Trial Report Cost Utility of Specialist Physiotherapy for Functional Motor Disorder (Physio4FMD), 2025, Hunter, Stone, Carson, Edwards et al

They divided the marginal cost relative to CAU by the marginal effect relative to CAU, and found that the cost per effect per person per year was slightly lower than the £20,000 limit that is commonly used by the government.

I briefly looked at the calculations (but I did not double check their sources) and I believe they add up except for one number I was unable to trace in the document.

The fatal flaw in their analysis is that they assumed that any positive effect would be meaningful regardless of the size. That is obviously not true, which is why we have MCID.

So they are essentially arguing that the government should spend a lot of money on something that isn’t going to make any meaningful difference. I would not be surprised if you would have seen a better result of if you just gave the participants the money that the intervention cost.

 

I know it’s a joke, but that calculation would probably favour insulin due to the cost of the deaths that would occur in the other group.

Sorry, I’m an economist so I can’t stop running the numbers

 

It seems like the only books that matter are the ones from the insurance company

 

Exactly. It had a meaningful effect on them - they get paid!

 

Joanne Hunt and others argue that the state is committing democide through those mechanisms. I’m inclined to agree with them.

Edit: so yes, the calculations would be different for ME/CFS if we include all costs and not just the ones that the government wants to consider.

 

Even then, they have to operate on a limited time frame. Over 5 years it makes sense. At 10 it starts making a lot less sense. At the 20-30 years range, developing solutions so that the problem effectively ceases to exist because it's routinely treated makes much more economic sense.

They still end up paying a lot. It's a small fraction of it, but the alternative of paying nothing obviously makes more sense, but that would require long-term thinking and policies. Governments are just as much to blame about this. Where even in an industry that needs to have long-term stability, they still mostly focus on the next quarter earnings, governments are often even more myopic in that they know they will be out of power if, and when, by the time someone clues on what they did.

 

No question about that, IMO. It's very clearly eugenicist policies. And just because everyone enforces them doesn't make them any less than what they are. We are, in realpolitik terms, undesirables, disposable.

Even though it would pay for itself many times over to work out solutions instead. Humanity is just this constant struggle between doing good things and kicking others in the face on an industrial scale.

 

Agreed

 

Right, but beyond that...they added up so many various items and it's hard to tell how accurate these accounts are and also whether they really reflect whether or not someone did or did not get the intervention. It all seems pretty arbitrary.

 

Absolutely! If you ask an accountant what this year’s result is they’ll ask you «what do you want it do be?»

That’s a bit of a joke of course, but it’s very easy to manipulate an analysis like this. And even when you do your honest best, there’s no guarantee that the analysis is exhaustive or representative.

 

One big problem, IMO, is how they used a tiny marginal difference, and extrapolated from that:

This amounts to about a 4% difference, when the total costs will often vary by more than that. It's far too small to extrapolate anything, especially on a measly 0.03 QALYs. It's the kind of thing that academics should genuinely laugh at, if amateurs did it, use as an example of how to be bad at doing their job.

 

This difference showed that TAU cost less.

What they «showed», is that they can add 1 QALY at a rate of less than £20,000 per person per year, so they argue that it’s «profitable» from a societal economics perspective.

One thing that I just thought of now is that I don’t remember that they showed any actual savings in terms of lower expenditures etc. as a result of this increased QALY. So they are arguing that the government should pay for something with no meaningful benefit to the patient, and that doesn’t save them any money, just because the cost is low enough relative to the meaningless effect for the patients.

 

The cost-effectiveness article tried to explain why the CGI-I scale is better than SF-36 as a primary outcome. They say this:

"The secondary outcome, the patient-reported Clinical Global Impression Improvement score, allows for a broader assessment of potential impacts that specialist physiotherapy can have across various aspects of patients’ lives, including the cost impact, thus capturing a wider range of outcomes. It is notable that consensus recommendations for outcome measures in FND (published after this trial was planned) have recommended the patient-reported Clinical Global Impression of Improvement as the primary outcome measure in trials of interventions in FND.”

But that second sentence about consensus recommendations appears to be untrue. (The first sentence is an opinion, and it doesn't make a lot of sense to me--curious what others think.) The reference for these consensus recommedations is this article from 2020--several of the authors are also co-authors of the FMD trial. This consensus document does not say what the FMD cost-effectiveness paper claims, at least that I can find.

It lists multiple domains of outcomes with recommended measures for each domain. The CGI-I is listed as recommended for the domain of "core FND symptoms." But the SF-36 is the recommended one for studying "life impact." So to present this as if the SF-36 should be viewed as the "wrong" outcome to select for the primary outcome is misrepresenting the recommendations.

Does anyone see anything in the referenced consensus recommenndations that supports the claim that they have flatly recommended the CGI-I "as the primary outcome measure in trials of interventiosn in FND"???

 

There is so much wrong with that paper - I don’t even know where to start…

The tables make it pretty clear that CGI is about core symptoms, and SF-36 is better suited for life impact:

As far as I can tell, there are no recommendations about the primary outcome of all the primary outcomes.

And it’s certainly wrong to say that CGI is broader than SF-36!

Here are some of the relevant bits about SF-36:

I actually found this bit quite interesting:

This could possibly indicate that perceived symptomatic improvements doesn’t actually make them function any better.

 

https://virology.ws/2025/05/12/tria...-fnd-experts-prefer-their-secondary-outcomes/

 

https://virology.ws/2025/05/13/tria...king-a-correction-in-fmd-physiotherapy-paper/

 

How often do they actually end up making corrections after letters like this?

 

Not often. But it's worth letting them know someone is paying attention. I find best to focus on just one point that's a definite error rather than going on about the various other questions about the paper.