What Deserves a Retraction?

The current crisis in scientific publication is much based on improper peer review of an increasing number of often mediocre manuscripts submitted to an ever-increasing number of new scientific journals. Facing university libraries’ limited budgets, the global players among publishers have been advocating “open access” publishing for more than a decade. Once a more than welcome initiative for making research results immediately accessible for everybody, it’s now a business model of Wiley, Elsevier and, in particular, Springer for making more money. All have also announced “manuscript transfer” of rejected, in their hardcore journals, manuscripts (“if not too bad”) to newly established open access journals. They even suggest to reuse previous reviews. I have reported about that before.

Another dire consequence of “open access” is a still booming market for so-called predatory publishers which flood the internet with their shoddy and worthless “articles” after authors had paid a considerable amount of money. An updated list of these publishers can be seen here.

Improper or no peer review

Careless peer review of manuscripts is most evident when presentation of data contains obvious statistical flaws including numerical errors in tables or figures. The mere fact that this occasionally happens proves that peer review and editorial handling may be severely compromised. As careful reading of full articles seems no longer commonplace, one would expect many more letters to editors requesting proper corrections [1]. Nevertheless, they do occur albeit not so often as probably necessary.

When I recently wrote a letter to the editor of one of our hardcore journals, the paper had just gone online, had  not yet been assigned to a particular issue, and was never meant to be published in the printed version of the journal. My numerous concerns included (i) clinical irrelevance of the applied method of assessing thickness of soft oral tissue, CBCT; and either (ii) no statistical analysis of data acquired in bench experiments at all, or (iii) improper correlation analysis for indicating some sort of reliability. I supposed the paper had not undergone any peer review. What also stunned me was that,

[t]here seems to be a misconception when quoting a paper by our group.3 In the Introduction of their paper, Cao et al.1 write,

“Müller and colleagues applied a non-invasive ultrasonic device to measure mucogingival tissue [sic] thickness and found that both determining correct position and attaining reproducible results were difficult by this method.”

Authors may note that this was not a methodological study in which reliability and validity of ultrasonic measurements of mucosal thickness were reported. Instead, we presented results of systematic mapping thickness of the entire masticatory mucosa in periodontally healthy volunteers. Intra- and inter-individual variation of thickness of masticatory mucosa therefore refers to natural variation, not validity or reliability of measurements which had been found excellent and reasonable, respectively, in another paper.4 If the authors want to make their point about difficulties of ‘attaining reproducible results’ they may want to quote correctly, i.e. the 1999 paper by Müller et al.4

A false quotation of which, in times of Google Scholar, the author is quickly informed by email. When I suggested proper analysis of repeat data (for instance, Bland-Altman plots, 95% limits of agreement) I had a retraction in mind. Instead, the editor offered a published letter and advised to submit it through Elsevier’s manuscript submission system. Unsurprisingly, authors were grateful and did not really address any of my issues. A completely useless enterprise.

Plagiarism

Plagiarism may occur accidentally or unintentionally in required dissertations written by inexperienced researchers who are not properly guided by their supervisors. Plagiarism of senior academics may be regarded a signature of a personality trait in which cheating and stealing intellectual property is a prominent feature. I have been quite successful in uncovering widespread plagiarism in the work of a former colleague which ultimately led to retraction of five textbooks produced by a leading publisher of science [2]. The platform vroniplag.com has analyzed German dissertations, and revealing results have forced two federal government ministers to step down. The dissertation of current minister of defense, Dr. Ursula von der Leyen,  is being scrutinized by her former University in Hanover after Vroniplag had disclosed widespread plagiarism. In addition, Vroniplag have identified several medical and dental dissertations plagued with plagiarism. The most brazen case seems to be a Dr. med. dent. thesis submitted to Freiburg University in 1998 which contains about 50% pages where more than 75% of the text was copied from his supervisor’s PhD thesis of 1996.

Self-plagiarism may not be uncovered so easily. It may even be regarded a minor issue, and using again one’s own arguments should be totally okay when correctly referring to the previous instance(s). But sometimes, self-plagiarism may in fact indicate lack of fresh ideas and an inclination for quickly writing another paper. In a recent case, I had again repeatedly been informed by Google Scholar that a particular group had, time and again, quoted an old publication of mine in the wrong context [3]. Even that should deserve a proper correction. The erroneous quote led me to read at least five papers published by the group since 2007 which all deal with sampling strategies for the detection of subgingival periodontal pathogens. All papers were sponsored by companies which provided test kits and results apparently for free. The most recent article by Nickles et al. went online in July 2015 but had not yet been paginated. Introduction and Discussion sections were so similar in a previous publication by Wohlfeil et al. (2010) that a comparison of the sentences in a table was necessary.

