PACE Trial and PACE Trial Protocol

[Dolphin]

42% in the APT group improved by at least 2 points for fatigue and at least 8 points (isn't 8 impossible? doesn't it have to be 5 or 10?) for physical function at 52 weeks, compared with 59% for CBT etc.

Page 9, bottom left paragraph.

Tattoo'ed on this sad man's brain.

That's it.
I believe they filled in some missing values by averaging the rest of the entries so one could get final numbers that were not divisible by 5 and then the change could be 8. But you are right in general.

 

[Bob]

42% in the APT group improved by at least 2 points for fatigue and at least 8 points (isn't 8 impossible? doesn't it have to be 5 or 10?) for physical function at 52 weeks, compared with 59% for CBT etc.

Page 9, bottom left paragraph.

Tattoo'ed on this sad man's brain.

Oh, it's the stupid post-hoc analysis.

I didn't recognise it because those statements are so utterly misleading.

Yes, 8 is impossible, isn't it, but I know that was a rhetorical question.

And a score of 10 could be just one question answered differently which, apparently, is a significant and meaningful improvement.

 

[oceanblue]

42% in the APT group improved by at least 2 points for fatigue and at least 8 points (isn't 8 impossible? doesn't it have to be 5 or 10?) for physical function at 52 weeks, compared with 59% for CBT etc.

Page 9, bottom left paragraph.

Tattoo'ed on this sad man's brain.

Yes, this is one area where I think the measure of clinically useful is OK for SF-36 as it is 10 for any patient, equivalent to improving one level on 2 areas (eg from limited a lot to limited a little, or limited a little to not limited at all), or one question changing from limited a lot to not limited at all. Looking at SF-36 for other (non-CFS) clinical trials, 10 seems not a bad measure - though obviously is nothing like recovery. 5 would be very dodgy (and some non_CFS trials do use this) as it is the smallest possible improvement.

 

[Graham]

What bothers me though is that when I try to answer either the fatigue or physical function questions for myself, I find it difficult to decide with a few of them which level to tick. I can easily see that, although being at the same standard, but being in a more positive frame of mind, I could tick a "better" score. Now I like to con myself that I am an analytic sort, so if I have these uncertainties, how easy is it to massage the results in general simply through the positive assertions of the therapists.

 

[oceanblue]

What bothers me though is that when I try to answer either the fatigue or physical function questions for myself, I find it difficult to decide with a few of them which level to tick. I can easily see that, although being at the same standard, but being in a more positive frame of mind, I could tick a "better" score. Now I like to con myself that I am an analytic sort, so if I have these uncertainties, how easy is it to massage the results in general simply through the positive assertions of the therapists.

Interesting. When I've tried it there are only a couple of questions where there is any doubt (for me most are not limited at all eg bending, or limited a lot ie can't do at all). I posted a while a go on this and there were one, possible 2 questions where my mindset could change the answer by one degree ie 5 points easily changeable, 10 points just about possible (though unlikely) but more than that no way. But that's based on n=1 so good to get another perspective.

 

[Graham]

The trouble is that I am relatively fortunate. I can focus on things for up to an hour at a time (like maths tuition), and even longer when it is just single-channelled thinking. But I can only sustain two tutorials a week: I tried three, but ran out of steam after a few weeks. I can walk with the dog for up to an hour (and it is hilly around here), albeit at an amble. Now in comparison with what I was like when healthy, that is truly feeble, but in comparison with so many of you, I know how lucky I am. So I keep sitting in the borderline between 2 and 3 (or the equivalent), not knowing whether to answer in terms of a drop in percentage compared with when I was fit (I reckon around a 75% drop), or recognise how much worse it could be.

It seems to be people like me that produce the greater variations in scores from good to bad patches as well, and they are the ones that produce the statistics. Of course, that is just my interpretation: I have no real evidence to back it up. If the raw data were available, we would be able to pin down what type of patients produced the bigger changes, and we could compare them with the natural fluctuations that we found in our recent survey.

 

[Bob]

A change in just one answer in an individual's questionnaire could be the result of response bias. Perhaps response bias might have been accounted for by using a decent comparable control group, such as a 'positive thinking' group?

With regard to a clinically useful difference, the Guyatt paper* (cited in the PACE Trial paper, to justify the methodology) recommends using data from the general population, to avoid the pitfalls of the PACE Trial's methodology.

The PACE Trial defined a clinically useful difference as:

"A clinically useful difference between the means of the primary outcomes was defined as 0.5 of the SD of these measures at baseline, equating to 2 points for Chalder fatigue questionnaire and 8 points for short form-36."

But if, instead we look at 0.5 SD of the mean scores for the general population, as the Guyatt paper recommends, then the CUD (0.5 SD) is 9.7 for SF-36 PF scores, using the British ONS Survey 1992**. Or if we use the SD for the general population given in the PACE Trial (24 points), then 0.5 SD is 12 points for SF-35 PF scores.

Using the lower of these (9.7 points), would mean that neither CBT or GET would have resulted in a clinically useful outcome for average SF-36 physical function scores in the PACE Trial (i.e. a total FAIL!)

But those who understand statistics will be able to explain why it's not appropriate to use standard deviations for samples that do not have an even distribution, so even these figures are not appropriate.

Using the Norwegian normative data*** for the Chalder scores, 0.5 SD is 2 points for the chalder scale, which is identical to the PACE Trial paper.


Anyway, this is my understanding of all this, so far, but it's a work in progress, so there might be some glaring errors here.



