By David Tuller, DrPH
What does it mean that the top investigators in a field of research have collectively and consistently misrepresented a seminal figure in their purported domain of expertise? I’m talking about all those who present themselves as authorities on the topic of so-called “medically unexplained symptoms” (MUS) but have found it difficult to accurately cite a study that examined the purported costs of these conditions to the National Health Service. (This phenomenon was initially pointed out to me by an observant sleuth; I first wrote about it here.)
The study in question, Bermingham et al, is called “The cost of somatisation among the working-age population in England for the year 2008-2009.” It was published in 2010 in the journal Mental Health in Family Medicine. As one of its core findings, it estimated something very specific: NHS costs for patients of working age identified as “somatising” to even a minor extent accounted for around 10% of the total costs for that demographic category. In paper after paper, “experts” in MUS have translated that straightforward data point this way: Patients with MUS accounted for 10% of total NHS costs. The effect of that error is to more than triple the apparent impact of MUS on the NHS budget.
(Special Note for MUS experts reading this post: References to Bermingham et al’s age parameters are in mentioned in the title of the paper itself.)
It is unclear why these thought leaders have been so statistically challenged across so many years. I assume the mistake has been unintentional. Nonetheless, it has served to support the argument for desired policy outcomes, which could explain why it went unnoticed by the authors as well as any like-minded colleagues who peer reviewed the papers in question. If health officials wrongly believe that costs related to MUS account for 10% of their budgets, they would presumably be more amenable to ramping up psychological services along the lines of the ones proposed by those disseminating the false information.
I have recently been engaged in an exchange with Professor Robin Murry, one of the two top editors at the venerable journal Psychological Medicine. At issue is a paper that once again includes the false statement about NHS costs. I have also written separately to the senior author, Professor Anthony David, since a previous paper he co-authored was corrected for the exact same mistake 18 months ago.
I have not heard back about a correction from either Professor Murray or Professor David. I sent Professor Murray a follow-up message this morning. I have shared it below.
Two workweeks have passed since I contacted you about the significant error in “Characteristics of patients with motor functional neurological disorder in a large UK mental health service: a case-control study.” (The authors are O’Connell, Nicholson, Wessely, & David.) You indicated you would check with the authors.
In reality, it should take the authors no more than a few minutes to review the mis-cited study, Bermingham et al, as well as the other papers in which the exact same mistake has already been corrected. That is even more the case since Professor David, the senior author, was also a co-author on one of the previously corrected papers. Moreover, I cc’d the corresponding author of Bermingham et al in my e-mail communications, in case of any lingering confusion about the meaning of the study findings.
I am therefore surprised that I haven’t heard back from you yet about the correction. Perhaps this seems like an insignificant matter or a minor irritation to you and the authors. However, as I have noted, investigators who study so-called “medically unexplained symptoms” have spent a decade disseminating a false account of the findings of this seminal research in their own field of purported expertise. Coincidentally or not, this mistake aligns with these experts’ professional interests in promoting more services for people who would meet their diagnostic criteria for MUS.
When the peer review process fails repeatedly to catch the same substantive error in publication after publication, journal editors should be alarmed. Perhaps the lesson is this: If a school of researchers appears to be engaging in what might be called groupthink by, for example, collectively misrepresenting a key data point in the medical literature, it would be wise to extend peer review invitations to a wider circle of readers and to ensure that references are cross-checked against claims being made.
I look forward to a prompt update on the forthcoming correction to O’Connell et al. As I have suggested, I hope the correction also includes an explanation for the 18-month delay since the previous paper co-authored by the senior author was corrected for the exact same mistake.
David Tuller, DrPH
Senior Fellow in Public Health and Journalism
Center for Global Public Health
School of Public Health
University of California, Berkeley