Trial By Error: Professor Chalder’s PRINCE Has Arrived

By David Tuller, DrPH

*October is crowdfunding month at Berkeley. I conduct this project as a senior fellow in public health and journalism and the university’s Center for Global Public Health. If you would like to support the project, here’s the place:

So here’s yet another paper with Professor Trudie Chalder of King’s College London as the senior author. That usually means there’s a lot to dig into! In this case, the journal is BMC Family Practice and the paper is titled “Integrated GP care for patients with persistent physical symptoms: feasibility cluster randomized trial.” Professor Chalder’s co-investigators include her KCL colleague, Professor Rona Ross-Morris, and other leading members of the biopsychosocial ideological brigades.

The study is part of this group’s larger project to promote their favorite psycho-behavioral interventions for the category of conditions they refer to as “persistent physical symptoms” (PPS) or “medically unexplained symptoms” (MUS). That would all be fine if their science could withstand serious scrutiny. But most of it is based on unproven hypotheses and relies on tarting up unimpressive results to look better than they are.

These investigators are also prone to embarrassing mistakes—as happened in the first paragraph of the published protocol for this feasibility study. In that first paragraph, the investigators mis-cited a seminal study that sought to calculate the cost to the National Health Service of treating patients with MUS. The effect of the mistake was to more than triple this apparent cost—which bolstered their claim that addressing the issue would lead to NHS savings. 

The published protocol now carries a prominent correction.


A 90-Minute Training Session

The new study, dubbed PRINCE Primary, is premised on the notion that GPs can learn in a 90-minute session how to redirect the cognitive processes of patients with PPS. For the study, 18 GP practices, each including multiple physicians, were randomized to receive the training or not—eight received it, and ten did not. Patients registered with the practices receiving the training were given resources, like self-help materials, that presented information about the development and treatment of PPS, per the notions of the investigators. In addition to the intervention, the patients in these practices received standard treatment, known as “treatment as usual.” The patients in the other practices were offered only treatment as usual.

This was a feasibility study—that is, the main goal was to determine if a full-scale trial was possible and desirable. In addition to assessing the willingness of GPs and patients to participate, the investigators road-tested multiple possible primary and secondary outcomes, including work and social adjustment, severity of physical symptoms, depression, and other measures. They did not calculate statistical significance for the measures.

The underlying hypothesis is essentially the same hypothesis underlying the approach of the ideological brigades to conditions with no identified organic etiology, including irritable bowel syndrome, ME/CFS, and others. As usual, the investigators trot out their triple-P schema of “predisposing,” “precipitating,” and “perpetuating” factors for these chronic illnesses, with perpetuating factors including such constructs as fear of engaging in activity and an excessive focus on symptoms. Per the hypothesis, changing those cognitions and/or behaviors will help patients find a path toward resumption of normal activities.

In this paper, the investigators once again base their argument on the prevalence of PPS. Here is their first paragraph:

Medically unexplained symptoms (MUS), otherwise referred to as persistent physical symptoms (PPS), cannot be explained by organic pathology after medical examinations. The prevalence of PPS in primary care is high, ranging between approximately 11 and 65%.

Let’s stop right there. The first sentence demonstrates the unproven assumptions driving this field of research. The fact is that patients relegated to these categories have conditions that have not been explained to date by organic pathology after medical examination by one or more clinicians. To say they “cannot be explained” by such pathology implies that no organic explanation exists and that those who conducted the medical examinations knew enough to exhaust all possibilities.

(Northwestern University  law professor Steven Lubet and I recently published an essay in BMJ’s Medical Humanities that addressed many of the problematic claims about “medically unexplained symptoms” advanced by the ideological brigades.)


Questions from their own data

Beyond that, their own data raise some questions about the second sentence. In this case, they found a very low prevalence rate of PPS. In seeking study participants among the GP practices that agreed to participate, the investigators trawled the medical records database with an expansive array of terms.

“The bespoke search algorithm incorporated a range of Read Codes [clinical terms] that were related to PPS including specific and widespread back pain, chronic/multiple widespread pain, dizziness, fatigue, fibromyalgia, headaches, non-cardiac chest pain and medically unexplained symptom,” they wrote.

Of 193,824 patients registered in the recruited GP practices, only 2978 were identified through this very thorough search as potentially eligible for the study. That’s 1.5%–a far cry from the 11 % to 65% they cite in the first paragraph. (In the paper, they seem to have rounded this number up to 2%.)

It’s not that earlier studies are “wrong” about the prevalence of PPS and this data point is “right.” Without understanding how investigators have defined and assessed prevalence—for example, whether it is prevalence among registered patients, or prevalence among patient visits, or some other metric–it’s hard to assess what the numbers mean.

