Sometimes redundancy — the topic of our last post — is a failure of editors to adequately vet a manuscript. Other times, the blame falls more squarely on the authors.
Consider: In the August 2010 issue of Anesthesia & Analgesia, a highly regarded specialty journal, five researchers from the University of Pennsylvania, led by Andrew Ochroch, made a remarkable confession.
Their article in the May issue of A&A on ventilation of patients recovering from bariatric surgery plagiarized a 2009 paper in a competing publication, Anesthesiology — written by the same group:
We sincerely apologize for the inappropriate and unacceptable intellectual overlap and self-plagiarism of our paper … published in Anesthesiology.
Sincere apologies are better, we suppose, than insincere ones. But, never mind. They go on:
In presenting the results of the two studies, we duplicated much of the Introduction, structure of the Methods and Results, and the Discussion sections. Despite the data being original in each paper, this is a clear violation of the policies [of both journals].
To which we say, ouch!
Here’s an example from the crucial methodology section in the papers’ abstracts:
Anesthesiology 2009:
All subjects had continuous positive airway pressure initiated 30 min after extubation in the postanesthesia care unit via identical noninvasive ventilators.
A&A May 2010:
All patients had continuous positive airway pressure initiated 30 minutes after extubation in the postanesthesia care unit (PACU) via identical noninvasive ventilators.
Other than the fact that the wording is nearly identical, the methods themselves are suspiciously similar—a fact the authors admit in their letter:
We also fully understand that the difference of delivery of postoperative airway support in the two papers … would have potentially been found by the editors to be insufficient to justify independent papers.
In other words, the Penn group pursued parallel publications—it’s publish or perish, after all—and got burned. Blaming
delays and overlap of our attempts to publish both sets of information, we neglected to reference the paper in Anesthesiology in the final revision of the A&A paper. Had this citation appeared, we now realize that the incremental nature would have likely resulted in its rejection from A&A.
True enough, said Steven Shafer, the top editor at A&A, who told Retraction Watch by email that he praised the repentant authors for their handling of the incident:
The teaching point of this particular case is that authors must let editors know about related submissions. The obvious reason is to avoid sorting out a problem after publication, as in this case. The other reason is that editors can help authors figure out how to best handle parallel submissions.
Although the letter writers assert, and Shafer confirms, that the data used in the two manuscripts were different, the supporting matter was too close for comfort.
Shafer offered two suggestions here: Submissions could be merged into a single manuscript, or perhaps the editor can offer suggestions to see that the parallel submissions have a different focus to justify two manuscripts.”
Shafer is no stranger to retractions. His journal was forced to pull no fewer than 10 papers tainted in a sweeping research scandal involving anesthesiologist Scott Reuben. Reuben is now serving six months in a federal penitentiary in Devens, Mass. for fraud.
His message: In the Internet age, finding fraud, plagiarism and other forms of intellectual dishonesty — or, in the case of the latest incident, sloppiness — is simply too easy.
Authors should assume that any potential issues with a manuscript will eventually be discovered. Published manuscripts reach a huge global audience. Every day 10,000 Anesthesia & Analgesia pdfs are downloaded from HighWire Press. Additionally, automated search engines scan the literature looking for plagiarism, duplicate publication, and similar forms of misconduct. The vetting that manuscripts undergo during peer review is just a fraction of the scrutiny that papers get after publication. Authors should assume that any questionable publication will be caught eventually.
Why is it so wrong to duplicate one’s previous boilerplate explanation of the methods of a study.
Should have the following sentence “All subjects had continuous positive airway pressure initiated 30 min after extubation in the postanesthesia care unit via identical noninvasive ventilators” have been reworded simply for the sake of rewording one’s previous words?
Right. I’ve been so tempted to self-plagiarize for this same reason. If much of my experiment was the same as the previous experiment, why would I describe that part differently? Self-plagiarism can also be inadvertent; if the same author is describing the same event, then it shouldn’t be too surprising if they end up choosing the same words.
Normalizing this kind of redundancy also allows for the dilution of prestigious journals like that of Anesthesia & Analgesia. I do agree that self-plagiarism can be inadvertent, however, I’d challenge anyone who posits that the authoring was absent of copying and pasting.
I think the point of controversy, in this case, is that the paper could have been merely amended or republished. Instead, the authors chose to publish an entirely new manuscript with the potentiality of receiving twice the credit for reusing their original work.
I was the lead author on these papers. They were entirely different studies. One study looked at preoxygenation with a novel CPAP device, made by Vygon, in obese patients undergoing bariatric surgery. The other looked at preoxygenation with what is colloquially called “BiPaP” or pressure support ventilation, in the same patient population. The first study, using CPAP, showed better lung mechanics post op. We followed this is with the BiPaP study – which uses inspiratory pressure support (a completely different approach). Lung mechanics were slightly better than CPAP.
There was a third study, that was never sent for publication.
We were forced to retract because of self plagiarism in some of the discussion. That was a mistake of mine because I used a very similar document for the IRB for paper 2.
The publication in Aneathesiology was rush published as it was presented at the ASA conference. The original, retracted study, spent a lot of time with the editors. Honestly I had forgotten about it because I was moving from the USA to Europe after many years (to a job where there was no requirement to publish anything ever).
The mistake was not telling the editors of A&A about the prior publication in Aneathesiology. Shafer, who was being burnt by Boldt at the time, refused to hear the argument that CPAP and BiPap are entirely different airway support strategies. They insisted on retraction. There was no problem with IRB/ethics, recruitment, data, data analysis or outcomes. These were entirely honest studies.
I have read hundreds of papers and review articles over the years that self plagiarized like this. I don’t actually see the problem. The kicker for me was that I could easily have rewritten either paper, had I had thought of it (I obviously cut and pasted segments of the IRB document). But I didn’t think of it, and, aside from moving, I was editing a textbook at the time.
I don’t believe it was necessary to retract that paper – it leaves too much of a stain on the characters of the investigators particularly in an era where there is so much fabricated research.
I agree with you. They should have allowed you to correct those sentences containing self-plagiarism with a simple erratum. Journals knowingly let the worst forms of data fabrication slip, but suddenly become sensitive to the least malignant form of plagiarism, which is limited self-plagiarism. They should have corrected this issue, not to retract it.