Nickles et al. 2015

Well, when busily publishing, year after year, data collected on a daily basis, and then taking a few sentences verbatim from a previous, own, publication may not be a big issue. But quoting the same references after five years in an introduction to another, if similar, experiment would in fact signal a standstill in science. And, apparent self-plagiarism in the paper was widespread. Nearly 40-50% of both Introduction and Discussion sections in Nickles et al. (2015) were identical in respective paragraphs in the paper by Wohlfeil et al. (2015); and self-plagiarism could be traced in at least three other, similar, papers published by the group since 2007 [4].

Sloppy research – sloppy peer review?

The question remains, where should we (readers, reviewers, editors, publishers) draw a line and call for retraction. As a matter of fact, editors and publishers are hesitant to retract, and retractionwatch.com has rarely reported on dental papers which had been sunk for violation of scientific standards. Anyway, brazen plagiarists must and possibly will be hunted down [5]. Editors and publishers have nowadays to apply special, but not commonly available, software to expose plagiarism.

The common problem with unpaid reviewers who are asked by editors to evaluate frequently mediocre manuscripts, which are usually unwelcome, is well-known. Although reviews are submitted under the condition of anonymity, once an editorial decision has been made one may easily check the reviews submitted by other reviewers, which sometime point to issues with the paper missed by oneself; but sometimes indicate rather sloppy review which may lead to an editor’s suggestion of minor or major revision rather than rejection. As a rule, in case of eagerly publishing authors one should generally check their previous publications for similarities. If self-plagiarism is apparent and reported during the review process, the editor is usually hesitant to further review the manuscript. If self-plagiarism without proper referral to the previous paper is uncovered in an already published article by a reader, a correction would probably not suffice.

So, what to do with the paper? If large parts are plagiarized from the authors’ previous work (possibly including very similar conclusions) one may actually question the scientific value of the new research. Then, a decision as to retract the paper may indeed be justified.

Notes

[1] If that occurs in a field unrelated to one’s own subject, one would not automatically be inclined to inform the editor ut rather trust the scientific community more familiar with the issue to take action. There might be exceptions of that rule. In a recent case which I became aware of, the PhD candidate, her supervisors, one or several reviewers of the journal to which the material was submitted, and its editor-in-chief; and even opponents when the thesis was eventually defended all failed to identify grave numerical (conceptual) errors in at least one table. This is a devastating sign of current shoddiness in scientific publishing. Post-publication peer review is slowly gaining importance but nowadays papers may stay largely unnoticed among scientists as the main purpose of publishing is no longer gathering new information but rather author promotion. While this case might eventually be solved by a correction (with a need for amending most of the conclusions), a retraction would be more appropriate.

[2] Nevertheless, an official, i.e. public, retraction note by the publisher is still missing. Only after having had insisted, I had received an email stating the very fact. The publisher’s webpage does no longer mention her works. Respective books had been sold to book sellers, though, so they are still available at, e.g., Amazon. Bashful silence of a leading publisher of science had apparently encouraged the author to use seeming availability of her books when applying, unsuccessfully, at a university in Norway for senior academic positions.

[3] Part of my PhD thesis in 1993 was a longitudinal study in which patients with chronic and aggressive periodontitis had undegone non-surgical and, if indicated, surgical periodontal intervention and adjunctive systemic minocycline administration. Patients had been screened microbiologically for presence of subgingival Aggregatibacter actinomycetemcomitans which was, at that time considered so special that the mere presence of the organism would justify a name change, A. actinomycetemcomitans associated perodontitis. As there was no control group involved, the purpose of the study was not to establish that the chosen treatment regime was superior to another regime. It was, at that time and in the absence of proper randomized controlled clinical trials, considered very comprehensive aiming at resolving the periodontitis problem once and forever. The clinical results of this 2-yr study were very convincing. Subgingival plaque samples after therapy of a limited number of sites did not yield high number of the target organism. A new and interesting finding was, however, that A. actinomycetemcomitans could be detected frequently from extracrevicular locations, i.e. buccal mucosa, tongue and saliva. And the number of colony forming units in buccal samples were associated with the number of persisting pockets. I speculated that, due to the limited number of subgingival sites sampled for A. actinomycetemcomitans, the organism was certainly overlooked in subgingival plaque most of the time but, as it was most probably transferred from subgingival locations via gingival exudate and saliva to other areas of the oral cavity, can be detected from, e.g., easy to sample buccal mucosa.

[4] A letter to the editor with full documentation of self-plagiarism (and other issues) may be seen here. The editor has confirmed that he has received the letter and that he would take action. In a telephone call he informed me that the authors should get the opportunity to respond to the letter. I will report about their response, which has not yet been received, if and when available.

[5] The revealing work of the German platform vroniplag.com is most valuable but as they focus on dissertations (including those of high-ranking politicians) written in German language, their scope is rather limited.

17 October 2015 @ 8:39 am.

Last modified October 17, 2015.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s