Last week, Nature published a commentary written by two researchers from the University of Texas Southwestern Medical School, who analyzed tens of thousands of abstracts indexed in Medline looking for duplicate papers, either by the same authors or by different ones. Mounir Errami and Harold Garner estimated, based on analysis using text-mining software and manual inspection, that 1.35 percent of the 62,000 Medline abstracts they analyzed were duplicates written by the same authors and 0.04 percent were from different authors. They’ve posted a list of the suspected duplicates here.

The two authors of this study give plenty of qualifications on their work: without inspection of the full text, it was hard to see if proper attributions were made and if the duplications were from innocent intentions. But, they argue that this may be a growing problem and that journals and the scientific community need to better monitor and deal with the issue of plagiarism and self-plagiarism.

I posted a brief blurb about this in the Publishing in the New Millennium forum and it touched off quite an interesting discussion about whether they see this as a growing problem and whether there are legitimate reasons for an author to publish similar versions of his/her work in different places. One of the UT researchers, Errami, posted a short reply hinting that news of some fallout of their analysis (eg retractions) was on the way.

Well, here is one such development. On Friday, the Boston Globe reported that a Harvard/Beth Israel Deaconess physician/researcher was being investigated by Elsevier for allegedly duplicating in his 2004 review about half of the text from a review written by another author in 2003. The two papers were flagged in Errami’s and Garner’s analysis. The researcher, Lee Simon, was first named last week in the Dallas Morning News in its report of the Nature commentary.

I’ve received confirmation from an Elsevier spokesman that the review by Lee Simon is being retracted by the journal, Best Practice & Research Clinical Rheumatology.