A glance at the first of the three references, a US study, indicates that the 11% figure was calculated from responses from more than 200 primary care and other physicians about how many people with MUS they saw on their most recent day of clinical work. But that finding related to use of a broad definition of MUS that included cases in which doctors were continuing to pursue tests and consult with specialists to figure things out.

When doctors included only cases in which they were 95% certain patients’ complaints could not be explained through organic pathology, the MUS prevalence dropped to 4%. That number was even lower—3%–for those identified as having multiple or chronic unexplained symptoms that caused “significant psychosocial impairment.” Those numbers look less impressive than 11% if you’re trying to present the burden of MUS on primary care as a major problem. But they’re certainly closer to the 2 % that the investigators found through their own methodology.

The 65% figure is also questionable. It is from a German study that involved in-depth interviews with around 300 primary care patients. The investigators asked participants about dozens of symptoms–65, if I’m reading this right. At least one participant reported experiencing as many as 37. The investigators then asked the treating clinicians if the symptoms mentioned by patients in the interviews were explained, unexplained, or ambiguous.

The investigators reported that participants had an average of 9.2 symptoms each, and almost two-thirds had at least one that was considered medically unexplained. But here’s the thing: If investigators offer participants dozens of possible symptoms to choose from, it isn’t surprising that some or many people might choose quite a few. But the investigators provide no evidence that the interviewees were bothered by or sought treatment for the specific symptoms said to be unexplained. The high percentage seems more like an artifact of the methodology than a real-world metric of morbidity.


“Eminence-based medicine” at work

Finding their own low PPS prevalence rates should have triggered a glimmer of interest among investigators grounding the importance of their work in big claims about the high prevalence of PPS. Why are they ignoring this discrepancy? Why don’t they explain why they only found a prevalence of 1.5%, and discuss any implications for their framework?

In fact, they don’t seem to recognize the apparent contradiction. Recruitment estimates for the study were based on prevalence assumptions that do not seem to be borne out by their own data. Here’s what they wrote:

Although this [recruitment] method was time-efficient, recruitment of patients was relatively low and therefore further refinement of electronic patient searches is needed given that all patients who showed an interest gave consent if eligible. We anticipated that we would recruit approximately 20 patients from each GP practice given the high prevalence of PPS in primary care.”

They don’t raise an obvious possibility: Perhaps recruitment was “relatively low” because the prevalence of these patients is not as much of a burden on GPs as these experts have been arguing for years.

So what’s the paper’s main conclusion? Despite the recruitment issues, the trial is feasible, according to the criteria outlined by the investigators. The catch is that the intervention showed minimal evidence of efficacy across multiple outcomes at 12 and 24 months. So whether or not it is feasible seems beside the point. After touting the success of the feasibility measures, the investigators note in the last sentence of the abstract that “estimated effect sizes on potential clinical outcomes were small.”

Hm. How small? Well, here’s what they write:

None of the outcomes provided evidence for trial arm differences; all estimated effect sizes were small…The intervention needs to be optimised, possibly with therapist input, to bring about meaningful, enduring change in participant outcomes...Our low intensity intervention requires adaptation before we can test the efficacy in a fully powered randomised controlled trial.”

Given the poor outcome results, why would they conduct a follow-up trial? Why would anyone provide them with more funding when they have already shown that their intervention has little or no effect?

Under normal circumstances, that would be the end of it. But these disappointing results probably won’t stop this crew, if past behavior is any guide. No matter how many times members of the ideological brigades conduct studies that undermine their claims of efficacy, they declare success and move forward. Then they receive further funding for further study of whatever it is. “Eminence-based medicine” at its best.

{ 5 comments… add one }
  • CT 15 October 2020, 2:09 am

    I love the title…

    ….but this is so important –
    “Given the poor outcome results, why would they conduct a follow-up trial? Why would anyone provide them with more funding when they have already shown that their intervention has little or no effect?”

    Whether the funding comes from a charity/charitable foundation or directly out of the public purse, how can a further trial possibly be justified?

    Are these people science experts or sales experts? Do they ever read past the abstracts of the papers that they cite? In relation to the MUS/PPS primary care prevalence rates, I looked up that German (Steinbrecher et al) study and some of the papers that have cited it. Alan Carson (who incidentally was the Chair of the Programme Steering Committee ‘PSC’ for the PRINCE feasibility study) and Jon Stone wrote a paper* with other authors in which they reference the German study for its primary care prevalence rate (percentage of patients who have symptoms but no ‘pathophysiological’ explanation has been found for them). But THEY seem to think that the prevalence rate as per this German study lies between 16% and 34%, not the 65% that Chalder and her colleagues have touted. Surely Professors like Chalder, Moss-Morris, Mc-Crone, Hotopf, David and Landau should be across all this stuff, and especially those who are supposedly experts in this field?

