The European Journal of Cardiothoracic Surgery has retracted a 2007 article by Chinese researchers after the senior author decided he liked the data so nice he’d publish them twice. And he appears to have done so without the knowledge of the corresponding author.
Here’s the notice for the paper, titled “Open-heart surgery in patients with liver cirrhosis”:
This article has been retracted at the request of the Editor-in-Chief of this journal after a reader brought to our attention that the data described in this article were the same as those included in the article ‘Open-heart surgery in patients with liver cirrhosis: Indications, risk factors, and clinical outcomes’ by An et al., published in Eur Surg Res 2007;39:67–74 (doi:10.1159/000099145).
When submitting this article to the European Journal of Cardiothoracic Surgery, Dr Young An declared explicitly, on behalf of his co-authors, that the work was original and that it had neither been submitted nor published elsewhere.
The senior author, Dr Ying-Bin Xiao, has admitted to infringing submission and copyright policies of the two journals.
We apologize to the readers of the European Journal of Cardiothoracic Surgery and to the Journal European Surgical Research for this redundant publication and any inconveniences this retraction may cause.
Each version of the study has been cited 14 times, according to Thomson Scientific’s Web of Knowledge — meaning the duplication doubled the authors’ number of citations, at least until now.
The paper trail (well, the pixel trail) here is pretty straightforward. Xiao was hedging his bets. According to the ESR, their version of the article was received on Sept. 5, 2006 — five days before the EJCS received its manuscript. Interestingly, ESR accepted the revised manuscript less than a month later, on October 2. The EJCS, on the other hand, received its revision in late January 2007, and accepted the paper the following day.
The ESR EJCS clearly was right to retract the paper. But we think the EJCS ESR could make the case that it, too, should retract the article. After all, Xiao was working on the revised manuscript long after the first paper had been accepted — indeed, almost to the week before the journal published his paper. That hardly suggests oversight or an administrative error.
Meanwhile, Circulation is retracting a 2002 paper by researchers in Lyon, France, who sought to double-publish their data. Here’s the notice for the paper, “Multiple Atherosclerotic Plaque Rupture in Acute Coronary Syndrome“:
Notice of Retraction
For the paper by Rioufol, et al., “Multiple Atherosclerotic Plaque Rupture in Acute Coronary Syndrome: A Three-Vessel Intravascular Ultrasound Study,” Circulation. 2002;106:804–808, the editors have discovered that the authors violated Circulation’s Ethical Policy, as stated in our Instructions for Authors:
“Manuscripts are considered on the understanding that they contain original material, that the manuscript and material within the manuscript have not been published and are not being considered for publication elsewhere in whole or in part in any language, including publicly accessible web sites or e-print servers, except as an abstract. The authors also certify that any and all other work in preparation, submitted, in press, or published that is potentially overlapping either in the actual data presented or in the conceptual approach is enclosed along with the original submission. Any material within the manuscript that has appeared elsewhere must be cross-referenced and permission to use or adapt the material must be received, in writing from the copyright holder.”
The authors violated the policy stated above in two ways:
The amount of original material in the Circulation manuscript is limited given the preceding publication, “Ruptures multiples de plaques d’ atherosclerose dans les syndromes coronaires aigus,” Archives Des Maladies Du Coeur Et Des Vaisseaux. 2002;95:157–165, from which original data has been taken.
The previous publication of some material in the Archives Des Maladies Du Coeur Et Des Vaisseaux’ paper was not acknowledged within the Circulation manuscript, and no permission from the original copyright holder for the material’s use in the Circulation paper was provided.
The editors of Circulation, therefore, retract the paper.
As it happens, the French study made a splash when it first appeared — in Circulation, that is — with a release on Science Daily and a story in American Medical News. The paper has been cited an impressive 382 times. The American Medical News article states that
The study is believed to be the first to use high-resolution pictures available with intravascular ultrasound to get three-dimensional views of the inside of the three major coronary arteries during the month following a heart attack.
Um, not so much. …
That last sentence of Circulation‘s retraction notice, by the way, initially was behind a $20 pay wall (an error, we’re told). At $2.5 a word, that’s a pretty steep price for a something the headline already told us!
Hat tip on Circulation: Clare Francis
Maybe I shouldn’t be, but I’m sympathetic to the group that got a splash of attention after they republished in English. I certainly wouldn’t be capable of reading and citing an article in French, if I could even find it.
I think that all papers should be in French.
English-speakers and the Latin-derived language-speakers would be able to understand
nearly all the abstract words without any special training. That is my opinion.
The rest is about the scientific record, which is actually meant to mean something, to be a reflection of reality, and to be useful.
You win some, you lose some.
http://icvts.oxfordjournals.org/content/14/6/915.long
Interact CardioVasc Thorac Surg (2012) 14 (6): 915.
RETRACTED: A 5-year clinical experience with bipolar radiofrequency ablation for permanent atrial fibrillation concomitant to coronary artery bypass grafting and aortic valve surgery
[Interact CardioVasc Thorac Surg 2008;7:777–80]
“the corresponding author, on behalf of his co-authors, declared explicitly that the work was original and that it had neither been submitted nor published elsewhere”.
