Ghostwriting: the importance of definition and its place in contemporary drug marketing

I thank Claire Barton for her considered response. I note that in contrast with the trade associations, Barton quite rightly uses the word “ghostwriter” as a description of writers who compose manuscripts, but who are not authors.

Barton provides further evidence of the perversity of excluding writers from coauthorship. She states: “Even when writing on very familiar topics, a straightforward trial report can take around 40-50 hours and a review article around 100-150 hours to write.” This is not clerical work, but that of highly intelligent and qualified individuals. The ICMJE rules denying them byline recognition as coauthors are insupportable.

I disagree with Barton, however, that my primary article paints a “murky picture of conspiracy and manipulation.” Rather, I view the shortcomings of industry literature as a cultural problem for which all major stakeholders are responsible. No objective reader could conclude that bias in respect of content or attribution is not a difficulty in some commercial articles, and my article considers how it gets into manuscripts. I identify several potential routes, one of which is writing, but this does not mean that writers are generally biased or dishonest. Rather, my article argues that writers are in general honest and professional. My article is highly critical, however, of campaigning by the publications and marketing trade in the domain of publication ethics, as illustrated by its efforts to change the definition of ghostwriting. Such activities do not portray the trade in an encouraging light.

With respect to Barton’s other points, she states that I fail to identify the driving force behind the use of ghostwriters, which she claims is “time.” In fact, my primary article states: “Ghostwriting .. remains widespread .. for reasons that have not changed— it delivers articles to order and on schedule, and ensures they are commercially useful, professionally finished, and, most importantly, attributed to academics.” Barton is right that time is an issue, but time is only one of several reasons why ghostwriting remains widespread.

I agree with Barton that company employees should be included as coauthors, as they currently are, although sometimes in fewer numbers than is warranted. As my article points out, however, company coauthors provide another potential conduit for commercial perspectives to enter some manuscripts. Furthermore, their presence is underplayed by generally placing them in the least conspicuous, mid-byline locations, whereas the first author in the large majority of cases is an academic. With regard to Barton’s discussion of “sources of support” statements, which are required by the ICMJE and used in some journals, these are better than nothing, but these statements overlook what should be critical disclosures – notably, whether a company instigated the work, whether it owns the data, and what product is being marketed.(1) Furthermore, companies are often not merely “sources of support” and do not just provide “funding” or “sponsorship”. They are often heavily involved in every stage – in effect they are the lead and corporate authors of these projects. Thus, the ICMJE’s own language of “support” and “funding” downplays their true role (1).

Barton disagrees with my claim that the academics who participate in industry projects are contingent. However, if the company has a drug entering clinical trials, it will look to select from a range of potential clinical research centres to partner with for the study, and the research will be done whichever centre happens to be picked. The design and analysis are frequently done or substantially steered by the company or its commercial contractors, sometimes with regulator input, and the company generally owns the data. The choice of academic centre has little bearing on these considerations, and the study will proceed and generate similar results whichever one is chosen. Similarly, if a company decides it is necessary to develop a review article on an “educational” theme related to its product, this decision may well be made before a particular academic author is selected. The academics who work on industry projects are generally honest and assiduous, but the instigation and framing of projects often comes from the companies. That is why it is unacceptable that published articles typically portray the academics as the leaders of the enterprise, and the companies as mere funders and junior colleagues.

Finally, Barton is correct to point out that not all writers work for industry. Many of them do, however, and I think in this respect Barton misunderstands the nature of marketing. Even when industry literature is wholly unbiased, the use of expert textual composition or editing and the scheduled production of journal articles constitute marketing, because they deliver the vendor’s data and clinical argument to the right customers, at the right time and in clear and logical, and therefore persuasive language. The ability of pharmaceutical companies to spend enormous sums generating such literature creates what I have referred to as “discourse bias” in distribution of published data and opinion.(1) Additionally, however, subtle bias may sometimes creep into this literature at the level of content, for instance in the framing of disease concepts and clinical challenges, and the consequent highlighting of certain product attributes and points of differentiation. Clinical trials designs, analysis and interpretation may also contain biases that feed into the process of manuscript development. Finally, attribution is often biased because it emphasizes the role of academics and underplays that of industry. Without any intent among writers to be biased, therefore, trade literature through these processes is shaped by marketing on many levels to serve the interests of commercial products.

I close by reflecting again that if Barton, a highly qualified pharmaceutical physician, is spending 40-50 hours writing manuscripts on subject matter she is familiar with, and 100-150 hours on review articles, it is absurd that the ICMJE seeks to deny her coauthor status. I disagree with some of Barton’s points, but commend her for using the term “ghostwriting” in the correct way, and believe her authorial work deserves byline recognition.

Reference

1. Matheson A. The ICMJE Recommendations and pharmaceutical marketing--strengths, weaknesses and the unsolved problem of attribution in publication ethics. BMC Med Ethics 2016;17:20