Legal remedies for ghostwriting
Everyone agrees that ghostwriting in the medical literature is a bad thing. The question is what can be done to eradicate it.
Professional medical writers' organisations such as EMWA, AMWA, and ISMPP have done their bit by publishing guidelines and position statements and educating their members about ethical publication practices, and there is some evidence that those efforts are pushing things in the right direction, but it's clear that they are not going to solve the problem by themselves and that more needs to be done.
So what more can be done?
Well a paper recently published in PLoS Medicine by Bosch et al argues that the best way to deal with it is through legal action. Of the paper's 3 authors, 1 is a lawyer and another declares an interest in doing consulting work for a law firm, so they would say that, wouldn't they? It appears to be hard to talk about ghostwriting without conflicts of interest rearing their ugly head.
Now, I don't want to completely pour scorn on the idea of legal action. It's an interesting idea, which I suspect could be appropriate in some of the most egregious cases of publication malpractice, and would certainly make people sit up and take notice if it were successful. The deterrent effect of a couple of high profile cases could be quite valuable. Bosch et al suggest that authors of dishonestly written papers could be named as defendants in any legal action relating to harms to patients that result from physicians relying on those papers in their prescribing decisions, and if harms do result in that way, then it's hard to argue that the authors of the papers aren't responsible. I wonder if the action could even be extended to the journal that published the paper, although Bosch et al don't mention that possibility.
However, I think suggesting that legal action is a sensible way forward in the majority of cases is misguided (unless, of course, you're a lawyer or a consultant to a law firm, in which case I dare say you'll do very nicely thank you). Legal action is time consuming and extremely expensive, and is going to be completely impracticable as a way of dealing with most unethically written papers.
I know I've said this before, but I think it bears repeating: we are never going to get rid of ghostwritten papers unless journal editors start to take the problem seriously (and by taking the problem seriously, I mean actually DOING something about it, not just publishing articles moaning about how terrible the problem is). But as Bosch et al rightly point out, journal editors have not shown a willingness to tackle ghostwriting. And that leaves us with a problem.
I don't know how we can encourage journal editors to do their bit (although if they did find themselves on the wrong end of a couple of high profile lawsuits involving ghostwritten papers that they published in their journals, I suspect it might focus a few minds), but the question of editorial responsibility is something that's cropped up in a couple of other contexts recently, and I think it's worth drawing some parallels.
The first context is my unpleasant experience of finding defamatory material published about my company in the Journal of Bioethical Inquiry. It turned out that, unless I was willing to go down the legal route of suing for libel, which I wasn't (did I mention that legal action is expensive and impracticable in most cases?), my attempts to have the record corrected were dependent entirely on the good will of the journal editor. If that good will was absent, as it turned out that it was, there was nothing I could do about it as there is simply no mechanism for enforcing ethical behaviour on journal editors who misbehave. Although COPE claim that they investigate complaints against journal editors, their total lack of transparency in their complaint procedure makes it hard to have confidence in the process.
My friend Karen Shashok wrote a detailed article analysing the Journal of Bioethical Inquiry case, in which she argued that a mechanism for making journal editors accountable for their actions is badly needed.
So, if there were a body that was able to hold editors of biomedical journals accountable for their actions, would they be able to take action against journals who publish ghostwritten papers? I don't know the answer to that, as it's all a bit hypothetical. But doesn't it sound like an idea worth pursuing?
It's also hard to talk about editorial accountability without thinking of the Leveson Inquiry, currently ongoing in the UK. This inquiry is focused on mass media, rather than scientific journals, but perhaps some of the ideas of editorial accountability that will come out of that inquiry could be applicable to journals too?
I have to take issue with your first sentence,
"Everyone agrees that ghostwriting in the medical literature is a bad thing"
mainly because this is an issue that is complicated by semantics and definitions, and leads to parties being unfairly blamed for things for which they cannot be held responsible.
The definition of a "ghostwriter" or of "ghostwriting" can ONLY ever be made a posteriori. It cannot be used to define what someone does in the present or to describe their current occupation, because it is only through the actions (or inactions) of others that it can exist at all.
Any writer who is uncredited for a piece of work is by definition a ghostwriter (rightly or wrongly), but they only take that name by virtue of being uncredited at a time subsequent to the act of writing. Someone else has to take that credit from them, and to appear on the byline in their place for that act to have occurred.
The problem is that, by calling the writer a ghostwriter, the issue is transferred from the parties that commissioned them to do the work or failed to acknowledge them to the writer themselves. The ghostwriter becomes the demon to be exorcised, and unfairly, in my opinion.
I would maintain that the practice of allowing someone to have their writing credit taken away from them and assigned to someone else in an opaque manner is the thing we should be demonising. But that's fake authorship, not ghost writing.
Hi Lucy
Some interesting thoughts there.
I agree that fake authorship is indisputably a bad thing and that it's a hugely important problem to address. Oddly, that was mainly what the article in PLoS Medicine was about, despite the title giving the impression it was about ghostwriting (as Karen Shashok pointed out in her response to the paper).
But does that leave the ghostwriter as an innocent party in this? I think that really depends on the details of what happens. I can imagine cases where a paper is ghostwritten and no real blame attaches to a ghostwriter, but I also think we shouldn't assume that ghostwriters can always be excused so easily.
Let's imagine 2 hypothetical scenarios. In scenario 1, a medical writer is hired to write the first draft of a manuscript. She discusses the results with one of the authors, and writes the first draft in good faith, assuming (but not actually ensuring) that she'll be credited for her work. The paper is then taken out of her hands, a guest author is added to the paper, and the medical writer's name is not mentioned. Like you, I wouldn't necessarily blame her for being the ghostwriter in this situation. Perhaps you could criticise her for not insisting that acknowledgement for her role be written into the contract, but maybe that's a counsel of perfection. I'm sure there are many medical writers who don't insist on that.
In scenario 2, a medical writer is hired by an evil pharmaceutical company to write a paper. The medical writer writes the paper with no knowledge of who the authors of the paper are to be, and puts a deliberate spin on the results to make the company's drugs look good. Guest authors are then added to the paper and the medical writer is not mentioned.
In that latter case (which fortunately I suspect is rare in practice), the medical writer has behaved entirely unethically. I would have no problems if the ghostwriter herself were to face being on the wrong side of legal action. Obviously other people are to blame too in this case, but that doesn't give the ghostwriter a free pass to knowingly take part in the wrongdoing.
For me, the most important thing in deciding whether the medical writer has done anything wrong is how the paper was written. Was it written in collaboration with named authors? If not, then that's unethical. But more importantly, did what was written in the paper accurately reflect the underlying data? If it did, then probably no great harm has been done, even if the authorship attributions are less than completely honest. But if the results are distorted in any way, then that is completely indefensible. Any medical writer asked to do work like that has a duty to refuse, and should face consequences if she decides to participate.
OK - but realistically - how much control does the medical writer have over who gets added as guest authors? I suspect they have very little control, and therefore they should not be held to be culpable for the actions of others at a subsequent stage of publication.
When I worked at The Lancet, there was an overwhelming sense that ghostwriting, and by implication, all medical writing, was wrong (you may not like the conflation, but I'm pretty sure that's what lies behind your problems with JBI too). It was journal policy that all named authors had to have actually written the paper. Any external writing help was viewed with suspicion, presumably because it is mainly pharmaceutical companies that use this service - and fair enough, IMHO, they understand that to stand the best chance of publication, your research needs to be effectively communicated (call in the experts). I suspect that the situation at The Lancet is still the case, although I'm not sure - you'll have to ask Richard Horton.
Personally, I don't have a problem with ghostwriting. Yes, I agree that it's clearer if writing help is acknowledged, but what's more important here is accuracy and avoidance of fraud. Simply requiring the writer to be named will not remove the problem of fraud or misreporting in scientific research. In your second scenario, the spin will still be there whether or not the writer is named, and researchers are just as capable of adding spin as medical writers. The main problem is that having your name on a paper is now more important than the research itself because of the way in which funding works. Combine this with the insistence by journals that researchers should also be experts at writing, then things start unravelling. Researchers don't want to admit that they didn't write the paper for fear of being left off the byline and therefore losing precious funding, even though they actually did the work behind it.
One simple thing would change this: for journals to stop stigmatising writing help and allow it to be properly acknowledged without penalising the researchers.
Medical writers seldom control who gets added as authors, but they may influence it. And unless an author is added at a late stage, they will know who the authors are. What a medical writer should never agree to do is develop a draft of a manuscript without knowing who the authors are. But yes, if they do that in good faith and someone else is added later, then there's probably not much the medical writer can do.
I think things have probably moved on since you were at the Lancet. They now accept the role of medical writers and have specific instructions about what to do if they were involved.
Nonetheless, there are still some who conflate medical writing and ghostwriting, and that's a shame.
