《 PLoS Medicine 》杂志不久前（4月24日）曾发表的社论提出，科研人员在发表文章时披露有关合作者之间的利益冲突，在实际收效上可能会适得其反。《科学家》杂志6月12日发表评论文章对此提出了不同的见解，认为应该肯定“利益冲突（COI）”的好处，不能随意对COI横加指责。

http://the-scientist.com/2012/06/12/opinion-whats-wrong-with-coi/

By Thomas P. Stossel | June 12, 2012

Advances in medical and surgical care are hard-won. They require rigorous, carefully interpreted laboratory research. Equally important is the painstaking clinical work to translate basic discoveries into useful diagnostics, drugs, and devices. Despite the odds, the achievements made in the past half century are unmistakable: a 50 percent reduction in cardiovascular mortality despite an epidemic of obesity; a dramatically decreased cancer mortality rate; and the conversion of AIDS from a death sentence to survival with good life quality.

The key to such success has been the growing number and complexity of collaborations between academics, physicians, regulatory agencies, and—not least—industry. Unfortunately, over the past 20 years, a mania has taken hold that discounts the social value of collaboration and has mounted an inquisition against it, encapsulated by the epithet “financial conflict of interest (fCOI).” Critics’ unwarranted allegations that such conflicts cause bias have limited the sources of intellect that can contribute to a given project.

Medical journals have taken a leading role in promoting this mania. A study recently published in the April issue of Nature Biotechnology documents its pervasiveness: a content analysis of 108 articles in four highly cited medical journals (The New England Journal of Medicine, JAMA, Lancet, and Lancet Neurology) found that 89 percent of the publications emphasized what they considered risky or problematic with industry collaborations.

But what is the basis for this assertion? Approximately half of these articles presented no evidence whatsoever for their conclusions. They merely postulated them as self-evident. When provided, the evidence was weak, and the interpretations one-sided: fewer than 15 percent even mentioned any alternative interpretations, and only 3 percent bothered to discuss them. In contrast, the comparatively few papers emphasizing benefits of collaboration all cited evidence, and recognized and attempted to rebut opposing viewpoints.

The condemnation of industry relationships is pervasive and particularly potent in the academic community. Many universities reflexively ban industry-sponsored peer-to-peer speaking events, pejoratively dubbed “speakers’ bureaus,” that physicians attend often and voluntarily. Sheer prejudice, not fact, prompts the censorship.

Glaringly illustrative of this ignorant viewpoint is a commentary entitled “Does conflict of interest disclosure worsen bias?” authored last month by the editors of PLoS Medicine and discussed recently in The Scientist. The editors referenced an article published in their journal documenting that a large and increasing fraction of psychiatrists who contributed to updating the American Psychiatric Association’s (APA’s) Diagnostic and Statistical Manual of Psychiatric Disorders (DSM) have fCOIs.

The DSM is an evolving effort to codify diverse mental behaviors to better inform research and treatment of psychiatric disorders. Why would the APA predominantly select DSM consultants with industry relationships? According to the PLoS editors, it evidences a commercial conspiracy: an “explosion of diagnostic categories and new diagnoses (and thus markets) is…a virtual bonanza for the pharmaceutical industry.” They savage the APA for not banning physicians from receiving unrestricted industry grants or from participating in industry-sponsored peer-to-peer speaking. The PLoS editorial blames this laxity on the APA’s financial dependency on industry to which the society therefore panders by allowing “conflicted” consultants to make recommendations based on industry interests.

But if they are searching for bias, shouldn’t the PLoS Medicine editors first look in the mirror? That DSM contributors have relationships with industry partners might actually be praiseworthy. Why wouldn’t industry seek out the best, brightest, and most productive experts for collaboration? It’s in their financial interest to do so. And how could unrestricted grants, ideal for research that follows up serendipitous findings, possibly be problematic? The money leads to better research that can benefit patients.

Furthermore, the DSM does not recommend specific treatments, meaning DSM consultants are not able to actively push their products in this medium. On the contrary, it could be that the increasing number of diagnoses to which the PLoS editors refer may actually lead to more patients partaking in talk or cognitive behavioral therapies, which do not involve drugs, and thus cannot benefit the drug industry.

The PLoS editors’ confidently assert: “It is widely established that conflicts of interest impair objectivity and integrity in medicine.” But “widely reported” should not count for “widely established.” Even a voluminous Institute of Medicine report on COI, itself riddled with speculation as to its dangers, admitted that there is no empirical evidence that fCOIs have any impact on what should count most—patient outcomes.

The one-sided messaging by medical journals, led by editors arguably motivated to market their publication, has inflamed the fCOI mania that in turn has reduced physician-industry collaboration and industry-subsidized medical education. It also drives a prosecutorial racket that forces companies to pour money into settling dubious allegations under threat of debarment penalties that would prevent them from selling any of their products to Medicare or Medicaid. That money could be better applied to finding more effective treatments and cures for the suite of ailments that continues to plague the human species. Patients and industry should resist this blatant intellectual dishonesty.

Thomas P. Stossel is the Director of Translational Medicine at Brigham & Women’s Hospital and an American Cancer Society Professor of Medicine at Harvard Medical School.