PROFITS AND PLAGIARISM: THE CASE OF MEDICAL GHOSTWRITING

ABSTRACT

This paper focuses on medical ghostwriting in the United States. I argue that medical ghostwriting often involves plagiarism and, in those cases, can be treated as an act of research misconduct by both the federal government and research institutions. I also propose several anti-ghostwriting measures, including: 1) journals should implement guarantor policies so that researchers may be better held accountable for their work; 2) research institutions and the federal government should explicitly prohibit medical ghostwriting and outline appropriate penalties; and 3) a publicly available database should be created to record researchers' ethics violations.

INTRODUCTION

A growing literature¹ is documenting the practice of medical ghostwriting, whereby pharmaceutical companies draft favorable scientific articles, send them to academic physicians or researchers who sign on as author, and publish the articles in medical journals. Often the pharmaceutical company contracts the work out to a medical education and communications company, which executes key steps in the process such as drafting the article, inviting academics to be author, paying these honorary authors typically $1,000 to $2,000 per article,² and a range of other ‘publication planning’ services.³ Medical ghostwriting is a serious breach of medical ethics and a violation of the standard of excellence that readers have come to expect from journals. An adequate response to the problem requires firm action from different institutions within the biomedical community. This paper argues that honorary authorship is plagiarism and thus subject to federal and academic policies on research misconduct. Also, several measures are proposed to limit the publication of ghostwritten papers in medical journals.

Honorary authorship in science is not a small or localized problem; it has involved many researchers and a wide range of medical journals.⁴ A 2005 survey estimates that 10% of researchers funded by the National Institutes of Health have engaged in inappropriate assignment of authorship credit (which can include but is not limited to honorary authorship).⁵ A decade ago, several high-profile journals (including JAMA, The New England Journal of Medicine, and Annals of Internal Medicine) acknowledged the presence of industry-ghostwritten articles in their publications.⁶

Although it is difficult to estimate the frequency with which ghostwritten articles are published in journals, all signs indicate this is a continuing if not growing problem. The publication planning industry continues to flourish. During key marketing phases, up to 40% of published articles for a particular drug may be ghostwritten or have their messages closely managed by a medical education and communications company.⁷

Examples such as Fen-Phen and Zoloft make the social costs of industry ghostwriting abundantly clear. Wyeth's Fen-Phen and Pfizer's Zoloft were marketed in part through ghostwritten journal articles that downplayed and even omitted evidence of serious life-threatening side effects which included primary pulmonary hypertension, valvular heart disease, and suicidality.⁸ Ghostwritten journal articles have also been produced for GlaxoSmithKline's Paxil and Parker Davis's (now a Pfizer subsidiary) Neurontin.⁹

GHOSTWRITING AND PLAGIARISM

Current rules on authorship

Industry ghostwriting in medical journals violates norms of transparency and proper conduct in science. It also violates authorship guidelines published by the International Committee of Medical Journal Editors (ICMJE).¹⁰ The ICMJE states that authorship credit should be given to those who fulfill three requirements: ‘1) Substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the draft for publication.’ The document further states: ‘All persons designated as authors should qualify for authorship, and all those who qualify should be listed.’¹¹ Finally: ‘All contributors who do not meet the criteria for authorship should be listed in an acknowledgments section.’

What happens when the guidelines are violated? The ICMJE stops short of recommending disciplinary actions against those who make dishonest authorship claims. It is largely up to individual journals to decide how to respond. The ICMJE ‘strongly’ encourages editors to develop policies that define their journals' requirements for authorship. ICMJE also encourages editors to have a policy on guarantorship, which would require that at least one author be identified as the guarantor, who assumes responsibility for the integrity of the work as a whole.

The World Association of Medical Editors (WAME), also offers advice to journal editors. The organization has suggested various actions a journal can consider once a ghostwritten article has been detected. Actions include: ‘1) publish a notice that a manuscript has been ghost written [sic], along with the names of the responsible companies and the submitting author; 2) alert the authors' academic institutions, identifying the commercial companies; 3) provide specific names if contacted by the popular media or government organizations; and 4) share their experiences on the WAME Listserve and within other forums.’¹² In the end, though, journals are free to decide whether to follow any of the ICMJE or WAME guidelines.

