What Would It Take to Prove Harm from Detailing?

I hope to do a few posts about a recent exchange in the American Journal of Bioethics centering on a paper by Dr. Thomas S. Huddle, a general internist at Universityof Alabama-Birmingham. Dr. Huddle objects to the recent report by the AAMC, "Industry Funding of Medical Education" (see http://brodyhooked.blogspot.com/2008/04/aamc-has-spoken-more-or-less-pharm-free.html). AJOB ran 10 peer commentaries in response to Dr. Huddle's paper, one by yours truly, to which Huddle then replied.

In this post I want to focus on only one point Huddle raised. He takes issue with the AAMC for trying to turn the presence of drug detailing on an academic medical campus into an ethical issue. So far as he is concerned, it ought to be treated strictly in prudential terms: does it cause more good vs. harm or vice versa? And in both of his commentaries he regards the record as straightforward--there is some substantial evidence that good results from drug detailing, and no reliable evidence whatever of harm having resulted, so prudential considerations require that detailing be allowed.

I am struck by this "on what planet do you live?" disagreement between those of us who take the perspective usually represented on this blog-where just about every post is about some harm that presumably has resulted from today's system of drug marketing--and those people we have labeled as pharmapologists, and who have now formed the organization ACRE (though I see no clear link between Dr. Huddle and that organization). The latter group has dismissed as merely anecdotal all the evidence that the former group believes to be overwhelming. By way of trying to tease out the presumptions underlying that radical difference in worldview, I will address here what would count as a truly definitive study of harms resulting from drug detailing.

We would have to select a cohort of entering medical students and randomly assign them to two groups. The first group would be encouraged to maximize their contact with, and acceptance of gifts from, the pharmaceutical indistry, while the second group would be absolutely prohibited from any such contacts or goodies. We would follow both groups through medical school and residency and into practice. After each group had been in practice for (let's say) five years, we would do an in-depth health assessment of all their patients. If we could demonstrate worse health outcomes among the patients of the maximum-contact group, compared to those of the no-contact group, then we would have proven that drug detailing is harmful. Absent that result, we'd have no reliable evidence on the matter.

I used to say that we'd have to examine the prescribing patterns of the two groups of docs and prove that according to some set of objective criteria for "rational" prescribing, there was more irrational prescribing among the maximum-contact group. But that would not do the trick, because the buzzword in evidence-based medicine today (to which I try my best to adhere) is patient-oriented outcomes. Even if drug detailing caused physicians to prescribe irrationally, it would mean nothing unless that irrational prescribing caused worse health among the patients.

It seems to me that folks like Dr. Huddle can say with a straight face that we have no evidence that docs' hanging out with drug reps leads to any harmful consequences for patients only so long as they adhere to this sort of standard for what would count as good evidence. I trust it's clear that no such study could or will ever be done, and so it is a foregone conclusion that there will never be evidence that meets the criteria set forth by the pharmapologists.

Do us pharmascolds then simply abandon the field? There are many areas of medicine where we can show that the ideal randomized controlled trial cannot practically be done, and evidence-based medicine requires then that we consult the best available evidence, not some ideal evidence that will never exist. I believe that websites such as Drug Promotion Database and Healthy Skepticism have done an excellent job of compiling the best available evidence. What is striking to me about that evidence is its univocal nature. We do not see some studies saying that detailing provides benefits and others that detailing causes harm. Virtually every study shows some sort of negative outcome as a result of drug marketing to physicians, and the sorts of harm shown in one study are usually replicated in similar studies. So I take the evidence (as I attempted to summarize in HOOKED, and which has only grown since that book was published in 2007) as quite clear in raising serious concerns about the consequences of drug marketing to physicians in the form of standard detailing practices.

I continue to believe that the best statement of the current evidence is the formula generally employed by our colleagues at Healthy Skepticism--we currently know of no way for physicians to receive either benefits or information from representatives of the pharmaceutical industry, in such a way that the exchange results in more good than harm for patients.

Huddle TS. The pitfalls of deducing ethics from economics: why the Association of American Medical Colleges is wrong about pharmaceutical detailing. Am J Bioethics 10(1):1-8, 2010.

Huddle TS. Response to open peer commentaries on "The pitfalls of deducing ethics from economics: why the Association of American Medical Colleges is wrong about pharmaceutical detailing." Am J Bioethics 10(1):W1-3, 2010.