    * Shipston-Sharman O, Hoeritzauer I, Edwards M, Reuber M, Carson A, Stone J. Screening for functional neurological disorders by questionnaire. J Psychosom Res. 2019 Apr;119:65-7

  • Lady Shambles 15 October 2020, 5:47 am

    The BPS authors, and Chalder in particular, seem to consistently fall into the traps of…

    “The Hasty Generalization Fallacy”:
    ‘This fallacy occurs when someone draws expansive conclusions based on inadequate or insufficient evidence. In other words, they jump to conclusions about the validity of a proposition with some — but not enough — evidence to back it up, and overlook potential counterarguments. ‘ …

    ..married with…

    “The Texas Sharpshooter Fallacy”:
    ‘This fallacy gets its colorful name from an anecdote about a Texan who fires his gun at a barn wall, and then proceeds to paint a target around the closest cluster of bullet holes. He then points at the bullet-riddled target as evidence of his expert marksmanship.
    Speakers who rely on the Texas sharpshooter fallacy tend to cherry-pick data clusters based on a predetermined conclusion. Instead of letting a full spectrum of evidence lead them to a logical conclusion, they find patterns and correlations in support of their goals, and ignore evidence that contradicts them or suggests the clusters weren’t actually statistically significant.’

    (Definitions thanks to Marketing Hub Spot).

  • CT 16 October 2020, 4:59 am

    Why is the PPS feasibility study by Chalder and co important for ME patients? Because the BPS crew have absorbed ME/CFS into ‘MUS’/’PPS’, or rather built their whole MUS/PPS approach on their management model for ME. ME is at the centre of it. In the UK, ME services, poor as they were, are closing and being replaced by MUS services (sending ‘MUS’ patients- including ME/CFS patients – to IAPT and other psychological therapy services), and this is a model that they are promoting abroad -(

    The political situation regarding how ME patients are treated isn’t static – these people ensure that it is constantly changing and developing, and patients need to keep up with those changes or face the consequences.

  • jimells 18 October 2020, 9:26 am

    I wonder if the rest of the medical industry understands just how fraudulent this work is, and how bad it makes all of them look. Absolutely *all* of the BPS work fails at step one when they ignore the non-BPS publications. For example, a PubMed search on “histamine and IBS” yields 62 hits. All of that work is completely ignored. They don’t even try to argue against it.

    Meanwhile the real investigators with actual labs studying actual biology remain silent. Do they not realize the psychobabblers are wrecking the reputation of the whole industry? When magic CBT becomes the official treatment for everything, there will be no need to fund any real investigators. They might want to think on that for a minute, for their own economic survival, if not for the patients.

  • CT 19 October 2020, 1:17 am

    Jimells said, “When magic CBT becomes the official treatment for everything, there will be no need to fund any real investigators. They might want to think on that for a minute, for their own economic survival, if not for the patients.”

    The same applies to biomedical doctors - as well as researchers. Integrated care sounds very attractive on paper, ‘parity of esteem for mental health’ and all that jazz, but with limited healthcare resources it becomes a takeover bid with good ‘biomedical’ care being ditched for CBT and good ‘biomedical’ doctors being substituted with psychiatrists, psychotherapists, or the cheaper option – self-help packages and computer apps. Why do ‘biomedical’ doctors think they’re immune from market forces? They should see ‘liaison psychiatry’ moving into their hospitals and realize that it means fewer ‘biomedical’ doctors in the mix and a greater workload on them. Instead they see it as a way to offload their more difficult patients. With fewer colleagues to share their burden, they suffer more workplace stress, regard more of their patients as sources of their stress, and offload more of them onto psych services – a truly vicious cycle. Seriously ill patients are gaslighted, and the physical health needs of mentally ill and vulnerable patients are overlooked to a greater extent than they were before. Is that parity of esteem for mental health?

    Nobody wants mental illness to be stigmatized or for mental health patients to be driven to the periphery and treated as if they are alien beings. But the push for more and more ‘liaison psychiatry’ is taking medicine down a very dark path in my opinion. And here’s the thing – if ‘dualism’ is so wrong then why can’t those ‘biomedical’ doctors brush up on their mental health training and then we could dispense with psychiatrists in our hospitals and healthcare system altogether. Now that really would be fully integrated and holistic care, wouldn’t it, instead of the psych takeover I outlined above?

Leave a Comment