Note: two links were given with the question above, and they have been run together. They are:

http://download.thelancet.com/pdfs/journals/lancet/PIIS0140673613619339.pdf

and

http://exclusive.multibriefs.com/content/experiment-reveals-ugly-side-to-open-source-journal-industry?goback=%2Egde_2006026_member_5796611479251730433#%21

Sorry about that

Yes, last year I caught a (conference) paper "written" by a Head of Department, which had been accepted, subject to a final language edit. I was asked to edit it and discovered that it was a cut, shake and paste job from the Web. The paper was patently flawed as it described damp, cold living conditions in Scotland in a paper on housing in hot underdeveloped countries!

P.S. You can edit your own RG question if you notice an error, and even delete an RG comment of your own, if you later regret it. Hint: hover your mouse over the top right-hand corner of your comment.

Thanks Ian, and for the tip on editing. I hadn't realised it is possible.

There seems to be some interest in this question but people are reluctant to comment or answer. On reflection, perhaps this is not surprising. An answer might compromise or create difficulties close to home. I was more interested in general information and will try and reformulate a less risky version of the question.

Again for conference papers, I attended an International conference, where a paper was presented that had exactly the same title and contents (and author) as did a paper presented by the same author elsewhere at another conference! I stopped contributing to that conference.

Good for you, Ian, for reacting in the above manner. I have just come back from Cosmed 5th International Conference at RECSAM, Penang. RECSAM is a body under Southeast Asian Ministers of Education Organisation. Things went well, nothing 'spectacular' in the way that Jim has asked. (But again, I have come back with my head full of questions and I have put some on RG. Please share your wisdom, Ian and friends.)

Dear Jim, unfortunately I am not sure if I understand the final goal of your question correctly. Are you looking for evidence that it is possible to get a ridiculous paper submitted, even in high ranked journals, or are you looking for examples of journals accepting and publishing flawed papers?

If you are looking for the latter, i can provide you with some examples where a high ranked journal did so. While the editorial board had the stamina to publish critical comments, an official statement of the journal following the comments was lacking (or is lacking so far). In one case even despite the fact that the author of the criticism would easily have been able to cite the journals own rules for publication who were in discordance with the manuscript they broadcasted. Due to critical timing, I have to ask you to contact me personally for more detailed information.

If time is ready, I will be glade to update this answer with the according evidence/references.

Bruno. I am looking for the type of thing you describe and also looking for other examples of unacceptable behaviour in science. Examples of the sort of thing include

-frank dishonesty about data (eg making it up or massaging it)

-frank and significant abuse of the authorship system

-frank and significant abuse of thr refereeing system

And so on. There are multiple other "sins" that can be included, including those you mention. I have encountered all of these and more, and in one to one discussions find many others can echo this experience. However there is a great reluctance to bring discussion of the matter into the public domain, or to alert graduate students to this "dark side of science" in courses on research methodology. See the publication on the Dark Side of Science on my page, which was published in a philosophy volume.

https://www.researchgate.net/profile/Jim_Malone/contributions/?ev=prf_act

With this clarification I can answer your question with a solid yes. I was thinking about other/systematical fraud, so it took me a day until 'I heard ringing the bells'. As they did, I think I have to stand my man:

I, myself have been involved in abusing the authorship system. Probably I was even responsible for it, – at least I was surely in the position where I could have avoided it – my data, my hypothesis, me who initiated the publication. Yet, was this abuse frank and significant? While I had crafted the hypothesis and harvested the data, another author was substantially helping me in the data's interpretation and a third author was substantially involved by adding additional background information and helping me straighten down the manuscript to a small but publishable work. Two additional authors were involved by sharing a marker, commercially available and surely not worth more then a few $100s, the other by countercheck the manuscript once. 

Given the journals policy: "Authorship credit should be based only on 1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; 3) final approval of the version to be published.", it is highly arguable wether the two last mentioned authors fulfilled these criteria. I think it was mutual friendship and not scientific contribution that gave these authorships.

