Science and ethics in publications in Psychiatry: the role of conflicts of interest

INVITED EDITORIAL

Science and ethics in publications in Psychiatry: the role of conflicts of interest^*

Mauro V. Mendlowicz

Associate professor, Department of Psychiatry and Mental Health, Universidade Federal Fluminense (UFF), Niterói, RJ, Brazil. Associate editor, Jornal Brasileiro de Psiquiatria. E-mail: [e-mail]

The New York Times published it! An American parliamentary commission of inquiry, led by senator Charles E. Grassley, is not only investigating the financial relations between renowned researchers, such as Alan Schatzberg, Charles Nemeroff and Joseph Biederman, and the pharmaceutical industry, but also the very funding structure of the American Psychiatric Association (APA). In 2006 the pharmaceutical industry accounted for 30% of the APA budget, especially through sponsoring of scientific journals, congresses and symposiums. In a letter to the APA, the congressman manifested his concern about such state of affairs, claiming that "he had finally understood how funds from the pharmaceutical industry could shape the practice of non-profitable organizations that intend to be independent in their points of view and actions." Dr. Steven S. Sharfstein, former president of the APA, recognized the severity of the situation: "At each new discovery, our credibility with patients is compromised and we have to protect it at all costs."

Opposed to what may seem, at first, the heart of the matter does not lie in the pure and simple existence of economic links between researchers, medical institutions and the pharmaceutical industry. Such association emerged and has gradually strengthened over the past decades, partly as a result of new laws, such as the Bayh-Dole Act (1980), which provided universities, small companies and non-profitable institutions with the benefits of their own inventions, to the detriment of the federal government. It also facilitated transference of their technologies to the private and public sectors. The increasing capacity and availability of the latter in funding research studies has also contributed to narrowing the relationship between university and industry. In the USA, research funds of pharmaceutical companies increased from US$ 1.5 billion in 1980 to US$ 22 billion in 2001 (an 1,367% increase), of which US$ 6 billion were invested in clinical drug trials.¹ Such partnerships became today an extremely important element, sometimes crucial to the scientific progress in many areas.

The true problem concerns the inadequate form with which conflicts of interest that emerged from such relationship have been managed by certain individuals and institutions. Conflicts of interest (COI) are sets of conditions in which professional judgment of a primary interest (such as, for example, the well being of patients or the validity of a study) tends to be improperly influenced by economic interests (such as personal prestige or economic gains).² Conflicts of interest can be manifested at personal level - physicians and researchers - or institutional - universities, research institutions and scientific journals.

There are three basic strategies to deal with conflicts of interest, each having its own advantages and limitations. The first strategy is stating the existence of conflicts of interest. Editors of scientific journals demand disclosure of COI from authors who submit their manuscripts. Lecturers should do the same in relation to their audience. Similarly, researchers should inform support institutions or research ethics committees about their COI. Although disclosure of COI is considered by some as "the golden rule for COI," it has some limitations. Firstly, disclosure of COI is a voluntary act based on an honor system that does not include further safeguard. In the cases mentioned above researchers failed to disclosure massive amounts received from the pharmaceutical industry for years (whether as payment for consultancy services or as research funds) to their universities, which in turn were peculiarly incompetent to detect and repair such irregularities.

In addition, studies suggest that disclosure of COI may not be as effective in terms of correcting the target audience's judgment as common sense suggests. In general, people tend to underestimate the role of situational influences and overestimate the influence of character and values on the behavior of third parties. As a consequence, they have difficulties judging the real extension of the influence COI have on judgment of individuals that recognize being in a situation of conflict of interest. Disclosure of COI may have the paradoxical effect of increasing the audience's reliability on the source of information, reducing the ability to correct their judgment as to a possible lack of objectivity and exemption. On the other hand, it is suspected that sources of information tend to disclosure more biased information after disclosing COI. Therefore, once existence of COI is publicly revealed, some authors and lecturers may feel morally authorized to disclosure biased information or to overemphasize their conclusions and recommendations as a form of deducting the effects of their disclosure. Cain et al.³ warned that disclosure of COI could paradoxically bring more harm than benefits, given people's inability to deal with information whose quality is known to be compromised by conflicts of interest.

The second strategy is prohibition (or elimination) of COI. This is a drastic procedure that should be used judiciously, especially applicable to severe cases of economic COI (for example, when the individual is a shareholder or manager of a pharmaceutical company, manufacturer of medical equipment, or has the possibility of buying shares from the company). Researchers under such situations should totally abstain from participating in critical activities, such as regulation and supervision of studies (for example, participation in research ethics committees), or working as reviewer of manuscripts or research grants.

