Medical journals and industry tiesBMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g7197 (Published 28 November 2014) Cite this as: BMJ 2014;349:g7197
All rapid responses
Thank you for these responses.
There seems to be some confusion about the extent of The BMJ’s new policy as set out in our recent editorial.(1) The policy applies only to editorials and clinical educational articles designed to guide patient care and does not extend to other types of article published in The BMJ. Moreover it precludes only those authors with a relevant commercial interest—one that relates directly to the topic of the article. Authors with relevant ties to commercial companies will still be welcome to write articles that are not editorials or clinical education articles for the journal. They may also be invited to write clinical education articles if their commercial interests have nothing to do with the article’s topic. In answer to Richard Smith,(2) GPs will still be invited to write articles on clinical care, and if Smith happened to be an expert on vaccines, the fact that he is also chair of Patients Know Best (a commercial company that provides software for patients to manage their own health data) would not preclude him from writing an editorial or review article on a new vaccine for us. He might also write non-editorial opinion pieces on the rights and wrongs of access to data, with full declaration of his interests.(2) But we would probably not invite him to write an editorial on how best to give patients access to their data, as readers might have difficulty in perceiving such a piece as truly independent of his affiliations.
And yes this is about perception of bias as much as reality. The research on perceived bias is, as Forsyth points out,(3) limited. But there is longstanding evidence of actual commercial bias in reviews and commentaries.(4) (5) We greatly value our authors but we make no apology for prioritising the needs of our readers and their patients. Authors with industry ties have many other outlets if they wish to publish clinical educational articles on topics related to those ties.
In his full response online, Smith raises concerns about academic prejudice as a potent source of bias, which our policy does not address. And Coffey and Prendergast make a similar point—that authors may take controversial rules in order to get published.(6) We agree that this is a problem. Our defences against it, on behalf of readers, are pre-publication peer review to ensure as far as possible that articles are properly evidence based, and the fact that The BMJ’s authors are accountable to readers through our rapid responses should they fail to keep their prejudices in check.
Coffey and Prendergast also ask why the NHS should pay for what may become a marketing opportunity for manufacturers. I would answer that the NHS should be funding, and in this case probably is funding, good independent science regardless of the results.
Yes, some authors will not properly declare their interests, as the New England Journal of Medicine found when trying to enforce a similar policy in the 1990s. But we are using rigorous processes(1) that we hope will make this less likely. We hope too that authors will choose complete honesty up front rather than risk post-publication embarrassment in the journal’s rapid responses. This is one of several aspects of this experiment that we will audit and report back on in the journal.
As for our own conflicts of interest, The BMJ sits within a commercial company, but the journal’s editors do not own shares or have other interests in industry as defined in the editorial. The BMJ gets some of its revenue from pharmaceutical advertising and sponsorship but as with all good journals there is strict purdah between the editorial and commercial teams.
Finally, Forsyth chides us for taking the law into our own hands.(3) To my mind, this is a journal editor’s job, within reason: to set policies that in small ways push and prod the worlds of clinical and academic medicine towards what we judge will be a better future.(7) And yes, I hope this will eventually mean no more promotional pens.
Competing interests: None declared.
1 Chew M, Brizzell C, Abbasi K, Godlee F. Medical journals and industry ties. BMJ 2014;349:g7197. (28 November.)
2 Smith R. Reasons to be cautious about competing interests. BMJ 2015;350:g7754.
3 Forsyth S. The BMJ should not take the law into its own hands on competing interests. BMJ 2015;350:g7766.
4 Stelfox HT, Chua G, O’Rourke K, Detsky AS. Conflict of interest in the debate over calcium-channel antagonists. N Engl J Med 1998;338:101-6.
5 Barnes DE, Bero LA. Why review articles on the health effects of passive smoking reach different conclusions. JAMA 1998;279:1566-70.
6 Coffey S, Prendergast BD. Public healthcare systems should not pay for private profit. BMJ 2015;350:g7774.
Competing interests: No competing interests
The recent Editorial in The BMJ (1) announcing the new policy for commissioned educational content is welcomed. Some of us and other epidemiology colleagues have already responded (2) to point out that the concerns addressed through the new policy of The BMJ are as relevant for public health as for clinical content.
However, for public health the dilemma of how far upstream to go hunting for commercial or other vested interests, which might constitute a significant conflict, is a daunting challenge. The challenges are not new to those working to secure better health for populations all over the world.
