Are Conflict of Interest Policies Fair and Effective?
Conflicts of interest are one source of bias. No reviewer, author, investigator, or clinician is without bias. Many forms of bias are either unconscious or unintentional. Further, bias is frequently unrecognized because it is psychologically justified by the individual as a fair and objective view. The recognition and control of bias is important. Unfortunately, although it is sometimes easy to draw the connections between bias and the opportunity for personal gain, the association is often more subtle.
To date, journal editors, newspapers, and other societal authorities have focused on areas perceived as most obvious and blatant, such as monetary reward for promoting one pharmaceutical product over another.
Industry-related financial conflicts of interest certainly need to be scrutinized given the role of industry in funding medical research. The need for disclosure is reinforced by recent reports of improprieties. Penalties should be severe for companies and scientists guilty of this misconduct because their actions impair advancement of medical science and damage public trust in medicine and scientists.
But there are other sources of bias beyond financial greed. For example, aggrandizement — a desire to advance one's own career and achieve success, fame, and power — is a common cause of a conflict of interest in research.
Should disclosures then be limited to industry and patent-related conflicts? Is there evidence that industry-related bias has a greater impact, much less a sole impact, compared to other sources? A moment of reflection might lead to other conclusions.
Consider, for example, the case of Eric Poehlman, PhD, a researcher on obesity and the effects of menopause on women's health, who received approximately $2.9 million in research funding from the NIH, Department of Defense, and US Department of Agriculture based on grant applications containing false and fabricated research data. The case highlights the financial impact and conflicts of interest that can exist for federally funded researchers.
Other conflicts might emerge in research that supports expensive clinical procedures, or promotes an individual's claim of authority in a lucrative clinical specialty. Individuals performing surgery or other diagnostic procedures (imaging and electrophysiology, for example) can generate large salaries. Yet, the present disclosure rules give the appearance that this type of conflict creates no bias.
DISPARITIES IN RULES OF DISCLOSURE
So why is there a disparity in applying the rules of disclosure, which focus almost exclusively on industry-related conflicts? First, there is the view that non-industry funded research and clinical expertise do not create major conflicts of interest. This is unlikely. If an investigator receives major salary support from a NIH-funded project involving a specific hypothesis on a therapeutic or diagnostic procedure, discrediting that approach might have substantial impact on the likelihood of future funding. If a published article supports a particular diagnostic or therapeutic approach from which the author may garnish income, how can the authors be considered unbiased?
A second possible explanation for the focus on industry-related financial conflicts is that it is too difficult to assess financial bias from grants that are not industry-related. However, it is possible to quantitate the total dollars of a grant, the amount of personal salary, and the specific aims and hypotheses of the grant. In fact, this is done routinely in the “Other Support” section of NIH grant applications.
The third possible explanation is that present policies may be more concerned about appearance than true bias. The new rules seem to react solely to the news media's focus on pharmaceutical company improprieties, ignoring other improprieties related to federally funded research. Policies on research ethics should not be based simply on perception. Rules on disclosure of bias should be fair, effective, and without bias themselves.
Most disclosure rules have little if anything to do with objective evidence of bias; rather, they relate more to the possible appearance of bias. For example, the AAN requires disclosure of industry-related conflicts of interest even if the subject of the abstract is not related to industry.
The Academy also prohibits investigators from speaking “socially” to industry sponsors during a session in which a study funded by that company is presented. Further, universities, scientific organizations, and journals require disclosures and actions of its members that these very organizations often do not follow themselves. If an organization receives funding from private industry, they will often cite operation and even survival of the organization as justification, that is, they depend on that support to survive. Yet, is that not the epitome of conflict?
The American Association of Universities (AAU) stated: “All research projects at an institution, whether federally funded, funded by a non-federal entity, or funded by the institution itself, should be managed by the same conflict of interest process and treated the same.”
We are not adhering to balanced standards of equality set by the AAU, however. Universities, journals, and scientific organizations commonly do not require disclosure, much less consider financial conflicts of interest related to NIH funding, clinical income, medical malpractice expert witness testimony, royalty income from books on a specific research topic, or other types of non-industry funding.
Directing all our concerns about conflicts of interest toward one source will not reduce and control the many forms of bias. Although we cannot assess all bias, disclosure policies should be equitable. To do otherwise is hypocritical.
Consultants who could best provide academic expertise to industry are discouraged from industry interactions. Likewise, the appearance of conflict of interest might deter collaboration between academics and industry. Although these relationships can lead to conflicts, they can also create a confluence of interests, which is important in bringing new drugs and devices to the public as well as further defining their risk/benefit ratios.
There is great interest in improving the pipeline for new medical therapies. And translation of any medical discovery to the public requires interaction with the private sector, through whose hands more than 90 percent of effective therapies ultimately travel.
Our approach to conflicts of interest needs to be fair and balanced. Policies that ignore other sources of bias will not only fail to control it, but will also hamper our best collaborative efforts to develop new therapies.