In the face of these relatively weak recommendations, offenders – pharmaceutical companies, medical education and communications companies, scientists – often act with impunity. Can we institute a stronger system of policies and enforcement to combat industry ghostwriting? Universities have been a leader in developing policies that govern issues of authorship. The next section briefly examines how US universities are managing these issues.

University policies on plagiarism and honorary authorship

University policies are quite vocal when it comes to plagiarism. Plagiarism occurs when an author takes credit for writing or ideas that have come from someone else. University students who are found guilty of plagiarism may face academic probation, suspension, or even outright expulsion in cases of extreme forms or repeated acts of plagiarism. Faculty researchers are not exempt either. Take Brown University for example, which has policies on plagiarism and other types of research misconduct. At Brown, research faculty and department administrators found guilty of research misconduct may face probation, suspension, or termination of employment.¹³

Policies at Brown and other universities do not draw an explicit connection between plagiarism and honorary authorship. Yet honorary authorship is a form of plagiarism because it entails claiming authorship for work that was done by others. The twist is that, here, the corporation (the willing ‘victim’ of plagiarism) actually wants the scientist to ‘steal’ its work and publish it as his or her own. But surely this one difference makes industry ghostwriting no more acceptable. Honorary authorship should be frowned upon by the research community in the same way that other types of plagiarism are frowned upon. Yet it remains unclear from their policies whether universities view this practice as being as serious as other forms of plagiarism.¹⁴

Once the view that honorary authorship is plagiarism gains currency, honorary authorship can be fought more effectively (particularly because universities already have policies on plagiarism). Universities should therefore be more explicit about the close link between honorary authorship and other forms of plagiarism. For example, university policies could explain that honorary authorship is a serious form of plagiarism and will be treated as such.

Academic honorary authors are recruited by pharmaceutical companies because these intellectuals are widely recognized experts. These experts should have access to the data and be involved in conceptualization of studies – this is the only way their expertise can honestly be used. Pharmaceutical companies of course are trying to have their cake and eat it too: they want their studies published with big experts named as author, but withhold from these experts the tools they actually need to apply their expertise in a sound manner. This is why pharmaceutical companies have carefully restricted scientists' involvement in publication production, to the point of blocking scientists' access to raw data.¹⁵

To address this problem, national governments might pass legislation that prohibits attempts by pharmaceutical companies and medical education companies to obfuscate authors' full access to data. This may promote integrity of corporate publication practices.

GHOSTWRITING DETERRENCE

Previous authors¹⁶ have recommended a number of compelling strategies designed to prevent ghostwritten journal articles. In addition to the above recommendations for universities and national governments, this section suggests other strategies to make would-be ghostwriters and honorary authors think twice. Some of these recommendations are best suited to the US and other similar policy environments that feature federal research grant monies and a mixture of public and private universities and research centers. However, the principles guiding these recommendations (principles such as addressing ghostwriting as a form of plagiarism and enforcing anti-ghostwriting measures for academics and corporations) may be applied in many different international settings.

Implement guarantor policies

The ICMJE recommends that journals have a guarantor policy; ideally, all journals would implement these policies so that they become standard practice. A journal with a guarantor policy would publish articles only if one or more authors (or contributors) sign a form stating that they are responsible for the integrity of the work as a whole, from inception to published article. Published articles should also contain the names and contact information of guarantors. The aim of such policies is to hold researchers to an appropriate standard of accountability.

Explicitly prohibit industry ghostwriting via federal sanctions and institution-specific policies

Second, sanctions should be enforced against companies and scientists who subvert scientific research by engaging in ghostwriting.

