Effectiveness of a symptom-clinic intervention ... multiple and persistent physical symptoms, 2024, Burton, Deary et al

"At baseline, PHQ-15 scores were 14·9 (SD 3·0) in the group receiving usual care and 15·0 (2·9) in the group receiving the intervention (figure 3)."

These are I think moderate scores on this scale. Max score is 30 indicating higher symptom presence.

The PHQ-15 is made up of 15 items, and each item is rated on a 3-point scale, with 0 indicating "not bothered at all", 1 indicating "bothered a little", and 2 indicating "bothered a lot". The total score can range from 0 to 30, with higher scores indicating greater symptom severity.

"At the primary-outcome point of 52 weeks, PHQ-15 scores were 14·1 (SD 3·7) in the group receiving usual care and 12·2 (4·5) in the group receiving the intervention."

This differences may well be statistically significant but so what These are tiny changes

Gobsmacked that this work of colossal unseriousness has been published in The Lancet

Whow.

 

I was just skimming through this issue of the Lancet, and in addition to the Reme commentary there's also an editorial ("Taking persistent physical symptoms seriously") referencing the trial:

Link | PDF

 

This is pure nonsense and it has zero relevance to the issue of suffering from chronic symptoms.

There's so many layers of nonsense to this, including the fact that pain doesn't have to come from injury. There is no injury to regular headaches. Or migraine headaches. Or many types of pain.

All they're doing is taking something that could be possible, in a very limited fashion and uncommon circumstances, and deciding that it has to be the explanation. Zero scientific validity.

And it's not even true that "years, the term medically unexplained symptoms has been used to describe patients with persistent physical symptoms but no detectable disease", in several ways. There are several dozen terms that mean exactly the same thing, which is usually a big tell that it's all made-up. But it's not just used in cases of "no detectable disease", and of course medicine isn't even close to being able to detect all diseases.

Reasoning like this would flunk an intro philosophy class. It reminds me or creationist arguments from homeschooled people who manage to get to a university and are faced with having to argue nonsense beliefs but have never had to actually reason through the arguments.

Actually, it reminds of The waterboy: "Mama says that alligators are ornery... 'cause they got all them teeth but no toothbrush." This is the level of intelligence I'm seeing here, it's just a bit more flowery but it has the same intellectual level. And boy do I wish I were kidding.

 

Looks like a highly problematic study.

It has the design A +B versus A. The control group got no intervention at all, only the usual care that the intervention group also received (in addition to the intervention itself).

Patients and GPs were not blinded and they only used a subjective outcomes (questionnaires).

They excluded patients with more severe symptoms as 'pilot work' indicated it would not work for this group:

"Eligible participants were adults aged 18–69 years with multiple and persistent physical symptoms as evidenced by a Patient Health Questionnaire-15 (PHQ-15)20 score between 10 and 20. This score represents moderate or moderately severe symptom burden. We excluded people with PHQ-15 score more than this range, as pilot work indicated that they were likely to be too severely affected by their symptoms to benefit from this moderateintensity intervention"​

The primary outcome (PHQ-15) was measured at three time points: at 13 weeks, 26 weeks, and 52 weeks after randomisation. For the first two the difference was not statistically significant. For the third measurement it was but the adjusted difference (1.82 points) was lower than the minimally clinically important difference (MCID) for this questionnaire which they themselves report to be 2.3 points. They write:

"Although the mean treatment effect that we observed was less than the published clinically important difference of 2·3 points,23 the 95% CIs around the point estimate included this value. Furthermore, the skewed distribution of treatment effect (appendix p 13) suggested that some participants benefitted particularly highly, such that the number needed to treat for twice the clinically important difference was five."​

The intervention itself just seem to consist of 3-4 short GP consultations where the GP gave some reassurance and a possible (made-up) explanation for the patient symptoms.

The only strength about this study is that it was quite large and complex with 108 GPs participating but of course that does make the design any less flawed. Strange that the Lancet published this. PACE was flawed as well but at least it had some interesting data.

 

Another flaw: just noted that the statistical analysis section says: "...no adjustments were made for multiplicity."

 

This is the primary outcome, the PHQ-15

 

Apparently, the symptoms are no longer medically unexplained. That seems to be the main message that they want to spread.