Yet, was it significant? On one side, yes it was significant because authors did at the utmost scratch the journals conditions for claiming authorship. On the flip side, no, as I knew that both colleagues did not get profit from this additional authorship, i.e. it did not significantly change their scientific career. For one author it was a hearty bye bye, as I knew he left science for good and this additional authorship did not give him additional merits. Concerning the other, I knew that he published enough solid work at his own, so the additional publication did not change much in his portfolio.

Why did I do it? Because I was younger, less experienced, and less stringent? Because there is this 'publish or vanish'-mentality, undermining medical practice at tertiary care facilities in the university-setting? Probably a mix of it.

To finish it, allow me to respond to your question, why we are reluctant to talk about the matter: 1st, If you talk about your own flaws you or close friends of you risk to get into problems. 2nd, if you talk about flaws of colleagues you know about, you have to precisely balance the pro's and con's, to make it public. Selfishly, you could loose profit of a well running 'giving-and-taking'-system assuring a maximised authorship-output in a lab. More philosophically thought, human by its kind, fortunately, is not eager to blame friends for misconduct, as long as it is considered 'acceptable' – many con's, but the entire flip-side would be a 'watch-your-neighbour-closly'-system, we blamed the eastern block for decades to act like this, and nowadays we are blaming the NSA rightfully for doing so… 3rd, If you talk about systematical scientific fraud in the public domain, you have to have courage and a broad back. As a scientist, you have to take into account, that your reputation might suffer. Sadly but not far-fetched.

Let me endorse the 3rd point with a non-scientific, yet suitable example. In Switzerland, back in the 80ies, when you were talking about our countries misbehaviour during world war II (WWII), many older citizens blamed you for defamation. In fact, the swiss government or its affiliated services stored such and other whistleblowers systemically (at least 700'000 entries in a country with 6.5 million citizens in the late 80ies). This was done until it became public in 1989 (Schweiz. Parlamentarische Untersuchungskommission (EJPD): Bericht der Parlamentarischen Untersuchungskommission (PUK) vom 22. November 1989: Vorkommnisse im EJPD. Bern 1989). It took us nearly another decade until the class actions in the U.S. forced the officials of Switzerland to admit such misconduct like gold transfers between the National Bank of Switzerland and Germany during WWII, and systematical abuse of name-less property of jews, killed during the holocaust (http://www.uek.ch/de/schlussbericht/synthese/uekd.pdf).

I think the discussion can be brought into running. But to bring the topic beyond philosophical discussion level, a way has to be found to gather data systematically (yet to gather hidden fraud will not be easy), evaluate it, and present the results like any other biomedical study – big work, hard to achieve. Hence, it was not for nothing I zoned out into swiss politics, where the shear power of the U.S. was needed to trigger a systematical work-up of the obvious evidence that finally brought a reluctant country to take responsibility for its misbehaviour it willingly ignored for halve a century.

@Bruno: I do not classify you as an author abuser. Remember that only certain fields have _recommendations_ about authorship today, and such recommendations have only _recently_ become availablefor certain fields.

I have examples from my own history where I hardly added a word to a paper with my name on it, and conversely where a "co-author" (sadly) hardly added a word to my paper. Usually once an agreement to co-author has been taken orally, it sticks. I do now favour a co-author contract to formalise the proces.

Bruno and Ian

Here is an excellent book you may both be interested in. It is The Trouble with Medical Journals by Richard Smith. He is a former editor of the British Medical Journal (BMJ), which is the fourth most cited medical journal (impact factor ~17) and deals with all aspects of clinical practice. He writes well, is very accessible and covers many topics including, for example, the peer review system and the financial model on which journals are run. In the latter case his analysis is devastating, as he makes the case that the main product journals sell is knowledge. Yet they contribute little or nothing in a financial sense, to creating new knowledge, or to the peer review or editorial processes. This leaves them as potentially very profitable activities for publishing houses and professional bodies. The link below is to the Amazon entry for the book (in which I don't have any commercial interest!) and although it was published in 2006, it is still relevant.

