Cognitive Behavioural Therapy for chronic fatigue and CFS: outcomes from a specialist clinic in the UK (2020) Adamson, Wessely, Chalder

[Esther12]

Due in the Journal of the Royal Society of Medicine

Abstract
Objectives
Cognitive Behavioural Therapy (CBT) is commonly used to treat Chronic Fatigue Syndrome (CFS) and has been shown to be effective for reducing fatigue and improving physical functioning. Most of the evidence on the effectiveness of CBT for CFS is from randomized control trials (RCTs) but there are only a few studies in naturalistic treatment settings. Our aim was to examine the effectiveness of CBT for CFS in a naturalistic setting and examine what factors, if any, predicted outcome.

Design
Using linear mixed effects analysis we analysed patients self-reported symptomology over the course of treatment and at three-month follow up. Furthermore we explored what baseline factors were associated with improvement at follow-up.

Setting
Data was available for 995 patients receiving CBT for CFS at an outpatient clinic in the UK.

Participants
Participants were referred consecutively to a specialist unit for Chronic Fatigue or CFS.

Main outcome measures

Patients were assessed throughout their treatment using self-report measures including the Chalder Fatigue Scale, Short Form Health Survey (SF-36), Hospital Anxiety and Depression
Scale (HADS), Global Improvement and Satisfaction.

Results
Patients fatigue, physical functioning and social adjustment scores significantly improved over the duration of treatment with medium to large effect sizes (|d|=0.45 – 0.91). Furthermore 85% of patients self-reported that they felt an improvement in their fatigue at follow-up and 90% were satisfied with their treatment. None of the regression models convincingly predicted improvement in outcomes with the best model being (R2 =0.137).

Conclusions
Patients fatigue, physical functioning and social adjustment all significantly improved following CBT for CFS in a naturalistic outpatient setting. These findings support the growing evidence from previous RCTs and suggests that CBT could be an effective treatment in routine treatment settings.

https://kclpure.kcl.ac.uk/portal/files/132955332/pdf_proof_re_submit.pdf

https://kclpure.kcl.ac.uk/portal/en...fs(67b43c92-f30b-4e21-a7b6-1b94e2ead525).html

 

[Esther12]

The version currently on-line seems to be missing a lot of information that should be in the final paper.

Pretty sure this study was another one using silly criteria for 'recovery':

Flo and Chalder (2014) conducted a study within routine practice and found that after CBT treatment, just under 40% of patients no longer met Oxford or CDC criteria for CFS, and just under 20% were recovered, similar to rates of recovery reported in the Netherlands 13.

I'd just seen someone here posting about whether age affected prognosis:

Research suggests that older age has been consistently shown to be a predictor of poor outcomes 11,12,15.

According to self-reported accounts of their symptoms, 754 (76%) participants met Oxford criteria for CFS and 518 (52%) met CDC criteria for CFS.

edit: As Trish notes: "All met NICE criteria for CFS."

Drop-out was defined as those patients who did not complete any questionnaires at the end of treatment, or any follow-up and was 31% in this naturalistic setting.

So presumably even more than 31% was missing from follow-up?

There were no statistically significant differences in demographic characteristics between those who dropped-out from treatment and those who completed. We did find a statistically significant difference in self-reported physical functioning at baseline between the groups (t(763)=-3.74, p<.001) with those that dropped out reporting lower physical functioning scores with a mean difference (MD) of -7.38. Furthermore, those who dropped out had more severe WSAS scores at baseline (t(987)=2.48, p<.05) with a MD of 1.55. However, no difference in level of fatigue was observed (t(975)=1.61, p>.05). Those who dropped out also had higher scores on self-reported depression symptoms at baseline (t(981)=2.35, p<.05) but no difference in anxiety scores (t(978)=1.4, p>.05).

. Interestingly, initial gains in fatigue between the start of treatment and session four was where we saw the largest improvement in fatigue and WSAS scores. ... In this large chort, changes occurred within the first 4 sessions.

I think other data has also indicated that the improvement in self-reported outcomes come very early on.

Data accessibility - We will assess requests for data on a case by case basis

 

[Amw66]

Could charities/ APPG group request data ?
Though not researchers it would surely set the wrong tone to refuse such groups?

Or am I being incredibly naive on low sleep?
ETA spelling

 

[Trish]

To be published in the Journal of The Royal Society of Medicine. Authors Chalder and Wessely - President of RSM - Wessely. Chalder has conflicts of interest listed. Showing their true colours. I thought he said he'd given up CFS research.