In the introduction the authors write:

"People with persistent physical symptoms commonly have poor experiences of health care8,9 and are often told that their symptoms cannot be explained.10 However, developments during the past 10 years in interoception,11 symptom perception,12 and biopsychosocial integration13 mean that persistent symptoms can be increasingly understood as entities in their own right5 and usefully explained."
​

Similarly the editorial by Reme says:

"scientific developments since 2015 challenge the accuracy of the term medically unexplained symptoms, suggesting that these symptoms are both explainable and treatable"
​

I wish I had their confidence...

 

Gobsmacking overconfidence.

They do not demonstrate in any objective manner that they made a real world difference to anyone.

Even the reductions in PHQ-15 are tiny. Unlikely to be relevant to patients.

Excluding the more severely affected is a red flag. Why would such an approach 'work' only on moderately affected patients but not those more severely affected. And then claim the symptoms are treatable. Unfathomable logic going on.

 

Is there an established minimal clinically significant difference for the PHQ-15?

EDIT: I see ME/CFS Skeptic raised this issue above that that the findings are below the MCID. This is like the REGAIN study--they claimed success and even called the intervention "clinically effective" even though the findings were below the MCID.

 

What specifically happened in 2015? Interestingly, the same year of the IOM report on ME/CFS.

 

This is I think mentioned in the paper

"The PHQ-15 has been reported as having an MCID of 2·3 points" Ref 23

23.Toussaint A Kroenke K Baye F Lourens S
Comparing the Patient Health Questionnaire-15 and the Somatic Symptom Scale-8 as measures of somatic symptom burden.
J Psychosom Res. 2017; 101: 44-50

This was published after the trial was designed.

'Intervention' achieved reduction of 2.8 points

 

I've scanned the paper and they mention fatigue, ibs and fibromyalgia. Cannot see ME/CFS specifically though it might be in there.

 

"Although we excluded people with organic disease in the design of this trial....."

Now there's a claim.

Boom. (Apologies it's a Friday...)

 

"GPs with an extended role spent substantial time actively listening to the participant describe their illness and proposing and negotiating explanations and actions."

So they listened and then what? From the paper it's clear the GPs and researchers don't think anything organic going on, which is bold.

So, did they literally sit there undermining their patients reality?

Negotiating and proposing what?

Did they openly state they 'believed' no organic and that they 'believed' the patients symptoms were psychosomatic? Or was it all some fudge..... which means informed consent was not gained.

More honest and ethical response would be to admit they simply don't know what is causing the symptoms.

No need to propose and negotiate, whatever that might mean and entail.

 

It's also a strange thing to study: whether making up an explanation for symptoms makes the patient feel less ill.

I assume that the main message from this paper is to tell doctors that it's ok to use some made-up explanation for symptoms they can't explain. It's not only allowed, it's now the evidence-based approach!

 

Here's a PDF for the Toussaint et al paper: link - the SEM was 2.3 for the PHQ-15m. The adjusted between-group difference was -1.82 (95% CI –2.67 to –0.97), which although statistically significant (p<0.0001) does not meet that MID for clinical significance.

A few additional thoughts from a very brief skim through: 65% of patients did not meet the cutoff for "sufficient" health literacy (including 47% "problematic" and 14% "inadequate"). I haven't looked at this HLS-EU6 questionnaire but this seems high? Also looking at Supplementary Table 8 the secondary outcome results look unimpressive (3/10 achieving statistical significance). It's a relatively homogeneous study population (82% women, 93% white). The inclusion criteria specified a PHQ-15 score between 10 and 20 which seems to exclude both the mild and the most severe symptom levels. The control group is a "usual care" one; it would've been better to use an attentional-placebo type control. Also: "proposing and negotiating explanations and actions" - negotiating explanations?

 

I think you are correct in that it risks normalising making explanations up and rubber stamping this as OK.

A much less energetic approach might be just to listen to patients and let them know 'we simply don't know what is causing your symptoms' tiddily pom.

No expense needed.

 

they are making a rather large assumption that those patients being fobbed of with bs explanations are not going to do any research of their own .resulting in a huge loss of confidence in their gps and the entire medical industry . personally i prefer that a gp has the common sense and decency to admit that there needs to be far more rigorous research into human biology before they can claim m u s means anything other than i don't know.