Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of PACE trial

[Dolphin]

This thread is for discussion of the paper itself

There's a separate thread for discussion of the media coverage, etc. here: http://forums.phoenixrising.me/index.php?threads/new-pace-paper-more-smc-spin.34924/

If you want to write a letter, here are the details:
http://www.thelancet.com/lanpsy/information-for-authors/article-types-manuscript-requirements

Types of article and manuscript requirements

[..]

Correspondence

• Letters should be written in response to previous content published in The Lancet Psychiatry
• Letters for publication in the journal online must reach us within 4 weeks of publication of the original item and should be no longer than 500 wordss
• Letters of general interest, unlinked to items published in the journal, can be up to 400 words long
• Correspondence letters are not usually peer reviewed, but we might invite replies from the authors of the original publication, or pass on letters to these authors
• Only one table or figure is permitted, and there should be no more than five references and five authors
• All accepted letters are edited, and may be shortened in the interest of space. Proofs will be sent out to authors before publication

 

[Dolphin]

This paper is open access (requires free registration)

Note there's also a supplemental file which is free:

http://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(14)00069-8/abstract

Rehabilitative therapies for chronic fatigue syndrome: a secondary mediation analysis of the PACE trial

Summary

Background

Cognitive behaviour therapy (CBT) added to specialist medical care (SMC), or graded exercise therapy (GET) added to SMC, are more effective in reducing fatigue and improving physical function than both adaptive pacing therapy (APT) plus SMC and SMC alone for chronic fatigue syndrome.

We investigate putative treatment mechanisms.

Methods

We did a planned secondary mediation analysis of the PACE trial comparing SMC alone or SMC plus APT with SMC plus CBT and SMC plus GET for patients with chronic fatigue syndrome.

641 participants were recruited from six specialist chronic fatigue syndrome clinics in the UK National Health Service between March 18, 2005, and Nov 28, 2008.

We assessed mediation using the product of coefficients method with the 12 week measure of the mediators and the 52 week measure of the outcomes.

The primary outcomes were fatigue measured by the Chalder fatigue scale and physical function measured by the physical function subscale of the SF-36.

We included confounder covariates and used treatment by mediator interaction terms to examine differences in mediator–outcome relations by treatment group.

Findings

The largest mediated effect for both CBT and GET and both primary outcomes was through fear avoidance beliefs with an effect of larger magnitude for GET (standardised effects ×10, CBT vs APT, fatigue −1·22, 95% CI −0·52 to −1·97, physical function 1·54, 0·86 to 2·31; GET vs APT, fatigue −1·86, −0·80 to −2·89, physical function 2·35, 1·35 to 3·39).

Increase in exercise tolerance (6 min walk distance) was a potent mediator of the effect of GET (vs APT, fatigue −1·37, 95% CI −0·76 to −2·21, physical function 1·90, 1·10 to 2·91), but not CBT.

Interpretation

Our main finding was that fear avoidance beliefs were the strongest mediator for both CBT and GET.

Changes in both beliefs and behaviour mediated the effects of both CBT and GET, but more so for GET.

The results support a treatment model in which both beliefs and behaviour play a part in perpetuating fatigue and disability in chronic fatigue syndrome.

Funding
UK Medical Research Council, Department of Health for England, Scottish Chief Scientist Office, Department for Work and Pensions, National Institute for Health Research (NIHR), NIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust, and Institute of Psychiatry, Psychology, and Neuroscience, King's College London.

 

[Dolphin]

It should be noted that what is discussed is the mediators of two subjective outcomes.

Here's one of them:

Chalder Fatigue Questionnaire

We would like to know more about any problems you have had with feeling tired, weak or lacking in energy in the last month. Please answer ALL the questions by ticking the answer which applies to you most closely. If you have been feeling tired for a long while, then compare yourself to how you felt when you were last well. (Please tick only one box per line).

Q. Do you have problems with tiredness?

0. Less than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you need to rest more?
0. Less than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you feel sleepy or drowsy?
0. Less than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you have problems starting things?
0. Less than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you lack energy?
0. Better than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you have less strength in your muscles?
0. Better than usual ____
1. No more than usual ____
2. More than usual ____
3. Much more than usual ____

Q. Do you feel weak?
0. Less than usual ____
1. Same as usual ____
2. More than usual ____
3. Much more than usual ____

A. Do you have difficulty concentrating?
0. Less than usual ____
1. Same as usual ____
2. More than usual ____
3. Much worse than usual ____

Q. Do you find it more difficult to find the correct word?
0. Less than usual ____
1. No more than usual ____
2. Worse than usual ____
3. Much worse than usual ____

Q.. Do you make slips of the tongue when speaking?
0. Less than usual ____
1. No more than usual ____
2. Worse than usual ____
3. Much worse than usual ____

Q. How is your memory?
0. Better than usual ____
1. No worse than usual _____
2. Worse than usual ____
3. Much worse than usual ____

 

[Dolphin]

Here's the other subjective outcome measure that was used:

SF-36 physical function sub-scale

Please tick one box on each line
The following questions are about activities you might do during a typical day. Does your health limit you in these activities? If so, how much?