Smith also writes a provocative Blog in which he constantly presses the scientific community to improve how it deals with its own affairs. The entry of 9 Sept 13 starts as follows: "Most scientific studies are wrong, and they are wrong because scientists are interested in funding and careers rather than truth". A little alarming but he goes on to discuss it and explain his basis for the statement.

http://www.amazon.co.uk/Trouble-Medical-Journals-Richard-Smith/dp/1853156736

If you google hoax research and bogus conferences, you would be amazed of how many reports are documented. The researcher's ethics should act as a barrier against erring and even the thought of it.

If you wanna skip this monologue you can find a letter with references enclosed on my profile: https://www.researchgate.net/profile/Bruno_Strebel2/?ev=hdr_xprf (dataset temporarily removed 12-06-2013)

In March 2013 Guan et al. published online in the Journal of Clinical Oncology (JCO) a study comparing lapatinib (a drug by GlaxoSmithKline - GSK) vs placebo in HER-2-positive metastasised breast cancer (HER2+MBC). Both arms also got paclitaxel. The trial started recruiting patients in 2006.

The authors claimed to stick to the Helsinki-Declaration version 1996. Yet, this version was out-dated. In the 20-10-2013 issue of the JCO, Dr. Ades pointed out, that the declaration 1st was out-dated, yet 2nd already recommended that “the interest of science and society should never take precedence over considerations related to the wellbeing of the subject”. In 2006, it was known for 5 years that trastuzumab is beneficial in HER2+MBC. The sole reaction on Dr. Ades critical comment was a replay by Dr. Guan arguing that "unfortunately, anti–human epidermal growth factor receptor 2 drugs were not universally available at that time in developing countries."

Does GSK who funded this n=444 trial not have the recourses to provide the control arm trastuzumab? Or were they not willing to do so? Maybe they wanted the best advantage for their own lapatinib? And they did right to do so (from an economic point of view), because they risked the direct comparison in October 2008. With a comparable design, lapatinib's efficacy was compared vs trastuzumab… and failed. This second study was stopped early in 2012 (officially presented at the ASCO-meeting), halve a year before Guan et al. reported their results.

So, not only did GSK sponsor a study with a questionable ethical design, Guan et al. reported the results of this study in March 2013, halve a year later then it was known (at least for those who attended the ASCO- meeting 2012) that lapatinib failed the direct comparison to the standard of care treatment. A treatment that was not provided in Guan's study design. Did Guan et al. comment this fact? No! Was none of them aware of it? I don't believe so. Was the sponsor of both studies not aware of it because GSK lacks in-house-communication? I don't think so. My only sarcastical conclusion is, that the sponsor, i.e. GSK was sticking that hard to the ethical principle, that a sponsor should not interact with its researcher, that it did not tell a word to Guan et al. about the 2nd mentioned trial that failed.

Five days after Ades comment, it was October 25th, I contacted the JCO by mail, and asked how they want to discuss this discrepancy with me? No Answer. So, at October 29nt I submitted a comment to get at least a response. It was thanksgiving and, imagine, what happened after thanksgiving at December 2nd? I got my answer ASAP: stay calm, we are investigating the issue already. Today they declined my comment. A decision I can accept. Yet, as the manuscript is my own property again, its posted for download on my profile: (https://www.researchgate.net/profile/Bruno_Strebel2/?ev=hdr_xprf – dataset temporarily removed 12-06-2013). You also can find there, the network action in ResearchGate that took place at December 2: hits from China (Guan), Belgium (Ades) and US (JCO).

Now, dear Jim, I can tell you another story about reluctance. All mails were declared CONFIDENTIAL, as was all the data. While I can not remember that I willingly agreed to any contract agreeing into confidence, I can not say for sure, that I overlooked such a statement. And at December 2nd the first, and at December 3rd, the other scientist who initially agreed to the issue, and were in contact with me stopped interact (one because he was 'kindly asked' to do so by his superior, the other, I don't even know).

