Saturday, June 23, 2012

Those Persistent Ghosts: The Key Loophole

Our old friends Jonathan Leo and  Jeff Lacasse are at it again, this time in the Pharmalot blog:
--on the subject of ghostwriting. As we've detailed in many previous posts, there has been a great deal of huffing and puffing over the evils of ghostwriting, where a hired hand for the drug industry writes almost all the article, academic experts then affix their names to the author byline, and the hired gun disappears into obscurity when the paper appears in a journal. Yet there's no good evidence that the practice has diminished, and indeed defenders of the medical communications firms that employ the ghostwriters have become even more brazen in defending these practices, insisting that it's not ghostwriting at all. What Drs. Leo and Lacasse add in this more recent post is a clue as to the large loophole that people jump through to keep ghostwriting alive and well.

The ICMJE, the International Committee of Medical Journal Editors, has been forceful (it would seem) in moving against ghostwriting and similarly unethical practices by formulating a set of criteria for authorship that are now widely adopted as authoritative. They say that to be an author you must:
  • Make substantive contributions to the research work
  • Draft the article or revise it critically
  • Give final approval to the version to be published
Now, these criteria do a pretty good job of eliminating one common abuse of authorship, honorary authorship, in which senior physicians insist on having their names added to the publications of their juniors even when they actually did nothing to contribute to the study. Drs. Leo and Lacasse indicate how the criteria fall down when used to police ghostwriting.

The common practice now is that a hired-gun medical writer will write the crucial first draft of the paper, making sure to organize and spin the paper in a way favorable for drug marketing. The academic guest authors (who meet the first criteria by having been involved in conducting the research) then meet the second criteria by doing some edits--often really minor and trivial, but enough so that they can claim that they had something to do with the manuscript. The hired gun then turns the manuscript over to the academics, who approve the final version. The hired gun is named in an acknowledgement as having provided writing or editorial "assistance." In this scenario, the ICMJE guidelines have been followed, and the hired gun cannot be listed as an author because he/she did not meet the 3rd criterion.

What has changed as a result of this is that the old practice of eliminating all mention of the ghostwriter has given way to the newer practice of listing the ghostwriter in an acknowledgement. No doubt the drug firm preferred the old way, keeping all of its involvement in writing the paper hidden. But the new way works well enough, as the actual key role of the ghostwriter in shaping the paper the way the company wants it is still quite well concealed, and no one who doesn't read the fine print at the end of the article knows of the ghostwriter's existence.

Drs. Leo and Lacasse note that if either medical journals or academic medical centers truly wanted to police this practice, they could do so. As an example of a journal that has adoped more stringent standards than ICMJE, they cite Neurology, that has cut to the chase to ask the really key question, "who influenced the content?" When other bodies continue to swear by the ICMJE criteria, despite widely published accounts of their limitations, we can only conclude that people don't really want to get rid of ghostwriting-- which is highly lucrative for the medical journals through sales of reprints, and also for academic centers, as those faculty who are asked to be ghostwriters in all probability bring in hefty funding from Pharma.

As I explained in HOOKED and in many previous posts, ghostwriting, despite recent efforts to sanitize it, has to be the most egregious case of unethical practice at the medicine-Pharma interface. If you cannot believe that the listed authors actually wrote an article, what can you believe? Maybe in the world of celebrity kiss-and-tell memoirs, no one is shocked to discover that the celebrity "author" really did no writing at all (and more than likely couldn't if he tried), and some hired hack writer churned out the text. But the entire system of trust in science is built on the reader's confidence that the authors, with their academic reputations, are vouching for the results stated in the article and especially for the interpretation they attach to those results. If academic medicine cannot put a stop to this, it is time to call in Senator Grassley and his troops and turn the control of academic medicine over to Congress or some other outside agency, because we've failed.

No comments: