Yesterday, we reported on a retraction in the Journal of Clinical Microbiology involving plagiarism and author issues. Well, it turns out we only had half the story.
Thanks to a comment on Derek Lowe’s In the Pipeline blog, which picked up our item yesterday, we’ve learned of a remarkable pair of letters in the journal about the paper. (We missed the letters because we didn’t originally see the “This article has been cited by other articles” section of the notice page, and the people involved, who might have made reference to it, haven’t returned our requests for comment.) At the core of the matter is whether — as the authors of one letter strenuously argue — the publication owes its readers the same kind of apology it served up to the scientist whose work was plagiarized in the offending article. The answer they received is an equally vehement no.
We think the exchange is noteworthy enough that we’re posting it below. Before we do, though, we’ll state that journals and editors frequently apologize to their readership in retractions, so that’s not really what’s at stake here. Rather, what the debate drives at is, in a sense, whether journal reviewers have a sort of fiduciary responsibility to the scientific community.
We also need to correct the record. In our original post, we surmised that we knew who the plagiarizing author was (although we did not name that person). Turns out, as letters below indicate, our hunch was off base.
Now to the letters:
Why Should the Journal of Clinical Microbiology Apologize for Publishing a Paper Plagiarized from Its Archives?
In the March 2011 issue of the Journal of Clinical Microbiology (Journal), a retraction was published in which the authors stated the following: “we realized after our article had been published that major parts of the text had been plagiarized almost verbatim from Colombo et al.” (3). The plagiarists apologized “to Prof. Colombo . . . and to Journal of Clinical Microbiology readers” (3). However, no explicit word of apology came from the Journal itself. There are at least two reasons for which we believe the Journal should also have apologized to its contributors and readers.
First, the article by Colombo et al. (4) was published in the Journal of Clinical Microbiology in 2006. Yet, in 2010, plagiarists were able to copy not just parts but the entire article and have their plagiary published in the same Journal without notice. Clearly, the review process in 2010 completely failed the layers of editors, editorial board members, and ad hoc reviewers (1) who evaluated the Colombo et al. paper just 4 years earlier.
Second, the plagiary (2) underwent a peer review that was at best complaisant and not reflecting the reputation of the Journal (http://jcm.asm.org/) and at worst an exercise that amounted to no review. In reading the plagiary to see if anything should have been revised, even though plagiarism would have remained undetected, we found at least the following six internal inconsistencies:
●The incidence rates were expressed only per 1,000 admissions but not per patient days, as indicated in “Statistical analysis.”
●Males comprised 58.64% of cases (as stated in Table 1 of the article) and not 64.02%, as stated in the text.
●Cancer was documented for 311 cases (31.61%), and 20 of these cancers (2.03%) were hematologic malignancies (Table 1), as opposed to 195 (19.84%) and 35 (17.94%), respectively, as stated in the text.
●At the time of candidemia, 120 patients were receiving a systemic antifungal agent (fluconazole, 78; amphotericin B, 31; itraconazole, 3; voriconazole, 3; echinocandins, 5) and not 122 patients, as they stated.
●Candidakrusei comprised 2.13% (Table 2) and not 5% of cases, as stated in the text.
●Table 2 is titled “Species distribution and incidence . . . ,” with no incidence data in the table.
More comments could be made, such as regarding the discussion of data from Italy using a citation from Chile (page 4203, reference 40) (2). But, while we recognize that reviewers did not have the benefit of hindsight we now have, the above inconsistencies were so obvious that the submitters should have been queried about and/or asked to reconcile their inconsistencies, whether their submission was a plagiary or not.
If the Journal of Clinical Microbiology can gladly accept apologies from others who plagiarized an article initially published elsewhere (5), then the Journal certainly can extend its own apologies to its contributors and readers for publishing a poorly reviewed plagiary of its own archived article.
REFERENCES
1. American Society for Microbiology. 2010. Instructions to authors. J. Clin. Microbiol. 48:1–21.
2. Cisterna, R., et al. 2010. Nationwide sentinel surveillance of bloodstream Candidainfections in 40 tertiary care hospitals in Spain. J. Clin. Microbiol. 48:4200–4206.
3. Cisterna, R., et al. 2011. Retraction. Nationwide sentinel surveillance of bloodstream Candidainfections in 40 tertiary care hospitals in Spain. J. Clin. Microbiol. 49:1193.