Diagnosis CF or CFS. So not ME/CFS research then, and irrelevant to the NICE guidelines. Self reported measures and only 3 months follow up. Useless.

All met NICE criteria for CFS. According to self-reported accounts of their symptoms, 754 (76%) participants met Oxford criteria for CFS and 518 (52%) met CDC criteria for CFS.

How can you meet NICE criteria and not Oxford criteria. I thought NICE criteria included 6 months fatigue, which is all that is required for Oxford criteria.

So only half actually had CFS by the Fukuda criteria which doesn't require PEM.

Drop-out was defined as those patients who did not complete any questionnaires at the end of treatment, or any follow-up and was 31% in this naturalistic setting.
We did find a statistically significant difference in self-reported physical functioning at baseline between the groups (t(763)=-3.74, p<.001) with those that dropped out reporting lower physical functioning scores with a mean difference (MD) of -7.38. Furthermore, those who dropped out had more severe WSAS scores at baseline (t(987)=2.48, p<.05) with a MD of 1.55. However, no difference in level of fatigue was observed (t(975)=1.61, p>.05). Those who dropped out also had higher scores on self-reported depression symptoms at baseline (t(981)=2.35, p<.05) but no difference in anxiety scores (t(978)=1.4, p>.05)

So nearly a third dropped out from giving follow up data, and they were the sickest ones. It gets worse and worse.

SF-36 significantly improving over treatment; start of treatment M = 47.81 (SD=25.74); end of treatment M = 59.56 (SD=27.3).

So even those who were milder on average to start and gave follow up data were still on average under the entry criteria for PACE on SF36 at the end.

After up to 20 sessions over up to a year, the outcomes are truly pathetic.

 

[dave30th]

I don't understand this:

"According to self-reported accounts of their symptoms, 754 (76%) participants met Oxford criteria for CFS and 518 (52%) met CDC criteria for CFS."

Who were the 24% that didn't meet the Oxford criteria? Did they meet CDC criteria but not Oxford criteria? That would be interesting, since in PACE the CDC criteria group was a subset of the overall Oxford group, based on the apparent presumption that anyone who met the CDC criteria would automatically have met the Oxford criteria as well. They made a lot of hay out of saying that PACE applied no matter how the illness was defined.

Now something else seems to be happening.

What percentage had PEM? I didn't see that. Maybe none of them. So it wouldn't be surprising if CBT helped people with generalized or idiopathic fatigue.

 

[Esther12]

I don't understand this:

"According to self-reported accounts of their symptoms, 754 (76%) participants met Oxford criteria for CFS and 518 (52%) met CDC criteria for CFS."

Who were the 24% that didn't meet the Oxford criteria? Did they meet CDC criteria but not Oxford criteria? That would be interesting, since in PACE the CDC criteria group was a subset of the overall Oxford group, based on the apparent presumption that anyone who met the CDC criteria would automatically have met the Oxford criteria as well. They made a lot of hay out of saying that PACE applied no matter how the illness was defined.

Now something else seems to be happening.

@Trish caught the sentance I missed "All met NICE criteria for CFS."

I think it came before the bit I quoted, so I just forgot about it.

 

[wigglethemouse]

Interesting. SW has given indication that he got out of CFS research a long time ago (re: his personal website - not going to link, trackbacks and all that). And here he is. Again.

My takeaway is that they have realised the GET boat is sinking and they are trying to shore up the need for continuing to use CBT with a large and recent study.

EDIT: If in fact 100% of participants met the NICE criteria, then I'm assuming that this study is for the benefit of NICE guidelines. None of the major ME/CFS papers say they diagnose with NICE guidelines, so this paper is much more "useful" for that purpose.

 

[Trish]

What percentage had PEM? I didn't see that. Maybe none of them. So it wouldn't be surprising if CBT helped people with generalized or idiopathic fatigue.

It seems likely those with ME were in the 31% who dropped out. There's no data on PEM. What amazes me is how pathetically low the improvement was even for this ill defined group with CF. Well within therapist/placebo effect, I think.

 

[dave30th]

@Trish caught the sentance I missed "All met NICE criteria for CFS."

After I posted that, I saw what Trish wrote. But I still don't get it. NICE requires four months, not six months. So that means a quarter were people who had four to six months of fatigue, since by six months they would have qualified for Oxford? Does that make sense at all?