Limitations of COI prohibition are patent when analyzing the experience of some scientific journals using this strategy. In the 1990's, the editors of the New England Journal of Medicine decided that a mere disclosure of COI by authors of manuscripts would not be enough in some cases. They were concerned that editorials and review articles, because they represent personal selections and interpretations by the author(s) based on the existing scientific literature, could be particularly vulnerable to manipulations with commercial purposes. As a consequence, the New England Journal of Medicine started demanding from authors of editorials and reviews not to have any financial interest in the company manufacturing the product discussed in the article or in their commercial competitors. The unexpected consequence of such well-intentioned policy was that, over a 2-year period, the editors of the New England Journal of Medicine could not publish a single review of pharmacotherapy involving a new drug. The editors concluded that, if high-level scientific journals no longer published review articles written by the most qualified authors, their readers would be entirely at the mercy of pharmaceutical companies to obtain the information they need.⁴

Faced with this situation, the editors were forced to abandon their policy of prohibiting COI and adopting a COI management, the third and last strategy: the New England Journal of Medicine started demanding from authors of editorials and reviews not to have any significant financial interest in the company manufacturing the product discussed in the article or in their commercial competitors. Inclusion of the word significant represents an implicit recognition that not all financial links are the same: while it is unlikely that perception of pharmaceutical company fees for occasional lectures compromises a researcher's independence and exemption, having the possibility of buying shares from such company would be ethically unacceptable. The editors of the New England Journal of Medicine determined that US$ 10,000 a year would be the highest amount an author could receive from a pharmaceutical company not to have financial interest considered as significant. According to the journal, such guidelines should be considered more general principles than fixed and immutable rules. Editors should analyze known facts about each potential author and make the decisions they believe will bring the best medical and scientific information to the journal.⁴

COI management is an intermediate strategy between the relative passiveness of COI disclosure and the purism of COI prohibition, although it does not exclude its moderate use. One of the basic procedures for COI management consists of identifying critical areas in the production and disclosure of knowledge that deserve a more active protection than a simple COI disclosure and from which researchers with potentially severe COI should be preliminarily removed (for example, recruiting and obtaining informed consent from participants of clinical drug trials). Another essential procedure is the creation of committees in charge of reviewing the research project to ensure it is not ethically compromised. The monitoring process by the review committee should contemplate "critical points of control," such as request for research funds, submission of projects to research ethics committees, submission of reports to specific agencies (for example, Food and Drug Administration) and publications. In thesis, COI management is the ideal strategy to deal with COI. However, the number and complexity of judgments involved in the research process nowadays demand the creation of a huge bureaucratic machine that might block and suffocate scientific progress. Therefore, it is unlikely that plans of COI management are able to completely prevent occurrence of undesirable influences on research projects.

Conflicts of interest permeate nearly all human activities, including Science and Medicine. There is nothing intrinsically nonethical in the interactions between physicians, researchers and the industry, as there is nothing inherently nonethical when a physician, researcher or author is in a situation of conflicts of interest. The situation becomes a problem when there is refusal in recognizing the existence of a conflict of interest and the opportunity to satisfactorily solve it is lost. Acknowledgement of the existence of a conflict of interest does not correspond to the admission of a severe moral failure, but represents a realistic evaluation of the potential impact of secondary interests. Relying on the personal integrity of physicians and researchers is absolutely indispensable, but it has not proven to be enough. None of the three main strategies currently available to deal with COI seems to meet the criteria of effectiveness and feasibility. Elimination of COI only seems applicable to cases involving particularly severe conflicts. Psychological research has recently questioned the effectiveness of disclosure of COI. Complex judgments involved in scientific research also seem to compromise the effectiveness of COI management plans. A flexible approach, adapted to the specificities of each situation, seems to be the most appropriate conduct considering the limitations of currently available strategies.

At a critical time in which some of the most important researchers and psychiatric institutions in the world proved to be incompetent to autonomously manage their COI and, as a consequence, the American government shows their intention of taking over, it is important, more than ever, that physicians, researchers, editors of scientific journals and institutional managers are aware of the importance of early identification of COI and familiar with strategies to deal with them.

References

• 1. Bodenheimer T. Uneasy allianceclinical investigators and the pharmaceutical industry. N Engl J Med. 2000;342(20):1539-44.
• 2. Thompson DF. Understanding financial conflicts of interest. New Engl J Med. 1993;329(8):573-6.
• 3. Cain D, Loewenstein G, Moore D. The shortcomings of disclosure as a solution to conflicts of interest. In: Moore D, Cain D, Loewenstein G, Bazerman M, eds. Conflicts of interest: cand solutions in business, law, medicine, and public policy. New York: Cambridge University; 2005. p. 104-25.
• 4. Drazen J, Curfman G. Financial associations of authors. New Engl J Med. 2002;346(24):1901-2.

*

Relatório apresentado no Fórum de Publicações – Ciência e Ética em Publicações em Psiquiatria, XXVI Congresso Brasileiro de Psiquiatria, Brasília, realizado em outubro de 2008.

Publication Dates

• Publication in this collection
  17 Mar 2009
• Date of issue
  Dec 2008