4. Colombo, A.L., et al. 2006. Epidemiology of candidemia in Brazil: a nationwide sentinel surveillance of candidemia in eleven medical centers. J. Clin. Microbiol. 44:2816–2823.
5. Deng,T., et al. 2011. Retraction. Association of three bacterial species and periodontal status in Chinese adults: an epidemiological approach. J. Clin. Microbiol. 49:2082.
M. Jacques Nsuami*
Louisiana State University Health Sciences Center
New Orleans, Louisiana
Nsanga Besa
University of North Texas Health Science Center
Fort Worth, Texas
Yanyi K .Djamba
Auburn University at Montgomery
Montgomery, Alabama
Emmanuel Sikabwe
Troy University
Troy, Alabama
Wato Nsa
Oklahoma Foundation for Medical Quality
Oklahoma City, Oklahoma
[contact information redacted]
Editor in Chief’s Reply
In the foregoing comment letter, Nsuami et al. draw attention to a 2010 publication in the Journal of Clinical Microbiology (JCM) in which Cisterna et al. (1) were found to have plagiarized large segments of text from a paper that had been published by Colombo and colleagues 4 years earlier (3).
Regarding this matter, a concern for plagiarism by Cisterna et al. had been brought to our attention by a reader of the Cisterna et al. paper shortly after it had been published. A thorough examination of the Cisterna et al. publication was immediately undertaken using the CrossCheck electronic plagiarism screening system. CrossCheck was developed by CrossRef (http://www .crossref.org/) in collaboration with iThenticate (http://www .crossref.org/crosscheck/index.html). The system provides a report that compares the text in question to text in a continuously updated database of published articles. The CrossCheck report was analyzed by an ethics panel of the American Society for Microbiology (ASM) Publications Board, and a determination was made that Cisterna et al. had indeed extensively plagiarized the Colombo et al. paper.
On this basis, Cisterna et al. were instructed to draft a retraction of their paper. Their retraction was subsequently published in JCM (2). In addition, Cisterna was informed that he would be banned from publishing any material in any ASM journal for 2 years. This sanction was not imposed on the coauthors of the Cisterna et al. paper, as they provided evidence that they had not participated in the study or had any involvement in developing the manuscript that was published. The coauthors claimed to have learned of the paper only after it had been submitted for publication. And lastly, a correspondence was sent to Colombo et al. in which I expressed our regret that this unfortunate incident had occurred.
In the foregoing comment letter, Nsuami et al. contend that JCM should apologize to its readers for not having recognized the plagiary of Cisterna et al. prior to publication of their paper and, further, that the peer review process that had been applied to the Cisterna et al. paper had been flawed and inadequate. As Editor in Chief of JCM, I enthusiastically reject both of these assertions.
Approximately 2,500 manuscripts are considered annually for publication by JCM. There simply does not currently exist a cost-effective plagiarism screening process that could be applied proactively to this number of manuscripts that would yield reliable information in a timely manner. For this reason, the ASM and JCM apply the CrossCheck plagiarism screen selectively whenever concerns for plagiarism arise.
The position of the ASM and JCM regarding matters of plagiarism is explicitly articulated in the JCM author instructions (http://jcm.asm.org/misc/journal-ita_edi.dtl#02). Therein, we state clearly and emphatically that plagiarism will not be tolerated and, further, that by virtue of their participation as authors, authors of submitted papers understand and accept this dictate. We believe that this position is consistent with the ethical precepts of current publication practices in the peer-reviewed medical scientific literature.
I find the second assertion of Nsuami et al., namely, that peer review of the Cisterna et al. paper was inadequate, at best curious; uniformed and lacking in common sense might be better terms. The Cisterna et al. paper was carefully and thoroughly reviewed by two individuals with primary expertise in the field of the epidemiology of antifungal resistance, the focus of the Cisterna et al. paper. While it is true that this review process failed to identify certain data presentation inconsistencies by Cisterna et al., a fact that we very much regret, overall the reviews were comprehensive and instructive. That the review process did not recognize the plagiary of Cisterna et al., something we also regret, is perhaps understandable insofar as the paper that was plagiarized had been published 4 years previously and none of the reviewers had participated in that investigation. As with most things, the peer review process is simply not perfect, nor will it ever be.
