Monday, June 15, 2009

More Zyprexa Documents Released--Ghostwriting at Lilly

Once again, it takes court-ordered release of documents related to a lawsuit to shine daylight on previously hidden drug company practices. The present relevations come from Bloomberg News and are based on court-ordered release of internal Eli Lilly documents as a result of litigation over their antipsychotic olanzapine (Zyprexa):

Both in HOOKED and on this blog I have documented previous accusations of ghostwriting, so I'll try to comment here mostly on what is new. Of most interest perhaps is a guide called "Medical Press: Pre-Launch Feature Outline" (unfortunately, undated) which was an internal how-to primer on preparing ghostwritten articles. It covered both what the article should say to boost Zyprexa sales and how to find a compliant putative author to put his or her name on the final published paper. As the Bloomberg reporters recount, "The guide also offered tips on how to find authors by identifying a 'key opinion leader' and providing them either an outline of the article or a finished copy. Authors could include a study investigator, an advisory board member or 'Lilly-friendly' doctor, according to the documents."

The main importance of this passage is the frank admission that the puitative author might be sent a finished copy of the article. Up until now, the company and the putative author, when confronted, have routinely defended themselves by insisting that all that was provided by the company was a rough draft and that the author listed in the journal then made extensive changes in it. Even if that were true, it is still a violation of standard authorship ethics guidelines not to list to "ghost" author who actually wrote the first rough draft. But it is also more in keeping with the assertions of insider-whistleblowers like psychiatrist David Healy that in fact, few if any modifications to the draft are routinely made.

Sadly this "guide" does not state the going price paid to the putative author whose name appears on the published article. I have only been able to document payments of as high as $1000. Colleagues in the know tell me that figure is woefully low, but I still cannot find written documentation of anything higher. It's well documented that the "ghost" writer might be paid $20,000 for an article that is successfully placed in a top-tier journal.

Another insight is provided by a passage that recounts an exchange over a paper putatively authored by Dr. John Buse of UNC-Chapel Hill, a former president of the American Diabetes Association. The paper, actually authored by Lilly staffer Patrizia Cavazzoni, took issue with another paper that stated that there was a substantial risk of Zyprexa causing diabetes--the side effect (along with weight gain) that is today prompting the spate of lawsuits against manufacturers. The skeptical paper defending Zyprexa was submitted to the Journal of Clinical Epidemiology. When the paper was not promptly published, Lilly folks became concerned and in November, 2002, Lilly's senior scientific communications coordinator, Suraja Roychowdhury, wrote the journal. The editor, Andre Knotterus (here quoting the Bloomberg article) "replied in an e-mail that it was 'a bit strange to be contacted by the Lilly product team. Dr. Buse and coauthors can contact us directly next time.'"

Okay, let's get this straight. When journal editors are criticized on the issue of ghostwriting, they routinely reply that they are not private detectives, and that if Dr. Bigname sends them a paper and swears up and down that he actually wrote it himself, how is the poor little bitty journal to know different? So here, apparently, we have a paper submitted to a journal editor, and then a production team from a drug company calls the editor to complain that the article was not published as fast as their business plan called for. Instead of wondering, "Hmm, I wonder if this could have been a ghostwritten article, which good journal ethics requires that I refuse to print?" the editor instead says, "Next time, by the way, have the fake author send me the message instead." There is here at least the appearance of full journal complicity in the ghostwriting arrangement.

For the record, Dr. Buse denied to Bloomberg that he had ever put his name on an article that he did not write. He then proceeded to tell the reporters 1) that he had no recollection of who said what to whom regarding that particular paper on diabetes; 2) that his experience showed how biased an academic physician can become who works regularly for the drug companies; and 3) that in his opinion ghostwriting used to be very common but is becoming much more rare. I must say that for somebody who has never personally been involved in ghostwriting Dr. Buse seems to know a lot.

Finally the article makes the point that several of us critics of Pharma have stressed--we know of no case where an academic physician has been fired or has received any serious sanctions after being caught putting his name to a ghostwritten paper. Given that everyone agrees that ghostwriting is a clear and egregious violation of scientific publication standards, the current track record of academic medicine in policing itself is abysmal.