This happens quite a bit and editors need to be vigilant. At Computers in Biology and Medicine (Elsevier), one of my first days on the job in January 2013, my managing editor writes to me that she has noted changes to the data table when the author sent her the galley proof corrections. This was a single author paper, the author a postdoc who was switching institutions. I had nothing otherwise to do with the paper as it was refereed and accepted before my time. I noticed that he had added 5 points to all of the accuracies without even changing the decimal part of the number. I confronted him, and mentioned the paper could not be published. As he was young, I advised him to redo the entire study then send to another journal, I would not accept a de novo submission. In other instances I have found significant portions of a manuscript copied from other authors' work. For those I also reject, stating why, without the possibility of de novo submission. The latter happens quite frequently. At this point I have never contacted an offending author's institution for data fabrication, or the plagiarized group. I think under some circumstances I would, as in the case when there could be real harm to patients, but it has not arisen yet.

Edward, I think your response was a proportionate one and may be enough to persuade an ordinary person to consider what they have done and be more careful in future. In our responses to these problems being able to be proportionate in how we respond is an important part of the apparatus we have to create for dealing with unacceptable behaviour which is damaging to science and possibly to individuals. We must not ostracise people for minor misbehaviour, and also we must not ignore such behaviour and let it pass.

Two excellent references on the unacceptable face of science in medicine and in the pharmaceutical industry are:

Bad Science by Ben Goldacre

http://www.amazon.co.uk/Bad-Science-Ben-Goldacre/dp/000728487X and

Bad Pharma also by Ben Goldacre

http://www.amazon.co.uk/Bad-Pharma-How-Medicine-Broken/dp/000749808X

Both provide a lot of well researched background information. The story of the MMR vaccine told in Bad Science is truly shocking. More recently this was treated in a series of special articles in the British Medical Journal (BMJ). The first one is at:

http://www.bmj.com/content/342/bmj.c5347#alternate

and I will try and download it below.

http://www.bmj.com/content/342/bmj.c5347#alternate

Thank you Jim for the attachments provided. Sharing such cases is creating awareness.

As a journal editor (Clinical Obesity) yes. A paper that was copied from one already published but data claimed to come from a different country. Most journal management programmes now have ready plagiarism checking software which provides significant defence against plagiarism, but not fraud.

As a referee I have rejected a paper due to design flaws and then been sake to referee the same paper for a other journal where the same study suddenly had an n of subjects doubled, but the mean data identical. This sort of issue is probably not uncommon and much harder to detect.

Thanks for that Nick. I did not know about the plagiarism software and the resubmission incident is the type of thing that must be a temptation for many at an early stage in their career. Multiple authors should be a check against it, but unfortunately co-authors often do not read the paper, let alone check the statistics. Even if they read it and check the statistics, it does not mean the situation will be corrected. I know of a situation in a top rank journal where the number of patients involved was overstated by a factor of almost 10, and it was pointed out by a junior author. Yet the senior author went ahead and the publication was accepted.

Hi all,

Well, the next one:

https://www.researchgate.net/publication/259498754_When_the_Abstract_Reports_a_Well-Tolerated_Treatment_the_Data_Should_Support_It

Barlesi et al.published in the JCO the AVAPERL trial (JCO 2013) (http://jco.ascopubs.org/content/early/2013/07/03/JCO.2012.42.3749.short?rss=1), reporting a survival benefit by adding pemetrexed to bevacicumab in patients with advanced nonsquamous Non-Small-Cell-Lung-Cancer (3.7 v 7.4 months). While the benefit of 3.7 months can be questioned by itself, the study reported a good tolerability. The two endpoints to access this finding were adverse events and discontinuation of the treatment due to it (both twice as much in the 'well tolerated' combined treatment arm. The other endpoint was quality of life (HRQoL) – in the main body of the publication – this was covered by two sentences: “A preliminary report on health-related quality of life found no significant differences between patients in the maintenance arms; full results will be published elsewhere.” The data cited to endorse this statement referenced to an interim analysis of HRQoL in the AVAPERL-trial itself, presented 2011at the ECCO-congress (European CanCer Organisation) which concluded (by the same author-consortium: “there might be a trend toward improved HRQoL with bevacizumab alone compared to bevacizumab plus pemetrexed." And of course, the promise, that more detailed data will be published later. Finally, about two months after the publication of the AVAPERL-trial, the HRQoL-subtopic was really published elsewhere (http://www.ncbi.nlm.nih.gov/pubmed/24077452?dopt=Abstract). I did not expect that, yet I am still curious, why it was not published together with the more favourable endpoints in the main publication. Do I have to answer this question? I don't think so.