According to The BMJ’s new policy, authors of educational content commissioned by the BMJ must have no links with companies or organizations that have commercial or other conflicts, which would, presumably exclude pharma and device manufacturers, for example. But, what about educational content sponsored by companies or entities that, at first glance, may meet the policy requirements, but which many might think have conflicts of interest of a different sort (for example, the company itself has either close or more distant associations with the tobacco or arms industries, or with anti-vaccine activism)?
Would The BMJ be concerned if its commissioned educational content came from authors twice or more times removed from a source of income or a vested interest that is explicitly at odds with public health values and goals?
How far should The BMJ’s policy go in this regard? Our purpose is not to be judgmental or critical of BMJ policy, whose intentions we welcome. We ask in the context of the development of a Conflict-of-Interest and Disclosure position statement currently in preparation by the International Joint Policy Committee of the Societies of Epidemiology and would welcome the views of The BMJ’s editorial team and of its readers to inform our deliberations.
Fiona Sim*, Amy Davis, Eduardo Franco, Bruce Lanphear, Leah Phillips, Colin Soskolne
1. Chew M, Brizzell C, Abbasi K, Godlee F. Editorials Medical journals and industry ties. 28/11/14. http://www.bmj.com/content/349/bmj.g7197.full.pdf+html
2. Soskolne Cl, Al-Delaimy WK, Burns K, Finch MR, Gaudino JA Jr, Lanphear B, Oremus M, Phillips L, Ruff K, Weiss SH, Wing S. Medical journals and industry ties: Competing interests in epidemiology. 8/1/14. BMJ 2015;350:g7744
Competing interests: All the authors are members of the International Joint Policy Committee of the Societies of Epidemiology Working Group on Conflict of Interest and Disclosure. Fiona Sim is Joint Editor in Chief, 'Public Health'.
When you hear a politician from UKIP start a sentence with "I'm not racist, but...", you probably expect that the rest of the sentence is going to contain a reference to Bongo-Bongo Land or similar. This is a classic example of what discursive psychologists call a "disclaimer" . A disclaimer is a device used when you know that what you are about to say is likely to be considered unreasonable, in an attempt to soften the blow and dissociate yourself from such an unreasonable attitude.
So it is when Chew et al write "Nor do we want to be anti-business just for the sake of it. "
It is hard to see what this policy is based on other than being anti-business for the sake of it. There is, of course, an attempt to present this policy as being evidence based by including lots of sciency-looking references. However, none of those references provides evidence that banning authors with financial ties to industry will improve the quality of articles.
References 8 and 9 in Chew et al's paper show that readers believe articles written by authors with industry ties to be less believable, but it doesn't mean that the articles really are. All those studies show is a bias in readers, not a bias in authors. To continue the UKIP analogy, this would be somewhat akin to a UKIP politician stating that many voters believe that immigration harms the economy (despite the well documented benefits of immigration on the economy) and using that as a justification to curb immigration.
I am not aware of any evidence showing that papers written by authors with industry ties are more likely to be biased. And even if such evidence existed, would it justify a blanket ban? There is research showing that papers with a first author from a low or middle income country are more likely to be fraudulent . However, I'm sure Chew et al would agree with me that that does not justify a ban on authors from low or middle income countries. It's interesting to think about the differences between those two situations.
I am pleased to see Chew et al write "We will study progress and report back." I do hope that their study of progress will include an assessment of whether the policy has achieved its stated aims of making articles less biased.
But until I see such evidence, this policy looks to me more like an ideology-based policy than an evidence-based one.
1. Hewitt JP, Stokes R. Disclaimers. American Sociological Review 1975;40:1-11.
2. Woolley KL, Lew RA, Stretton S et al. Lack of involvement of medical writers and the pharmaceutical industry in publications retracted for misconduct: a systematic, controlled, retrospective study. Curr Med Res Opin 2011;27:1175-82
Competing interests: I work for a company that provides contract research services to pharmaceutical and medical device companies. For many years I ran my own company that did likewise.
We congratulate The BMJ! You were among the first medical journals to have implemented a policy on declaring financial competing interests. Now you are a pioneer in recognizing that transparency through such declarations is not a panacea, but requires some adjustments in your policy.