Yes, Limited A Lot

Yes, Limited A Little

No, Not Limited At All

Vigorous activities, such as running, lifting heavy objects, participating in strenuous sports

Moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf

Lifting or carrying groceries

Climbing several flights of stairs

Climbing one flight of stairs

Bending, kneeling, or stooping

Walking more than a mile

Walking several hundred yards

Walking one hundred yards

Bathing or dressing yourself

Scoring is basically:
Yes, Limited A Lot: 0 points each
Yes, Limited A Little: 5 points each
No, Not Limited At All: 10 points each

 

[charles shepherd]

Statement from The ME Association - issued to health journalists and editors on 12 January 2014

The ME Association believes that energy management, which involves both physical and mental activity, is the most important aspect of managing ME/CFS.

Consequently, we welcome research which aims to improve our knowledge of how this can best be achieved.

Energy management programmes should be individually tailored. And they must take account of the wide range of clinical presentations and disease pathways that come under the ME/CFS umbrella.

We are therefore very critical of over-simplistic exercise regimes which are based on deconditioning, fear of activity and the assumption that people just need to 'try harder' in order to get better.

This approach does not acknowledge the muscle, brain and immune system abnormalities in ME/CFS, which help to provide a physical explanation for the debilitating central (brain) and peripheral (muscle) fatigue that occurs in this illness.

Feedback from patients, who are not normally treated as part of research studies, consistently indicates that around 25% find exercise programmes to be of no value. Around 50% report that their condition worsens as a result.

So the real reason why cognitive behaviour therapy and graded exercise therapy are not producing significant benefits across the whole ME/CFS population is that they are based on a flawed model of illness causation. They are are not taking account of the medical complexities involved.

In our experience, patients with ME/CFS are highly motivated to get better.

They are very willing to take part in energy management programmes that establish a safe baseline of activity, take account of the stage and severity of their illness, involve increases in physical and mental activity that are gradual, flexible and do not result in the person going beyond their limitations and causing symptom exacerbation.

We fear that the results of this study will be interpreted to reinforce the over-simplistic view that rest is bad and exercise is good for people with ME/CFS.

And without discussing the medical complexities involved, they also infer that any reluctance to progressively increase activity levels is far more related to fear or avoidance behaviour than any underlying disease process.

What we desperately need are high quality research studies that are aimed at producing a range of individual energy management programmes based on clinical presentation, stage, and severity of illness.

Dr Charles ShepherdHon Medical Adviser, MEA

Background information:
The MEA submission to NICE, regarding their recommendations relating to the use CBT and GET in their current (2007) guideline on ME/CFS:

CBT, GET and Pacing

Our principal reason for requesting a fundamental review of the NICE guideline on ME/CFS relates to the recommendation that CBT and GET should be automatically offered to everyone with mild or moderate ME/CFS.

This is coupled with the continuing failure of NICE to take note of highly consistent patient evidence, dating back to evidence that was published in the 2002 Chief Medical Officer’s report on ME/CFS, regarding the efficacy and safety of these two behavioural treatments.

The largest ever survey of patient evidence relating to all aspects of the management of ME/CFS was carried out by The ME Association and published in 2010 (ME Association). The report provided important evidence regarding concerns over the efficacy of CBT and the safety of GET.

For CBT (997 responses)
Greatly improved: 2.8%
Improved: 23.1%
No change: 54.6%
Slightly worse: 11.6%
Much worse: 7.9%

For GET (906 responses)
Greatly improved: 3.4%
Improved: 18.7%
No change: 21.4%
Slightly worse: 23.4%
Much worse: 33.1%

For Pacing (2137 responses)
Greatly improved: 11.6%
Improved: 59.6%
No change: 24.1%
Slightly worse: 3.5%
Much worse: 1.2%

The MEA is currently in the final stages of preparing a further report covering the use of CBT, GET and Pacing – but this time in much greater depth. The report will be based on the answers to questions on the above three treatments that were provided through 3142 responses given by 1429 respondents during 2012.

Overall, the patient evidence contained in this new MEA report is very similar to the evidence contained in the 2010 report. The two MEA surveys show a total of 6599 responses about the effect of treatments on symptoms, and a total of 6838 responses about appropriateness of courses, effectiveness of self management and helpfulness of consultations and general satisfaction.

However, to date NICE has failed to consider any of this patient evidence and both MEA reports support the findings from patient surveys referred to in the Chief Medical Officer’s Working Group report into ME/CFS.