Regulate biomedical companies

Companies such as Wyeth and GlaxoSmithKline have launched public relations schemes which involved ghostwritten articles. In recent litigation against Wyeth, patients have sued for ailments caused by Fen-Phen – but the Wyeth-sponsored ghostwriting itself has not been cast as a punishable offense. I propose that companies be penalized for ghostwriting (Elliott and Moffatt¹⁷ have made a similar argument.). For example, industry ghostwriting could be codified as a punishable federal offense, resulting in fines. Similar legal consequences already exist for pharmaceutical companies that commit the federal offense of promoting drugs for off-label use.¹⁸

Regulate scientists who honorarily author for industry

Public and private institutions – such as governments, medical journals, universities, or research foundations – should have clear rules for dealing with researchers who honorarily author for industry. Let us first look at the government, which sponsors publicly funded research; then we will discuss privately funded research.

The Federal Policy on Research Misconduct prohibits federally funded researchers from engaging in research misconduct. The Policy defines research misconduct as ‘fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results’[emphasis added].¹⁹ In other words, plagiarism is a specific type in a larger class of research misconduct. Honorary authorship is not mentioned in the policy. But, because honorary authorship is a form of plagiarism, it would also fall under the larger rubric of research misconduct. This implies that federal policy should be applied to cases of honorary authorship just as it is applied to other forms of research misconduct. Administrative actions recommended by the federal policy include: appropriate steps to correct the research record, suspension, and permanent debarment from receiving future federal grants. These consequences could be applied in cases of honorary authorship as well.

What about privately funded researchers employed at universities, hospitals, and foundations? Employers should have clear policies on honorary authorship that describe the process for investigating cases and the possible disciplinary actions. Many private universities have rules regarding research misconduct that can simply be extended to cover honorary authorship as well. For example, Yale University policy adopts the federal definition of research misconduct and holds all its researchers – publicly and privately funded – to the federal standard of misconduct-free research.²⁰ At institutions like Yale, then, there is a policy in place so that researchers who honorarily author for industry can be disciplined (provided that Yale and other institutions adopt the stance that honorary authorship is in fact a form of plagiarism).

Compile database for tracking offenders

Finally, there could be a publicly available, online database of researchers' ethics violations. These records could be set up and maintained by individual national governments. The records should contain biographical information on the offenders, including place of employment, a citation of all their publications, and detailed information regarding their ethics violations. Visitors could browse the listing alphabetically by name or search for specific individuals' names. The system would be similar to the Excluded Parties List System, which is maintained by the US General Services Administration and provides a comprehensive list of individuals and companies barred from receiving federal grants for reasons including conviction for research misconduct. Goals of the database include deterring researchers from accepting honorary authorships and providing information about cases where researchers accepted such authorships.

Some of these proposals are tougher than the ICMJE and WAME recommendations, but they seem warranted in light of the serious social repercussions industry ghostwriting can have. Until strong policies are implemented to fight industry ghostwriting, the practice will most likely continue to be a significant problem.

OBJECTIONS AND REPLIES

Skeptics could raise a number of objections to the arguments of this paper. For one, skeptics may argue that ghostwriting is largely prevented when journals require conflict-of-interest disclosures. Such disclosures inform readers of any financial relationship that may exist between the author and a biomedical company. Thus, readers armed with this knowledge are able to assess the trustworthiness of the paper. But this argument has major flaws. First, conflict-of-interest policies are not effective against industry ghostwriting; the whole point of industry ghostwriting is to hide the involvement of the industry. In ghostwriting, it is common for honorary authors simply not to admit to any conflict of interest.²¹ The second flaw is that even in the rare instance that an industry-ghostwritten article declares its author's conflicts of interest, the reader cannot know exactly what part of the article was biased. Detecting biases in articles can be extremely difficult – and when companies take the extra step of hiding negative results or deliberately biasing experiments, bias detection becomes impossible. What, then, are readers to make of the published article that admits conflict of interest? Is any part of the article trustworthy or should it be disregarded wholesale? Only the researchers responsible for the study have enough information to answer these questions. As a result, it often seems that conflict-of-interest policies are nothing more than empty formality because in the end readers are still unsure what part of the article is credible. Or, as Carl Elliott puts it:²²

Disclosure does nothing to fix the underlying problem of pharma funding, which is not secrecy but power. It does patients no good to be told that doctors, researchers, and regulators are all in pharma's pocket if there is nothing they can do about it.

