Monday, October 24, 2016

Are Conflicts of Interest In Nutrition Less Worrisome?

Today's guest post comes from public health ethicist Daniel Goldberg. It has to do with conflicts of interest in nutrition. I've found it fascinating to watch otherwise evidence based individuals suggest that conflicts of interest in nutrition don't matter. That somehow, unlike the risks proven to exist with conflicts in medicine - where highly paid physicians have been shown to be influenced by cheap lunches and plastic pens - that the suggestion that conflicts in nutrition might influence opinions, research, attitudes, and more, is ridiculous. It's not.
The Academy of Nutrition and Dietetics ("AND") recently held its annual meeting. According to registered dietitian Andy Bellatti (and others), during her opening talk, the President of the AND, Lucille Beseler, opined on the rising concerns over financial conflicts of interest among nutritionists and dietitians:
"I'm not so weak-minded that I would make a decision on receiving a pen."
In the course of researching, writing, and teaching about conflicts of interest among health professionals for over a decade, what I have come to marvel at the most is the apparent ease with which leading health professionals insist on a kind of willful ignorance regarding the cognitive science underlying concerns over COIs. This is amazing to me because of the irony: members of professions presumably dedicated to basing their practice on the best evidence proceed to engage concerns over COI with almost no awareness whatsoever of what the best evidence actually suggests regarding the impact of conflicts of interest on human behavior.

This is why, in December of 2015, I published a small chart with explanation in BMJ entitled "COI Bingo." I grew so exasperated with the same tired justifications for financial COIs that I categorized the standard responses into a Bingo chart (up above)

Reasonable people of good conscience can, of course, disagree on whether the behavior of partiality that occurs in the presence of COIs are morally justified, and on the appropriate remedies, if any, for such behavior. But the argument should proceed with all stakeholders fully aware of what the cognitive science actually suggests regarding the impact of COIs on health professional behavior.

What does that evidence show? Beyond a shadow of a doubt, gifts almost certainly do influence health professionals' behavior, at least in the aggregate. Not only have we documented this finding itself ad nauseum, we also have powerful causal explanations that elucidate the mechanisms by which even gifts of de minimis value influence health professional behavior. Virtually all human societies exchange gifts -- they promote social cohesion and are therefore a critical adaptive mechanism. One of the ways gifts accomplish such cohesion is because they tend to automatically, unconsciously create a desire to reciprocate on the part of the recipient.

Commercial industries are well-aware of this phenomenon, which is why they provide such gifts. The gift exchange also cements the relationship with industry, which that in which commercial industry is most interested. The tighter the relationship that exists between commercial industries and health professionals, the more likely it is in the broad run of cases that behavior of partiality will occur. COIs have to be understood iteratively -- the existence of a financial COI does not imply that bad behavior will necessarily take place in any given case. But over the long run of cases, the existence of financial COIs makes shenanigans much more likely -- a conclusion which is -- again -- extremely well-documented in both experimental and uncontrolled (i.e., real-life) conditions.

That these kinds of gifts "work" in the health professions to serve the interests of the "donor" is therefore beyond dispute. Moreover, some of the more darkly amusing findings in the COI literature document our own immunity bias: while we imagine ourselves much less likely to be influenced by pens and mugs, we have serious concerns that the professional sitting next to us may have their judgment clouded if they accept gifts from commercial industry:

Steinman, Shlipak & McPhee 2001
Maybe any given health provider will indeed remain entirely uninfluenced by deep entanglements with commercial industry. But the evidence establishes beyond all doubt that the odds are forever not in your favor.

Ultimately, far too many stakeholders seem willing to wade into the fray with a perfect, almost studied indifference to the significant evidence base regarding COIs. This is itself an ethical problem -- mistakes themselves are not ipso facto morally blameworthy -- but mistakes made because health professionals did not bother to examine an available evidence base and ground their practice in that evidence come much closer to moral failure.

Daniel S. Goldberg is a faculty member in the Center for Bioethics & Humanities at the University of Colorado Anschutz Medical Campus. He is trained as an attorney, a historian, and a public health ethicist.