We are therefore looking at a consistent picture from patients with regard to all three approaches to management going back over at least a decade and the picture has not improved.

As a result of growing concern amongst people with ME/CFS about the efficacy and safety of CBT and GET, we will be making a number of radical recommendations regarding the future use of CBT and GET in ME/CFS in this report.

This is clearly important new evidence that cannot be ignored by NICE.

The PACE trial and the March 2011 surveillance review

Finally, in relation to CBT and GET and Pacing, we assume that the NICE guideline surveillance review that took place in March 2011, and which followed publication of the PACE trial results in February 2011, simply ‘rubber stamped’ the 2007 NICE guideline recommendations on the basis that the PACE trial had supported the recommendations relating to CBT and GET.

However, there has been widespread and valid criticism about the way in which the PACE trial was carried out, as well as the way in which the results were presented and reported.

In addition, it should be noted that the cost effectiveness paper by McCrone et al reported that take up of state sickness benefits had increased during the PACE trial for all four treatments (ie CBT, GET, Pacing and Standard Medical Care). The MEA report will also contain similar information on benefit status.

 

[Dolphin]

Note that buried in the paper they report that CBT & GET didn't result in improvements on fitness/step test

There were no effects on HADS anxiety, physical fitness, or the adjusted perception of effort measure (Borg scale).

This can be seen in Figure 3.

 

[lansbergen]

I think this is a good statement. Even journalist must be able to understand it.

 

[Bob]

Note that buried in the paper they report that CBT & GET didn't result in improvements on fitness/step test

There were no effects on HADS anxiety, physical fitness, or the adjusted perception of effort measure (Borg scale).

This can be seen in Figure 3.

Well spotted, Dolphin.

There's an interesting passage about "fitness" (step test) vs "physical activity" (6MWDT)...

We found that fitness and perception of exertion did not appear to mediate treatment effects, but that timed walking distance, assessed for the first time in our study, mediated the effect of GET. This suggests that increasing tolerance of physical activity might produce benefit without improving physical fitness. These findings are consistent with those of previous studies.[8,9,32] A randomised controlled trial of GET for chronic fatigue syndrome [8,16] found that those who rated themselves as better were no fitter or stronger than the rest. A second trial of GET also suggested that physical reconditioning was not a mediator of the effect of treatment, but that a reduction in symptom focusing and increased exercise tolerance (as assessed by maximum heart rate achieved with exercise) mediated change in mental and physical fatigue.[9] It must be acknowledged, however, that the walking test might not reflect activity or exercise levels in everyday life and might provide an explanation for why Wiborg and colleagues[33] found no evidence that actometer-measured physical activity mediated the effect of CBT.

I'm not certain but they seem to be rebutting/trashing their own deconditioning hypothesis (while propping up their fear/avoidance hypothesis), as also suggested by these extracts...

Fitness measures did not mediate the effects of the treatments.

We confirm that fear avoidance beliefs partially mediate the effects of CBT and GET on physical function and fatigue outcomes, and that fitness measures do not appear to mediate the effects of either treatment.

The results suggest that GET might be more specific in its effects than CBT, with two strong mediators, fear avoidance beliefs, and timed walking distance. The increase in exercise tolerance (walking distance) without an increase in exercise capacity (fitness) might have been facilitated by the mediating effect of reduced fear avoidance beliefs.

This mediational analysis strengthens the validity of our theoretical model of CBT and supports the idea that a similar model is valid for GET by confirming the role of fearful beliefs and avoidance behaviour.

 

[Bob]

They've carried out an analysis of the 6MWDT for the first time, and they've managed to portray GET as having a significant effect in the 6MWDT:

In both types of comparisons, GET significantly increased the number of metres walked, whereas CBT did not (GET vsAPT 0·43, 95% CI 0·25–0·61; GET vs SMC 0·46, 0·28–0·63; CBT vs GET p=0·001).

 

[Bob]

Interesting to note:

We did not find much evidence for symptom focusing as a mediator in this study.

 

[Sean]

Note that buried in the paper they report that CBT & GET didn't result in improvements on fitness/step test.

Is this the first time they have reported on the step test outcomes?

GET significantly increased the number of metres walked,

The increase was barely statistically significant, and didn't reach the much more important and relevant measure of clinical significance. PACE patients (average age 40 years old at trial end) still ended up only scoring the same as people twice their age, and scored worse than almost every category of serious cardio-pulmonary disease, including terminal ones.

BFD.