Skeptics might also argue that if ghostwritten articles were biased, the scientific community would weed them out through peer review and scientific reproducibility; after all, this is how inaccurate findings in the past have been discovered and corrected. It is true that peer review was designed to screen out weak studies and studies are called into question when scientists are unable to reproduce their results. But this argument overstates the effectiveness of peer review and the timeliness of scientific reproducibility. Not only is peer review ineffective in detecting cases of scientific misconduct, but it is also a process prone to bias and abuse.²³ And faulty studies are not always caught by attempts at reproducibility. Many studies, such as large randomized clinical trials that require millions of dollars to conduct, can only be done once because they are so costly. Further, it may take years before scientists attempt to reproduce a study; in the meantime, a seriously biased study might have already led to patients being treated with dangerous drugs. The best way to avoid bias from ghostwriting is to fight ghostwriting directly, not rely on faulty conventions that may or may not detect bad studies.

Skeptics may claim that if a scientist thinks an article is legitimate enough to sign onto, then the article must have some merit and should go up for publication. This argument presumes that scientists can do no wrong when it comes to scientific integrity. But scientists often have many incentives to sign as author. Beyond the cash payment to sign, academics have professional incentives to publish such as greater professional distinction, promotions, and tenure.²⁴ These incentives may undermine an academic's ability to be impartial and intellectually honest when reviewing a ghostwritten paper before he or she signs it. Furthermore, many scientists have actually authored ghostwritten articles that were seriously flawed,²⁵ making it clear that the voluntary participation of scientists cannot vindicate the practice of industry ghostwriting. On the contrary, the fact that so many scientists are implicated in ghostwriting schemes reveals the seductive power of ghostwriting agendas and the need to combat them with renewed effort.

CONCLUSION

According to the chair of the WAME's editorial policy board, Robert Fletcher:²⁶

Ghost writing [sic] initiated by industry is a big concern. When it is undetected, it distorts the scientific record, substituting marketing and persuasion for the balanced exchange of views and the search for sound answers that characterize the contents of medical journals at their best.

Deliberately biased studies and inappropriate authorship claims constitute important obstacles to scientific progress. This paper offers a fresh perspective and new proposals to combat medical ghostwriting. The hope is that these ideas will stimulate productive dialogue and lead to the eventual diminishment of medical ghostwriting.

Acknowledgments

I thank Dr. Daniel Wikler and three anonymous reviewers for their thoughtful comments on earlier drafts of this article.

Tobenna D. Anekwe is a doctoral candidate in international health economics at Harvard School of Public Health. He received his Master of Public Health from Columbia's Mailman School of Public Health.

Footnotes

• 1

  R. Smith. Maintaining the Integrity of the Scientific Record. BMJ 2001; 323: 588; P. Baty. 2005. When Access to Data is a Real Bone of Contention. Times Higher Educ Suppl. Available at: http://www.timeshighereducation.co.uk/story.asp?storyCode=199922&sectioncode=26[Accessed 27 July 2008];

• 2

  B. Moffatt & C. Elliott. Ghost Marketing: Pharmaceutical Companies and Ghostwritten Journal Articles. Perspect Biol Med 2007; 50: 18–31.

• 3

  S. Sismondo. Ghost Management: How Much of the Medical Literature is Shaped Behind the Scenes by the Pharmaceutical Industry?. PLoS Med 2007; 4(9): e286.

• 4

  Multiauthor collaborations in science are fairly common, which may explain why some academics do not see the signing of a ghosted article as problematic. In his book chapter ‘Rights or Rewards? Changing Frameworks of Scientific Authorship,’ Mario Biagioli writes that‘[S]cientific authorship is losing (or has already lost) its role as the containment vessel for credit and responsibility (and the vast problems posed by their definitions). The development of large-scale multiauthorship is directly responsible for that.’ In Scientific Authorship: Credit and Intellectual Property in Science. 2003. M. Biagioli & P. Galison, eds. New York: Routledge: 253–279.]

