1 Boring Old Man

Justifying conflicts of interest in medical journals: a very bad idea

by Robert Steinbrook, Jerome P Kassirer, and Marcia Angell

British Medical Journal. 2015 350:h2942

[full text online]

A series of articles in the New England Journal of Medicine has questioned whether the conflict of interest movement has gone too far in its campaign to stop the drug industry influencing the medical profession. Here, three former senior NEJM editors respond with dismay

A seriously flawed and inflammatory attack on conflict of interest policies and regulations appeared recently in a most unexpected location: the venerable and trusted New England Journal of Medicine [NEJM]. In a series of rambling articles, one of the journal’s national correspondents, Lisa Rosenbaum, supported by the editor in chief, Jeffrey Drazen, tried to rationalise financial conflicts of interest in the medical profession. As former senior editors of the NEJM, we find it sad that the medical journal that first called attention to the problem of financial conflicts of interest among physicians would now backtrack so dramatically and indulge in personal attacks on those who disagree.

Physicians and the public rely on journals as unbiased and independent sources of information and to provide leadership to improve trust in medicine and the medical literature. Yet financial conflicts of interest have repeatedly eroded the credibility of both the medical profession and journals. As the Institute of Medicine explained in its 2009 report, a conflict of interest is “a set of circumstances that creates a risk that professional judgment or actions regarding a primary interest will be unduly influenced by a secondary interest.” The key issue is that “a conflict of interest exists whether or not a particular individual or institution is actually influenced by the secondary interest.” The report drew heavily on a 1993 NEJM article by Dennis Thompson, not cited by Rosenbaum, which made clear that the rules “do not assume that most physicians or researchers let financial gain influence their judgment. They assume only that it is often difficult if not impossible to distinguish cases in which financial gain does have improper influence from those in which it does not.”

The NEJM has now sought to reinterpret and downplay the importance of conflicts of interest in medicine by publishing articles that show little understanding of the meaning of the term. The concern is not whether physicians and researchers who receive industry money have been bought by the drug companies, as Drazen writes, or whether members of guideline panels or advisory committees to the US Food and Drug Administration with ties to industry make recommendations that are motivated by a desire for financial gain, as Rosenbaum writes. The essential issue is that it is impossible for editors and readers to know one way or the other.

Judges are expected to recuse themselves from hearing a case in which there are concerns that they could benefit financially from the outcome. Journalists are expected not to write stories on topics in which they have a financial conflict of interest. The problem, obviously, is that their objectivity might be compromised, either consciously or unconsciously, and there would be no easy way to know whether it had been. Yet Rosenbaum and Drazen seem to think it is insulting to physicians and medical researchers to suggest that their judgment can be affected in the same way. Doctors might wish it were otherwise, but none of us is immune to human nature.

Straw men
Rosenbaum’s language is colorful, but her arguments for the purported harms of conflict of interest policies and regulations are fanciful and data-free. No one is proposing that “we prevent the dissemination of expertise, thwart productive collaborations, or dissuade patients from taking effective drugs,” or allow “true experts to be replaced?on advisory panels, as authors of reviews and commentaries, in other capacities of authority by people whose key asset is being conflict-free.” Where is the evidence of “a loud chorus of shaming,” or “a stifling of honest discourse,” or that “the license to trample the credibility of physicians with industry ties has silenced debate?” Silliness and fear mongering about straw men are masquerading as scholarly analysis.

In 2014, under the Open Payments program [the Physician Payment Sunshine Act which is part of the Affordable Care Act], the Centers for Medicare and Medicaid Services in the United States published 4.45 million financial transactions from healthcare industries to physicians and teaching hospitals over just the last five months of 2013; the total value was nearly $3.7bn [£2.4bn; €3.4bn]. When full data for 2014 are reported later in 2015, the amounts may well exceed $9bn. Drug and device companies are investor owned businesses that are required to maximize profits by any legal means. These companies are not charities, so they expect to get something in return for all the largesse; the evidence is that they do, and it is naive to explain the situation otherwise.

Put simply, financial conflicts of interest in medicine are not beneficial, despite strained attempts to justify them and to make a virtue of self interest. Unmistakably, collaborations between academia and industry can speed medical progress and benefit patients. Such partnerships, however, can flourish with far less money in aggregate flowing from drug and device manufacturers to physicians and their institutions, and without the web of other lucrative ties between industry and physicians that lack a clear scientific or medical purpose. There are few reasons for physicians and other investigators to have financial associations with industry other than research support and bona fide consulting related to specific research programs and projects. Physicians who develop products and hold patents or receive royalties should not evaluate the product. Other types of payments, such as speakers’ and other personal fees, payments to be ghost authors of review articles, and ill defined consulting arrangements, distort physicians’ work and undermine our independence, as has been repeatedly documented. And there are no excuses for outright gifts, such as meals, travel, lodging expenses, and entertainment.

Editorial responsibility
In 1984, the late Arnold S Relman, then the NEJM’s editor in chief, instituted the first conflict of interest policy at any major medical journal.10 The policy required authors of research papers to disclose all financial ties they had to health industries, and if the ties were deemed significant they were published. In 1990, Relman extended the policy to prohibit authors of editorials and review articles from having any financial interest in a company [or its competitor] that was discussed in the article, since these types of manuscripts do not contain primary data but rely exclusively on the authors’ judgment in citing and interpreting the literature.11 As Relman’s successors, two of us [JPK and MA] continued these policies. We found that it was sometimes difficult, but nearly always possible, to find outstanding authors with the needed expertise and without a conflict of interest to write editorials and review articles. In 2002, however, after Drazen succeeded Angell, the policy was weakened, so that it only applied to authors with “any significant financial interest in a company [or its competitor] that makes a product discussed in the article.” To its credit, The BMJ has taken the opposite approach and implemented a zero tolerance policy on educational articles by authors with industry ties.

