Ghostwriters in the Sky

Reader Responses to this Editorial | Add your response  

Once again, concern over the commercialization of medical research has hit the lay press. (1,2)  This time the specific topic of interest is the thorny issue of medical authorship and the phenomenon known as “ghostwriting”. As described by CBC Marketplace group in somewhat sensationalistic, but accurate terms, “…you may not have heard of it, but you'll probably want to know about it. It's a world that could make your doctor prescribe the wrong drug…”.

For those of you who are unfamiliar with the term, ghostwriters are wordsmiths who are commissioned to write a paper that is ultimately credited to another individual. In the context of the pharmaceutical literature, ghostwriters are typically hired by industry to write a favourable research or review article about their drug product based upon selective information provided to them.  The outcome? As described by David Healy in a recent interview with CBC, “…what appear to be scientific articles are really infomercials”. (1)

Why do sponsors hire ghostwriters?

The reasons for this practice are obvious. Efficiency and control over investment. The worldwide industry research and development (R&D) expenditures are estimated to be $40-60 billion US per year. (3,4)  Industry is now the principle sponsor of all biomedical R&D, accounting for about 62% of expenditures in 2000, as compared to only 32% in 1980. (5)  While the estimates vary widely, the cost of bringing a new drug to market is now being quoted at $400-800 million US, and it is estimated that about 11 years is required to take a product from the bench to marketplace. (4,6)  Combine these costs with the recognition that drugs have a relatively short marketing half-life, and financial success is predicated on sending just the right message at just the right time to health care professionals and consumers.  Add to this mix the fact that good investigators are busy investigators, and busy investigators are not necessarily good (or efficient) writers and you have most of the ingredients necessary for industry to justify the practice of employing ghostwriters. As expressed by one member of industry, “…in the best of worlds all authors would write their own papers and write them well.  Alas, the best of worlds eludes us and formal education is rare, especially for scientists…”. (7)  If this were the case, I would have to question how these sponsors could permit investigators who are unable to express themselves in writing to assume the responsibility for the conduct of a clinical trial.

So why, you may ask, would industry still abide by the practice of adding the names of respected clinicians to papers already written by ghostwriters? Simple economics again. The underlying objective of many publications is to enhance sales of pharmaceuticals. To enhance sales, you need to convince the reader to use your product. Would you be more likely to modify your drug recommendation and/or prescribing habits in response to reading an article written by a presumably objective group of peers, or a paid scribe? In this case, the messenger is just as important as the message.

What is the extent of ghostwriting?

The true extent of this practice is unknown and I suspect we will never know the actual extent of ghostwriting that is occurring because the parties involved (i.e. the sponsors, the writers and those investigators who should have been the writers) have little desire to reveal these facts.  While the International Committee of Medical Journal Editors (ICMJE) standards for authorship clearly delineate authorship criteria for all persons designated as authors, they recommend that contributors who do not meet the criteria for authorship (i.e. a persons who provided writing assistance or purely technical assistance) simply be listed under "Acknowledgements", provided that these individuals first give written permission for this recognition. (8)  And while most journals require submitting authors to assume responsibility for the content for the paper, they do not routinely attempt to ascertain whether any other parties were involved in actually writing the text.  Consequently, it is very difficult to ferret out those papers written by the investigators from those that are primarily crafted by ghostwriters. Compound this with the fact that study data is often analyzed by the sponsor and company employees may function as co-authors, and the waters get even muddier.