While I think medical writers should always be acknowledged for the sake of transparency, I do agree with you that that's less important than whether the paper itself is presenting honest and unbiased results. I also agree with your analysis of why things aren't working very well.
I think things are moving in the right direction, however. Some journals really don't stigmatise writing help. Nonetheless, there is still a lot of hostility to medical writers out there. Hopefully that will wane over time as more evidence accumulates about the benefits that professional medical writers can bring to publication.
Hi Adam & Lucy:
As I see it, the "Litigation Solution" would only discourage the publishing of manuscripts that are fraudulent or misleading, since those are the characteristics that are legally actionable. It does nothing to address the practice of a) misattribution; b) lack of attribution - the key elements of what are commonly conflated (to use Lucy's term) into the common perception of "Ghostwriting".
I agree with Lucy's bottom-line statement that journals [and lay press, and - in the States - Congress (in the person of the likes of Senator Charles Grassley), and trial lawyers] stop stigmatizing professional writers, whose contributions do much to clarify information and add quality and integrity to publications.
Adam, I am not sure what you expect journal editors to do. Most journals have detailed author declaration forms which require information on medical writing assistance and a declaration by each author that they fully meet authorship criteria. Are you expecting journal editors to audit each declaration?
I believe the issue is with the authors and their institutions. Institutions put in place incentives for authors to rort the system (promotion dependent on number of publications). They have in the past turned a blind eye to guest/fake authorship - especially where it involved leading researchers who attract large amounts of funding. As a recent article in the Scholarly Kitchen blog stated - 'academia has been corrupted by careerism.' Cases of scientific misconduct and papers being retracted appear to be a daily occurrence.
This issue will not be solved until it becomes widely accepted by authors that guest/fake authorship is ethically unacceptable and institutions and scientific societies are prepared to take the appropriate sanctions
Hi Chris
You make a fair point in that journals probably can't be expected to audit every paper they receive. However, I do think they need to do much more when it comes to asking questions about how manuscripts were developed. You say "Most journals have detailed author declaration forms which require information on medical writing assistance", but are you sure? Do you have evidence for that?
There's a paper by Liz Wager (slightly out of date now, so I suppose it's possible things have changed) which shows that journals are on the whole pretty bad at making these things clear in their instructions for authors. Top journals like the Lancet & BMJ have clear policies on this sort of thing, but most journals are not top journals. I've just had a look at the instructions to authors from your journal, and I see that it doesn't mention the role of professional medical writers anywhere. Perhaps you should revisit that?
But even the top journals are not necessarily routinely asking questions. The BMJ apparently has a policy of sometimes asking authors whether they had help from a medical writer, but they do so only at the discretion of individual editors. I can't see why the question isn't asked routinely of all authors. I wrote about this previously here.
So I'm not suggesting that journals do a detailed audit of every paper they receive. Clearly that would be unreasonable. But I do think they should ask authors of every paper they receive questions about medical writing assistance. That really shouldn't take much in the way of extra resources. I've even published a little checklist to help journal editors here, but to my knowledge not a single journal anywhere is using it, which is disappointing.
Now, some journals do ask questions about medical writers when you submit a manuscript. If Pharmacoeconomics do, then kudos to you for doing so. But sadly, many other journals don't ask. While journals are never going to be able to eradicate deliberate fraud, as you simply don't have the resources to be able to verify that every author statement is accurate, deliberate fraud is much less common than people simply failing to disclose information that they should have disclosed but no-one asked them about it. So while asking these sorts of questions routinely wouldn't eradicate ghostwritten papers, I do suspect it would make them a lot less common.
And, of course, if journals could find the resources to do the occasional audit of a randomly selected paper every now and again, that would help too. That's not too much to ask, is it?
I totally agree with you about academic institutions, both on the question of the inappropriate incentives they put in place and their lack of appetite for appropriate sanctions of those found to have been dishonest. They also need to do their bit here.
Thanks for your reply Adam. I accept your point that the editors and journals I come into contact with may not be representative of all journals.
We do mention in our guidelines (Section E of Appendix D) that those who provide writing or editing assistance must be named in the acknowledgements section. However, I agree we don't ask a direct question such as you suggest. I am on a panel at TIPPA next month with editors from Neurology and the Cleveland Clinic of Medicine. Given the stance of these two journals it should be an interesting debate.
Having been a medical writer I appreciate the value they bring especially to a complex field like mine. As Lucy alludes to, journals are becoming more hands off with research post acceptance and as such are requiring papers to be more polished with respect to language. This will only increase the need for professional writers especially amongst non native English speakers. The key is to be transparent about this