 

[Dolphin]

We have new 6-minute walking test data in the supplement. These are the scores at 24 weeks.

n (%) complete for walk test data

APT: 113 (71)
CBT: 122 (76)
GET: 119 (74)
SMC: 118 (74)

Metres walked

Baseline
APT: 322 (306 to 338)
CBT: 336 (320 to351)
GET: 321 (308 to 335)
SMC: 332 (316 to 347)

24 weeks

APT: 342 (322 to 361)
CBT: 361 (345 to 377)
GET: 381 (366 to 395)
SMC: 350 (333 to 368)

Here are the results at 52 weeks:

APT: 334 (117) i.e. -8m compared to 24 weeks
CBT: 354 (106) i.e. -7m compared to 24 weeks
GET: 379 (100) i.e. -2m compared to 24 weeks
SMC: 348 (108) i.e. -2m compared to 24 weeks

So doesn't look like one can conclude that they are gradually getting better after an initial slight improvement all around

They have data at baseline and 52 weeks for:

APT: 111 (70%) i.e. -1% compared to baseline/24 weeks
CBT: 123 (76%) i.e. no percentage change compared to baseline/24 weeks
GET: 110 (69%) i.e. -5% compared to baseline/24 weeks
SMC: 118 (74%) i.e. -1% compared to baseline/24 weeks

The 5% drop off (8 people) for GET makes me wonder whether people in GET were more likely to overpush themselves at 24 weeks and then have payback and so decide not to do the test at 52 weeks.

Knoop and Wiborg themselves speculated this might be happening:

"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."

 

[Bob]

Is this the first time they have reported on the step test outcomes?

Yes, I think it is...

So, as @Dolphin said on the other thread, it can now be said that: "Most objective measures showed no difference in the PACE Trial."

 

[Dolphin]

Note that buried in the paper they report that CBT & GET didn't result in improvements on fitness/step test.

Is this the first time they have reported on the step test outcomes?

I believe so. (I think it might have been mentioned at a lecture (Bristol?) but not published)

 

[Bob]

GET significantly increased the number of metres walked...

The increase was barely statistically significant, and didn't reach the much more important and relevant measure of clinical significance.

Yes, like you say, it was a barely significant result and I'm pretty certain that it did not reach a moderate effect size.
I'm not yet sure what their newly published outcomes indicate e.g. GET vs SMC 0·46, 0·28–0·63
I wonder if "0.46" is supposed to be an effect size, or what? It looks far too large to be a cohen's d effect size.

 

[Sean]

"Most objective measures showed no difference in the PACE Trial."

I think we can safely say something much more definitive and punchy:

No objective measure showed any clinically significant improvement.

n (%) complete for walk test data

APT: 113 (71)
CBT: 122 (76)
GET: 119 (74)
SMC: 118 (74)

I don't know if that dropout rate is normal, let alone acceptable, for clinical studies. But given the very meagre 6MWT 'gains', presumably it could easily confound that result.

 

[Dolphin]

n (%) complete for walk test data

APT: 113 (71)
CBT: 122 (76)
GET: 119 (74)
SMC: 118 (74)

I don't know if that dropout rate is normal, let alone acceptable, for clinical studies. But given the very meagre 6MWT 'gains', presumably it could easily confound that result.

I edited the following into the post above at some stage around when you posted.

White et al (2011) has the data at baseline and 52 weeks for:

APT: 111 (70%) i.e. -1% compared to baseline/24 weeks
CBT: 123 (76%) i.e. no percentage change compared to baseline/24 weeks
GET: 110 (69%) i.e. -5% compared to baseline/24 weeks
SMC: 118 (74%) i.e. -1% compared to baseline/24 weeks

The 5% drop off (8 people) for GET makes me wonder whether people in GET were more likely to overpush themselves at 24 weeks and then have payback and so decide not to do the test at 52 weeks.

Knoop and Wiborg themselves speculated this might be happening:

"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."

 

[Bob]

I don't know if that dropout rate is normal, let alone acceptable, for clinical studies. But given the very meagre 6MWT 'gains', presumably it could easily confound that result.

Just noting that they do say something about the drop-out rates (and dismiss its importance, naturally)...

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.

 

[Sean]

So, on the 6MWT, the GET arm results are:

1. A barely statistically significant improvement, (a gain of just 35.3 metres, as mean difference from SMC),

2. no clinically significant improvement,

3. patients still scoring worse than almost all of one of the sickest disease groups in medicine, (with any ambulatory function left – the cardio-pulmonary disorders, including terminal end-stage forms),

4. the outcome data for the (objective) 6MWT is only available for 69-74% of patients, compared to the range for (subjective) self-report measures of 89-97%, and

5. the highest drop out rate of all arms.

Correct?

 

[anciendaze]

I'd say it wouldn't have made much difference if patients had supplied no data at all. This bunch is used to ignoring what patients say or do. Perhaps the next paper in this series will analyze the patterns of tea leaves left in cups patients used to show that there will be great improvements for these patients at some unspecified time in the future.