 

[Trish]

I think this is service evaluation, not research.
I think it's ok as a public thread. We have nothing to hide. If they see us picking it apart, perhaps they will have second thoughts about publishing with such ridiculous claims of success.
I wonder who peer reviewed it.

 

[Esther12]

I wonder who peer reviewed it.

I saw the journal says it has open peer review, so when it's published we can have a look.

 

[Peter T]

Interesting that they did not include any changes in hours worked or benefit claimed. I seem to recall that somewhere an evaluation of the UK/ME specialist services showed service users ended up working less and claiming more benefit on follow up.

 

[Wonko]

More of the same, from the same.

When are they going to stop beating their partners, show some truly innovative thought, really dig deep, to try and further understanding of this 'complex' condition, and use sock colour as a primary outcome measure?

 

[Trish]

CFQ
Start: 977 patients, mean CFQ 24.19
Session 4: 392 patients, mean CFQ 19.45
Follow-up: 503 patients, mean CFQ 18.60

SF-36
Start: 768 patients, mean 47.6
Follow-up 404 patients, mean 58.51

So is there something dodgy going on with their comparison of means? The initial figures are calculated on close to double the number of patients as the final figures, and they say themselves the ones lost to follow up had worse initial scores. So most or all of the 'improvement' in the averages could be simply down to the sickest patients no longer being included.

To take a simple example, if you have 10 patients, half of whom score 5 points on a scale and half score 10 points, the mean at the start is 7.5. If all the ones scoring 5 points drop out, and the rest don't change their scores, the mean at the end is 10. Nobody has increased their score, but the means have improved from 7.5 to 10. Magic.

Surely they couldn't have done such a trick could they? What am I missing Ms Chalder or Mr Wessely if you're reading this? I think you need to show your working.

 

[Trish]

This is ridiculous, the more I look at it the more it's full of holes. They say 31% loss to follow up. That may be true overall, in that maybe 31% gave no follow up information, but:
CFQ 48.5% not followed up.
SF36 47.4% not followed up.

 

[NelliePledge]

Circling the rickety wagons of eminence

 

[Sean]

So it wouldn't be surprising if CBT helped people with generalized or idiopathic fatigue.

Except for the exclusive reliance on subjective self-report without adequate blinding or objective measures thing.

All this sort of paper demonstrates, yet again, is that they can induce changes in self-report scoring behaviour in some patients.

They have yet to demonstrate that this phenomenon delivers any actual meaningful practical benefit to those patients.

 

[chrisb]

If a patient was missing 25% or less of the data from any questionnaire, a prorated score was calculated (this used the mean of the remaining items from the same individual to calculate a prorated score).

Why not just make it all up as you go along? What science is there to validate this as a means of creating data?

 

[Hoopoe]

I have often heard about various tricks the clinics of this kind use to make the treatment look successful. Presumably they do this to survive, because of the expectation set by the PACE trial and other studies that CBT/GET are effective. A clinic without results would simply be closed.

For example, one of the tricks is to remove patients from their programme at the first sign of problems. That hides failure or harm.

It's circular thinking all the way. CBT/GET were declared to be the solution, then the trials were done to prove it (with methodology as poor as was required to obtain the desired results). The clinics then do their best to pretend these results are really obtainable. Now the data from clinics is said to add to the evidence that CBT/GET are effective.

 

[Trish]

For example, one of the tricks is to remove patients from their programme at the first sign of problems. That hides failure or harm.

It reminds me of schools who only let the pupils sit an exam if they are sure they will pass - which makes the pupil achievement scores look great.

The thing that surprises me is how poor the outcomes are at 3 month follow up. It looks like most of the so called improvement happens in the first 4 sessions - so still well within the first flush of gratitude that they are getting help and persuasion to see the questionnaires differently and declare themselves improved. And some may improve in that time from idiopathic CF if they are getting advice on things like sleep hygiene. After that, the so called improvement is very small.

To go on again about my point earlier about the effect of drop outs on the data:

We did find a statistically significant difference in self-reported physical functioning at baseline between the groups (t(763)=-3.74, p<.001) with those that dropped out reporting lower physical functioning scores with a mean difference (MD) of -7.38.

So if I interpret that correctly, the nearly half who dropped out scored on average 7.38 points lower than those who stayed on SF-36 physical functioning.

with SF-36 significantly improving over treatment; start of treatment M = 47.81 (SD=25.74); end of treatment M = 59.56 (SD=27.3). The effect size was medium (|d|=0.45) with a mean difference of 9.91 [95% CI (7.94,11.88)].

And at the end, there was an average improvement of 9.91.