A prominent (yet controversial) journalist in Australia has admitted to duplicating three images that were part of her PhD thesis — a practice outside experts agreed was acceptable, if not ideal, at the time, according to a report released today.
As part of an inquiry, the University of Adelaide convened an expert panel to investigate 17 allegations of duplication and/or manipulation in Maryanne Demasi’s 2004 thesis. Duplication is a common reason for retractions, such as when researchers use the same image to depict the results of different experiments.
After earning her PhD in rheumatology, Demasi became a journalist who got headlines for more than just her reporting. In 2013, she produced a controversial series about cholesterol and fat (which suggested they have been unfairly villainized, and which cast doubt on cholesterol-lowering drugs known as statins). A few years later, Demasi was fired from the science program Catalyst, after it aired an episode alleging wi-fi could cause brain tumors.
Regarding the allegations of misconduct in Demasi’s thesis, the originals of the images in question were long gone, so in 14 instances, an expert concluded it was not possible to conclude whether or not duplication had occurred. But in the remaining three instances, Demasi admitted she had “duplicated or probably duplicated” the images:
However, she understood that duplication was acceptable at the relevant time for the particular category of images at issue. The relevant experts agreed, albeit they considered duplication was not best practice. The Panel accepted this evidence. They could not be satisfied that the duplication constituted a deviation from the applicable standards at the relevant time.
As a result, the report concluded:
In summary, on the basis of all of the evidence before the Panel, the Panel could not be satisfied that any research misconduct had occurred.
The report does not specify which images Demasi admitted to duplicating.
Last year, the Journal of Biological Chemistry issued an expression of concern for one of Demasi’s papers that arose from her 2004 thesis, noting “credible concerns have been raised regarding some of the data and conclusions in the article.” Also last year, the journal Arthritis & Rheumatology, which published this 2004 paper, told us it had received an inquiry from the University of Adelaide, and “may publish a correction or retraction pending the conclusions reached by the University.” We contacted both JBC and Arthritis & Rheumatology, but did not receive an immediate response.
We reached out to Demasi via the contact form on her website; according to the inquiry report, she “firmly denied any research misconduct.”
As part of the inquiry, the expert panel commissioned reports from an expert in forensic image analysis, as well as two experts “in the standards applicable to the presentation of Northern and Western blot images at the relevant time.”
The report notes:
The evidence established that the original x-ray films which corresponded with the impugned figures had been discarded, in accordance with appropriate records management requirements, or lost in the fourteen years since the completion of the Thesis. The expert in forensic image analysis warned that it would be “dangerous” for the Panel to make any findings of duplication in the absence of these original films. The Panel accepted this evidence. They could not be satisfied that duplication had taken place in respect of the 14 specific allegations where the Respondent had denied duplication.
Recently, Demasi raised questions over whether doctors were over-prescribing coronary stents.
Update, May 10 2018 1:11 UTC: We’ve received a response from Demasi:
I thank the independent tribunal for its careful consideration of this matter.
The tribunal’s finding demonstrates an unequivocal rejection of each and every claim levelled against me, all of which were found to be unsubstantiated.
It has taken almost two years, and considerable legal expense, to defend myself against allegations that had no basis.
Moving forward, my case demonstrates that serious questions need to be asked about the current framework for making and investigating claims of research misconduct to ensure that careless or vexatious claims cannot qualify for investigation.
Update, May 10, 2018, 17:36 UTC: We’ve heard from a spokesperson for the University of Adelaide, who told us:
The process followed by the University was compliant with the Australian Code for the Responsible Conduct of Research. The panel comprised former South Australian Supreme Court judge the Honourable John Sulan, QC (chair) and three distinguished professors from New South Wales and Victoria.
The panel determined that allegations of research misconduct were not substantiated.
The University has accepted the panel’s findings. No further action will be taken.
The panel’s findings can be found on the University’s Research Ethics, Compliance and Integrity website.
Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
I fail to understand what is meant here by the term “duplication of images”.
Is it the use of the same picture in more than one thesis or article? If so I can recall many instances, all of which openly declare that it’s done.
Is it using a picture from one experiment to illustrate another? Is duplication the correct and sensible term for that?
Or is it something else entirely?