In summary, as is true of all journals published by the ASM, JCM takes the matter of plagiarism very seriously. Further, we strongly believe that the process we have in place for assessing publications for plagiarism, while not infallible, is rigorous, practical, and nearly always works.
REFERENCES
1.Cisterna, R., et al. 2010. Nationwide sentinel surveillance of bloodstream Candida infections 65 in 40 tertiary care hospitals in Spain. J. Clin. Microbiol. 48:4200–4206.
2.Cisterna, R., et al. 2011. Retraction. Nationwide sentinel surveillance 66 of bloodstream 67 Candida infections in 40 tertiary care hospitals in Spain. J. Clin. Microbiol. 49:1193.
3.Colombo, A.L., et al. 2006. Epidemiology of candidemia in Brazil: a nationwide sentinel 69 surveillance of candidemia in eleven medical centers. J. Clin. Microbiol. 44:2816–2823.
Gary V. Doern
Editor in Chief
Clinical Microbiology Laboratories
University of Iowa Hospital and Clinics
Iowa City, Iowa
[contact information redacted]
Here’s the Crosscheck fee schedule:
http://www.crossref.org/crosscheck_fees.html
which lists a $0.75 per document checking fee for up to 5000 documents per year, plus an annual fee based on publisher revenue. I don’t know what this publisher’s revenues are, so I don’t have a total cost to estimate; in addition, you’d have to figure out how to add this into the publisher’s workflow, which might have additional costs beyond the out of pocket costs.
The Editor-in-Chief doesn’t address the key issue, that an even cursory examination of the manuscript would have revealed serious problems with the text as published. I haven’t read the manuscripts in question, but Nsuami et al. have raised what appear to be significant issues. Detecting plagiarism is not the key question, but rather a peer review that appears to have been slight to the point of non-existence. Expressions of the Editor like “carefully and thoroughly” and “comprehensive and instructive” seem to be wildly aspirational rather than accurate. I know that in reviewing manuscripts I would not catch plagiarism, but I hope to catch inconsistencies as revealed by Nsuami et al.
Also, I _love_ the word “plagiary”.
1. Multiple inconsistencies between text and tables… pretty obvious if you review the article carefully, at least I would think (not being well informed, I only guess at this, but “duh”)
2. They don’t have to use crosscheck on all 2500 submissions, just the ones that they intend to publish…”duh”
I think an apology is called for but truly inadequate. It has become a scene of “caveat emptor” for readers of any journal.
I think the response from the editor is quite good, and yet I agree with an above poster that an apology is called for but truly inadequate. The only thing I take exception to is the editor’s last sentence. Does he really think “the process we have in place for assessing publications for plagiarism, while not infallible, is rigorous, practical, and nearly always works”?? There is currently no process for assessing publications for plagiarism, except when someone flags a suspicious paper that has already been published. (It is unreasonable to expect that reviewers have instant recall of the text of every paper in a field, which is what is required to have peer reviewers detect plagiarism prior to publication.)
Try eTBLAST. It’s free.
“Approximately 2,500 manuscripts are considered annually for publication by JCM. There simply does not currently exist a cost-effective plagiarism screening process that could be applied proactively to this number of manuscripts that would yield reliable information in a timely manner.”
Rubbish. With every paper I review, I copy an introductory sentence or two into Google and hit search. You’d be amazed how easily (and often) you see plagiarism. How hard is it for a sub-editor to do that before sending the paper our for review?
As a journal editor myself, I sympathise with most of what Dr Doern says. However, I do find astonishing the lines “coauthors of the Cisterna et al. paper, . . . provided evidence that they had not participated in the study or had any involvement in developing the manuscript that was published. The coauthors claimed to have learned of the paper only after it had been submitted for publication.”
First, does JCM not require that every author indicate their consent to authorship by a signed covering letter? Second, when the supposed coauthors “learned of the paper”, which if any of them immediately contacted the journal to have their name(s) removed.
The former seems to me the sloppiest part of JCM’s editorial process. The latter, suggests that several people were happy to have their names on a paper to which they made no contribution, until the faeces hit the fan, and therefore have a case to answer of publication misconduct.
No one comes out of this affair well. But, their but for the grace of god . . .