Expiration dates for conflicts of interest?

“The past is never dead. It is not even past…” wrote William Faulkner. Should something similar be said of conflicts of interest?

While this blog has addressed future COIs it has never previously done so with past ones. The latter was suggested to me by a recent posting in MedPage Today by Milton Packer MD, which posed the question: “Does a financial conflict of interest ever expire?” Doctor Packer – writing about COIs in the medical research realm – noted: “All organizations that worry about conflicts of interest have a ‘sunset’ provision. It is the identification of [a] date before which the influence of a prior relationship is deemed to be irrelevant. You can argue about whether it should be 1, 3, 5 or 20 years. But at some point in time, the influence of that relationship becomes negligible.”

However, formal sunset provisions of this sort do not necessarily exist in all COI management regimes. For instance, it would be rare to find one in a corporate code of conduct, although presumably organizations without such provisions would take the time factor into account in applying more general COI standards in their respective codes. The same might be the case regarding various professional services and other ethical standards.

So, what criteria should those handling conflicts of interest – either in drafting or applying COI policies – consider in determining whether a given COI is really “past”?

First, one should assess whether the COI at issue is based purely on the economics of the relationship or if “substance” comes into play. As a general matter, the logic of having an expiration date for a COI of the former sort seems sound, since the impact of receiving such a benefit would indeed tend to diminish over time. By contrast, where the COI is more qualitative – meaning based more on the substance of such work– then its influence is less likely to be negligible, particularly if the prior work is related to the contemplated opportunity.

Second, size matters. The larger the financial benefit in question, the further back one may need to go to reach a point where its influence is negligible.

Third, appearance matters. As a general matter, some types of COIs will seem more worrisome than others – particularly when they are difficult to evaluate by key constituencies.

Fourth, one should consider in these deliberations – as Doctor Packer’s post does – the implications of a given sunset provision vis a vis recruiting the most able individuals for the task at hand. I.e., the maximum ethical approach does not always yield the best results. While this consideration is of perhaps of most obvious relevance in designing or applying medical research COI regimes, it can come up in other contexts too.

Fifth, I’ve lumped a lot of things together in this short post, but want to emphasize that whether a COI should be deemed to be in the past may be a narrower test than what needs to be disclosed in the first instance. This distinction may be necessary to ensure that the party with the putatively past COI is in fact applying the applicable expiration date appropriately.