In capitalist and in communist societies, industry has a responsibility to conduct research on health and safety concerns. For corporate research to be credible, it should be methodologically sound, transparent, and available. So should non-corporate research.
In the United States and in Europe, much important research is done only by private corporate sponsors. Of course, private funding of research raises questions about potential conflicts of interest (COIs), but political frenzy over such COIs is a serious diversion often motivated by a desire to live in a faith-based world in which industry and chemicals are demonized far beyond what even precautionary principles would support. Susan Sarandon’s superstitions about the herbicide Round Up come to mind.
Although members of the Lobby, the Litigation Industry, and the environmental groups of the less rational kind frequently find their knickers in a knot over corporate scientific COIs, the fact is that publicly funded and self-styled “public interest” research is often afflicted by non-financial COIs that are more mind numbing than the anticipation of money.[1] Some groups, such as the Society of Toxicology, have implemented more complete definitions of COI to include advocacy and positional conflicts.[2]
Joseph Huggard recently posted an interesting piece at Innovative Science Solutions’ blog, on the need to “Follow the Science Not the Money,” to remind us of the first principle, that research should be evaluated primarily on its merits, and not on its perceived or imagined COIs. Huggard likens the current situation of proliferating ad hominem attacks to less talented footballer who approaches the game thinking “If you can’t play the ball, play the man.” (Or if you are Donald Trump, then play the ref.)
Huggard cites an interesting meta-observational study in which researchers attempted to obtain research protocols from epidemiologic studies on phthalate exposure. Not surprisingly, researchers who published studies that purported to find adverse associations involving phthalates were three times less likely to share their study protocols.[3] A request for study protocols is hardly an intrusive or difficult request to meet. Of course, there are “reasons,” such as researchers’ desire to privilege their methods when so-called positive studies will serve as stepping stones to funding for future studies, but future studies should be conditioned on making past protocols available, and the failure to share protocols generally is pretty dubious scientific behavior.
As grim as the situation has been in the United States, Huggard suggests that an upcoming Luxembourg Chamber of Commerce conference this week, on October 10th, will seek to redress the imbalance in European COI rhetoric by calling attention to the importance of non-financial conflicts and biases.[4] Let’s hope so, but the more likely outcome is that the Chamber of Commerce’s sponsorship will disqualify any conference recommendations among the “political scientists,” those who practice science to achieve political aims.
[1] Simon N. Young, “Bias in the research literature and conflict of interest: an issue for publishers, editors, reviewers and authors, and it is not just about the money,” 34 J. Psychiatry & Neurosci. 412 (2009) (positional conflicts, based upon prior beliefs, can create much more intractable bias than financial rewards). See also “Conflict Over Conflicts of Interest” (July 12, 2015); “Conflicts of Interest in Asbestos Studies – the Plaintiffs’ Double Standard” (Sept. 24, 2013); “Conflicted Public Interest Groups” (Nov 3, 2013).
[2] See, e.g., Society of Toxicology, Conflict of Interest, Bias and Advocacy: Definitions and Statements.
[3] Gerard M.H. Swaen, Miriam J.E. Urlings, and Maurice P. Zeegers, “Outcome reporting bias in observational epidemiology studies on phthalates,” 26 Ann. Epidemiol. 597E4 (2016)
[4] “Managing Bias and Conflict of Interest: Ensuring that Policy-Makers and Regulators Access the Best Quality Scientific Advice,” at the Chambre de Commerce Luxembourg, at 7, rue Alcide de Gasperi, Luxembourg (Kirchberg).