Contacted by the Bloomberg reporters, Lilly protested that its legal opponents in the current wave of lawsuits are cherry-picking damaging documents, quoting them out of context, and creating an erroneously negative picture of Lilly's corporate practices. There is of course the danger of this occurring whenever tens of thousands of pages of documents are released and reporters have to paw through the pile for the juiciest bits. That's why it is so important to know the background--what has been demonstrated in previous accounts of whistleblowing that are not dependent on this same stack of documents. As HOOKED and this blog will demonstrate, the facts contained in the Bloomberg News account are believable because reasonably consistent with what we have known previously about ghostwriting practices.

A tip of the hat to Roy Poses' Health Care Renewal blog for heads-up on the Bloomberg News article.

NOTE ADDED 6/15/09: The paper by Buse that I referred to above, on further inspection, turns out to fall very questionably under the "ghostwritten" category. John B. Buse is listed as the first author, but the Lilly staffer, Patrizia Cavazzoni, is clearly listed as the second author. The paper as a whole is also clearly labeled as funded by Lilly. If indeed the paper was primarily written in-house by Lilly staff, then the most serious ethical infringement was the listing of Buse as a first author.

Buse JB, Cavazzoni P, Hornbuckle K, et al. A retrospective cohort study of diabetes mellitus and antipsychotic treatment in the United States. J Clin Epidemiol 56:164-70, 2003.


Anonymous said...

This article conflates two quite different things and smears them both. The key question is, what is a ghost author? I would argue that a "ghost author" is a person who writes an article without being named as an author. This is neither illegal nor unethical in most circumstances, though it probably should be discouraged. What is unethical and should be illegal is what can be referred to as guest authorship. A "guest author" is a person who did not write an article but nevertheless is named as author. This is akin to plagiarism or theft and should not be permitted.

Anonymous said...

My concerns with the issue of ghost-writing is that it ignores the more important issue, which is the content of the article. Does a ghost-written article automatically mean that the data within it is falsified? Does a ghost-written article mean that the content is any less accurate or is in some way misleading? This may very well be the case, but none of these questions are addressed.
It seems that the blog author is merely upset with being misled by the name at the top of a publication. But so what? Who cares who actually wrote the publication if it actually includes facts. Does it really matter if Dr. Buse actually wrote the article or just endorsed it? James Patterson puts his name on countless books that someone else writes and they become best-sellers. Sometimes, academia's antiquated rules are unnecessarily applied to real world situations.
The truth is, ghostwriting happens in a lot of industries, including journalism. So decide on the facts of the content, leave the author out of it and stop complaining.

Howard Brody said...

Anonymous #2 above raises an interesting question. The ethical rules of scientific authorship are in the end conventions that we could agree to change. We could make ghostwriting cease to be an ethical problem simply by changing the conventions.

The problem with the advice from this commentator is that ultimately good science is based on trust--both to the fellow scientist and the consumer. The reader of a scientific paper has to be able to trust that the methods described were the ones actually used, that the results were reported as observed and not as wished for, etc.

The problem as I see it with the position, "Forget whose names are at the top and concentrate on whether the data are accurate," is that we already know in a case of ghostwriting that the author of the paper has violated our trust by failing to adhere to the ethical publishing conventions. I do not know how I am then supposed to regain my trust so as to be able to believe that the data are accurately presented in the rest of the paper.

Michael S. Altus, PhD, ELS said...

Anonymous (June 16, 2009 9:56 AM) incorrectly defined “ghost author” as “a person who writes an article without being named as an author.” “Ghost authoring” is the unethical practice of making substantial contributions without being identified as an author. For example, a pharmaceutical company with hire a MECC (medical education and communication company) to prepare articles for signature of key opinion leaders (thought leaders). The role of the pharma and the MECC in the authorship are concealed. Contribution to content is the issue, not who actually put pen to paper (or fingers to keyboard).

Anonymous continued by incorrectly defining “guest author” is a person who did not write an article but nevertheless is named as author. "Guest authoring" refers to being named as an author without having made substantial contributions. Again, it’s not an issue of who writes the article. It is an issue of who has made substantial contributions to an article and is willing to take responsibility for it.

For a primer on the unethical “g-word” publication practices, see AMWA Ethics FAQs (, which I wrote.