• Welcome to Phoenix Rising!

    Created in 2008, Phoenix Rising is the largest and oldest forum dedicated to furthering the understanding of and finding treatments for complex chronic illnesses such as chronic fatigue syndrome (ME/CFS), fibromyalgia (FM), long COVID, postural orthostatic tachycardia syndrome (POTS), mast cell activation syndrome (MCAS), and allied diseases.

    To become a member, simply click the Register button at the top right.

"What makes a difference in chronic fatigue syndrome?": Commentary on PACE Trial mediators paper

Dolphin

Senior Member
Messages
17,567
(Note: There are two other threads on the PACE Trial mediators paper itself here: http://forums.phoenixrising.me/inde...ndary-mediation-analysis-of-pace-trial.34927/ and here: http://forums.phoenixrising.me/index.php?threads/new-pace-paper-more-smc-spin.34924/)

This is also free if one signs in: http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(14)00145-X/abstract

Note: these two psychologists are big CBT fans/promoters.
Abstract

Comment
What makes a difference in chronic fatigue syndrome?
Hans Knoop
Jan Wiborg

A substantial subgroup of patients with chronic fatigue syndrome (CFS) profits from behavioural interventions.1–3 A minority fully recovers after these interventions, but most do not,1,2 and quite a few patients do not profit at all.3 Therefore, outcome has substantial room for further improvement. Mediation analysis of intervention studies can help to identify the variables that bring on a reduction in chronic fatigue syndrome symptoms. With this knowledge, interventions could be improved and more patients could fully benefit from treatment.
 
Last edited:

Dolphin

Senior Member
Messages
17,567
Most objective measures showed no difference in the PACE Trial. One that did was the 6-minute walking test distance for GET (but not for CBT) which showed a small improvement although the results were still very poor for people of a mean age of 39 (esp. as people with lots of other chronic illnesses couldn't take part).

However, Knoop and Wiborg think that may be simply because the GET patients pushed themselves a little harder:

In the present study, no increase in fitness after GET was noted. An increase in distance walked during a test situation without an increased fitness suggests that patients walk more because of a change in cognitive processes (eg, daring to do more or an increased self-efficacy with respect to activity), not because of a change in physiological capacity.
 
Last edited:

anciendaze

Senior Member
Messages
1,841
Remember, we are talking about an increase of about 40 meters out of hundreds. This would not count as clinically significant even if we were talking about patients with a serious problem like heart failure, where change is notably difficult.

Also note that about 1/3 of the patients did not participate in the six-minute walk both before and after a year of treatment. I suspect that the original values were simply carried forward for those patients who did not take the second test. This is equivalent to stipulating that no patient suffered a setback, one of the hypotheses allegedly under test. Since the criteria for adverse events were changed during the study, to make it harder to register an adverse event, PACE provided absolutely no assurances about possible setbacks caused by GET beyond the observation that nobody died.

If the effect of GET was simply that patients with a fixed energy budget who were tired after a session declined the walk test, while those with a larger energy budget participated, this could produce the reported results without any deep psychological interpretation. The effort required to participate in therapy would simply do a better job of separating sheep from goats.

A change of some 20 meters in the distance of the group considered a control either means that similar modest improvement takes place without treatment, or that the one objective measurement left in the study was not properly controlled for test/retest familiarization.

The result was about as meaningless as it is possible to get, yet this has had no impact on researchers ability to spin hypotheses.
 

biophile

Places I'd rather be.
Messages
8,977
index.php
 
Last edited:

biophile

Places I'd rather be.
Messages
8,977
The extent to which an increased activity level and change in beliefs contribute to the outcome of behavioural interventions has clinical implications. Chalder and colleagues[3] conclude that future studies should focus on improving self-efficacy and increasing physical activity because these identified mediators had strong relations with the outcomes. We tend to draw a different conclusion with respect to the potential of improving outcome by increasing physical activity. We assume that an increase in physical activity is nothing more than a catalyst for the change in cognitions about activity and symptoms in patients with chronic fatigue syndrome. Future studies should focus on how these beliefs can be changed more rapidly and effectively. In our own protocol, we ask patients to gradually increase physical activity and present it as a way to increase your ability to become active. Once a patient is convinced that this is possible, irrespective of the actual level of activity, an important step towards recovery is taken.

Why do they keep talking about the importance of increasing activity levels when their own research disputes the notion that CBT/GET increases (multiple objective measures of) activity levels? Haha @ "an increase in physical activity is nothing more than a catalyst". If it doesn't or never happens how can it be a catalyst for anything? IIRC in previous papers they speculate about undetected increases in physical activity in the earlier stages of therapy which later get converted to social activities.
 
Last edited:

Sidereal

Senior Member
Messages
4,856
Also note that about 1/3 of the patients did not participate in the six-minute walk both before and after a year of treatment. I suspect that the original values were simply carried forward for those patients who did not take the second test.

It wasn't last observation carried forward. They just assumed that data were missing at random and carried out a completer analysis.

This was a complete case analysis, requiring an assumption of missing completely at random. This limitation was a greater concern for the step-test and walk-test mediators, where more data were missing, than for the questionnaire-based mediators. However, the results of the full information maximum-likelihood analysis did not differ greatly from the complete case analysis, suggesting serious missing data biases were unlikely.

"did not differ greatly"

How about presenting the maximum likelihood analyses so we can judge for ourselves?
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
It wasn't last observation carried forward. They just assumed that data were missing at random and carried out a completer analysis.

...

"did not differ greatly"

How about presenting the maximum likelihood analyses so we can judge for ourselves?
Would this analysis cover the scenario where people didn't complete the tests because they were unable/felt ill, where it would have been reasonable to assume their performance would be worse than average? That would seem to be key here.
 

Sidereal

Senior Member
Messages
4,856
Would this analysis cover the scenario where people didn't complete the tests because they were unable/felt ill, where it would have been reasonable to assume their performance would be worse than average? That would seem to be key here.

Well, in that case, data would not be missing at random which is an incredibly difficult problem to deal with.
 

Simon

Senior Member
Messages
3,789
Location
Monmouth, UK
Cause or effect?

At the heart of the mediation approach is showing a sequence of events, where the treatment afffects the mediator first, then the mediator affects the outcome. This appearst to be the case in this study, where mediators change first (12 weeks, or 24 for 6MWT) followed by an improved outcome at 52 weeks.

Treatment => Mediator => Outcome

Except, the outcomes improve at 12 and 24 weeks too. In other words, judging by figure 3 in this paper and the outcomes in the original PACE trial, 'mediators' and outcomes improve together. It may be that as people become less fatigued/able to do more, the 'mediator beliefs' change as a consequence. I'm not sure this study is able to show which happens first: the outcomes of fatigue and function may be changing the 'beliefs', not the other way around. Or the two may happen in tandem.

Added to this, the study says that the relationship between meditiators and outcomes were the same in all treatment groups, ie including APT and SMC, where beliefs and behaviours were not challenged.
 

A.B.

Senior Member
Messages
3,780
'm not sure this study is able to show which happens first: the outcomes of fatigue and function may be changing the 'beliefs', not the other way around. Or the two may happen in tandem.

Added to this, the study says that the relationship between meditiators and outcomes were the same in all treatment groups, ie including APT and SMC, where beliefs and behaviours were not challenged.

They really don't seem to care anymore. It's just spin, spin, spin.
 

anciendaze

Senior Member
Messages
1,841
It wasn't last observation carried forward. They just assumed that data were missing at random and carried out a completer analysis...
Again, this is tantamount to ignoring either the possibility that any patient for whom data were missing suffered a setback, or that patients were using energy from a fixed budget to participate, as has been found in both patient reports and other studies.

Both alternatives amount to assuming something which was supposed to be a result of this study. If I am allowed to assume results, when convenient, I can produce some really astonishing research.
 

Dolphin

Senior Member
Messages
17,567
This was a complete case analysis, requiring an assumption of missing completely at random. This limitation was a greater concern for the step-test and walk-test mediators, where more data were missing, than for the questionnaire-based mediators. However, the results of the full information maximum-likelihood analysis did not differ greatly from the complete case analysis, suggesting serious missing data biases were unlikely.

"did not differ greatly"

How about presenting the maximum likelihood analyses so we can judge for ourselves?
One will recall that in the PACE Trial's cost effectiveness paper
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0040808 , they said
No other sensitivity analyses (i.e. changing the value of informal care, lost employment and standardised medical care) had a large impact on cost-effectiveness.
when in fact valuing care at the minimum wage rate (one of the sensitivity analyses mentioned in the paper) made a huge difference (and valuing it at 0 which was a sensitivity analysis in their statistical plan made an even bigger difference).
 
Last edited:

Bob

Senior Member
Messages
16,455
Location
England (south coast)
This might have been posted before, but I'm catching up...
Here's an interesting extract from the Knoop and Wiborg commentary...
Knoop and Wiborg said:
In the present study, no increase in fitness after GET was noted. An increase in distance walked during a test situation without an increased fitness suggests that patients walk more because of a change in cognitive processes (eg, daring to do more or an increased self-efficacy with respect to activity), not because of a change in physiological capacity. This would also explain why patients are not more physically active after successful treatment of chronic fatigue syndrome:9 they do not necessarily have to move more, but their perception of effort and symptoms has to change.
They seem to be acknowledging that CBT in general fails to change the illness itself, and fails to improve disability, but that it only changes interpretation of symptoms or the patients' perception of the illness. That's probably a big thing if you consider CFS to be psychosomatic, but it's an interesting observation nonetheless.
 
Messages
15,786
.They seem to be acknowledging that CBT in general fails to change the illness itself, and fails to improve disability, but that it only changes interpretation of symptoms or the patients' perception of the illness.
Or they're suggesting that there's no disability in the first place, despite all evidence to the contrary. No reduction in movement means no need to improve.
 

Dolphin

Senior Member
Messages
17,567
Or they're suggesting that there's no disability in the first place, despite all evidence to the contrary. No reduction in movement means no need to improve.
Although they can't really say that there is no reduction in movement when their own studies have found lower than average movement, both before and after treatment.
 
Messages
15,786
Although they can't really say that there is no reduction in movement when their own studies have found lower than average movement, both before and after treatment.
Yes, I think that's why they're not being completely specific and are just saying things which might imply several things to their audience. I think they're counting on readers to interpret it to mean that there's no disability to start with - after all, it wouldn't make sense for someone to say that reports of fatigue are more important than actual disability!
 

Bob

Senior Member
Messages
16,455
Location
England (south coast)
Yes, I think that's why they're not being completely specific and are just saying things which might imply several things to their audience. I think they're counting on readers to interpret it to mean that there's no disability to start with - after all, it wouldn't make sense for someone to say that reports of fatigue are more important than actual disability!
I understand what you're getting at Val, but I can't see how they can deny that there is a disability, considering that all of their research demonstrates that there is a severe problem with physical function, physical capacity and physical disability etc. I think perhaps they might dismiss disability as unimportant because it is a indicator of illness beliefs rather than an indicator of an actual physical illness. But I'm not over familiar with the work of Knoop and Wiborg, so I don't really know what their stance is.