Are medical journals reliable sources of objective information, or do they, at times, act as shills for the pharmaceutical industry and other interests? We believe the latter after reading a Perspective article on drug pricing in the New England Journal of Medicine (NEJM) that presented the perspective of the pharmaceutical industry on what drugs should cost without explicitly revealing the industry ties of its authors.
Editorial content in journals is expected to provide objective information about medical science, care, and health policy. That objectivity is threatened by authors of editorials, review articles, and other “perspective” pieces with critical conflicts of interest due to financial associations with the topics being discussed. To prevent this, or at least limit it, the International Committee of Medical Journal Editors advocates transparency about such conflicts.
Many journals, including NEJM, have adopted such policies, and collect information from authors on their conflicts. For research articles, they judge that to be sufficient. For editorials and reviews, they specify their level of tolerance for relationships to companies with financial interests in the topic. Many journals also publish perspective pieces that are more akin to editorials but that are not always explicitly mentioned in conflicts disclosure policies. A particularly good test of whether these policies solve the problem of conflicts of interest occurs when they publish articles on topics with major implications for corporate players in health care.
The NEJM Perspective that caught our attention was “Drug-Pricing Debate Redux — Should Cost-Effectiveness Analysis Be Used Now to Price Pharmaceuticals?” published in November 2021. It showed us that the journal’s policy about transparency is not particularly transparent, and its policy about limiting the level of conflict of interest for writers of editorials is not enforced for at least some Perspective pieces.
In it, authors Peter J. Neumann, Joshua T. Cohen, and Daniel A. Ollendorf contend that the appropriate price “is a drug’s value-based price, measured appropriately.” This turns out to be unrelated to costs but rather to how much patients or society value any health benefit. The authors argue: “If we pay more than the value-based price, some of those resources could have been better spent elsewhere to improve health. But paying less… may adversely affect patients today (e.g., companies may walk away from the market, leaving patients with less effective alternatives) and may fail to stimulate new drug development in areas where it is most needed. Value-based pricing informed by cost-effectiveness analysis will not achieve the lowest prices, but it will help society to spend its resources more efficiently to improve health.”
This commentary echoes the perspective of pharmaceutical manufacturers, whose price increases regularly exceed the inflation rate and the actual rise in production and distribution costs. “Value-based pricing” frees them to justify higher prices by tying them to “the maximum payment that an individual (or an insurer or government) is willing to make … for the value patients impute to its effect” — such as the economic value of years of life gained from the treatment, and also to what one pharmaceutical company describes as “the investment required to fund further research and development.”
In support of this point of view, companies such as Precision Health Economics consult extensively with industry to identify the right “price points” for products, an approach criticized by others, including the late Princeton economist Uwe Reinhardt, who observed: “If you did value pricing and say it’s OK for the drug companies to charge up to what the patient values his or her life to be, you are basically saying that the pharmaceutical companies can take your savings.”
Drug pricing is a matter of considerable public policy concern. In highlighting it, NEJM has an obligation to readers to make clear any potential influences on the opinions expressed — like the backgrounds and disclosures of the three authors.
Like many journals, NEJM collects conflict-of-interest disclosures. Some journals, such as JAMA, include footnotes listing all disclosed associations. Others like NEJM do not publish disclosures in print but include online links. In neither instance do they disclose the dollar amounts.
We believe this is not sufficient.
If a journal did not publish commentaries or editorials by authors with any conflicts of interest regarding topics discussed, that would suffice. Thus, in 1990, NEJM’s editor-in-chief, Arnold Relman, wrote that readers of review articles and editorials “must rely on the objectivity of the author. When authors have a financial as well as a scientific interest in their subjects, questions inevitably arise that cast doubt on this presumption of objectivity. It is a problem that can and should be avoided by selecting authors who have no financial stake in the subjects they write about.”
The journal retreated from that policy in 2002. Its editorial policy now states: “The Journal expects that authors of such articles have no significant financial interests [emphasis is ours] in any biomedical company relevant to topics and products discussed in the subject they are reviewing or the article on which they are commenting. When prospective authors do have financial ties to disclose, Journal editors decide whether they are relevant to the subject and whether they are de minimis.” Although this statement does not explicitly mention Perspective articles, it is directed at pieces containing interpretations and opinions, a category that clearly includes Perspectives.