Surely people can remember signing such declarations?
Same editor as Eur J Cardiothoracic Surgery, I believe.
He is at least one person who thinks that original should mean what it says.
I hope my praise does not damn him.
Re “Multiple Atherosclerotic Plaque Rupture in Acute Coronary Syndrome“:
The retracted article was published in 2002. If it took the editor 10 years to detect the “problem”, I’d guess there was a late tip-off from a bilingual whistleblower (someone involved in some dispute with the author?) rather than any actual oversight by the editor.
The language of the retraction notice sounds more like publisher-speak than editor-speak. Copyright violation is a legitimate concern, but I don’t see who was actually harmed by secondary publication in English — other than the affront to the editor’s sense of power. Retraction in this case looks like overkill to me. A correction notice pointing out the overlap in content and clearly linking the secondary and primary publications (so that people know the same data appear in both the French and English versions) might have been enough.
But the trend appears to be for editors to resort to retraction for any violation of the Instructions to Authors, on the grounds that the authors promise to provide accurate info in their covering letter, authorship, copyright transfer and conflict of interest, so if they are caught later (or much later!) the editor can punish them. I guess this is now seen as the most “virtuous” option in the wake of all the attention to the long delay in Lancet’s Wakefield retraction. Editors rarely seem to consider the possibility of honest misunderstanding of the rules for manuscript preparation and submittal, which are getting longer and more technical by the day.
Was there an actual problem in this case (between publishers, perhaps, over copyright?), and did retraction solve the problem?
In reply to Karen Shashok June 15, 2012 at 4:14 am
As pointed out by many people duplicate publications (2 publications for 1 piece of original work) are not benign. You are not supposed to reuse data. You are not supposed to reuse text. Duplicates are most useful in promoting the careers of some (at the expense of others). Darwin at work.
You do not need to be bilingual, or even understand the languages concerned to detect duplicates/unwarrnted overlapping publications). No need to invoke more simple, or more complicated senarios.
The pair below were in the same year. Duplicates come in very handy for solving that perennial pronblem of having to have a publication each year. Many institutes are hung up on this idea.
It is quite simple:
http://spore.vbi.vt.edu/dejavu/
http://spore.vbi.vt.edu/dejavu/duplicate/48958/
Open the webpage above,
click on the number in the ID box,
scroll down to see the side-by-side comparisons.
The overlap can be more than the blue shows.
you can see all the hyperlinks, for example the medline entries.
Arch Mal Coeur Vaiss. 2002 Mar;95(3):157-65.
[Multiple ruptures of atherosclerotic plaques in acute coronary syndrome. Endocoronary ultrasonography study of three arteries].
[Article in French]
Rioufol G, Finet G, André-Fouët X, Rossi R, Vialle E, Desjoyaux E, Convert G, Huret JF, Tabib A.
Source
Service d’hémodynamique, hôpital cardiovasculaire et pneumologique Louis-Pradel, hospices civils de Lyon, BP Lyon Monchat, 69394 Lyon. [email protected]
Abstract
The aim of this study was to assess the three coronary arteries systematically by endocoronary ultrasonography in patients with unstable angina to check the hypothesis of global destabilisation of atherosclerotic plaques in acute coronary syndromes (ACS). Sixty two coronary arteries were examined (2.6 per patient). Fifty plaque ruptures were diagnosed (2.08 per patient). Rupture of a plaque of the culprit lesion of the ACS was clearly detected in 9 patients (37.5%). At least one ruptured plaque on a site other than the culprit lesion was observed in 19 patients (79%), on another artery in 70.8% of cases and on two other arteries in 12.5% of cases. A complete endocoronary ultrasonic examination of the three coronary arteries in patients with a first ACS demonstrated that: multiple atherosclerotic plaque rupture may be detected by endocoronary ultrasonography; these multiple plaque ruptures occur simultaneously with the culprit lesion; they are frequent and can be situated on the three main coronary vessels and multiple plaque rupture other than the culprit lesion are less severe, non stenotic and less calcified. Thus, although a single lesion is clinically symptomatic, ACS seems to be associated with global coronary instability.
PMID: 11998329
and
Circulation. 2002 Aug 13;106(7):804-8.
Multiple atherosclerotic plaque rupture in acute coronary syndrome: a three-vessel intravascular ultrasound study.
Rioufol G, Finet G, Ginon I, André-Fouët X, Rossi R, Vialle E, Desjoyaux E, Convert G, Huret JF, Tabib A.
Source
Department of Hemodynamics, Cardiovascular Hospital and Claude Bernard University, Lyon, France. [email protected]
Abstract
BACKGROUND:
To test the hypothesis of general atherosclerotic plaque destabilization during acute coronary syndrome (ACS), the present study sought to analyze the 3 coronary arteries by systematic intravascular ultrasound scan (IVUS).