It looks like it’s duplicating images from one part of the experiment to anther. The pubpeer thread on her JBC paper shows bands that looks like they’ve just been copied and pasted: https://pubpeer.com/publications/2AC1B6191D3B4ACD01A41C4AF45E5C
Combined with her admission that she duplicated or probably duplicated the images, I don’t understand the panel’s failure to find misconduct (but then again, they never do find misconduct). It almost seems as if the panel was intimidated into finding no misconduct – being told that it would be “dangerous” to do so.
Not quite sure how image duplication was ever an acceptable practice…
No hint of the Expression of Concern
http://www.jbc.org/content/292/38/15993.full
on the main article page
http://www.jbc.org/content/278/40/38607.full
nor in the PDF of the article, even though the EoC was issued many months ago.
Let me get this right. Dr Demasi admits to duplications https://pubpeer.com/publications/2AC1B6191D3B4ACD01A41C4AF45E5C
The experts in the independent tribunal say this is acceptable.
Yet Dr Demasi says in her response that all of the claims were found to be unsubstantiated.
I don’t understand how she can admit to the duplications, but say allegations that figures contained duplications were unsubstantiated.
-seems like the physics of Schrodinger’s cat: the duplications simultaneously exist, and yet have not substance.
Also last year, the journal Arthritis & Rheumatology, which published this 2004 paper
Here is the relevant Pubpeer thread: https://pubpeer.com/publications/08CFB7DEED5038D5ED645B192DE6E1
The Thesis is credited to Dr Demasi, yet the 2004 paper is credited to Maryanne Demasi , Leslie G Cleland, Rebecca J Cook-Johnson , Michael J James.
I cannot find the division of labour, but I did find
Acknowledgements
We are grateful to Dr. Greg Goodall for the helpful discussions regarding hypoxia and the use of the hypoxia chamber, and to Jodie Nitschke of the Renal Laboratory at the Hanson Institute for her assistance in Northern blot analysis.
It seems likely to me that in defending her thesis, Dr Demasi may merely be guilty of refusing to throw others under the bus, 10 years after the event. After all, she knows attention has only been drawn to the blots due to her own recent willingness to risk controversy in other areas, and has nothing to do with her collaborators.
The same duplications are present in both the thesis and the papers. All authors bear some responsibility for the contents of papers that bear their names, but for a PhD thesis, it is hard for me to see that anyone other than the author should be held responsible. Regardless of who produced the duplications, those figures can’t be used to support the conclusions of the papers, so they should be retracted by the authors or the journals.
I agree with Mr. Briggs, how does image duplication not lead to retraction? Presenting the same data (image) as the results of two different experiments automatically invalidate any discussion or conclusion that can be drawn from at least one if not both experiments. Unless originals can be provided for a correction, such paper has no place in the literature.
Another paper from the same group mention on PubPeer:
https://pubpeer.com/publications/0C11C249612DEB7B759439A8676A7D#1
Regardless of whether the blots are identical or not (and I think some are very suspicious), the true test would be whether or not her work can be replicated. Has anyone tried this?
JBC editorially retracted the paper in December 2018:
This article has been retracted by the publisher. An investigation by
the Journal determined the following. In Fig 4, the “no LPS” lanes in
the GAPDH Northern blots were duplicated between normoxic and
hypoxic conditions. In Fig. 6A, several bands were duplicated in the
COX-2 immunoblot. In Fig. 10, the first lanes between normoxic and
hypoxic conditions were duplicated in the phosphorylated cPLA2
immunoblot. Additionally, in Fig. 10, lane 3 of the normoxic panel was
reused in lane 2 of the hypoxic panel.
Arthritis & Rheumatology published an expression of concern in August 2019:
Expression of Concern: Demasi M, Cleland LG, Cook-Johnson
RJ, James MJ. Effects of hypoxia on the expression and activity of cyclooxygenase 2 in fibroblast-like synoviocytes: interactions with monocyte-derived soluble mediators. Arthritis Rheum 2004;50:2441–9.
https ://onlin elibr ary.wiley.com/doi/10.1002/art.20429
Arthritis & Rheumatology (formerly Arthritis & Rheumatism) and John Wiley & Sons are issuing an Expression of Concern to inform readers that questions regarding the similarity of images used in Figures 1, 2, and 6 of the above-referenced article were raised, which may be pertinent to the data and conclusions of the article. An inquiry convened by the University of Adelaide, SA, Australia did not find research misconduct.