Thursday, December 08, 2011

A Logical Fallacy Affecting Selection of Panelists on an FDA Advisory Committee

An old argument used to defend against criticisms of conflicts of interest was just employed in a disturbing context. 

Expert Removed from FDA Advisory Committee for Having an Opinion

As first reported on the PharmaLot blog, and later by the Newark Star-Ledger, a panelist was just disqualified from voting on a US Food and Drug Administration (FDA) panel for having previously expressed an opinion about the safety of the drug up for re-evaluation.  Per the Star-Ledger,
Federal drug regulators have notified Sidney Wolfe, one of the nation's leading advocates for drug safety, that he would not be permitted to join a committee of experts asked to review new dangers associated with a group of birth control pills, including Bayer Healthcare's top-selling Yaz.

The Food and Drug Administration scheduled a meeting Thursday of two advisory committees — one on drug safety and risk management and the other on reproductive health drugs — after new information emerged on the safety of oral contraceptives containing the synthetic hormone

Why Was Dr Wolf removed from the committee?
The agency recently learned that Public Citizen, a non-profit consumer advocacy organization, had placed one of the contraceptives, Bayer’s Yasmine — a predecessor to Yaz — on its list of 'Do Not Use Pills' in 2002.

'He did not volunteer this information,' said agency spokeswoman Erica Jefferson. 'It was brought to our attention.'

The FDA offered Wolfe two options: He could present information to the advisory committee like other members of the public or he could sit on the committee, participate in the discussion but refrain from voting.

Logical Fallacy: False Dilemma

We frequently post about conflicts of interest affecting health care decision-makers.  It is now clear (e.g., look here) that leading health care academics often have significant financial relationships with drug and device companies and other health care corporations which could potentially influence their clinical research, clinical teaching, health policy recommendations, or direct patient care.  These relationships are frequently defended, often with logical fallacies used by those who themselves have conflicts. 

One common argument is based on the assertion that conflicts of interest are ubiquitous and everyone is conflicted.  Therefore, if one were to ban people with conflicts from responsible positions, there would be no one left to fill these positions, so such a ban would be untenable.  This seems to be an example of the false dilemma.  It is often employed by people who themselves have conflicts of interest.

One way to make it appear that everyone has conflicts of interest is to broaden the concept of conflicts of interest to "intellectual conflicts of interest."  Doing this facilitates the assertion that everyone who has an opinion on a subject has a conflict of interest, so this argument implies that all sentient beings have important conflicts.  This argument would make equivalent a doctor who would not use a particular drug because his or her reading of the clinical research literature about this drug suggests its benefits do not outweigh its harms, and a doctor who advocates using the drug, and is paid $100,000 a year by the marketing division of the company that makes this drug as a marketing consultant. 

The decision to prevent Dr Wolfe from voting on this committee seems to be based on this logical fallacy. As Dr Wolfe said,
In his statement, Wolfe said if intellectual conflict of interest means being informed and subsequently having opinions on a drug, then 'many more members of advisory committees would have to be excluded.'

'For members of a scientific and technical advisory committee, possessing information and expert views on matters within the purview of the committee is not a conflict of interest,' Wolfe wrote. 'To the contrary, qualified experts are likely to have developed views on a variety of subjects based on their professional experience.'

As Larry Husten wrote on the CardioBrief blog,
do we really want to choose advisory committee panelists who have never expressed opinions about the topics they are reviewing? Are we reaching the point where potential FDA panelists will be required, like Supreme Court nominees, to have avoided any discussion of all important issues at every point in the past?

Thus they point out the absurdity of banning people with "intellectual conflicts of interests," that is, with relevant opinions, as if they had real conflicts of interest. (But wait for someone to argue that if Wolfe were allowed to serve, it would be unfair to ban anyone with financial conflicts of interest from serving.)

What is most distressing about this case is that the sort of fallacious arguments usually employed by the conflicted to defend conflicts of interest are now being employed by leaders of government agencies, who are supposed to not have their own conflicts, and to serve the people, and in this case, to be dedicated to improving the health and safety of the population and of individual patients.