The adjustments you have made relate specifically to articles commissioned by The BMJ and apply to medical education articles that will, in future, be commissioned solely from authors without industry funding (see http://www.bmj.com/content/349/bmj.g7197.full.pdf+html).
Reducing the influence of financial competing interests in clinical education and practice guidelines is essential to promoting best practices that put the public interest ahead of personal gain. This also protects the ability of scientific researchers and health practitioners to legitimately inform policy and the public based on the best possible evidence. We look forward to learning of your experience in having taken such a bold step to mitigate biased information resulting from competing interests.
What you have done extends to applied research of the kind conducted by epidemiologists where bias must also be vigilantly evaluated and mitigated. Indeed, there has been increased awareness of the adverse impacts of biased science through financial competing interests, culminating in a compelling 2013 report on this issue by The Center for Public Integrity (www.publicintegrity.org).
Although work on ethics in epidemiology has been ongoing since the 1980s, more intense deliberations and actions are now underway by both the International Society for Environmental Epidemiology (ISEE) and the International Joint Policy Committee of the Societies of Epidemiology (IJPC-SE).
The ISEE is focusing on the transparency dimension of the problem by developing stronger accountability criteria for declaring financial competing interests. The IJPC-SE is developing a broad-based Position Statement on Conflict-of-Interest and Disclosure to be completed in 2015 (www.ijpc-se.org/initiatives.php). The IJPC-SE’s focus extends beyond company funding and affiliations.
We will pay particular attention to news on your experience as we advance our own initiatives.
Competing interests: No competing interests
We disagree with the BMJ’s new policy of restricting educational article to those without perceived competing interests, as detailed in the editorial “Medical journals and financial ties”.
Our experience highlights an unintended consequence of this line of thinking. We are performing a population cohort study examining the prevalence of valvular heart disease. As a substudy, we wish to determine the performance of auscultation as a predictor of valve disease found on echocardiography. To avoid having to declare a competing interest and because we were lucky enough to have charitable funds available, we decided to purchase the equipment (electronic stethoscopes) ourselves. However if we do show that auscultation with this brand of electronic stethoscope picks up significant valve disease, then we have provided the manufacturer with an excellent advertising soundbite. Why should the National Health Service, which funds this study via the National Institute for Health Research, pay for marketing opportunities for private companies?
There are other reasons to oppose the ban. There will be a disproportionate effect on researchers in lower income countries without the levels of governmental and charitable funds available for research in higher income countries. Even in higher income countries, it is virtually impossible to bring a new treatment to clinical use without industry support due to the cost of doing so. Is the BMJ going to ban any researcher successful enough to develop a clinically useful treatment?
Links with industry are all around us. Is the advertising that appears in the BMJ having a conscious or unconscious effect on the editors, who ultimately decide what appears in the journal, and who, despite a financial and professional interest in the BMJ remaining profitable, have declared no competing interests? The unwritten competing interest for most researchers, especially early career researchers, is that their future financial livelihood is entirely dependent on having articles published, especially controversial articles that are likely to receive more citations. It is rare for an author with a link to industry to have such a direct benefit from publication.
1. Chew M, Brizzell C, Abbasi K, Godlee F. Medical journals and industry ties. BMJ 2014;349:g7197.
2. Coffey S, D’Arcy JL, Loudon MA, Mant D, Farmer AJ, Prendergast BD. The OxVALVE population cohort study (OxVALVE-PCS)--population screening for undiagnosed valvular heart disease in the elderly: study design and objectives. Open Hear 2014;1:e000043.
Competing interests: SC is an early career researcher.
You asked for my opinion on this policy before you introduced it, and as you are encouraging debate I thought that I would share some of the points I made.
I agree with your general point that disclosure on its own is not an adequate response to competing interests. There has to be a point at which the competing interest is so extreme that the person has to be ruled out of the article, review, decision, or whatever the activity. But defining that point is notoriously hard.
1. The subheading of the editorial says “zero tolerance,” but you presumably don’t mean zero tolerance in that that would rule out an author who had been given a pen, a pad, or lunch by a company. That would be virtually all doctors.
2. I asked how you would define “industry,” and you’ve now done so as “companies producing pharmaceuticals, devices, or tests; medical education companies; or other companies with an interest in the article topic.” I, for example, am the chair of Patients Know Best, a start up that is making great progress, as reported in the BMJ, in giving NHS patients control over their own records. I asked if this was industry in your definition, and clearly it is. This may be a loss as many young doctors are starting companies, it’s very fashionable.
3. I wonder too if you haven’t excluded yourselves from writing editorials. The BMJ (the company as opposed to the journal) may be wholly owned by BMA but is very much a for profit medical publisher and educator and provider of decision support products and services.
4. You have chosen to concentrate on “industry,” but I and others have argued that impartiality does not exist. (1) For example, I suggest that GPs, who are in the private sector and for profit, have a huge financial interest in how general practice is organised. They don’t seem to qualify as “industry,” but I suggest that their financial competing interest is much sharper than most individuals working for companies that sell to general practice—because their compensation is unlikely to be directly affected by changes in general practice. So will general practitioners no longer be able to write editorials and other pieces on general practice?
5. I’ve been reading The Big Fat Surprise (2) for an article in the Christmas BMJ, and it makes clear how industry has had big effects on nutrition science and policy. But the prejudices of individual researchers and of the whole “nutrition profession” did much more damage.
6. You mention how the NEJM abandoned its policy, which, as I remember, was more generous than yours in that it allowed some funding from industry. One reason for abandoning it was that they had trouble finding authors for certain subjects, as you mention, but another was that they were often deceived. The LA Times published articles on NEJM authors who supposedly didn’t have conflicts of interest but in fact did. In other words, it’s hard to police the policy.
7. A broader point is that your policy might disengage the BMJ from companies and more importantly from the many academics closely linked with them. I’ve heard it argued that modern universities should be about not just scholarship, teaching, and research but also entrepreneurship—and many of the top universities (Stanford, Cambridge, MIT) already are. The point, as Marx said, is not just to study the world but to change it. Scaling up innovations (as I’m trying to do) usually means moving to a “business model”—because it’s a sustainable model in a way that grant funding and sponsorship are not. I don’t think that it would be good for the BMJ to be seen as “anti-business.”
I’m a “yesterday’s man” so what I think matters little, and I hope that your policy succeeds.
1 Smith, R, Feachem R, Feachem NS, Koehlmoos TP, Kinlaw H. The fallacy of impartiality: competing interest bias in academic publications . J R Soc Med 102(2): 44-45; doi:10.1258/jrsm.2009.080400
2 Teicholz N. The Big Fat Surprise. Melbourne: Scribe, 2014.
Competing interests: I'm employed by and have shares in UnitedHealth Group. I'm the chair of and have shares in Patients Know Best. I'm on the advisory board of Eliaso and have shares. I'm the chair of the board of trustees of icddr,b, and when I travel to Bangladesh I am offered a per diem, but so far I've always signed it over to the Dhaka Hospital. I'm the chair of the Cochrane Library Oversight Committee and belong to several other boards, but these are all unpaid positions. Sometimes I get paid for writing or speaking, usually not very much, but I was generously paid to speak at the centenary of the Medical Journal of Australia. I have a pension from the BMA, which has substantial financial input from the BMJ.
Why should researchers or clinicians whose professional interest leads them to collaborate with industry, be forever stigmatised by journal editors on the assumption that their educational contribution will be influenced by the needs of industry rather than reflect their own professional opinion?1. This proposal appears to be based on the personal beliefs of the BMJ Editorial team and those of some of their readers rather than explicit evidence. It could be construed that their action reflects a prejudice against industry rather than an act of altruism.
Industry provides the backbone to healthcare; it is responsible for the design and development of hospitals, operating theatres, imaging equipment, and sophisticated surgical instruments and many other key elements of health care provision, including medicines. It is imperative that scientists and clinicians collaborate with industry on these developments if we are to continue to have a health service that is innovative and meets the need of new health challenges. The major global health crisis at present is the Ebola epidemic and many researchers and clinicians will be working with industry to develop an Ebola virus vaccine which will reduce the risk of widespread mortality and morbidity caused by this potent virus. Will those researchers and clinicians now have educational articles on the management of Ebola virus infection rejected by the BMJ because they have ties with industry?
The evidence provided by the authors relates to a paper in which the previous Editor and other BMJ staff were cited as co-authors and the research was funded by a BMJ research grant2. The key finding of this study, which had a 59% response to a questionnaire, was that reader’s validity and believability ratings were significantly lower in the “financial statement” group than in the “none declared” group. The second reference had a similar design with a 53.5% response to a questionnaire and it was reported that physicians were less willing to prescribe drugs that were studied in industry-funded trials compared to drugs that were studied in NIH-funded trials3. Is it surprising that readers are more likely to be critical of an industry funded study if they are asked to compare papers where there is no funding source mentioned, funding by the National Institutes of Health (NIH), or funding by a pharmaceutical company, with financial involvement in that company on the part of the lead author. This does not appear to be a level playing field and the US paper correctly refers to the “perception” of industry related trials and the effect that this may have on the translation of clinical research into practice.
It has recently been acknowledged that research misconduct can occur at all levels, including the individual researcher, the department, the institution, the journals and the funding bodies4. Bias in educational articles may reflect many influencing factors such as associations with research funding organisations, government bodies, politics, religion, culture, or simply being employed by a university that adopts a policy of “publish or perish”. It seems that the key message is that all authors need to declare all potential conflicts of interest, editorial teams and peer reviewers need to continue to carefully review papers for their scientific content, validity and relevance to the journal, and readers should be allowed to make their own judgement on the value of the article to them. All contributors to medical journals need to adhere to professional standards, including those relating to probity, and if there is evidence that these standards are being breached referral to our professional bodies or our law masters should be the way forward.
However, it is concerning that the BMJ has decided to take the law into its own hands and the principle appears to be that you are guilty until proven innocent. What is especially confusing is that the back cover of the current edition of the BMJ has a full page advertisement from an international pharmaceutical company, which makes significant health claims. Surely this represents potential organisational conflict of interest for the BMJ as it will be receiving payment from the pharmaceutical company for the advertising space, however, each of the BMJ authors do not declare any conflicts of interest.
1. Chew M, Brizzell C, Abbasi K, Godlee F. Medical journals and industry ties. BMJ 2014;349:G7197.
2. Schroter S, Morris J, Chaudhry S, Smith, R, Barratt H. Does the type of competing interest affect readers’ perceptions of the credibility of research? Randomised trial. BMJ2004;328:742.
3. Kesselheim AS, Robertson CT, Myers JA, Rose SL, Gillet V, Ross KM, et al. A randomized study of how physicians interpret research funding disclosures. N Engl J Med 2012;367:1119-27.
4. Sarwar U, Nicolaou M. Fraud and deceit in medical research. J Res Med Sci 2012;17:1077-1081
Competing interests: Have previously been involved in national planning of child health services in Scotland and have a longstanding research interest in infant and child nutrition with funding from government, charitable organisations and industry. Have recently been undertaking consultancy work and at present have contracts with the UK Department of Health and also with DSM, a company that produces nutritional ingredients.
I applaud the BMJ for this decision. I for one will place greater trust in educational articles published in the BMJ in the future. Many specialists refuse industry funding so they can maintain their independence and trust in their professional and academic communities - there are lots of clever, learned researchers and academics but independence is worth it's weight in gold (i.e. prior to the recent commodity crash). Well done!
Competing interests: No competing interests
Very eloquent argument for stated intention.
But it is important that academic journals themselves do not unwittingly become corrupted and thus impaired by the very virus they seek to control/eliminate;namely 'bias'.
Bias o the part of the journals against authors with suspect links causing their peripheralisation may lead in effect to
' casting pearls into the wilderness ' thereby robbing readers and potential users of a valuable resource having thrown the baby away with t he bath water.
The challenge really is how to get rid of the dirty bath water without tossing out the bridge to a valuably scientific future
The bias by 'corrupted ' authors is not any less harmful in my view than the 'over correction' bias by journals to the wholesome process of intellectual transactions.
The one defiles potential knowledge, the other quarantines potential knowledge, net victim in both approach being potential knowledge.
It is not unreasonable to imagine that industries court and commission the best in the fields to do these kinds of jobs and to have publishing houses keep such bests at arms length as a rule rather than discretion is too un selective a filter ; t he sensitivity and specificity inherent in such a path is too abysmally low to recommend it in the very kind of science at hand.
Let the authors write; interests or not! And let anyone from the other side with any meritorious counter argument ' cast the first stone' to alert to any dangers in protagonsts' viewpoints.
And let the stronger pugilist prevail.
What is intellectualism if not contention between all ideas with unfettered entry into the market place, the wise checkmating the unwise.
Competing interests: No competing interests