While ghostwriting was once generally confined to highly lucrative journal supplements, the practice appears to have encroached into the realms of original research and review publications.  Healy estimates that as much as 50% of the drug literature may be ghostwritten. (1)  In 1998, Flanagin et al (9) published the results of a survey of corresponding authors in which they found that 11% of 809 articles appearing in six prominent medical journals involved ghost writers, an additional 19% had evidence of honorary authors and 2% had involved both.  For the purpose of this study, ghost authors were defined as those individuals not listed as authors, but made contributions that merited authorship. In contrast, honorary authors were defined according to ICMJE criteria as those individuals who were named on the paper, but did not play a substantial role in the conduct and/or writing of the study and manuscript. Of those articles meeting the criteria for ghost authorship, there was no difference by type of article (i.e. research vs. review vs. editorial), although small-circulation journals tended to have a higher incidence of ghost writing for reviews (16% vs. 4%; OR 4.2; 95% CI 1.5-13.5) and a lower incidence for research articles (9% vs. 17%; OR 0.49; 95% CI 0.27-0.88). Similarly, a more recent survey of corresponding authors by Mowatt and colleagues (10) revealed that 9% of 362 reviews published in the 1999 issues of The Cochrane Library had evidence of ghost writers, 39% involved honorary authors and a further 2% had evidence of both.  In this study the higher incidence of honorary authors was felt to be related to the stricter definitions (i.e. the criteria was applied to all authors, not just the corresponding author) that were employed.  While other studies have been published on the topic of authorship, most focus on the issue of the inappropriate awarding of honorary authorships, rather than the concealment of ghost writers. (7)

Are there any examples of ghostwriting?

Probably the most well publicized examples of ghostwriting are the articles promoting the use of dexfenfluramine (ReduxR) for obesity. In this particular case, the patentee apparently paid a medical communications company to write several articles on this weight loss product under the names of prominent clinical researchers before the drug was removed from the market in 1997. (11)  A Texas woman was awarded $23.3 million US in one the first verdicts related to the heart-valve damage and primary pulmonary hypertension associated with the use of fenfluramine and dexfenfluramine.  Her lawyers successfully argued that the patentee had hidden the drug risks. (12)

Is ghostwriting a lucrative business?

Medical ghostwriting indeed appears to be a lucrative business.  According to the CBC Marketplace group, an article published in one of the high impact journals (e.g. New England Journal of Medicine, Lancet, British Medical Journal) can earn the writer up to $20,000. (1)  Larkin reports that Excerpta Medica Medical Communications received about $200,000 for the ReduxR articles. (11)  A GoogleR search on the term “ghostwriters” yields about 8,100 hits, including many websites advertising medical ghostwriting services. (13)  To quote one of these services, “you are the author, we do the writing and the best part is that we remain in the wings while you get to claim authorship…we do the work, you get the praise. (14)

What is wrong with ghostwriting?

After all, wouldn’t you rather read a well-written article than not?  From my perspective, this is analogous to students making the same argument after submitting a plagiarized paper to their course instructors.  It’s misrepresentation at best; fraudulent and unethical practice at worst.

The fundamental problem with ghost writing is objectivity.  Provided their professional livelihood is not threatened, independent investigators can generally be relied upon to write objective papers containing a complete and accurate description of the outcomes of a clinical trial.  On the other hand, most ghostwriters are usually not experts in the field, can only write about the selective data provided to them, are commissioned to put a positive spin on this information, and finally, future opportunities are dependent upon providing the client with what they are looking for. And what is that?  A good return on investment. Nothing more and nothing less.  In other words, writer objectivity is about as elusive as the timeliness and writing eloquence that sponsors apparently claim researchers themselves are unable to delivery. (7)

So what’s the solution?

From the perspective of the submitting authors, we need to start with full disclosure.  As recommended by Greenfield et al (15), identification of all contributors (as opposed to simply authors) plus the naming of a guarantor for the integrity of the research work may help resolve some of the problems.  While I support the concept in principle, I'm not sure that this will make the problem of ghostwriting disappear.  I have to agree with Grossman (7) that medical writers are here to stay.   At the same time, I must also agree with Reidenberg (16) that it is time to expose the ghosts of medical writing and that readers are entitled to know who is responsible for all aspects of the entire research project, including the presentation of its findings.

From the perspective of the reader, as always, read with caution.  Always ask yourself; do the authors or the sponsors stand to gain by the message that is being delivered in the article?  Determine for yourself, if you can, who sponsored and who actually wrote the paper you are reading.  First impressions are often deceiving.  And as far as ghostwriting is concerned, assume that there is about a one in ten chance that the real writers of the article you are reading do not appear in the list of authors. Who knows, perhaps someone else even wrote this editorial for me.  You only have my word that it was not.

