Daniel S Goldberg: COI bingo

goldberg_danielI am fortunate to possess content expertise on ethics and conflicts-of-interest (“COI”) in medicine and science, and I teach it in a variety of health professional settings. After reading the latest news story on a commercial industry’s partnership with academic scientists, I grew tired enough of seeing the same (to my mind) poor rationalizations for the permissibility of such an arrangement to create this handy COI bingo chart.

coi_bingoThis chart is obviously crude and over-simplified, but as I explain in a post on my personal blog, what bothers me is not the legitimate disputes on the ethical permissibility of COIs in general. That is, reasonable people of good conscience can, to my mind, disagree on when and whether COIs pass ethical muster. What I find most frustrating is the extent to which leading physicians and scientists whose profession seems to require a commitment to some kind of evidence based practice are unaware of the best evidence on motivated bias. This literature is robust and well developed. And the comments that I often see and hear (which are also well-represented in the literature on COIs) indicate to me that the speakers are almost totally unaware of what the evidence on motivated bias actually demonstrates.

Among other well settled findings, the literature shows that relationships between commercial industry and physicians or scientists are extremely likely to influence physician/scientist behavior in a variety of ways. The claim that various barriers to such influence—i.e. individual virtue, institutional management, disclosure—are sufficient to prevent such influence is simply not an evidence based view. Human behavior changes depending on the relationships we maintain, and the extent of those changes tends to increase with the depth of the relationship (which explains why we tend to treat family members differently from perfect strangers). Financial exchanges are of course one indicator of the depth of a relationship, although they are not the only such marker.

If the physicians and scientists implicated in any particular COI were better aware of the literature on motivated bias, I find it difficult to believe that they would continue to rely on the same types of rationalizations for deep financial entanglements with commercial industry. It is not credible to claim that a clinician’s virtue is adequate to protect against influence on prescribing practices. It is not credible to claim that a scientist’s disclosure of relationships with commercial industry is adequate to prevent manipulation of the knowledge produced from a study. The evidence on motivated bias belies these defenses and many other standard claims reflected in the COI bingo chart.

Where physicians and scientists choose to enter into relationships with commercial entities, they should do so with the same commitment to evidence based practice that they honor in their clinical and scientific work. That commitment requires an admission that the relationship is in fact very likely to affect their clinical practice and/or the scientific study or inquiry being performed. As noted above, such an admission does not necessarily imply ethical impropriety. One might argue for a variety of reasons that even given the well documented finding that relationships with commercial industry are extremely likely to change the physician/scientist’s behavior, such relationships are nonetheless morally justified. But the factual premise of such claims ought to be grounded in what the best evidence on motivated bias actually shows. The COI bingo chart I made is intended as satire of attempts to justify such relationships ethically that remain, in my view, uninformed on the empirical basis for many of the ethical criticisms of such relationships.

Ultimately, if we are going to disagree on the ethical propriety of COIs in medicine and science, at a minimum we ought to agree to inform ourselves on what the best evidence shows regarding the impact of such relationships.

Daniel S Goldberg is an attorney, an historian of medicine, and a bioethicist. He is an assistant professor in the Department of Bioethics & Interdisciplinary Studies in the Brody School of Medicine, East Carolina University, USA.

Competing interests: None declared.

  • jtom

    I wish you would expand your analysis to include government research. I find it amazing that people believe that government scientists would not be biased by the positions taken by the politicians and policy makers who control their budgets, promotions, raises, and areas of research to be pursued.

  • Jeff Behrens

    Dr. Goldberg,

    You argue that COI’s exist as an inevitable consequence of industry/academic relationships. It seems to me that Industrial/financial COI’s could perhaps be viewed as an instance of a whole host of COI’s that researchers must confront throughout their careers. Could we define COIs as occurring any time any interest has a potential to conflict with pure, objective, platonic scientific judgement? These interests could include career interests (the need to publish), personal financial interests (the need to get, and keep a job), the need to get grants and other funding, personal interests (caring about the careers of students), romantic interests (appropriate or not), intellectual interests (pride in self-generated hypotheses and theories) and so many others.

    I was led to your post from the Retraction Watch blog, where the results of some of these other types of COIs are all-too evident. It seems to me to be a fairly uninteresting argument to state that industrial relationships lead to COI and possible bias since I believe that all relationships (and the very state of being human) inevitably leads to bias. I think it might be more interesting to try to somehow measure and compare the relative strengths of these biases – and the impact of their negative consequences. If COIs from industry are much stronger, and lead to measurably worse science (perhaps measured by retractions?) than COIs from the need to publish to obtain tenure, this might be a notable point. Conversely, if we find that COIs from industry and their consequences are demonstrably weaker than other types of COI then perhaps the concern – and policy efforts to address – are misplaced?

  • mlh

    What does C.R.E.A.M. in the free space stand for?

  • JDobson

    C.R.E.A.M is a reference to Wu-Tang Clan’s song: Cash Rules Everything Around Me

  • mlh

    Ah-ha! Thanks for the clarification.

  • First Officer

    Are any of us truly without COI’s then? It has been suggested that if you can’t call a donor the devil anytime you like, you have a COI.

    https://twitter.com/amy_harmon/status/640277179985739776

  • Tids1960

    All I need is the lower right corner
    and I’ve got a bingo