Monday, September 28, 2009

Why medical researchers often fail to disclose conflicts of interest

In response to the recent storm of publicity over the failure of many medical researchers to disclose the lucrative personal payments they get from the drug and medical device companies that also fund their research, many medical journals have developed written policies requiring disclosure of such conflicts of interest. That's the good news. The bad news is that many of these new policies lack the kind of scope and specificity that would really make a difference.

That at least is the conclusion of a rigorous new study, published online in the journal Accountability in Research, which found that more than 75 percent of the 227 medical and toxicology journals studied had minimal levels of specificity in their written conflict of interest policies. For example, most journals do not specify what they mean by a conflict of interest, so it's up to the researcher-author to decide if, say, the consulting gig they have with a drug company that is also funding their research, poses a conflict of interest or not.

"People have very different ideas about what constitutes a conflict of interest and what doesn't, so that gives authors a lot of latitude not to disclose," said Tufts University professor Sheldon Krimsky, who is the lead author of the new study.

Likewise, most of the journals did not specify what type of content researchers need to disclose conflicts of interest for. Whether, for example, they are required to disclose when they were submitting an original article, or a review or commentary article. And very few journals provided a monetary threshold for when disclosure is a must.

As a result of such lack of specificity, many researchers make their own decisions about when and what they should disclose. Such findings may explain why previous studies have found big gaps and inconsistencies in who discloses what to medical journals.

"This explains why many researchers disregard the requirement and simply don't report conflicts of interest," Krimsky said.

As a typical example, Krimsky cites the written COI policy of the International Journal of Cancer Research, which merely states: "Authors must include financial support received for research. Acknowledgements for financial support must be stated." The policy says nothing about consulting or speaking gigs. Nor does it specify what kind of financial support it's referring to or for what type of submissions the authors should disclose this support.

The solution? Journals have got to got more specific in what they mean by conflicts of interest and when authors should disclose these conflicts. As Krimsky notes, "Journals have to take responsibility for what is not being reported and should be reported."

The American public (who funds much of this research) deserves no less.


Bernard Carroll said...

The current approach of having authors just list the corporations for which they consult or speak is not adequate… it permits what I call hiding in plain sight.

Recently, a review article in the American Journal of Psychiatry talked up antipsychotic drugs for depression. The author provided all the pro forma disclosures. But he did not get around to telling readers or the editor that he had appeared in a multi-city road show pushing these drugs for that indication – with his honoraria from a drug company being laundered through a medical education communications company (MECC). Nor did he reveal that his name appeared, along with the names of several other well known key opinion leaders, in the authorship line of a commissioned article prepared by a professional medical communications company for a journal targeting primary care physicians. The same drug company paid for the preparation of this article.

That’s not disclosure. The problem is compounded when, as in this case, the review article is biased by talking up weak efficacy signals while glossing over toxicity problems and concerns. That is how the medical literature gets corrupted.

Evelyn Pringle said...

I agree with Dr Carroll that merely listing drug companies is meaningless.

Although the profiteering shrinks avoided conflict of interest disclosure for years, now they simply list every drug maker on the planet with no shame, without revealing the actual amounts they are raking in from each drug maker.

"Full" disclosure is a joke and has led to carrying on the corruption of medical literature in a business as usual fashion.

The by now well-known gang of suspects are addicted to greed and appear to be unstoppable.

Evelyn Pringle