P. Jewesson, PhD FCSHP
Co-Publishing Editor

J Inform Pharmacother


  1. Johnson E. Medical ghostwriting. CBC Marketplace. http://www.cbc.ca/consumers/market/files/health/ghostwriting/  Accessed April 4, 2003.
  2. Johnson E. Ghostwriting: the basics. http://www.cbc.ca/consumers/market/files/health/ghostwriting/faq.html  Accessed April 4, 2003.
  3. Anon. Look, no strings: publishing industry-funded research. (editorial) CMAJ 2001;165. http://www.cmaj.ca/cgi/content/full/165/6/733  Accessed April 4, 2003.
  4. Moses H 3rd, Martin JB. Academic relationships with industry: a new model for biomedical research. JAMA 2001;285:933-5.  http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=11180737&dopt=Abstract  Accessed April 6, 2003.
  5. Bekelman JE, Li Y, Gross CP. Scope and impact of financial conflicts of interest in biomedical research. A systematic review. JAMA 2003;289:454-465.  http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=12533125&dopt=Abstract  Accessed April 4, 2003.
  6. Moser J. Links between drug company profitability and investments in research: A fact sheet. July 2, 2002. http://www.galen.org/news/070202.html  Accessed April 4, 2003.
  7. Grossman L. Ghostwriting. The Lancet 1998;351:1741. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9734925&dopt=Abstract Accessed April 4, 2003.
  8. Anon.  Uniform Requirements for Manuscripts Submitted to Biomedical Journals http://www.icmje.org/ Accessed April 4, 2003.
  9. Flanagin A, Carey LA, Fontanarosa PB, Phillips SG, Pace BP, Lundberg GD, Rennie D. Prevalence of articles with honorary authors and ghost authors in peer-reviewed medical journals. JAMA. 1998;280:222-4. http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9676661&dopt=Abstract Accessed April 10, 2003. 
  10. Mowatt G, Shirran L, Grimshaw JM, Rennie D, Flanagin A, Yank V, MacLennan G, Gotzsche PC, Bero LA. Prevalence of honorary and ghost authorship in Cochrane reviews. JAMA. 2002;287:2769-71. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=12038907&dopt=Abstract Accessed April 7, 2003.
  11. Larkin M. Whose article is it anyway? The Lancet 1999;354:136. http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=10408500&dopt=Abstract  Accessed April 6, 2003.
  12. Anon. News in brief. US fen-phen payout. the Lancet 1999;354:658.
  13. http://www.google.ca/search?q=ghostwriters&ie=ISO-8859-1&hl=en&btnG=Google+Search&meta=  Accessed April 7, 2003.
  14. Anon. Ghostwriters. Information about ghostwriters. http://www.ghostwriters.org/Information/information.html   Accessed April 3, 2003.
  15. Greenfield B, Kaufman JL, Hueston WJ, Mainous AG III, DeBakey L, DeBakey S, Rennie D, Yank V, Emanuel L.  Authors vs.contributors: accuracy, accountability, and responsibility. JAMA 1998 279: 356-357. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9459466&dopt=Abstract Accessed April 6, 2003.
  16. Reidenberg JW. Unmasking ghost writers. Clin Pharmacol Ther 2001;70:208-9.  http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=11557907&dopt=Abstract Accessed April 6, 2003.

Reader Responses to the Editorial, "Ghostwriters in the Sky"

Read the Editorial | Add your response

To the Editors:

I believe this problem can be avoided by following only the advise provided by known researchers, and serious publications. Healthcare professionals should be very unconfident of what the pharmaceutical companies tell about their products. We have to lean only on known and proven authors. Strong legislation against this dangerous practice should be promoted.

Roberto Brenes, Physician

To the Editors:

Few doctors have the training, ability, time, or resources to conduct a proper clinical study. neither the medical practioner nor the employees of pharmaceutical companies should be trusted to .supervise drug testing on the general public. 'do no harm' is in direct conflict with the way that new drugs are introduced. 90 percent of the drugs now prescribed have been in the marketplace for less than 10 years- far too short a time for anyone to have determined their long term health consequences.

P Schmied

Read the Editorial | Add your response

Copyright © 2003 by the Journal of Informed Pharmacotherapy. All rights reserved.