The privilege to serve as an editor of a major medical journal is accompanied by the responsibility to provide leadership on the critical issues that define the profession. How medicine responds to conflicts of interest and earns the trust of the larger society in which we exist is one such issue. In 1990, it was a bad idea for authors of editorials, review articles, and other opinion articles in medical journals to have financial conflicts of interest. A quarter of a century later, it is a very bad idea. The articles by Rosenbaum and the supportive editorial by Drazen could presage a further weakening of the conflict of interest policy at the NEJM, or they could serve as a wake-up call for all medical journals and the profession. It is time to move forward, not backward.

Revisiting the commercial-academic interface in medical journals

by Elizabeth Loder, Catherine Brizzell, and Fiona Godlee

British Medical Journal. 2015 350:h2957

[full text online]

The New England Journal of Medicine goes on an ill advised journey
Public trust in the pharmaceutical and biotechnology industry is low. Many practising physicians share that mistrust and are inclined to discount the results of otherwise sound studies that are industry funded. There are good historical reasons to be sceptical. But has suspicion degenerated, as some have charged, into “mindless demonisation?” The New England Journal of Medicine [NEJM] seems to think so. It has published a series of commentaries and an editorial suggesting there have been serious negative consequences of strict, “oversimplified” conflict of interest and disclosure policies, including the development of a “hostile climate” and “loss of trust.” Editor in chief, Jeffrey Drazen, says the “divide” between academic researchers and industry is not in the best interests of the public because “true improvement can come only through collaboration.”

A close reading of Drazen’s editorial suggests he is having second thoughts about policies put in place by many journals—including The BMJ—that make it “harder and harder for people who have received industry payments or items of financial value to write editorials or review articles … Having received industry money, the argument goes, even an acknowledged world expert can no longer provide untainted advice.” These policies, he says, came about “largely because of a few widely publicized episodes of unacceptable behavior.” He urges revisiting of the reasons that “medical journal editors remain concerned about authors with pharma and biotech associations.”

We are deeply troubled by a possible retreat from policies that prevent experts with relevant commercial ties from authoring commentary or review articles. The pharmaceutical and biotechnology industries may well be our medical saviours, but that is not a good reason to return to past practices. Such policies were not motivated solely by a few events, as Drazen asserts, but by recognition of extensive, systemic problems. These problems are far from solved, including internationally, as shown by recent events in India and China.

Checks and balances remain important
We agree that people with industry affiliations may be capable of expressing impartial views about matters affecting the commercial interests with which they are associated. Journal readers and editors, however, have no reliable way of identifying which industry affiliated views are disinterested and which are inappropriately influenced by commercial considerations, particularly in subtle ways. Bias is not always overt or easily detected. Authors with industry ties may be likely to approach a topic from a perspective shaped by their associations, so that their views will reflect industry assumptions, priorities, and preferences. The existence of academic and non-financial conflicts of interest does not reduce the need to be wary about conflicts that arise out of the powerful economic incentives associated with industry connections.

In our view, no one has such superior knowledge that he or she is the only one qualified to write an article on a subject. Checks and balances are important in any system. In the case of medical evidence, they should be based on the assumption that it is a mistake to combine evidence production and appraisal functions in a single person or group. Some academics must work closely with industry to develop and commercialise new medical treatments, but they should not also author editorials, reviews, or guidelines that appraise them. These are different professional responsibilities, and they clash.

The stakes are high. Editorials, reviews, and guidelines legitimise medical knowledge and shape clinical practice. Society needs a group of people who can evaluate medical evidence completely free of the appearance of commercial taint. One goal of The BMJ’s zero tolerance policy on education pieces by authors with industry ties was to offer unconflicted authors “prominence and visibility.” The success or failure of this policy can be evaluated only after the distinction between these different responsibilities—developing treatments or evaluating their place in practice—has been established long enough to influence the career choices of young doctors.

Disclosure does not solve the problem of bias and might make it worse. Advisers who disclose conflicts may subsequently feel more comfortable giving biased advice, a phenomenon called moral licence. Those who receive advice from a biased adviser often do not discount it sufficiently. Finally, “requiring disclosure is much easier than changing the status quo … I’d rather tell you I’m on the gravy train than get off it.”

We don’t find much to agree with in NEJM’s anecdotal analysis, but we do agree that criticism of the pharmaceutical and biotechnology industry is often reflexive and unfair. In fact, industry leads academia in complying with trial registration and reporting requirements. Many companies have embraced the open data movement. These are good things, but improvements in obvious problems should not be a pretext for regressive change. Instead, we should encourage all medical journals to separate the functions of evidence generation from those of appraisal. Policies that prevent experts with commercial ties from participating in evidence evaluation institutionalise this protection instead of making it optional. They are an important safeguard against bias and a defence against the perception of a “trial-journal pipeline” in which “companies treat trials and journals as marketing vehicles.” We agree with Steinbrook and colleagues that journal editors have a responsibility to lead on this issue and that “financial conflicts of interest in medicine are not beneficial.” It is a mistake by NEJM to suggest that rigorous standards should be revisited. To do so would undermine the trustworthiness of medical journals and be a disservice to clinical practice and patient safety.

• NEJM reignites conflict-of-interest debate with reader poll
• Former NEJM editors slam “backtrack” on conflict of interest