Toward Credible Conflict of Interest Policies in Clinical Psychiatry

POINT
Lisa Cosgrove, PhD and Harold J. Bursztajn, MD

A recent letter to the American Psychiatric Association (APA) from Sen Chuck Grassley about the APA’s financial relationship with pharmaceutical companies raises concerns about undue industry influence.¹ By instituting a disclosure policy for DSM-V, the APA took a halting first step in restoring public trust in the most influential text on psychiatric taxonomy in the world. Unfortunately, the APA’s efforts at creating a conflict of interest (COI) policy have failed to ensure that the process for revising diagnostic and therapeutic guidelines is one that the public can trust. The need for more safeguards was evidenced when the APA reported that of the 27 task force members of DSM-V, only 8 reported no industry relationships.² The fact that 70% of the task force members have reported direct industry ties—an increase of almost 14% over the percentage of DSM-IV task force members who had industy ties—shows that disclosure policies alone, especially those that rely on an honor system, are not enough and that more specific safeguards are needed.

At first blush, it might appear that there is less reason for concern about industry influence on the development of guidelines on diagnosis than on treatment. But diagnosis informs treatment decisions. Hence, pharmaceutical companies have a vested interest in the structure and content of DSM, and in how the symptomatology is revised. Even small changes in symptom cri-teria can have a significant impact on what new (or off-label) medications may be prescribed. Public trust in the independence of clinical psychiatry is undermined if former DSM panel members are using—or are perceived as using—their participation on DSM to leverage lucrative consulting arrangements with the pharmaceutical industry or to funnel industry funding to their departments, associates, and programs (eg, exerting their influence on prescription practices through public speaking arrangements, such as industry-sponsored CME symposia).³

We need to remember that, as Louis Pasteur said, “serendipity favors the prepared mind.” That is, to the extent that DSM is constructed as a reliable diagnostic taxonomy that emphasizes relatively short-term, acontextual symp­toms that tend to be insensitive to characterological variability in expression, it encourages overuse of diagnostic checklists. Among other things, diagnostic checklists provide the basis for outcome measures in clinical trials conducted by industry-funded researchers.⁴ These researchers then maintain that the “evidence-based research” speaks for itself, that the disorder has been validated, and that the new drugs and medical devices have been proven to treat the disorders effectively and safely.

Diagnostic checklists thus become established as good science and contribute legitimacy to the proliferation of new diagnoses and new medications. Therefore, although checklists can facilitate diagnostic reliability, too often they become self-serving industry tools and lead to inflated statistics about how many people are “suffering” from a disorder.⁵ They also make it more difficult for researchers and clinicians to consider context and individual variability. The end result of a diagnostic manual with acceptable reliability but limited validity is that it allows pharmaceutical companies to bring what has been referred to as “me-too” drugs to market—medications that are just different enough from existing pharmaceutical agents that they can be patent-protected. Many existing and “new” agents can be helpful in the short-term for symptomatic relief but also have increasing risks of adverse effects with long-term use, leading to a decreasing benefit­to-risk ratio as time goes on. The relative neglect of psychodynamically informed narrative descriptions in DSM has led to it being of limited use in serving as a guide for psychodynamically informed treatment approaches whose benefit-to-risk ratio increases over time.^6,7

Clearly, if the APA is to continue to be entrusted with the task of creating unbiased diagnostic systems and treatment guidelines, it must refocus attention on the issue of diagnostic validity and move beyond its current exclusive self-monitoring practice. To best ensure that psychiatric guidelines are objective and evidence-based, we offer the following recommendations:

1. In May 2008, the APA began the process of restoring credibility by naming all of the DSM panel members and making public all of the disclosures from work group members. Unfortunately, there are still work groups with a majority of mem­­bers who have disclosed industry relationships. We recommend that the APA review the composition of these newly identified work groups and institute a policy that ensures that no panel has a majority of members with ties to the pharmaceutical industry. Also, a feasible and practical way to ensure bal­ance and avoid undue influence is to have critics of industry actively recruited to be on the various panels.
 

2. In addition to disclosing direct financial relationships (eg, consultancies, honoraria), individuals should be required to disclose indirect support. These include support received by one’s institution in the form of pooled funds for academic departments, hospitals, and medi-cal schools. Unrestricted research grants (eg, the ties of associated principal investigators on joint research projects or family members who have received industry support) should be revealed.
 

3. Despite a requirement to disclose potential COIs, many individuals fail to do so—as researchers and investigative journalists have demonstrated. The threshold of disclosure should be consistent with the current standard for informed consent.⁸ Specifically, the industry ties and financial benefits that must be disclosed are those that a reasonable patient would want to know rather than what professionals generally tell their patients. (If a reasonable patient being treated according to an APA treatment guideline wants to know about an industry tie, then the individual should disclose this information. Such information should be disclosed even if most professionals or colleagues do not consider such a tie to be a COI.)

Transparency of potential COIs is critical for the conduct of psychiatric teaching, research, and clinical care. To restore public trust and protect patients’ welfare, the field must require more than the mere reporting of blatant COIs. We need to expand the scope for disclosure, develop patient-centered policies, and eventually have diagnostic and treatment guidelines produced by an organization that is not beholden to industry funding.