METHODS AND RESULTS:
Seventy-two arteries were explored in 24 patients referred for percutaneous coronary intervention after a first ACS with troponin I elevation. Fifty plaque ruptures (mean, 2.08 per patient; range, 0 to 6) were diagnosed by the association of a ruptured capsule with intraplaque cavity. Plaque rupture on the culprit lesion was found in 9 patients (37.5%). At least 1 plaque rupture was found somewhere other than on the culprit lesion in 19 patients (79%). These lesions were in a different artery than the culprit artery in 70.8% and were in both other arteries in 12.5% of these 24 patients. Complete IVUS examination of all 3 coronary axes in patients who had experienced a first ACS revealed that multiple atherosclerotic plaque ruptures were detected by IVUS; these multiple ruptures were present simultaneously with the culprit lesion; they were frequent and located (in three quarters of cases) on the 3 principal coronary trunks; and the multiple plaque ruptures in locations other than on the culprit lesion were less severe, nonstenosing, and less calcified.
CONCLUSION:
Although one single lesion is clinically active at the time of ACS, the syndrome seems nevertheless associated with overall coronary instability.
Comment in
* Circulation. 2003 Mar 11;107(9):e65-6; author reply e65-6.
* Circulation. 2002 Aug 13;106(7):760-2.
* Circulation. 2003 Mar 11;107(9):e64; author reply e64.
PMID: 12176951
Yes, I understand the same data were published twice and that’s considered bad.
Why is it bad for two different, non-overlapping populations of readers to have access to the same research results? Why is it bad for cardiologists who cannot read French (most of them, I’d assume) to have access to the information?
The reason why the primary publication (in French, in this case), needs to be cited in the secondary publication (in English, in this case) is so that researchers know the same data are reported in two separate articles. What the authors did wrong was neglecting to mention that the same data had already been published (in a language most cardiologists probably do not understand) and perhaps making the false claim that the data had not been published previously and were therefore “original”. These actions presumably broke the ICMJE rules and the rules Circulation asks authors to follow.
I’m not convinced, though, that retraction is appropriate in cases like these, where there is no scientific fraud but rather a procedural failure. If the authors had let the editor of Circulation know up front that the same data had been published in another language, had referenced the earlier article, and had let the editor of Archives Des Maladies Du Coeur Et Des Vaisseaux know they wanted to submit the same findings to a second journal, it is likely that Circulation would have had no problems with secondary publication.
Is it clear that the authors set out intentionally to fool readers about the previous publication history of the data? And why has it taken the journal 10 years to notice the overlap? Where Circulation’s instructions to authors clear about primary and secondary publication at the time the authors submitted their manuscript?
In reply Karen Shashok June 16, 2012 at 3:44 am
In the modern world, and perhaps in the ancient (before 2000) world, people are lazy and will only see the abstracts, both of which are in English. People are easily confused and might think there were two studies, which would be more convinicing. The titles in the example given, are not exactly the same, the beginnings of the abstracts are not exactly the same. When the similarities are pointed out to you, e.g. by the “deja vu database”, with the blue highlighting of the overlapping keywords and the same details, it is obvious, but otherwise one article might lodge in your mind and 5 months later another. Try comparing 2 articles when you cannot see them both at the same time. It is not so easy, you try to fill in the gaps by thinking that they get very similar results becasue they are studyin the same subject, some people suffer from the notion that the world is the most perfect it can be and “of course there must be legitimate reasons for any publicaitons”. I make no claim about the superiority of one language over the other. Where would you draw the line? 2 languages, 3, 4,…? Parallel universes?
Many people in the U.K. who have gone through high school (I believe this is necessary to become a medical doctor) can read French.
Most of the higly overlapping publications in French journals, and German journals, and English journals are NOT due to one being in one language and the other being in another. They are quite happy with English/English, French/French, or German/German. The argument that different languages for the same piece of work is permissible would give licence to these journals, and English journals (they are all languages) to merrily go on publishing near duplicates and unwarranted overlapping publications in one language. There is no need to be PC about giving the French or the Germans licence to break the rules. As soon as one mentions rules people take the libertarian view that they don’t like rules. The point about the rules is to stop the robber barons, and to have the same rules for everybody, otherwise they are not rules.
I don’t know if you are in the U.S., or another federal country, but how would you feel if people from some states, provinces, we allowed to publish twice, but you not? Of course you do not have to be French to publish in French journals, or German to publish in German journals, but most are.
Why should we forget arithmetic? Just as French (or another foreign language, 90% take French) is compulsary until the age of 16 in most of the U.K. (English is not the only language of education in parts of the U.K.), so is mathematics compulsary. I think that there may be a reason for that. Something to do with only giving the vote to the educated citizenary otherwise the authorities feared disorder.
An American example:
http://en.wikipedia.org/wiki/Boston_Public_Library,_McKim_Building
FYI : Massachusetts is a commonwealth, not a state.
“THE PUBLIC LIBRARY OF THE CITY OF BOSTON • BUILT BY THE PEOPLE AND DEDICATED TO THE ADVANCEMENT OF LEARNING • A.D. MDCCCLXXXVIII”; and on the north: “THE COMMONWEALTH REQUIRES THE EDUCATION OF THE PEOPLE AS THE SAFEGUARD OF ORDER AND LIBERTY”.