardiovascular Society.

"We found that overall ... the majority of guideline panel members and half of guideline panel chairs in our study receive some form of compensation from industry. And that indicates a potential risk of industry influence on guideline recommendations," lead author Dr. Jennifer Neuman said from New York, where she works at Mount Sinai School of Medicine.

The problem isn't just theoretical, said Dr. Gordon Guyatt, who has been studying the issue for years.

"It's part of why we're paying more for drugs than we should be and it contributes to the overuse of drugs," said Guyatt, who teaches at McMaster University's school of medicine. Guyatt was not involved in Neuman's study.

For the work, Neuman and three colleagues went through conflict of interest declarations by members of the 14 panels. But they didn't just take panellists at their word.

If someone declared he or she had no financial conflicts, the researchers searched to see if the expert had published any recent studies, then checked to see if the work was funded by relevant drug companies. They also did Google searches to see if they could find evidence that the experts had been paid for giving talks for drug companies or had earned fees for similar types of services.

Of the total of 288 panel members, 48 per cent declared they had financial conflicts of interest and 52 per cent said they had none.

But Neuman's searches suggested the figures were actually reversed — 52 per cent of experts had financial conflicts, with 12 experts who signed statements saying they were conflict free actually having relevant financial ties to industry.

Neuman found those non-disclosures disappointing. "It's surprising that some people just didn't declare conflicts, (and) that they weren't picked up," she said.

Financial conflicts of interests are a tricky situation in medicine. Just because someone has been paid for services or has accepted research funding from industry doesn't mean the advice they give is going to be biased on behalf of a particular drug, drug maker or course of treatment.

In fact, some experts acknowledge that if one excluded all people with conflicts of interest from expert panels, the level of expertise on those groups would drop.

"One suggested remedy is that guideline committees should be composed of only those who have no conflicts of interest," Edwin Gale, an emeritus professor of diabetic medicine at Southmead Hospital in Bristol, England, noted in an editorial that accompanied the study.

"There is a charming sense of unreality about this suggestion — money from drug companies is the oxygen on which the academic medical world depends."

Gale suggested practitioners in academic medicine have to come to recognize there is a problem that needs fixing. "What is needed is a change of culture in which serving two masters becomes as socially unacceptable as smoking a cigarette."

Guyatt has suggested experts with financial ties to industry may be included as witnesses on panels, but should not be involved in the drafting of recommendations.

It's also an issue that has preoccupied the U.S. Institute of Medicine and the World Health Organization.

"The reason why the Institute of Medicine and ... the WHO have made recommendations on how to manage conflicts of interest on guideline panels is that it poses a potential risk of industry influence on the recommendations. And it's important to recognize that it's a potential risk," Neuman said.

Her work found higher rates of conflicts of interests among Canadian panels than American panels. One struck in 2009 by the Canadian Cardiovascular Society was comprised of 23 members. All had financial conflicts of interest, though three members had not declared them.

Another Canadian panel, struck by the Canadian Diabetes Society in 2008, had 78 per cent of members with conflicts of interest, all but one of which were declared.

Neuman said she could not explain the higher rates on the Canadian panels. She said some people have suggested it might be due to the fact that Canada has a smaller pool of experts to draw from, but she said there is no evidence with which to prove that theory.

She suggests the fact that government-appointed panels had fewer experts with conflicts suggest that in fact it is possible to limit conflicts of interests on these panels. "So it is possible to convene very expert panels with minimal conflicts. It just might take a little bit more work."