Trial By Error: The Cost of MUS

By David Tuller, DrPH

In January, I wrote about the problematic online training course developed by the Royal College of General Practitioners and touted by Steve Brine MP as addressing the misconceptions about ME (or CFS/ME, as the online course calls the illness). The lead author of the course was Carolyn Chew-Graham, a professor of general practice research at Keele University in Staffordshire. Rather than addressing any misconceptions, the course repeats various bogus claims that the GET/CBT ideological brigades have been pushing all along.

Like the PACE authors and those in their orbit, Professor Chew-Graham is a leading proponent of psychological and rehabilitative treatments for patients identified as having medically unexplained symptoms, or MUS. (Journalist Maya Dusenbery and I discussed some of the problems around the MUS construct in this Q–and-A last year for In fact, Professor Chew-Graham has represented the Royal College of General Practitioners as co-chair of a joint effort with the Royal College of Psychiatrists to develop commissioning guidelines for MUS services.

One vehicle for advancing these treatments is a program called Improving Access to Psychological Therapies, or IAPT. NHS England has been rolling out IAPT for the past decade. (Different national branches of the NHS work differently. When NHS is mentioned below, it refers to NHS England.) In recent years, the program has grown to include more people suffering from long-term conditions and MUS. For IAPT’s purposes, the latter category includes what the program calls CFS/ME. The most common treatments offered through IAPT include cognitive and behavioral therapies.

Those promoting IAPT and a biopsychosocial approach to MUS argue that the strategy is evidence-based. But many of the same people argue that the rehabilitative treatment paradigm for ME/CFS is evidence-based, so their judgement cannot be considered reliable. IAPT advocates further suggest that the program offers financial advantages, since people with MUS have been identified as heavy users of medical services. If these patients are indeed experiencing untreated psychological problems, then linking them to mental health interventions makes sense, potentially alleviating their suffering while reducing their overall health care usage.

The promise of this idea rests on two assumptions: First, that all or most of those with MUS suffer primarily or to a serious degree from untreated psychological problems rather than from undiagnosed organic illnesses; and second, that the proposed interventions are effective for these untreated psychological problems. Both assumptions are questionable.

The current expansion of IAPT to those with MUS has been premised in part on the potential for savings in NHS expenditures. Given the significance of this factor in the debate, it is surprising that Professor Chew-Graham has demonstrated some difficulty in accurately citing the relevant literature. On at least a few occasions, a blog post, a published paper and a slide presentation–she has conveyed apparent misinformation about the estimated costs of MUS to NHS.

Here is the core of the issue: A 2010 study (Bermingham et al; more on this study below) documented the costs of health care services for working-age people with MUS. The authors concluded that these services accounted for around 10% of NHS expenditures for all services for those of working age. This figure appears to have been translated, by Professor Chew-Graham among others, to mean something a bit different–that spending on these working-age patients with MUS accounts for 10% of total NHS expenditures, not just expenditures for those of working age.

Here is what Professor Chew-Graham wrote about the cost of MUS in a 2017 blog post on her university’s website: MUS actually accounts for a considerably high proportion of NHS activity, with approximately 10% of total NHS expenditure being spent on services for the working age population in England with medically unexplained symptoms.

And a 2017 paper she co-authored in the British Journal of General Practice stated that annual NHS costs for MUS in adults of working age in England was estimated to be £2.89bn in 2008/9 (11% of total NHS spend). Professor Chew-Graham included the same statement on a slide called The scale of the problem, part of a conference presentation on the development of the MUS commissioning guidance. The co-author of the presentation was Dr Simon Heyland, a psychiatrist and Professor Chew-Graham’s co-chair on the collaboration between the RCGP and RCPsych to create this commissioning guidance. (The stated change in the proportion of MUS-related costs from 10 % to 11% of NHS expenditures was not explained.)

So according to Professor Chew-Graham’s public statements, the NHS spends about a tenth of its total budget on working-age patients with MUS. A commenter on the 2017 blog post explained the problem, that the 10% related to how much was spent on those of working-age, not how much was spent on everyone. Professor Chew-Graham has never responded. A few weeks ago, I sent Professor Chew-Graham a letter that mentioned the misleading statements in both the blog post and the journal article. I have not heard back. Neither the post nor the journal article has been clarified or corrected.

Professor Chew-Graham is a key player in the development of national health policy for people with MUS. Any sloppiness or ambiguity on her part is especially troubling, given that this kind of misinformation about such a salient issue could influence the opinions of others’ and impact their decision-making. But Professor Chew-Graham is not the only one to propagate this error. It has popped up repeatedly in academic discussions of MUS in the nine years since Bermingham et al was published.