In summary, the HRQoL endpoint was of good use threefold:

1) It gave a congress poster 2011 while focussing on the 9 significant sub-topics out of 16 ones concluding a trend towards bevacizumab alone.

2) The reference to this poster together with the promise that data will be published in detail later was obviously sufficient to cover (or bypass – i.e. not really discuss) the endpoint concerning HRQoL. Yet, as I am aware that the 2011 poster was an interim analysis, and a longer follow up COULD have changed data, it A) did not (see later) and B) the reference used was still pointing straight forward to the 2011-poster that ended with a completely opposite conclusion.

3) At the end, using all 16 HRQoL-topics for the first time, the data brought A) a second publication and B) while using all 16 topics, the statistical calculation was able to dilute the earlier reported trend into an non-significant over-all result. Of interest, the table that was used in this third and definite publication also picked out the 9 significant topics, 8 of them still being in favour for bevacizumab alone.

All the best,

Bruno

P.S.: The pipeline is not empty yet...

Article When the Abstract Reports a Well-Tolerated Treatment, the Da...

Hi Jim, its a long time. Maybe this would of interest for you: a statement of medscape-authors , commenting the withdrawal of a questionable 'editorial' published recently in the Journal of Clinical Oncology (JCO):  

http://www.medscape.com/viewarticle/835436?pa=7PWp0VDauotDCPv%2BRv1M%2BI2rDxrAAHgLI%2FkaaZ6YTdvNf6lvqBir2EZxOU920Ic2CHoDDyTRW0Y0TjD2Y1gUmA%3D%3D

I got aware of it by a manuscribt published by medscape at  2014-11-28, today the provided link leads to and version updated at 2014-12-16.  

The medscape comment  at Nov 28 reports that the JCO did publish a questionable article in the 'comments and controversies'-section on Nov 17, 2014 the article did question the conclusion of another study – the new EPOC-trial – published halve a year earlier in the Lancet Oncology), and withdraw it few days later. At Nov 28, I assumed that the JCO at least will re-publish the editorial explaining why they decided to withdraw it temporarily. Such a withdrawal, will surely not be explain by a typographical error... Interestingly, at the evening of Nov 28 (CET), the editorial was online again, unchanged. This information about the JCO's action, is hidden (you have to finde a small-fonted Link (below the link to download the file as a PDF).

The content of the JCOs explanation lacks any back-bone – I am citing theEditor's comment: The commentary entitled, "Should the Results of the New EPOC Trial Change Practice in the Management of Patients With Resectable Metastatic Colorectal Cancer Confined to the Liver?" (J Clin Oncol doi: 10.1200/JCO.2014.58.3989), was temporarily unavailable on our Web site in order to give the authors an opportunity to respond to issues raised by a reader. Journal of Clinical Oncology has received the authors' response and concluded that the original article should be made available for the benefit of our readers, in accordance with our policy for peer-reviewed, accepted manuscripts. We encourage readers to submit their comments on published articles to the Correspondence section of JCO, so that they are available for all to consider.

By publishing an article, in my eyes a journal takes a port of responsibility of its content. This is one of the reasons, a journal asked for time to internally review the article, and also asked for external opinions, the peer-review-process. To argue, that possible discordance can be sorted out by commentaries of 3rd party readers, is a bit fewer then expected.

Given the JCOs own declaration about handling conflict of interests: (http://jco.ascopubs.org/content/31/16/2043), stating: "

V. RESTRICTIONS ON CERTAIN RELATIONSHIPS

This policy calls for disclosure and management of financial relationships with Companies and does not create a presumption of impropriety in the existence of these relationships. However, ASCO's policy is that certain financial relationships give rise to conflicts of interest that are not capable of being effectively managed and are, in fact, inconsistent with actual and perceived independence. Individuals who are free of these relationships should play a key role in the authorship of original research submitted to ASCO meetings or to JCO and JOP. Although the contribution of every author is important, ASCO looks to these individuals primarily to ensure the balance and objectivity of the research as presented."

What do we need more then three of three authors with relevant conflicts of interest to question the independence of these authors?

http://www.medscape.com/viewarticle/835436?pa=7PWp0VDauotDCPv%2BRv1M%2BI2rDxrAAHgLI%2FkaaZ6YTdvNf6lvqBir2EZxOU920Ic2CHoDDyTRW0Y0TjD2Y1gUmA%3D%3D

http://jco.ascopubs.org/content/31/16/2043

It seems that most of the problems mentioned apply to medical research. Are you aware of any in sciences?

Hussin, I don't follow the issue in physics or maths, but have heard colleagues mention problems in a similar vein that occurred with graduate students.  One colleague feels that the problem is there in physics but on a lesser scale.  The real answer is we don't know, but physicists seem inclined to believe their community is more committed to the values of science than those from other disciplines?  My personal opinion is that this is not likely to be true.

Jim, I think medical cases are getting wide audience, as there is NOT ONLY science; it is our HEALTH and, in oncology, OUR LIFE!

On my behalf, for biology, I had several submissions which I checked, and after I  asked author to provide raw data or excell files, they dissappeared. I think later I will look for these titles in the paid OS journals - they for sure will take the bait.

And, best 2015 to you all! 

I agree Linas. The impact on individual and community health is big, and means that much that would otherwise be ignored comes to light.

Have a good 2015 and enjoy joining the euro!

are you sure about the last 4 words, Jim?

"One would think correcting this was a trivial matter"

Why not review the topic in question, including a critique of your thesis?  You will end up with a useful accessible publication in a journal rather than a PhD thesis which very few will read and even fewer cite, no matter how good or bad it was.

Your advisor supports you, but the external examiners do not. A tiebreaker is needed.

They appear to criticise your thesis-book on the grounds

that you are not building on the giants who have gone before

you (a common problem). So, if you want your PhD, you have to

retrieve it with as much grace as possible. Obviously you cannot

start a new thesis title, although it may be necessary to modify

your original title slightly in the light of your findings and even

the comments of the external examiners.

The solution is to take each criticism of your external examiners,

and use each, either to support what they say or to refute what they say IN FOOTNOTES.

This does not disturb your pagination or indexing:

___________

1. According to the foremost authority in the field, Wagner (1970) believed that A=B.

However, this research shows with a 95% confidence interval that AB.

2. Bach (2015a) has results which confirm my hypothesis.

3. There is doubt whether Offenbach(2000) had the same timeframe in mind when he reported on his experiment.

Then arrange for your university to convene an ad hoc committee to examine whether you have have satisfactorily responded to the criticisms of your external examiners. Good luck.

There is a website, http://retractionwatch.com, that looks at journal retractions, whether due to fraud, flawed methodology or other reasons.  They mainly cover the life sciences, as that is their professional field of interest, but will cover others if articles come to their attention.  An article that springs to mind with your question is one by three serbian scientists who had a phony paper accepted by a Journal.  Some of their prose was wonderful :P  "Bugs in our system caused the unstable behavior throughout the experiments.  On a similar note, witlessness in supporting theory caused theunstable behavior throughout the experiments."

http://retractionwatch.com/2013/09/23/a-serbian-sokal-authors-spoof-pub-with-ron-jeremy-and-michael-jackson-references/