By David Tuller, DrPH
Earlier this week, I wrote to the journal Psychological Medicine about a significant mistake in a paper on functional neurological disorders. The mistake involved a misquotation of Bermingham et al, a key 2010 analysis of the National Health Service costs associated with care for working-age people found to be “somatising.” One of Psychological Medicine‘s two editors-in-chief, Professor Robin Murray of King’s College London, responded promptly and indicated that he was checking with the authors about the matter.
I subsequently followed up with Professor Murray to point out that the senior author of the paper was also co-author of another paper that had been corrected for the same mistake in October, 2019. He again responded promptly and indicated that he would check on it.
In the meantime, today I sent the following to Professor Anthony David, a neuropsychiatrist and the senior author of the Psychological Medicine paper. He previously was at KCL, along with Professor Murray and a co-author of the paper in question, Professor Sir Simon Wessely. Professor David is now at University College London.
Dear Professor David–
As you are likely aware, I have approached Psychological Medicine about a mistake in O’Connell et al, an article for which you are the senior author. (Other co-authors are Professor Tim Nicholson and Professor Sir Simon Wessely.)
The article is called “Characteristics of patients with motor functional neurological disorder in a large UK mental health service: a case-control study.” In communicating with the journal, I also noted that another paper on which you were a co-author was corrected for the exact same mistake 18 months ago.
In addition to seeking the correction, I am reporting on the issue for Virology Blog, so I am reaching out to you directly to see if you would like to offer any comments about what happened. (I have cc’d my colleague Vincent Racaniello, Higgins Professor of Microbiology at Columbia, who hosts Virology Blog.)
In brief: In 2010, Bermingham et al reported that the costs of care for working-age people who were purportedly “somatising” accounted for 10% of National Health Service spending for working-age people. Many subsequent papers about so-called “medically unexplained symptoms” (MUS) or “persistent physical symptoms” (PPS) have garbled that statistic, declaring instead that these costs represent 10% of total NHS spending. In fact, they represented about 3% of total NHS spending in the year covered in Bermingham et al.
Any data point about NHS costs has possible policy implications–even more so when a mistake has the effect of tripling the apparent impact, as this mistake does. Certainly the false budgetary claim would seem to bolster an argument for the expansion of specialized services supposedly effective in treating MUS/PPS (i.e. cognitive behavior therapy and related approaches). That is generally the position of those disseminating the false claim in the first place.
Here are my questions:
Were you aware that another paper you co-authored was corrected for this same misrepresentation of Bermingham et al in October, 2019? If not, why not?
If so, why didn’t you move to correct O’Connell et al? In the forthcoming correction, will you take responsibility for this 18-month delay since the same false claim was corrected in your other paper?
Why do you think so many investigators in the MUS field have repeated this false claim in paper after paper across a decade? If thought leaders in this domain can spend years collectively misquoting a seminal paper in their area of expertise, should that raise concerns about their overall scientific literacy and about whether they are basing other assertions about MUS on similarly distorted, misunderstood or biased data?
Thanks very much. I will of course include any comment or explanation you choose to offer, at any length, in a post on Virology Blog. In the meantime, I also look forward to hearing back from Professor Murray about the timing of the correction itself.
David Tuller, DrPH
Senior Fellow in Public Health and Journalism
Center for Global Public Health
School of Public Health
University of California, Berkeley
8 responses to “My Letter to Professor Anthony David Asks Why MUS Experts Keep Misquoting a Major Study”
It’s impossible to attribute this to error at this point, it’s been repeated and pointed out so many times by now it’s clearly deliberate disinformation. And it frankly shows how utterly broken the peer review system is, a mere formality to some. It offers the same level of quality assurance as a sign that says “DO NOT ENTER” provides actual security to an open door with no camera and a direct path to the ventilation system that blows the whole thing up.
To repeat part of my comment from a previous blog:
“….by my counting, 9 of the authors are/were professors at the time of publication. In my day, professors were learned people who knew their stuff. From time to time, they wrote carefully crafted papers that were highly regarded. What happened?”
