One of the most powerful images of modern times is of Winston Smith, the anti-hero of George Orwell’s dystopian novel 1984, rewriting history for the totalitarian regime he serves by dropping original but no longer convenient documents into the “memory hole” of an incinerator.
Now the editors of JAMA appear to have adopted Winston Smith as their role model. An earlier version of an editorial that had been published online has now been removed from the JAMA website, in stark defiance of the rules and ethics of medical and scientific publishing.
First a little background (you can read about it in more detail using our chronology of the controversy): Jonathan Leo, a professor of neuro-anatomy at Lincoln Memorial University, published a letter on the BMJ website. The letter pointed out that a study on Lexapro in stroke patients published in JAMA failed to disclose a conflict of interest of one of the authors, Robert Robinson. Leo had initially reported his finding to JAMA. After waiting several months without any action from JAMA he sent his letter to BMJ. The story’s focus soon shifted to the JAMA editors for their pitbull-like response to events, especially after they had a series of contentious conversations with their critics and Wall Street Journal reporter David Armstrong.
In March the JAMA editors published the original combative editorial, which took a strong stance against both conflict of interest and free speech. The editors wrote that it was acceptable to accuse JAMA authors of conflict of interest, but it was unacceptable to tell anyone except the journal editors until the editors themselves reported their final determination of guilt or innocence. The editorial also included combative remarks about Leo and denied reports by the Wall Street Journal’s David Armstrong of abusive language and harrassing behavior stemming from DeAngelis.
Following a great deal of negative publicity, the AMA announced that its Journal Oversight Committee would review the affair. On Tuesday of this week JAMA published a highly edited version of the March editorial. The new version contains no direct reference to the Robinson and Leo affair, and includes no reference to the previous version. It also fails to disclose that the remaining text is almost identical to text in the previous version published online. (Most journals have explicit policies against the reuse of previously published content. But perhaps those rules don’t apply to editors.) The authors also fail to disclose that the editorial is apparently a response to recommendations from the the AMA’s Journal Oversight Committee.
At some point following the publication of the new editorial on Tuesday the original March version of the editorial disappeared from the JAMA website. I first learned about this on the blog of medical ethicist Udo Schuklenk. As Schuklenk wrote:
“No retraction notice was published, no erratum of any kind. As one of my colleagues pointed out: what does this mean for the substantial commentary (overhelmingly critical in nature) that was published in various fora on this now non-existent article?”
Schuklenk raised another point about the implications of removing a document with a doi number, since it should “not be changed in any print version or on-line without proper errata, withdrawal notes etc.; basically an on-line paper with a doi number ought to be treated just like one would treat a print article.”
Schuklenk has posted an archived copy of the original March editorial on his website.