Efficacy of web-based cognitive–behavioural therapy for chronic fatigue syndrome: randomised controlled trial (2018) Knoop

Here is an example of what they recommend in their manual (this manual isn't specifically mentioned in this study). I think it is completely reckless but imagine that very few patients actually try to stick to it for very long:

 

The harms data is on page 2 of this file:
https://static.cambridge.org/resour...0007125017000228:S0007125017000228sup002.docx

 

Yes, the text confirms this:

 

I find it interesting to see the individual data in studies, like they have given here. Though I would have been more interested in other outcome measures than fatigue.

 

These people were a lot healthier at baseline than for example the people in the PACE trial.

Average SF-36 physical functioning (0-100) score:
PACE trial: 38.05
This trial: 62.5

Average Chalder fatigue questionnaire (0-33) score:
PACE trial: 28.2
This trial: 24.1

 

I wondered about that. It felt like there was something missing there.

Missing data for appx 25% of the protocol-driven group and the control group, isn't great, but it was down to 15% missing for the feedback on demand group.

Maybe there is some sub-group for whom a minimal form of CBT, with little involvement of a therapist, could be useful? Maybe that would make it easier for people to make use of any bits they found useful, and dismiss the rest?

Also - maybe these studies are getting increasingly unrepresentative samples that include those more likely to benefit from some sort of CBT-style intervention? Or maybe a wait-list control is demoralising and is worse than nothing?

So hard to say. These researcher's history of hyping and misrepresenting CBT as a treatment for CFS gives rise to some justified scepticism of any purported positive results. [edit: but if were were to take them at face value and ignore possible problems with a wait list control, the actimeter results for the on-demand CBT group look better than what we tend to see].

Yes - would be interesting to see the acitgraphy data.

 

Thanks Dolphin. I was just looking for the missing tables, and didn't think to check the other tables for that data. There was some other data I thought seemed missing too, so I'll go back to check if that got put in the wrong place too.

 

The Dutch fatigue clinic I went to rated me as "passive" based on my actometer data, but I was also crashing a lot like a boom-and-bust "relatively active" patient. So the psychologist and relaxation therapist (who used stealth-GET) were telling me I needed to do more, while the occupational therapist and physiotherapist were more focused on stabilization. They really don't have any conception of disease severity - it's just one type of crazy (passive) or a contradictory type of crazy (boom-and-bust) which somehow has the same cause and needs pretty much the same treatment

I think these are both factors. My observation was that most patients at Dutch fatigue clinics do not have ME/CFS. Out of the 1-2 dozen I saw regularly in the waiting rooms, only 2-3 of us were moderate or worse. The rest were either very mild or had a different form of fatigue which was not physically disabling them at all.

And giving worse scores on questionnaires may be a desperate plea for immediate help from waiting list patients who think there is an effective treatment available.

 

I also wonder how they ascertain from a distance how closely people are adhering to what they are actually being asked to do. How do they know, for instance, that the CBT sessions aren't in fact resulting in people learning how to pace themselves better.

 

This study is a bit of a head scratcher for me. One of the main (highly legitimate) criticisms of these trials is the lack of objective measures. Yet here the authors appear to have a statistically significant result when using activity monitors but tend to downplay that in the text. I find that puzzling - if I were the authors that would be plastered all over the page.

(I suspect the positive result is due to selection criteria - as others have already noted above - but need to spend more time with the paper. Though as I’m being dragged into the office today against my will there’s a good chance that’ll have to wait until after my resulting crash )

 

The Dutch BPS group engage in some interesting shenanigans at times, like declaring completely untreated patients cured by CBT. They might be downplaying actometer results to avoid further scrutiny.

 

(Not very exciting) From the protocol paper:

 

From the protocol paper:

 

From the protocol paper:

Firstly, I doubt a lot of these are truly recovered.
Also there is no sign that participants hear from patients who believe some or all of the programme made them worse.

 

From the protocol paper:

 

That's interesting.

According to the protocol, somebody with a low level of activity is not supposed to go to a stabilisation period even though they could be fluctuating, but just at a lower level of activity.

From the protocol paper:

 

From the protocol paper:

 

It would be interesting to have data on these goal sheets and for an independent person to assess whether patients had reached such goals. I don't believe the Dutch CBT school ever discuss these goal sheets in terms of outcomes in published papers.

 

They would appear to be totally barking.