Saturday, November 14, 2009

COI: Further Response from Stell

Folks, I trust this will be the last post on this particular exchange; any further discussion will be relegated to comments to posts. However, I think the topic is well worth the headlines because the definition of "conflict of interest" is central to our discussion--if we do not know what COI is, then much of the rest of our ethical concerns fall down. (Hence the shrewdness of a counterstrategy that attacks the definition directly.)

Lance Stell has kindly replied to my most recent post (one down) as follows:’s how I summarize your [Brody's] response.

· Langbein is sensible, but is not obviously relevant to the question at hand –
o Comment: How could Langbein not be obviously and directly relevant where the underlying ethical issue is fiduciary trustworthiness?.
· Langbein really provides more support for your (prohibitionist) view than my view (a management/disclosure view).
o Comment: Huh?
o With regard to overlapping relationships, Langbein makes “management/disclosure” the default. He is not a prohibitionist.
o Langbein does allow that some relationships are so toxic that prohibition is the right prescription. But even in proven toxic circumstances, he argues for a “reasonableness, best-interest” defense.
o You probably assume that your book provides ample empirical proof of severe toxicity.
o But no. Influence, even when established, is not thereby “undue-influence,” let alone toxic influence.
o There is ample evidence that doctors chronically fail to intensify therapy as they should (esp. in DM and HTN).
o Doctors’ clinical inertia is a chronic problem. Good quality CME may offset it somewhat.
o What may be said at CME events is tightly scripted, especially where drugs are concerned.
o Doctors have been criminally indicted for off-label use promotion. Free-speech anyone?
o At industry-sponsored CME, the corporation’s legal department screens the Power Points for staying well-within the FDA’s PI.
o The issue w/ CME is educational quality, not provenance.
o IMHO, Pharmascolds tend to have a provenance-fixation, rather than a quality-fixation.
o Huddle and Stossel have completely shredded the pharmascold COI research as having somehow proven “toxicity-with-patient-harm-resulting.”
o Even Wazana, a lead pharmascold, candidly admits that the COI research that she has reviewed hasn’t even examined patient outcome.
· As a card-carrying pharmascold you suggest (and by your book title imply) that doctors who consult for industry or attend sponsored CME should bear a “black box” warning. “Patients beware, your doctor maybe have acquired drug-company dependency.”
· This is an ad hominem (circumstantial), isn’t it?
· Finally, your working diagnosis for me: I naively “inhaled” at ACRE’s founding in July and am now either early-onset addicted or at-risk for – drug-company dependency myself.
o Comment: I was invited to speak at the ACRE meeting. No one paid my way. I attended on my own dime.

OK, that's Lance. Just a few comments in reply.

Prof. Stell appears to be astounded, first, that I sugggested that Langbein may not even be relevant to the Pharma issues; and second, he is even more astounded that I suggest that if Langbein is relevant, he actually defends the pharmascold view more than the views of pharmapologists like ACRE (which whom Stell has now thrown in his lot completely I gather).

The reason why Langbein is not (very) relevant is simple. His article is about trust law and the sole-interest rule. The criticism he lodges against that rule is quite dependent upon the detailed history of trusts and trust law that he provides in the article. So yes, it is about a fiduciary relationship, but one that has very different features (and a completely different set of empirical facts) from the impact of Pharma on the physician-patient fiduciary relationship.

The reason why despite that, Langbein offers more comfort to pharmascolds is more complex and I tried to give the detailed case in my previous post. But by way of a quick summation, the case is basically this. Stell and his ACRE colleagues realize that the only way to ward off the attack of the pharmascolds is to redefine 'conflict of interest' in such a way that it either disappears as a factor, or else that it becomes so widespread and omnipresent that one cannot distinguish financial COI with drug companies from a host of other COI's that we find acceptable in everyday life (which is just another way of making it disappear as an ethically relevant feature of the landscape). By contrast, Langbein offers no redefinition of COI and sees no need to do so. The pharmapologists see no ethical problem where the pharmascolds see a ton of them. Langbein sees exactly the same ethical problems as do his opponents who favor retaining the sole interest rule in trust law; he simply proposes a different solution.

Now, here is where we pharmascolds may have been sloppy in our arguments so as to have misled folks like Stell to think he has scored points against us. Stell seems puzzled that I would defend Langbein as reasonable, since my position against the pharmapologists is that most COIs between medicine and Pharma should be avoided, not accepted and "managed"; whereas Langbein obviously calls for managing COI in the area of trust law. (As a sidebar let me say that a "management strategy" makes more sense in trust law than in medicine/Pharma because the trustee is more directly under judicial control and review.) However, the argument I make in HOOKED has two parts: 1) a definition of COI that explains why it is a serious ethical problem that implicates patient trust; and 2) a further argument that most COI in medicine should be divested and not merely managed. If I were trying to claim that #2 followed directly from my definition, then indeed I would be open to all the criticisms that Stell throws in my direction.

It is important that we not expect the definition of COI to do more work for us than it can. Let's go back to Erde's fundamental analysis as I review in HOOKED. Erde insists that a statement that a COI exists is not a statement that a person has necessarily done anything ethically wrong. The person has, rather, placed herself in a position where she is at risk of doing things that are ethically wrong. The actual wrongdoing is a rebuttable presumption (which seems to be exactly what Langbein is saying).

Now suppose that in addition to the fact that a COI is present, we have some additional facts. First, significant patient harms have actually occurred as a result of those sorts of COIs in the recent past. (ACRE and Stell apparently deny that this is true in the Pharma case, and that is what would lead me to ask what he had been inhaling at their meeting--I made no suggestion whatever as to financial ties.) Second, the protections that have been offered to "manage" these COIs (such as disclosure) have been ignored or widely abused. Third, the COI is not a part of any arrangement that is necessary for medical practice or research to be properly conducted. Now, if all these things are true, it becomes more reasonable to blame the individual for entering into the arrangements that constitute the COI (such as signing up for a company speaker's bureau). But notice that the blame arises from the extra facts. It does not arise from the mere definition of COI. The charge of wrongdoing implicit in the label 'COI' remains a rebuttable presumption, by definition.

No comments: