Education And Debate For and against

Why journals should not publish articles funded by the tobacco industryForAgainst

BMJ 2000; 321 doi: http://dx.doi.org/10.1136/bmj.321.7268.1074 (Published 28 October 2000) Cite this as: BMJ 2000;321:1074

That the tobacco industry is allowed a platform through scientific journals is a controversial issue. The Cancer Research Campaign already withholds research grants from academic departments closely linked with the tobacco industry; here it argues that respected journals should not publish articles funded by the industry. Such a notion is challenged by two editors, who prefer a policy of transparency and disclosure rather than prohibition.

For

  1. Jean King (jking{at}crc.org.uk), director of education funding.
  1. The Cancer Research Campaign, London NW1 4JL
  2. a Western Journal of Medicine, San Francisco, CA 94105-1911, USA,
  3. b BMJ, London WC1H 9JR

    The arguments for scientists refusing to accept funding from the tobacco industry and, by extension, journals refusing to accept publications arising from such funding, fall into three main categories (box 1).

    Box 1: Arguments for refusing funding from the tobacco industry

    Harm to health

    The scale, range, and extent of the harm caused by tobacco far outweigh the effects of any other known product, legal or illegal

    Distortion of scientific findings

    Directly and indirectly the tobacco industry has systematically sought to undermine and misrepresent sound research on a massive scale, while itself funding studies and scientists of questionable credibility

    Other disreputable activities

    The tobacco industry's own documents show a uniquely discredited and disgraced sector that has sought to conceal evidence, recruit minors, and dupe the public and governments alike for many decades

    RETURN TO TEXT

    Harm to health

    The health risks associated with tobacco use have been detailed in many reports and were the basis of the justification for the Cancer Research Campaign's code of practice, which prohibits research groups in receipt of money from the tobacco industry from applying to it for funds (box 2).17

    Distortion and “distraction” of scientific research

    The tobacco industry has a long track record of seeking to cast doubt on good research, as shown in its own documents, now available on the internet owing to litigation cases in the United States. One tactic is to commission studies aimed at discrediting accepted findings: examples include studies that cast doubt on epidemiological findings, especially in relation to passive smoking, or which purport to identify other risk factors for ill health in active and passive smokers, such as a low intake of vegetables. 8 9 These studies are often open to criticism on methodological grounds: use of selective arguments and literature, inconsistency in the inclusion or omission of potentially biasing factors, playing with statistical inference rather than considering overall plausibility and weight of evidence, and drawing conclusions that are unwarranted by the evidence.1012 Such concerns may not prevent these studies from being published in international journals, where the peer review process has failed to address their methodological weaknesses. Often studies are funded through bodies that sound credible, such as the Centre for Indoor Air Research or the European Working Group on Environmental Tobacco Smoke, which have been set up by the tobacco industry. Editors and readers will be unaware of the source of their support.

    Frequently, studies that conflict with orthodox health messages receive prominence in the media far beyond what might be expected, given the relative lack of evidence to support their case or their relatively unknown source.13 This is often owing to public relations activities that have been carefully orchestrated by the tobacco industry. Once a story has appeared in a Western broadsheet, it is almost inevitably picked up in media around the world, again assisted by the efforts of the industry's public relations. The harm to efforts at tobacco control that such stories cause should not be underestimated, especially in developing countries, where the ability of health professionals to respond effectively is hampered by a lack of resources.

    Project “Whitecoat,” a plan to recruit scientists who would express views favourable to the tobacco industry, was also revealed in the industry's own documents.14 Cohen et al detailed further examples of how the tobacco industry has sought to use scientists and institutions to gain credibility and prestige, has supported “distracting” research, and has tried to censor the tobacco control activities of institutions in receipt of funds from the tobacco industry. The authors argue that independent research and internal papers from the tobacco companies both point to an industry that has systematically “suppressed, manipulated and distorted the scientific record” and continues to do so.15 Chapter and verse from the tobacco companies' papers are provided by Ong and Glantz for a multimillion dollar campaign to subvert science, which included: manipulating the media, commissioning “confounder” studies, using third parties to recruit scientists and infiltrate institutions, and, ironically, promoting their notion of “good epidemiological practice.”16

    Other disreputable activities

    Numerous activities by the tobacco industry, and important knowledge it has not acted on, have come to light from within the industry's documents.17 These range from concealing the growing evidence of the association between cancer and the addictiveness of cigarettes, concealing findings that so called “light” cigarettes deliver as much or more tar than ordinary strength brands, employing marketing strategies that specifically target minors, and being aware that a major proportion of their cigarette brands reaching the market, especially in developing countries, have been smuggled (see http://www.ash.org.uk/).

    The tobacco industry has also been active in promoting messages contrary to the health evidence in countries where tobacco control is currently poorly organised. “Hit squads” of so called experts paid by the tobacco industry can do untold harm, especially in developing countries, by delivering misinformation and creating confusion among government ministers and members of the public.18 This can set back legislation and undermine individuals' determination to quit; in other words, it costs lives.

    Although the editorial boards of some prestigious journals may consider that they are unlikely to publish studies funded by the tobacco industry because their quality of peer review will weed out poor methodology, they should recognise that not all publications can operate at such high levels of expertise. Given the harm that can be done by just one distorted study coming out in print, it is the view of the Cancer Research Campaign that editorial boards of respected journals have an obligation to take a lead and set the ethical standard. Some journals, for example, those of the American Lung Association and the American Thoracic Society, and the British Journal of Cancer, have already taken the decision not to publish papers funded by the tobacco industry.

    The “tip of the iceberg” principle does not apply here: although scientists must always be vigilant about potential pressure on them from corporate sponsors of research, the tobacco industry is unique in the scale and range of its dishonest and manipulative tactics. This debate goes beyond the hallowed principles of scientific freedom: the tobacco industry has lost all claim to any such high mindedness. Until it reneges on the promotion and defence of a product causing so much misery worldwide, its funds can have no place in reputable institutions or publications.—Jean King

    Box 2: Extract from Cancer Research Campaign's conditions for research grants

    From 1 April 1999, the campaign will not provide future research support—that is, new grant awards to any institution in which those who are or would be supported by campaign funds are working in such proximity to others supported by tobacco industry funding that there is any possibility or likelihood that facilities, equipment, or other resources will be shared. Funding in a quite different faculty or school of the university or institution is not covered by this protocol.

    Tobacco industry funding includes funds from a company or group of companies engaged in the manufacture of tobacco goods; and funds in the name of a tobacco brand whether or not the brand name is used solely for tobacco goods; funds from a body set up by the tobacco industry or by one or more companies engaged in the manufacture of tobacco goods.

    The following do not constitute tobacco industry funding for the purposes of this protocol:

    • Funding from subsidiary and associated companies, unless they bear the offending name or it is intended or likely that the parent or associated company with such a name will publicise the funding

    • Anonymous donation

    • Legacies from tobacco industry investments (unless the names of a tobacco company or cigarette brand are associated with them)

    • Funding from a trust or foundation no longer having any connection with the tobacco industry even though it may bear a name that (for historical reasons) has tobacco industry associations

    • Donations given to the university for general use by the university entirely at its discretion

    Funding falling within this protocol covers money provided or used for all or any of the costs of the research, including personnel, consumables, equipment, buildings, travel, meetings, and conferences, running costs for laboratories and offices, but not meetings or conferences unrelated to a particular research project.

    RETURN TO TEXT

    Acknowledgments

    I thank David Simpson and Tony Hedley for their helpful comments.

    Footnotes

    • Competing interests None declared.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.

    Against

    1. Gavin Yamey, deputy editora,
    2. Richard Smith, editor.b
    1. The Cancer Research Campaign, London NW1 4JL
    2. a Western Journal of Medicine, San Francisco, CA 94105-1911, USA,
    3. b BMJ, London WC1H 9JR

      Censorship is an easy way to deal with difficult issues. In a morally complex world, silencing the voice of those we distrust is simpler than grappling for the truth.

      Some biomedical journals, such as the Journal of Health Psychology and the two journals of the American Thoracic Society, will not publish research papers funded by the tobacco industry. 1 2 But does this kind of censorship serve the interests of science, journalism, or a free society? We believe not.

      The BMJ chooses papers for publication on the basis of scientific merit, originality, and public health importance, not on their source of funding. We will not impose a blanket ban on research papers from authors funded by any profit making industry. King cites three main reasons why we should. We will firstly respond to her concerns, before outlining the BMJ's policy.

      Harm to health

      The harmful consequences of tobacco are indisputable. But studies sponsored by the tobacco industry could feasibly produce valuable health research data. Although its product is harmful, industry data might have something to add to our understanding of disease processes and should not be censored.

      Distortion of scientific findings

      King believes that the tobacco industry has attempted to manipulate research findings, and there is a wealth of evidence to support her claim. But refusing to publish tobacco sponsored research is equally unscientific and antidemocratic. It would be a type of “publication bias” in itself and would have its own distorting effect. We believe that all study data—regardless of their funding source—should be placed in the public domain, allowing their assessment and use by the international research community. All trials, including those supported by the tobacco industry, should be registered, to prevent suppression of results unfavourable to industry.3

      Other disreputable activities

      Banning research because of the poor reputation of a corporate sponsor would put the BMJ in the precarious position of being an arbiter of morality. We do not believe that every act committed by the tobacco industry has wilfully caused harm. Many research sponsors, including the pharmaceutical and baby food industries, have been involved in “disreputable activities.” Although we should remain alert to these during our editorial decision making process, we do not believe that they should automatically disqualify a paper from publication.

      Graphic Refusing to publish tobacco sponsored research is equally unscientific and antidemocratic

      The BMG policy

      Disclosure by authors

      Many academic institutions have accepted money from the tobacco industry. A survey of universities in Australia found that 30% accepted research funds from the tobacco industry in 1991 or 1992.4 An analysis of a research database in the United Kingdom, which included biomedical research from 1988-94, found that the practice of accepting money from the tobacco industry was widespread. Only one medical school (Glasgow) did not receive support from the tobacco industry for its health research activities.5

      Competing interests are widespread in the biomedical community, and it is folly to think that they can be eradicated from journals. The task is even more elusive if you include personal, political, religious, or academic interests. We therefore ask all authors of research papers, review articles, editorials, and selected letters to declare their interests, and we plan to extend this to other parts of the journal.

      Disclosure of interests is difficult to enforce. Indeed, a recent study showed that 70% of articles from journals with disclosure policies made no mention of potential conflicts.6 We clearly need to do better at prompting authors towards declaration.

      Disclosure in the peer review process

      The main problem with tobacco funded research, apart from its generally poor quality, is that it is rarely hypothesis driven. Instead, it is mainly performed to feed the industry's political agenda. It is no surprise, therefore, that review articles on passive smoking come to different conclusions depending on whether or not the author is affiliated to the tobacco industry.7 The conclusions of original research are also influenced by tobacco sponsorship.8

      What this means for the BMJ is that we need to take extra care when reviewing industry sponsored research. It would be easier to reject it from the start, but at what cost? If the research is of high quality, then it should be submitted to careful review. We ask referees to explicitly declare any financial ties. If a referee has a major competing interest, then we try to obtain another opinion from a reviewer with no declared ties to industry. All BMJ editors have now declared their own interests, which we will shortly publish on our website.

      Graphic All study data—regardless of their funding source—should be placed in the public domain

      But peer review is no guarantee of validity. This is why we believe that readers, the ultimate peer reviewers, should be allowed to further assess the quality of the work.

      Conclusion

      We commend the Cancer Research Campaign for its firm stand against tobacco. However, if we followed its advice to censor the industry's research findings, then we would be adopting the same kind of tactics as the industry itself. We are not alone among journals in adopting a policy of transparency and disclosure rather than prohibition. Even Tobacco Control is willing to publish industry sponsored papers, provided that the links are made explicit (S Chapman, personal communication). The BMJ rarely publishes work funded by the tobacco industry, mainly because we don't receive many submissions. This is in contrast to the huge amount of data in its pages that show cigarettes' lethal effects. Journal editors should fight the tobacco industry “not with censorship but with the abundant evidence on the serious harm that its product inflicts.”9—Gavin Yamey, Richard Smith

      Acknowledgments

      We thank Lisa Bero and Simon Chapman for their helpful comments.

      Footnotes

      • Competing interests Our salaries come from publishing. We are thus likely to be biased towards publishing rather than not publishing. But the BMJ receives about 6000 papers a year and publishes only 10%. The BMJ would not benefit financially from publishing tobacco funded research, and we are paid fixed salaries.

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.
      6. 6.
      7. 7.
      8. 8.
      9. 9.