Every fallacious argument made in support of financial conflicts of interest affecting health care decision makers suggests we need to do more to combat such conflicts.  At an absolute minimum, all such conflicts should be fully disclosed in detail in any context in which they possibly could influence medical research, medical education, clinical care, or health policy.  Furthermore, we need to work towards ending as many such conflicts as possible.  A good starting point would be the recommendations made by the Institute of Medicine committee reports on conflicts of interest, and clinical practice guidelines.

See also comments by Merrill Goozner on the GoozNews blog.

ADDENDUM (9 December, 2011) - In response to comments below, see two posts by Dr Howard Brody in the Hooked: Ethics, Medicine, and Pharma blog on the problems with the concept of "intellectual conflict of interest" - here and here.

ADDENDUM (9 December, 2011) -  See further discussion by posted by Dr Brody today here.

7 comments:

Melissa Raven said...

The exclusion of Sidney Wolfe is reprehensible. However, I do not agree that the concept of intellectual conflicts of interest is a spurious issue. But it should be applied equally to all panel members. How many of the other panellists have written/spoken/taught positively about these drugs? They should all be required to disclose this.
Intellectual conflicts of interest are an important source of potential bias that is routinely ignored. McLaughlin et al. (2011) recently expressed this forcefully in relation to monograph working groups of the International Agency for Research on Cancer:
We have argued that in restricting conflict of interest concerns to only financial issues, IARC cannot see the elephant in the room. IARC working groups are composed of investigators from various fields. In our field of epidemiology, the area of research given the most weight in IARC decision making, the working group is often composed of researchers who have a strong, if not lifelong interest in the potential carcinogenicity of the exposure under consideration.... the IARC Working Group Preamble makes clear that members will be chosen because of this qualification. To argue that researchers who have spent years publishing on the possible carcinogenicity of an exposure are the best judges of the validity and methodological soundness of their own and allied work strikes us as naive, if not anti-scientific. They are clearly not disinterested evaluators of the research evidence being considered, as much of it represents their own work. http://ije.oxfordjournals.org/content/early/2011/11/16/ije.dyr158.short
People with intellectual conflicts of interest should not be banned from panellists, but they should be required to declare them.
Furthermore, panels should include some experts who have no opinions and not a lot of knowledge about the drugs in question – methodological experts (e.g. epidemiologists, or people with a strong background in evidence-based medicine at a meta-level) who can bring fresh eyes to the evidence and issues, unencumbered by personal investment of time and energy and passion as researchers/clinicians/teachers in that specific clinical sphere. [Here I should declare that I am an epidemiologist, and therefore potentially biased in my opinion of the importance of epidemiology.]

Roy M. Poses MD said...

Ms Raven (if that is really your name) -

I fully agree that methodological experts, especially in clinical epidemiology/ evidence based medicine should be included in FDA panels, and other groups that make recommendations about use of specific drugs, devices, or tests. It would be helpful if those experts have some understanding of the clinical context, if not extremely detailed knowledge of the specific drugs/ devices/ tests to be assessed.

However, I am very concerned about pushing for disclosures of "intellectual conflicts of interest." I have yet to see a good definition of "intellectual conflicts of interest." It often seems to be used to mean opinions, often those with which the complainant disagrees.

See these posts on the Hooked: Ethics, Medicine and Pharma blog by Dr Howard Brody -

http://brodyhooked.blogspot.com/2007/12/intellectual-bias-latest-salvo-from.html

http://brodyhooked.blogspot.com/2008/03/more-on-intellectual-conflict-of.html

Anonymous said...

This concept of "intellectual conflict of interest" came up in 2009 when cardiologist Sanjay Kaul was excluded from an advisory committee meeting because he had published articles on the drug in question. In that case the drug's sponsor actually contacted the FDA to request his exclusion.

Of possible interest, here is a letter from Dr. Wolfe and a colleague in which they protested Dr. Kaul's exclusion.

http://www.citizen.org/hrg1857

InformaticsMD said...

"Intellectual conflict of interest", I hate to say, is starting to sound uncomfortably like George Orwell's "Thoughtcrimes."