The 4th edition of a textbook called Psychiatry in Primary Care, published by Cambridge University Press the year after Bermingham et al, mentioned the statistic. The textbook noted that some have estimated that up to 10% of the NHS budget is taken up with medically unexplained symptoms, although it also suggested that the estimate was probably too high. A 2013 article about MUS in The Psychologist, published by the British Psychological Society, cited Bermingham et al in noting that the cost of these conditions accounted for approximately 10 per cent of total NHS expenditure at that time.

The Winter 2016/2017 edition of the RCPsych’s Medical Psychotherapy Faculty newsletter included a section on MUS that stated this: Current political interest in MUS is more about the potential for cost-savings than anything else, but the focus on this group of patients is long overdue. They are estimated to account for around 10% of the annual NHS budget, yet rarely get access to specialist care. Even as it disseminated this untrue information, the newsletter touted the RCPsych’s role in developing a guidance for commissioning services for people with MUS.

And here is a statement about MUS-related expenditures from a paper published last November in the journal Frontiers in Psychology: At approximately £11.64 billion it cost around 10% of the whole English National Health Service (NHS) budget in 2015/16. The paper is called “Different Strokes for Different Folks: The BodyMind Approach as a Learning Tool for Patients With Medically Unexplained Symptoms to Self-Manage.”

Surprisingly, the paper does not cite a reference for the £11.64 billion figure, but it follows a mention of Bermingham et al. Absent other clues, what seems likely to have happened is that the authors derived the statistic themselves by calculating 10 % of the total NHS budget–based on their misunderstanding of the relevant data from Bermingham et al. This interpretation of events is consistent with the fact that the NHS budget for that fiscal year was £116.4 billion.

Helen Payne, a professor of psychotherapy at the University of Hertfordshire, co-authored the Frontiers in Psychology paper. And at an event called the Medically Unexplained Symptoms/Somatic Symptom Disorder: National Summit 2017, her lecture slides included one that noted this about MUS: Extremely costly approximately £11.64 Billion, 10% NHS budget 2015/16. Presumably none of Professor Payne’s colleagues at the MUS conference understood, or bothered to tell her, that the cost numbers were off, since she included the point in the Frontiers paper the following year.

Such misstatements by well-regarded researchers have likely helped foster an impression that providing services to people with MUS consumes a much greater percentage of NHS funds than it does. Over-estimating the budget impact of MUS in this manner would likely raise hopes or expectations that expanding a program like IAPT could result in substantial savings. That would be convenient for those with professional or financial interests in expanding IAPT.

Let’s take a look at Bermingham et al, the study that has triggered so much apparent confusion among renowned experts in this field. The cost of somatization among the working-age population in England for the year 2008-2009 was published in a journal called Mental Health in Family Medicine. According to Google Scholar, it has been cited 120 times. (Google Scholar does not provide data on how often Bermingham et al has been mis-cited.)

The study abstract includes these two sentences: The incremental health care cost incurred by somatising patients is estimated to be £3 billion. This represents approximately 10% of total NHS expenditure on these services for the working-age population in 2008–2009. (In this post I’m not addressing whether the study’s methodology or calculations were sound.)

The phrasing of the second sentence could be considered a bit confusing. Taken alone, it would be possible to interpret it as did Professor Chew-Graham and some of her colleagues apparently have. But even a quick review would reveal any such interpretation to be incorrect.

First, the meaning cannot be misconstrued in that way when the £3 billion figure is considered, since 10% of total NHS expenditure would have amounted to much more than that amount. Moreover, elsewhere the study states the conclusion about the costs for working-age people with MUS more precisely, noting that 10% of total healthcare expenditure for this age group is accounted for by somatising patients.

So Bermingham et al reported that 10% of the funds NHS spent on health care services for the working-age population went toward patients with MUS. It did not report that 10% of all NHS expenditures went towards working-age patients with MUS. The two statements are very different. The matter should not be that complicated for people used to reading medical and public health studies. Yet these researchers have repeatedly managed to get this wrong.

One conclusion to draw here is that members of the MUS/IAPT ideological brigades have misrepresented, whether intentionally or not, a major plank in support of their biopsychosocial agenda. It is unsettling that such a substantive error could percolate for years through a field of scientific inquiry, and then get cited to help justify ill-advised health policies. This untrue statistic doesn’t by itself invalidate other findings, of course. But if these folks can’t grasp such basic facts or don’t notice (or care) that their colleagues are making unwarranted claims, it would be wise to regard their other pronouncements about MUS, IAPT and anything else with a huge dose of caution.

(I want to thank the keen-eyed sleuth who uncovered this sleight-of-budget situation.)

Comments are closed.

Scroll to Top