No one knows what “management” means in practical terms. The Institute of Medicine report on conflict of interest says that managing COIs should involve . judicious practical judgment sensitive to the institutional context.” In other words, managing COIs should be a “slow-thinking” (Kahnemann 2011) process—balancing costs and benefits, considering relevant differences. But real-world COI management has tended to favor “fast-thinking,” zero-tolerance, categorical, heavy-handed administration, i.e., “banning and prohibiting.” That’s what “taking COI seriously” typically means.
When “slow-thinking” nuanced management review of an actual or proposed commercial relationship is undertaken, it would tend to be well, slow, and responsive to contextualized relevant differences. But by avoiding blanket judgments in favor of exceptions, the process risks perceptions of favoritism and arbitrary indulgences. Delayed, slow arrival at assessments of COI imposes a time-tax on relationship proposals that discourage collaboration. Slow thinking does not weigh risks symmetrically. Scandal risk (headline liability) is weighed more heavily than the risks of foregone benefit. The latter’s costs are uncalculated and invisible and will not make news in the morning paper. On the other hand, no administrator will be fired for nixing a proposed collaboration.
A policy of avoidance of “relevant” COIs creates a presumption, a default against relationships with medical product companies. This on the assumption that such relationships are of no value, or are so shot through with corruption that the outcome of the relationships must be tainted. This thinking is exemplified in the DeAngelis-Fontanarosa Rule (J Am Med Assoc DeAngelis and Fontanarosa 2008) that imposed additional “independent” statistical review-costs on author submissions reporting research sponsored by medical product companies. This COI management policy was anecdote based, not evidence based.
A study (Woolley et al. 2011) of manuscripts withdrawn for research misconduct (e.g., falsification, fabrication, and plagiarism) certainly did not confirm JAMA's discriminatory policy. On the contrary, “of the 463 retracted publications retrieved, 213 (46%) were retracted because of misconduct. Publications retracted because of misconduct rarely involved declared medical writers (3/213; 1.4%) or declared pharmaceutical industry support (8/213; 3.8%); no misconduct retractions involved both declared medical writers and the industry.” Vera-Badillo et al. (2013) have observed: “The pharmaceutical industry is increasingly influential in clinical trial sponsorship with data showing an increase industry sponsorship of phase III RCTs from 24 to 72% over a 30 year period. In our cohort of trials, 67% were industry sponsored, but we found no association between industry sponsorship and biased reporting of either efficacy or toxicity, and no association of for-profit sponsorship with change of the PE [primary endpoint] between that listed in trial registries and the final publication” (1241).
Full disclosure, another nostrum, is problematic. Cain et al. (2005), in a laboratory guessing-game experiment played by a sample of undergraduate students, found that disclosure of misaligned incentives, where there is no ongoing relationship to encourage reliance, was associated with somewhat reduced reliance on conveyed information but also even more reduced reliability of that information. Koch and Schmidt (2010), however, found that established relationships encouraged reliance, also greater reliability and more truthful information. Ben-Shahar and Schneider (2011) found little empirical support for disclosure efficacy across a wide range of commercial and healthcare settings. No one reads boilerplate disclosures nor knows what to make of the information. Is physician’s full disclosure of his commercial/ noncommercial interests a “trigger warning” to patients that they should take precautions, or is it a boast of his having noteworthy relationships? Which relationships really matter? Won’t the discloser’s characterizing the importance of his own relationship risks make him a “judge in his own case”? Isn’t the whole point of asking for the disclosure some evidence that we can’t trust the discloser to tell the whole truth or to give an unbiased interpretation? Are the most current relationships most important? What about favorably remembered ones from long ago? What about deeply felt indebtedness to mentors and colleagues from residency or a fellowship? What about distinguishing hated, hectoring, unreasonable but well-paying commercial sponsors from those lower-paying but fondly remembered? Is a current high- dollar grant unduly influential? Suppose the doctor sincerely regrets having gotten himself into it. Disclosures and omitted disclosures tell us none of these things.
“Nothing to declare” is also non-illuminating. Ideological, religious, ethnic, and personal biases are not disclosed. A doctor’s commitments to her various value- causes and her tolerance for or distaste of commercial behavior, matters to her judgment. Disclosure of “significant” or “relevant” COIs in the recent past is also wanting. Examples of “significant” are unhelpful because no one knows what they mean in an individual case. The IOM insists that even “small gifts” of trivial value may (may!) influence a physician’s prescribing behavior. By implication, any commercial influence counts as “undue influence.”