I guess I won't ever be asked onto an EHR review panel due to my intellectual conflicts of interest, i.e., writing about them based on the evidence, and on problems with the evidence base.

-- SS

Afraid said...

Yep Scot, your toast at those gatherings.

Michael S. Altus said...

I’m disappointed, Roy, in the scholarly shortcomings and the logical fallacy inherent your dismissive comment, “Ms Raven (if that is really your name).”

Had you done your homework by a routine Google search, you would have learned that Melissa Raven is a psychiatric epidemiologist and policy analyst, and adjunct lecturer in the Discipline of Public Health at Flinders University, a public university in Adelaide, South Australia.

The following letter would be of particular interest to you:Raven MK, Rogers WA, Jureidini JN. Partnerships between academic psychiatry and the pharmaceutical industry [letter]. Australasian Psychiatry. 2005; 13(1):84-85.

The letter concludes, “Psychiatric research has the potential to improve the lives of many Australians. To achieve this, we need a research agenda determined by health care needs rather than commercial interests”--a sentiment much like yours. That Ms. Raven’s and your sentiments coincide is not surprising, inasmuch as some of your blog entries referred to Jon Jureidini’s work to support your views.

Your blog entries have helpfully identified logical fallacies in pronouncements from others by referring to the Nizkor Project’s list of logical fallacies (www.nizkor.org/features/fallacies). I am grateful for your identifying and explaining logical fallacies and referring to websites for more information.

I went to the Nizkor Project’s website to try to identify the logical fallacy in the comment “Ms Raven (if that is really your name)”. One possibility is that the logical fallacy is a red herring, a fallacy in which an irrelevant topic, in this instance, questioning if Melissa Raven is the commenter's real name, is presented to divert attention from the original issue.

I hope that your future blog entries will maintain the high quality of scholarship by researching topics, exposing logical fallacies, and not using a logical fallacy yourself to advance your positions.

Roy M. Poses MD said...

Michael S Altus -

After I wrote the comment above, I got an email separately from Melissa Raven at what appears to be a legitimate email address in Australia. So the Melissa Raven above does now appear to be the Melissa Raven to whom you refer.

I apologize if the phrasing in my comment above was too flippant.

However, I do not think it reflected a logical fallacy.

At the time of my first response, the identity of the commentator with the electronic identification of "Melissa Raven" was not so clear.

Here is some background:

1 - There is no way for me to tell who actually wrote any given comment sent to us. The Google software does not give us any identifying information about commentators except for however they sign their name. So while I guess you are really Michael S Altus, anyone could post a comment and sign it "Michael S Altus."

2. I must confess I didn't know that there was a Melissa Raven who is an epidemiologist. If you do a Google search in the US for "Melissa Raven," you get these hits before I get one that pertains to that Melissa Raven:
http://www.modelmayhem.com/322944
http://www.facebook.com/mravenski
http://twitter.com/melissa_raven85

The first and third apparently relate to a 25 year old fashion model. The second has the picture of a woman hoisting an apparently alcholic drink.

I got that far in the search, and decided I could not tell who actually wrote the comment.

3 - This blog now has a history of attracting comments from "sock puppets." These are generally comments that appear to defend, some very cleverly, a position that is favorable to the powers that be, e.g., pharmaceutical companies. These comments sometimes are anonymous, and sometimes are provided with identities that cannot easily be verified. We suspect most come from people employed by public relations units or firms allied with such companies, but cannot prove same.

4 - So when we are concerned that we do not know who really made a comment, we generally ask for more identification, sometimes, I am afraid, perhaps too flippantly. True sock puppets usually evade such queries.

5 - As noted by the other comments above, the notion of "intellectual conflict of interest" has often been pushed by those who seem to defend certain kinds of financial conflicts of interest.

Therefore, I added the phrase questioning the identity of the commentator to a reply I believe was respectful of the content of the comment.

Hopefully, Ms Raven will respond substantively to my concerns about problems with the definition and operationalization of the concept of "intellectual conflicts of interest."