Medical School Leaders' Conflicts of Interest, Quantitated

We have posted frequently about the pervasiveness of conflicts of interest in health care, (for example, recently here). In this weeks JAMA was an important article that quantitated at least some species of such conflicts. [Campbell EG, Weissman JS, Ehringhaus S et al. Institutional academic-industry relationships. JAMA 2007; 298: 1779-1786, link here.]

Let me summarize the methods. The authors sent a paper-and-pencil survey to all chairs of medicine, psychiatry, two randomly selected clinical departments, microbiology, and one randomly selected non-clinical department at all 125 US allopathic medical schools and the 15 hospitals with the largest amount of NIH funding. The survey was sent in 2006. The response rate was 67%, and was almost the same for clinical and non-clinical chairs. The surveys asked about the chairs' personal relationships with industry in the previous year, and about the chairs' departments' institutional relationships with industry in the previous year.

The results were striking. The proportion of chairs reporting personal relationships were (in percentages, for all chairs, and when the difference between clinical and non-clinical was statistically significant, for that breakdown):

• Company officer or executive: 7% (same)
• Member of board of directors: 11%
• Paid consultant: 27%
• Member of scientific advisory board: 27%
• Member of speakers' bureau: 14% (21% clinical, 2% non-clinical)
• Founder of company: 9%

The proportion of chairs receiving particular types of funds or gifts was:

• Research funding: 21%
• Personal compensation for writing: 3%
• Personal compensation for participating in CME: 19% (26% clinical, 6% non-clinical)
• Personal compensation for participating in meetings, conferences: 28%
• Equity of stock options for professional service or intellectual property: 6%
• Personal royalties, patent licenses, etc: 11% (7% clinical, 20% non-clinical)
• Personal gifts: 1%
• Travel or other personal expenses related to meetings or conferences: 16%

A total of 60% of department chairs reported at least one of the above relationships with industry.

The proportion of chairs reporting that their departments received specific types of resources or funds from industry was:

• Research equipment: 14% (17% clinical, 10% non-clinical)
• Unrestricted funding for operations: 13% (19% clinical, 3% non-clinical)
• Research support: 9%
• Support for trainees: 14%
• Money from licensing or transferring intellectual property: 20% (16% clinical, 27% non-clinical)
• Support for research seminars: 28% (36% clinical, 13% non-clinical)
• Support for fellowship training: 25% (37% clinical, 2% non-clinical)
• Support for CME: 46% (65% clinical, 3% non-clinical)

The proportion of chairs reporting that their dDepartments received discretionary funds to support specific activities were:

• Faculty bonuses: 0.5%
• Food, berverages: 38% (51% clinical, 12% non-clinical)
• Travel and meetings: 22% (30% clinical, 8% non-clinical)
• Journal subscriptions: 6% (8% clinical, 2% non-clinical)
• Software: 3%
• Research equipment or infrastructure: 12%
• Clinical equipment: 3% (4% clinical,1% non-clinical)

A total of 67% of chairs reported their departments received at least one kind of support.

The article properly noted that social desirability bias may have prompted some people with extensive relationships with industry not to respond, and some respondents to under-report their relationships. Hence, the proportions above should be regarded as lower bounds on the actual proportions.

This article was covered in a variety of wire service reports, which contained some interesting commentary.

The study's lead author, Eric Campbell, said (per HealthDay, via the Washington Post):

There is not a single aspect of medicine in which the drug companies do not have substantial and deep relationships, affecting not only doctors-in-training, resident physicians, researchers, physicians-in-practice, the people who review drugs for the federal government and the people who review studies.

Drug companies have relationships with everyone. They're involved in every aspect of medicine. Someone has to decide which of these is OK.

I believe there's very little reasonable justification for why drug companies should be involved in the education of medical students. What knowledge do they have that is not currently available in the biggest medical schools in the country?

Dr David Korn, Senior Vice-President for Biomedical and Health Sciences Research of the American Association of Medical Colleges, was more circumspect, and somewhat defensive.

I think the paper is a very valuable contribution, in that it provides what's probably the first comprehensive documentation of the extent of relationships that involve department chairs, and department chairs are certainly the key agents of overseeing and maintaining the day-to-day operations of a medical school or teaching hospital.

The extent of the relations is not surprising.

Supported research is fulfilling a social mandate. After all, the public support of NIH [U.S. National Institutes of Health] funding is really driven by the desire to see practical results . . . and the only way those practical results can come to pass is by having productive relationships between the discoverers of new information and the organizations that our country has established to determine whether that information can be developed into useful products for public health.

Not surprisingly, Alan Goldhammer from Pharmaceutical Research and Manufacturers of America (PhRMA), defended industry involvement with department chairs and their departments, saying "the study results don't mean these relationships are a problem. He said it makes sense to reach out to academic heads because they have the most expertise." (per the Associated Press)

But Jerry Kassirer, formed editor of the New England Journal of Medicine, thought they were a real problem (again, per the Associated Press),

I was appalled by the results.

No one knew that so many chairs of medicine and psychiatry were paid speakers. We've never had that data before.

I agree with Dr Kassirer and Dr Campbell. There is no educational need for academic medical leaders to have personal financial relationships with health care corporations. There is a rationale for medical schools and their faculty to collaborate with employees of such corporations in the conduct of research. However, there is no need for faculty, their leaders, and their departments to be paid by such corporations in order to conduct valid research.

Instead, the financial relationships revealed in the study by Campbell et al threaten the tripartite mission of medical schools. Medical schools' missions include providing the best possible care for their patients, teaching future generations of physicians and other health care professionals, and doing biomedical and clinical research.

Providing the best possible patient care requires deciding which tests or treatments are most likely to help and least likely to harm individual patients. Such decisions should not be biased by physicians' relationships with the companies that make tests and treatments.

Teaching future generations of health professionals should involve preparing them to make such decisions. Such teaching, again, should not be biased by the faculty's or their supervisors' relationships with the companies that make tests and treatments.

Performing valid clinical research that properly respects research subjects should mean conducting research designed, implemented, analyzed, and disseminated so as not to be biased towards a particular test or treatment because of the researchers' relationships with the companies that make them.

The data in the study by Campbell et al suggests there is a huge risk that medical schools' missions are being compromised by their faculty leaders' relationships and by departmental relationships with health care corporations.

Furthermore, the extensive personal relationships of department chairs with such corporations may suggest to their faculty and students that they ought to be particularly friendly to certain health care corporations, and at least not criticize such corporations and their products, and certainly not complain about any related conflicts of interest on the part of their supervisors.

This may explain why medical schools have been so reluctant to teach about the ethics of health care, and thus fail to prepare physicians for the threats to their values that they may encounter in the real world (see post here).

To preserve their missions, medical schools should ban personal financial relationships among their faculty, particularly their leaders, and pharmaceutical, biotechnology, and device manufacturers and other for-profit health care corporations.

Until they enact such bans, patients, physicians, policy-makers and the public at large will wonder about the extent that medical schools' missions are compromised by faculty leaders' personal relationships and the schools' institutional relationships with these for-profit health care corporations.