Earlier this year, we wrote about how corporate funding of research is treated, asking if it deserves to be singled out in a way that other potential biases are not. A recent study may help provide an answer to that question.
Since it has been the subject of intense controversy, the Type 2 diabetes drug rosiglitazone (brand name Avandia) is a naturally fitting case to use when looking at the possibly biasing effects of corporate funding. Avandia’s status means that it has been the subject of numerous studies and opinion pieces, with varying views on its suitability as a drug. For this study, published online by the journal BMJ on March 18, researchers examined 202 published articles on the risk of heart attack associated with Avandia, including “guidelines, meta-analyses, reviews, clinical trials, letters, commentaries, and editorials.” They looked for potential conflicts of interest — that is, financial ties to Avandia’s manufacturer (GlaxoSmithKline), its chief competitors in this drug class (Takeda and Eli Lilly, manufacturers/marketers of Actos, a drug in the same class as Avandia), or another diabetes drug manufacturer — for each author within two years of the article’s publication, whether or not the conflict was stated in the article. They then had two independent analysts, unaware of any conflicts of interest, categorize each article as favorable, neutral, or unfavorable to Avandia.
The researchers found that authors with a favorable view of Avandia were more likely to have a financial link to a diabetes drug manufacturer, and even more likely to have a link to GlaxoSmithKline, than authors with an unfavorable view of Avandia. In addition, authors with an unfavorable view of Avandia tended to be “largely free of identifiable financial conflicts of interest.” This was true of articles published both before and after the US Food and Drug Administration (FDA) issued a special safety warning for Avandia and Actos in 2007. However, only half of the articles included in the study had a conflict of interest statement. The researchers found this rate “disappointing,” but noted that conflict disclosure rates seem to have improved over the last decade.
Given that ties to manufacturers of competing drugs were linked to more favorable views of Avandia, there appears to be more at work in this picture than a simple slant due to financial motivation. These results suggest, in fact, that links to drug companies might cause a more favorable view of drugs in general — or, conversely, that people with a favorable view toward drugs decide to work for drug companies. These scenarios cast doubt on the current system in which drug manufacturers hire researchers to carry out trials of their drugs. In a New York Times article on this study, the two scientists who analyzed articles for the study lamented that drug-company funding is currently the only way to advance research on new drugs, given the lack of funding available from federal agencies or other organizations with no direct financial ties.
What do you think — should medical publications tighten their rules for reporting of conflicts of interest? Or does this study show that the situation is more complicated than simple disclaimers would suggest? Is it possible that researchers with no ties to drug companies have their own special biases? Should funding for drug trials be transformed, possibly through expanded federal funding? Leave a comment below!