Retraction Watch

Tracking retractions as a window into the scientific process

Springer, BMC retracting nearly 60 papers for fake reviews and other issues

with 3 comments

springerIn a massive cleanup, Springer and BioMed Central announced today they are retracting 58 papers for several reasons, including manipulation of the peer-review process and inappropriately allocating authorship.

The papers appeared in seven journals, and more are under investigation.

In a release issued today, the publishers note:

After receiving allegations of plagiarism confined to two journals, we immediately commenced an extensive investigation across our entire portfolio. In doing so, two teams dedicated to investigating issues around research integrity, the Research Integrity Group at BioMed Central and the Springer Ethics Team, identified a group of papers across seven journals that raised concerns relating to a variety of issues.

BioMed Central has identified 28 articles that will be retracted and is investigating 40 more articles. Springer has identified 30 articles that will be retracted, with a further 9 articles that need further investigation.

The investigation found:

evidence of plagiarism, peer review and authorship manipulation, suggestive of  attempts to subvert the peer review and publication system to inappropriately obtain or allocate authorship.

The move has prompted the publishers to take corrective action:

Alongside investigation into the identified papers, we have taken action to ensure that no further compromised papers from this group can continue through to publication. In addition, we have revised our policies on changes to the authorship list.

As a matter of course, early this year we began an audit of BioMed Central journals to track adherence to editorial policy. This has now been updated to look for potential hallmarks of abuse, such as non-specific or shared email addresses, assess peer review quality and look for changes in manuscript titles, authorship and content.

In addition, Springer has initiated a task force to better understand current patterns in abuse cases based on these experiences, with the aim to further develop our checks and balances.

Click here for a full list of papers. The most affected journals are Tumor Biology (25 papers) and Diagnostic Pathology (23 papers). The other journals are Comparative Clinical Pathology (one paper), Journal of Parasitic Diseases (four papers), Cancer Cell International (two papers), Journal of Ovarian Research (two papers), and World Journal of Surgical Oncology (one paper).

This is not the first time the journals have done a massive sweep to clean up the literature — in March, 2015, BMC pulled 43 papers for fake reviews; a few months later, in August, Springer did the same for 64 papers.

We asked a spokesperson for the publishers why these newly retracted papers — some of which were published before March, 2015 — weren’t caught during that earlier sweep. She told us:

A much more complex manipulation has taken place from a different group of authors. The publication process is fundamentally based on trust and if deliberate attempts are made to circumnavigate the system, it is possible that they will succeed.  However, we are committed to working to identify problems and to reacting quickly but responsibly when concerns are raised.

As the industry comes up with smarter tools for identifying these issues (i.e. plagiarism detection software) those determined individuals will look for new ways around it. It is vital that we remain vigilant, and share information with other publishers to stop this happening elsewhere.

Here’s why detection software wouldn’t have caught the plagiarism in these papers, she added:

This is not as simple as copy-and-paste plagiarism as it involves complex manipulation of our submission and peer reviews systems. Plagiarism detection software is a useful tool, but publishers need to employ a variety of tactics to stay on top of these issues.

Journals receive many allegations of plagiarism; she explained why these particular charges sparked a wider investigation:

We take every allegation seriously, and investigate every allegation of plagiarism. In this case we identified irregularities which led us to suspect a broader pattern of manipulation. Sometimes it is possible for the publisher to identify patterns of abuse across the portfolio where an individual journal editor would not be able to.

She confirmed that one of the articles that launched the investigation was one we flagged in 2015 as being under investigation by BMC,  “Decrease expression and clinicopathological significance of miR-148a with poor survival in hepatocellular carcinoma tissues.” That paper — now retracted — had been pegged by librarian Jeffrey Beall for allegedly using a 2014 paper as a template from which to plagiarize.

The spokesperson told us the publishers chose to issue all notices at once because all were part of a joined-up investigation. Each will be posted individually throughout the week, and explain whether the paper was affected by plagiarism, authorship issues, fake reviews, or some combination thereof:

For Springer all the articles showed evidence of authorship manipulation and peer review manipulation, and 70% showed evidence of plagiarism.

For BioMed Central all showed evidence of authorship manipulation, 57% showed evidence of peer review manipulation and 93% showed evidence of plagiarism.

Not all of the papers are indexed by Clarivate Analytics’ Web of Science, formerly part of Thomson Reuters; among those that are, only one — “Protective and antidiabetic effects of extract from Nigella sativa on blood glucose concentrations against streptozotocin (STZ)-induced diabetic in rats: an experimental study with histopathological evaluation” — was cited more than 10 times (14 citations). The papers were first published online between May 2013 and January 2016.

The retractions affect roughly 200 authors, more than one-third of whom co-authored more than one retracted paper. Most of the authors are based in institutions in Iran. Several papers list one corresponding author: Javad Javanbakht, whose affiliation on a 2013 paper is listed as the University of Tehran. Another name frequently listed as a corresponding author is Aram Mokarizadeh, noted in 2015 as based at the Kurdistan University of Medical Sciences. The name that appears most frequently among the co-authors is Emad Yahaghi at Baqiyatallah University of Medical Sciences, also on the editorial board of the Annals of Medical and Biomedical Sciences. Mokarizadeh and Yahaghi are both co-authors on one of the papers that launched the larger investigation.

We asked for more details about the changes in policy mentioned in the release. The spokesperson told us:

Alongside investigation into the identified papers, we have taken action to ensure that no further compromised papers can continue through to publication. BioMed Central has changed our policy on suggesting peer reviewers so that authors may do so in a cover letter with evidence of the peer reviewers’ authenticity.

We have updated our policy on authorship changes, to make it clearer to authors that authorship changes are not allowed after editorial acceptance – we instituted the use of a ‘change in authorship’ form to ensure authorship changes that are approved have been made in compliance with [Committee on Publication Ethics] guidelines and to allow us to gather data on these changes for future audits.

In 2014, Springer retracted a number of papers after learning they had been created by SciGen, a random paper generator, prompting the publisher to install new detection software. There is no evidence any of the papers were generated using SciGen, the spokesperson told us.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

  • Debora Weber-Wulff November 1, 2016 at 1:57 pm

    Of course these papers were not generated by SciGen – these are biomedical papers (and what bothers me a lot, many about cancer). SciGen generates computer science-sounding papers. Cyril Labbé in Genoble has created a system to identify SciGen-generated papers.

  • herr doktor bimler November 1, 2016 at 3:09 pm

    (and what bothers me a lot, many about cancer)

    There is the reassuring thought that no-one is citing the papers, and perhaps no-one is reading them (certainly not by the fake peer-reviewers). In the case of the paper that started investigations, Jeffrey Beall’s commenters point out that it is basically unreadable — it was not designed to be read.

    • Sylvain Bernès November 1, 2016 at 7:15 pm

      (it was not designed to be read.)
      More precisely: it was designed with the hope that no one would read it.

  • Post a comment

    Threaded commenting powered by interconnect/it code.