One of those professors was Helen Payne who, according to an entry on the ResearchGate website*, appears to have ‘done it again’ too in this paper -https://journals.sagepub.com/doi/10.1177/1541344619883892 – that was published November 2019, despite Tuller alerting her to the problem 4 months earlier on July 1st 2019 – https://www.virology.ws/2019/07/08/trial-by-error-some-thoughts-on-mus-and-bermingham-my-letter-to-professor-payne/ .
WOULD ANYONE WITH ACCESS TO THE PUBLISHED PAPER LIKE TO CHECK?
This time, she appears to have not only inflated the Bermingham et al estimate (for the cost of somatization to the NHS) of £2.89billion to £11.64 billion, but to have also mistaken Bermingham et al’s overall estimated cost to the economy of £18billion to be additional indirect costs. Thus the readers of the paper might calculate that the total cost of MUS to the economy is over £29.64 billion, rather than the £18billion estimate of Bermingham et al. NB It’s possible that the published paper, rather than the one uploaded by Payne onto the ResearchGate site, is correct.
Bermingham et al’s estimates have also been relayed incorrectly in several UK parliament petitions regarding FND including a current one, as explained in this recent post written by an anonymous blogger -Opposing MEGA – Important post from a writer who would… | Facebook . A key point made here is that the cost estimates are being used in relation to FND, but FND proponents make clear that FND should NOT be lumped in with MUS because doing so can cause patients harm. If this is so, then surely the costs of FND shouldn’t be lumped in with the estimated costs of somatization/MUS either?
Apologies that link didn’t work.. I hope this makes it trough moderation (please copy and paste without the – either side of the link):
Ooo, that’s interesting Lady Shambles. So the implication is then that the O’Connell et al paper (that this blog is about) perhaps shouldn’t have cited the Bermingham et al study at all as it should be irrelevant to “motor functional neurological disorder” that the paper is about (because FND Action says FND shouldn’t be lumped in with FND)? It looks like the sentence about the Bermingham et al estimate immediately follows a sentence that states that functional disorders often result in high healthcare and social care costs, so any reader might think that the Bermingham et al study is therefore relevant to the motor FND under discussion. It seems that 2 of the authors – Nicholson and David – are currently on the Medical Advisory Board of FND Action – so do they agree with the statement put out previously by that charity -https://www.fndaction.org.uk/fnd-action-response-to-fnd-petition/ that FND should be regarded as a diagnosis in it’s own right and not lumped in with MUS? If so, why on earth did they reference the Bermingham et al paper?
I’m guessing that people from outside the UK may be wondering what all the fuss over these MUS/somatization cost estimates is about. David explains this very well above, but I think it needs emphasizing so here it is again and capitalized for what I consider to be the most important bit:
“Any data point about NHS costs has possible policy implications–even more so when a mistake has the effect of tripling the apparent impact, as this mistake does. CERTAINLY THE FALSE BUDGETARY CLAIM WOULD SEEM TO BOLSTER AN ARGUMENT FOR THE EXPANSION OF SPECIALIZED SERVICES SUPPOSEDLY EFFECTIVE IN TREATING MUS/PPS (I.E. THE COGNITIVE BEHAVIOUR THERAPY AND RELATED APPROACHES). That is generally the position of those disseminating the false claim in the first place. ”
In the UK, ME/CFS is being lumped in with MUS by the NHS and the threat is that this strategy could expand across the globe if BPS proponents have their way and if governments and health insurers think that there could be substantial savings from doing so.
Correction to the above –
It should read: ” … (because FND Action says FND shouldn’t be lumped in with MUS)? “
These junk papers are full of bogus numbers because the details don’t matter when one’s “research” is marketing dressed in a lab coat. Authors clearly don’t expect (or desire) anyone to actually read their gibberish.
Likewise abstracts aren’t supposed to be read. We are only supposed to listen to the party line as delivered by the Science Media Centre. Actual reading and thinking is not expected nor desired, and will soon be dispensed with altogether.
How come Robin Murray is responding promptly to this point when he’s ignored the inaccurate and misleading claims in the 2013 PACE recovery paper? This paper not seen as too big to fail?