To translate: Authors can have financial relationships below a “de minimus” threshold, initially defined as $10,000 annually per company. To paraphrase the late Senator Everett Dirksen, that could add up to a lot of money. Relman (who was NEJM’s editor-in-chief from 1977 to 1991) warned that such payments could be large enough to influence authors’ attitudes and “Editors are on safer ground when they prohibit such conflicts of interest altogether rather than attempt to manage them by establishing flexible guidelines and negotiating with authors.”
The Institute of Medicine has defined conflicts of interest as “circumstances that create a risk that professional judgments or actions regarding a primary interest will be unduly influenced by a secondary interest.” In this sense, all financial relationships constitute conflicts of interest. It should not be left to authors to judge which financial relationships merit disclosure.
In this context, how has NEJM’s policy worked out?
It takes work for readers to understand that the authors of the article in question were long-time consultants to pharmaceutical and biotech companies. The online version includes a link to their disclosure forms, but it is unlikely that many readers check out such links when reading online. Those reading the print version are even less likely to track them down. All readers likely would presume that NEJM’s editors had exercised due diligence in evaluating authors of this article.
That presumption would be wrong.
The link reveals scores of financial associations with manufacturers of pharmaceuticals, biologics, and medical technologies; professional associations of such companies; and companies that do “value pricing” for the industry.
Cohen had 48 financial disclosures, all from companies with substantial interest in pricing of pharmaceuticals, biologics, or other medical products, including AbbVie, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Genentech, Gilead Sciences, GlaxoSmithKline, Janssen Biotech, Johnson and Johnson, Novartis, Pfizer, PhRMA Foundation, and Precision Health Economics. Of Neumann’s 18 disclosures, 12 were of companies with obvious interests in drug pricing, including Bayer, Merck, and Sanofi Pasteur. Four of Ollendorf’s eight disclosures represented similar conflicts, including Amgen, Eli Lilly, and Sunovion. (Ollendorf once wrote in a First Opinion essay about achieving “some success as a research consultant to the pharmaceutical and biotech industries.”)
The disclosure forms do not report the magnitude of financial relationships. No payment information was available at Open Payments, a valuable resource for identifying conflicts of interest and their magnitude, because it includes only physicians, and none of the three authors are M.D.s.
We believe NEJM failed its readers by publishing this article without making it clear to readers that the authors had profound conflicts of interest directly relevant to the content involving corporations advocating for and benefitting financially from the point of view expressed. In this instance, the conflicts are so profound that a more appropriate decision would have been to reject the article or, at the very least, provide details of conflicts both in print and directly accompanying the article online, along with details of why the Journal chose to publish it despite the conflicts.
We sent a copy of this essay to NEJM’s editors and asked them if it contained any errors. They did not reply to our request.
It may not be practical to publish in the print version of journals details of all financial relationships for authors of every article, but all journals ought to do that when topics and opinions expressed clearly parallel financial interests of entities paying the authors. Furthermore, such disclosures should reveal the dollar amounts of relationships with these entities, which the forms currently lack.
The magnitude of so-called de minimus associations (those supposedly not large enough to influence what an author would write or say) needs to be explicit. However, we question the adequacy of having editors assess whether disclosed relationships are de minimus. Disclosure alone does not guarantee that submissions are free from critical bias. With true full disclosure, readers might expect that journal editors would not publish pieces in which conflicts of interest are profound and directly relevant to content. When they do publish such articles, as NEJM did with the Perspective commentary in question, they are misleading their readers.
There is no justification for failing to provide full transparency regarding the relationships of authors to companies with financial interests related to topics they write about. All journal editors could — and should — do that. Readers deserve no less. They could then see for themselves if the editors’ judgment was correct as they evaluate the points of view presented.
Martin F. Shapiro is a professor of medicine at Weill Cornell Medical College and distinguished professor emeritus and former chief of the division of general internal medicine and health services research at UCLA. Sidney M. Wolfe is an internist and the founder of and senior advisor to Public Citizen’s Health Research Group.