On the PharmaLot blog, Ed Silverman posted about a new study that purported to show that research subjects don't care about conflicts of interest affecting study investigators. But Silverman also noted that the apparent conflicts of interest of one of the authors, a lawyer for a firm that represents many pharmaceutical companies, were not disclosed, at least in the press release about the article.
Furthermore, Silverman's post suggested that the college students who were the study subjects were only asked about studies in which "the investigator was an employee of the company, ... an employee and consultant, or ... an employee and a patent holder." They were not asked about investigators who had conflicts at all. Thus, at best, the article suggested that college students thinking about being subjects of clinical studies were indifferent about whether study investigators employed by a company thad made the product being tested on the subjects were also company consultants or stock-holders. Maybe the students reasonably assumed that an investigator employed by the company was already so conflicted that adding consulting fees or royalties would merely ice the cake. That's hardly the same as the author's assertion that "this study suggests that disclosure of conflicts may not play a large role in decision-making." One wonders if the study author's willingness to make assertions not supported by his own data had to do with his own conflicts of interest?
Further, I'd add that the problems with conflicts of interest -- that is, the ways in which they influence human behavior -- are not in any way dependent of the perceptions that human subjects have of those effects.
ReplyDeleteSo while I agree that understanding what subjects think and feel about COI is independently important, it certainly does not imply that the litany of problems that flow from the extensive COIs are any less significant.
--Daniel Goldberg
www.medhumanities.org