Issue 29 / 8 August 2011

GHOSTWRITING has been in the news lately with the revelation that the publishers of Anh Do’s award-winning memoir, The happiest refugee, hired a professional writer to get Do started on the project.

Although it was quickly established that Do and his publishers had no case to answer, debate is simmering elsewhere about the much more pernicious topic of ghostwriting in the medical literature.

According to the World Association of Medical Editors: “Ghost authorship exists when someone has made substantial contributions to writing a manuscript and this role is not mentioned in the manuscript itself.”

On the face of it, this might seem fairly innocuous but in practice it means manuscripts can be written by people with commercial interests and passed off as the work of unaffiliated, often well respected clinicians and academics. End users — readers, future researchers, regulatory bodies, courts, etc — are thus deprived of the chance to assess the content with full cognisance of its origins.

There have been some high-profile cases of ghost authorship over the years. Most recently, in probably the first case of its kind to be exposed, it was alleged that GlaxoSmithKline, the makers of the selective serotonin reuptake inhibitor Paxil, employed medical writers to draft a whole textbook, which was then published under the authorship of two prominent US psychiatrists.

Ghost authorship is sometimes, but not always, associated with “guest” authors who lend their names to the published article despite not having written it.

Often, however, it’s not that simple.

The psychiatrists in the Paxil case were adamant that they did qualify as authors because they “scrutinised every page and rewrote and edited as [they] deemed necessary”. They also said they controlled the final draft.

Australian gynaecologist and researcher John Eden found himself embroiled in ghostwriting accusations a few years ago, when assistance from a medical writer in producing a review article for American Journal of Obstetrics & Gynecology turned out to have been part of a concerted effort by drug company Wyeth to publish a suite of articles downplaying safety concerns about hormone replacement therapy.

Eden was exonerated after an investigation by the journal’s editors, but was shocked to realise that he had been manipulated into helping push the company’s commercial agenda. “I feel stupid — angry”, he said in a newspaper interview. “I look back and ask myself: How could I have been so naive?”

This raises the interesting question of why academics lend their names to material largely written by people with vested interests. Pecuniary motivations come low on a long list of reasons which, from my own observations, include agreeing with (and contributing to) the content; wanting to get the message out; a need to publish; ongoing relationships with drug companies for research funding; concerns that fully revealing the authorship might damage the perceived value of the article; commercial naïveté; and publishing naïveté.

As a medical editor, I flatter myself that I can pick up when something is slightly “off key” with an article — a lack of objectivity, a slightly skewed research question, a strange choice in framing results, jarring mentions of a particular product. The editorial process and peer review weed out many such papers. But history shows that medical editors miss many cases of ghost authorship.

Journals that follow the recommendations of the International Committee of Medical Journal Editors also ask authors to fulfil three criteria of authorship. They must also state the funding sources and their role (if any) in the project, and declare conflicts of interest. Many journals also now require a statement regarding authors’ specific contributions to the paper.

These are all good ways of seeing off guests but they might not flush out all the ghosts.

There is another question that needs to be asked: is there anyone else who has made a substantial contribution to writing this manuscript who is not listed as an author? Recent examples indicate that acknowledging writing assistance is not enough because it can hide the writers’ commercial affiliations.

Journals need to ask this question, but authors also need to ask it of themselves when they are in the process of writing for publication. If we don’t, and recent Canadian commentators have their way, it might be taken out of our hands and dealt with by the legal system as fraud.

Something needs to happen because, in the wash-up, dishonest authorship practices do immense damage to the medical literature and to the practise of medicine. When our patients look us in the eye and ask, “Is this treatment safe, is it effective?” we need to have confidence that we are doing our objective, unencumbered best for them.

The doctor is in session — ghosts and guests not welcome.

Dr Ruth Armstrong is a medical editor with the MJA.

Posted 8 August 2011

3 thoughts on “Ruth Armstrong: Time for some ghost busting

  1. Ruth Armstrong says:

    Dear Peter
    Thanks for reading and commenting on my article. Like you, I take enormous pleasure in assisting authors to publish their work — and sometimes this takes a lot of rewriting. I think rewriting, or genuine writing assistance, is a fine and necessary thing. But if the writing “assistant” writes the first draft (or a significant part thereof), or contributes to significantly to the ideas or the design, then not adding them as an author is ghostwriting. I think it’s fine to acknowledge writing assistance as long as that is truly what it is. Having a writing assistant who, like you, is free of commercial interests is gold!

  2. Peter Arnold says:

    Dear Ruth, you say “acknowledging writing assistance is not enough”, but that is all I ask for. I have re-written dozens of papers which have appeared in peer-reviewed journals, including the MJA. My colleagues, as talented researchers or statisticians as they might be, are often inept at ‘Plain English’. I take professional editorial pleasure in turning rejected ‘pigs ears’ into accepted ‘silk purses’. All I ask for is an “acknowledgement” – I claim no academic authorship of the paper. Not all of us “acknowledged” helpers have a commercial interest. Peter Arnold (medical editor for his friends)

  3. Bryan Walpole says:

    Yes Ruth, an awkward issue.
    In 1990 I mentored a final year medical student, writing a review of trauma scoring, we spent hours developing it, I was highly impressed with his ability, and was second author, thinking I had made a substantial contribution, although most of the writing was his.
    The editor wrote back, showing us that it was 75% plagiarised from a long past J Trauma , and I was embarrassed, ( the student wasn’t) but on reflection, how could I know? Thats what editors are for.
    Now I put it through a program like “Turnitin” and the results are astounding. It has solved the plagiarism problem at uni, as it not only compares with the published literature, but also the uni database.

Leave a Reply

Your email address will not be published. Required fields are marked *