Friday, October 28, 2011

COI Update from the ASLME--Capsules

I'm in Pittsburgh just having attended the conference, "Conflicts of Interest in the Practice of Medicine," sponsored by the American Society of Law, Medicine and Ethics, which you can read all about at:
ASLME also plans to publish all the papers from the conference in its journal.

So what's new in the field? The quick answer for regular readers of this blog is, apparently not much. There were few new issues and virtually no possible solutions raised that have not already been discussed here and elsewhere. The presentations were nevertheless interesting and the interplay between the legal and medical viewpoints was illuminating. I have listed a few capsules below of possible interest and will do at least one more post on a particular issue.

>>The conference began with the usual ceremony of the highly placed official welcoming everyone, and in this case it was the Dean of the Pitt School of Law. She began by saying that COI is "almost impossible to eliminate" and that the conference would address the "challenge of managing" COI. At that point I was getting worried that we were going to hear a rehash of really old ideas. Fortunately the subsequent sessions all paid appropriate attention to the desirability of eliminating and not merely managing COI.

>>Christopher Robertson, JD, PhD of U-Arizona law school was given the task if laying out the evidence for the seriousness of COI in medicine. He began by reviewing the evidence relating to physician self-referral (e.g., sending patients to the imaging center that the physicians' group owns instead of to get their scans at the local hospital x-ray dept.) and said what we all know, that evidence shows that the rate of ordering tests and procedures jumps astronomically with self-referral. I had generally not thought to connect the self-referral data to issues of COI at the medicine-Pharma interface. But I think Robertson raised a good point. Physicians commonly deny that they overorder tests and insist that when they send patients to the testing center they profit from, that either the patient benefits greatly from the test, or they have saved the patient a trip across town or to another city, etc. So we have clear evidence that 1) money changes physician behavior (duh) and that 2) physicians commonly rationalize that association away. I think one can reasonably argue that such is presumptively relevant to other money-laden relationships, until proven otherwise.

>>Bernard Lo MD from UCSF, a chair of the IOM panel that wrote their report on COI in medicine, was first to raise this issue but it was echoed by other speakers. As we enter the era of heightened, required disclosure, many physicians fear the liability of multiple, perhaps inconsistent disclosures, as different forms and bodies have different rules (such as whether to report relationships for the past 2 years or 5 years or whatever), and then providing fodder to investigative journalists who can check out all these disclosures on line and play "gotcha" with any inconsistencies that are revealed. Now the extreme pharmascold might say, so much the better, yet another good reason to divest oneself from these conflicted relationships. But the point the speakers made seemed very reasonable, that it would be far superior to have a single, uniform and common disclosure process so that each individual had to disclose in one place only for all purposes.

>>One of the most appreciated speakers was Sunita Sah, MD, PhD of Duke, who has co-authored several well-designed studies of how disclosure practices might impact physicians and patients. During the Q&A the point emerged of how reluctant patients are to appear to criticize their own physician's COI, assuming the relationship to have been established before the COI becomes known. Some studies that are cited to show that COI is no big deal for patients and does not decrease their trust are studies in which patients are basically asked about their own physician's COI. A fairer study design would be to get patients to imagine that they are seeking a new physician, and can choose among several physicians on a panel, some of whom are disclosed to have COI. The outcome of interest would be how willing patients are to select the physicians with COI, and at least some preliminary data, I understand, would suggest that they'd be less likely to select those docs.

>>A couple of speakers commented on the recent pullback at NIH in backing off proposed COI policies, that would have required academic medical centers to post faculty COI on accessible websites ( Blame was placed on the White House, feeling under fire to do something about the supposedly out-of-control government regulations that the Republicans are raising Cain (no pun intended) about. The timing was bad; the time that the NIH stringent guidelines would have gone into effect was precisely the moment that the White House decided they had to make a show to scaling back on regulatory load.


Anonymous said...

It certainly would be easier if there was one central place to disclose COIs. Gordon Rubenfeld proposed this in a comment in The Lancet in 2007.

"I propose that the International Committee of Medical Journal Editors, the American Academy of Continuing Medical Education, and other interested parties endorse a single document with a single set of definitions for disclosure of financial conflicts of interest. This conflict vitae would not be revised for specific presentations or papers, but would simply be updated like an academic curriculum vitae and submitted in its entirety when requested. By maintaining a single document, the conflict vitae streamlines the process of disclosure and minimises the risk of accidental omission. It eliminates an arbitrary time window when conflicts are presumed to expire. The conflict vitae would not identify relevant conflicts for a specific paper or presentation—a process that itself is subject to conflict and moral hazard. By containing all potential financial biomedical relationships for a career, the conflict vitae places the responsibility of assigning relevance and potential bias squarely where it belongs, with editors and the audience.

Importantly, it would be easy to configure the standardised elements of the conflict vitae into a searchable online database maintained by an independent organisation such as This would facilitate an exciting and novel body of meta-research linking studies and conflicts."

The conflict vitae: a CV for the new millennium
The Lancet, Volume 370, Issue 9584, Page 318, 28 July 2007

Hulda Clark said...

ASLME should held This type of conference in every month or a two...