• 5

  B.C. Martinson, M.S. Anderson & R. De Vries. Scientists Behaving Badly. Nature 2005; 435: 737–738.

• 6

  A. Flanigan et al. Prevalence of Articles with Honorary Authors and Ghost Authors in Peer Reviewed Medical Journals. JAMA 1998; 280: 222–224.

• 7

  S. Sismondo, op. cit. note 3.

• 8

  B. Moffatt & C. Elliott, op. cit. note 2; C. Elliott, op. cit. note 1; G. Kolata. 1997. How Fen-Phen, A Diet ‘Miracle’, Rose and Fell. NY Times 23 Sept: 1; D. Healy & D. Cattell. Interface between Authorship, Industry and Science in the Domain of Therapeutics. Br J Psychiatry 2003; 183: 22–27.

• 9

  B. Moffatt & C. Elliott, op. cit. note 2.

• 10

  International Committee of Medical Journal Editors (ICMJE). 2007. Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Philadelphia, PA: ICMJE. Available at: http://www.icmje.org [Accessed 28 July 2008].

• 11

  The strict interpretation of this rule is not always followed. Cases where a beneficent senior academic allows his or her junior to take full authorship credit for a paper on which they collaborated are not generally seen as a breach of the rule as long as the junior has fulfilled all authorship criteria. In any case, there is a big difference, ethically speaking, between that case and the case where a medical ghostwriter cedes authorship credit to the academic researcher. The former case is an altruistic act whereas the latter stems from industry sponsorship that simultaneously pays the medical writer and sabotages the medical literature by using it for covert marketing purposes.

• 12

  The World Association of Medical Editors (WAME). Ghost Writing Initiated by Commercial Companies. J Gen Intern Med 2005; 20: 549.

• 13

  Brown University. Brown University Policy for Responding to Allegations of Research Misconduct. Providence, RI: Brown Univ. Available at: http://research.brown.edu/policies/misconduct.php [Accessed 28 July 2008].

• 14

  We need not compare honorary authorship to plagiarism to see that honorary authorship is problematic. There are larger reasons, such as medical ethics and scientific integrity, why honorary authorship for industry is problematic. However, the honorary authorship/plagiarism comparison may be very useful, particularly for institutions that already have explicit policies on plagiarism but lag behind when it comes to controlling honorary authorship. This paper points out that anti-plagiarism policies should also address the issue of honorary authorship because the issues are closely related.

• 15

  R. Smith, op. cit. note 1; P. Baty, op. cit. note 1.

• 16

  B. Moffatt & C. Elliott, op. cit. note 2; S. Sismondo, op. cit. note 3; S. Ngai et al. Haunted Manuscripts: Ghost Authorship in the Medical Literature. Account Res 2005; 12: 103–114.

• 17

  B. Moffatt & C. Elliott, op. cit. note 2.

• 18

  M. Larkin. Warner-Lambert Found Guilty of Promoting Neurontin Off Label. Lancet Neurol 2004; 3: 387.

• 19

  Office of Science and Technology Policy (OSTP). 2000. Federal Policy on Research Misconduct. Washington DC: OSTP. Available at: http://ori.dhhs.gov/policies/fed_research_misconduct.shtml [Accessed 28 July 2008].

• 20

  Yale University. Policies and Procedures for Dealing With Allegations of Academic Fraud At Yale University. New Haven, CT: Yale Univ. Available at: http://www.yale.edu/grants/policies_reg/acadfraud.html [Accessed 28 July 2008].

• 21

  Ibid.

• 22

  C. Elliott, op. cit. note 1.

• 23

  F. Godlee & T. Jefferson, eds. 2003. Peer Review in Health Sciences. London, UK: BMJ Publishing Group.

• 24

  B. Martin. Plagiarism: A Misplaced Emphasis. J Infor Ethics 1994; 3: 36–47.

• 25

  A. Berenson. 2005. Evidence in Vioxx Suits Shows Intervention by Merck Officials. NY Times 24 April: 1.

• 26

  L. Eaton. 2005. Medical Editors Issue Guidance on Ghost Writing. BMJ 2005; 330: 988.