* Methods to Explain the Clinical Significance of Health Status Measures
GORDON H. GUYATT et al
http://www.ncbi.nlm.nih.gov/pubmed/11936935

** The Omnibus Survey in Britain - British ONS Survey 1992 (ages 16+).
http://jpubhealth.oxfordjournals.org/content/21/3/255.abstract

*** Fatigue in the general Norwegian population: normative data and associations.
Loge JH, Ekeberg O, Kaasa S.
1998
http://www.ncbi.nlm.nih.gov/pubmed/9720855

 

[Dolphin]

With regard to a clinically useful difference, the Guyatt paper* (cited in the PACE Trial paper, to justify the methodology) recommends using data from the general population, to avoid the pitfalls of the PACE Trial's methodology.

The PACE Trial defined a clinically useful difference as:

"A clinically useful difference between the means of the primary outcomes was defined as 0.5 of the SD of these measures at baseline, equating to 2 points for Chalder fatigue questionnaire and 8 points for short form-36."

But if, instead we look at 0.5 SD of the mean scores for the general population, as the Guyatt paper recommends, then the CUD (0.5 SD) is 9.7 for SF-36 PF scores, using the British ONS Survey 1992**. Or if we use the SD for the general population given in the PACE Trial (24 points), then 0.5 SD is 12 points for SF-35 PF scores.

Using the lower of these (9.7 points), would mean that neither CBT or GET would have resulted in a clinically useful outcome for average SF-36 physical function scores in the PACE Trial (i.e. a total FAIL!)

But those who understand statistics will be able to explain why it's not appropriate to use standard deviations for samples that do not have an even distribution, so even these figures are not appropriate.

Using the Norwegian normative data*** for the Chalder scores, 0.5 SD is 2 points for the chalder scale, which is identical to the PACE Trial paper.


Anyway, this is my understanding of all this, so far, but it's a work in progress, so there might be some glaring errors here.



* Methods to Explain the Clinical Significance of Health Status Measures
GORDON H. GUYATT et al
http://www.ncbi.nlm.nih.gov/pubmed/11936935

** The Omnibus Survey in Britain - British ONS Survey 1992 (ages 16+).
http://jpubhealth.oxfordjournals.org/content/21/3/255.abstract

*** Fatigue in the general Norwegian population: normative data and associations.
Loge JH, Ekeberg O, Kaasa S.
1998
http://www.ncbi.nlm.nih.gov/pubmed/9720855

Just pointing out that Jane Giakoumakis' letter makes this point for anyone who missed it:

The Lancet, Early Online Publication, 17 May 2011
doi:10.1016/S0140-6736(11)60689-2

The PACE trial in chronic fatigue syndrome
Jane Giakoumakis a

In their randomised trial of treatments for patients with chonic fatigue syndrome, Peter White and colleagues (March 5, p 823)1 define a clinically useful difference between the means of the primary outcomes as 05 of the SD of these measures at baseline, equating to 2 points for Chalder fatigue questionnaire and 8 points for short form-36. They cite achieving a mean clinically useful difference in the graded exercise therapy or cognitive behaviour therapy groups, compared with specialist medical care alone, as evidence that these interventions are moderately effective treatments.

The source for this definition of clinically useful difference states that such a method has a fundamental limitation: estimates of variability will differ from study to studyif one chooses the between-patient standard deviation, one has to confront its dependence on the heterogeneity of the population under study.2 In White and colleagues' study, we do not have heterogeneous samples on the Chalder fatigue questionnaire and short-form 36 physical function subscale, since both are used as entry criteria.1

Patients had to have scores of 65 or less on short-form 36 to be eligible for the study.1 However, most, in practice, would probably need to have scores of 30 or more to be able to participate in this clinic-based study. Indeed, only four of 43 participants in a previous trial of graded exercise therapy scored less than 30.3, 4 Guyatt and colleagues2 suggest that an alternative is to choose the standard deviation for a sample of the general population, which White and colleagues have given as 24.1 An SD of 24 gives a clinically useful difference of 12; both graded exercise therapy and cognitive behaviour therapy fail to reach this threshold. Whether they moderately improve outcomes, as claimed,1 is therefore questionable.


I am chair of a myalgic encephalomyelitis support and advice groupan unpaid voluntary position.

References

1 White PD, Goldsmith KA, Johnson AL, et alon behalf of the PACE trial management group. Comparison of adaptive pacing therapy, cognitive behaviour therapy, graded exercise therapy, and specialist medical care for chronic fatigue syndrome (PACE): a randomised trial. Lancet 2011; 377: 823-836. Summary | Full Text | PDF(309KB) | CrossRef | PubMed

2 Guyatt GH, Osaba D, Wu AW, et al. Methods to explain the clinical significance of health status measures. Mayo Clinic Proc 2002; 77: 371-383. PubMed

3 Fulcher KY. Physiological and psychological responses of patients with chronic fatigue syndrome to regular physical activity. Loughborough: Loughborough University of Technology, 1997. http://hdl.handle.net/2134/6777. (accessed March 4, 2011).

4 Fulcher KY, White PD. Randomised controlled trial of graded exercise in patients with the chronic fatigue syndrome. BMJ 1997; 314: 1647-1652. PubMed

a xxxxxxxxxxxxxx, Glasgow xxxxxxxxx, UK

 

[oceanblue]

Origin of the Chalder Fatigue Scale (Simon Wessely)

He's commenting in general on one of his 'landmark' papers here:

There was no instrument available to measure subjective fatigue, so I simply invented one, which would later get modified into the Chalder Fatigue Scale, which also became a citation hit. And basically that was that.

Thread on the commentary

 

[biophile]

I noticed that the PACE Trial manuals call CBT "Complex Incremental Pacing" and GET "Simple Incremental Pacing". No wonder some people think elements of CBT and GET have been stolen from pacing. APT is called "Simple, non-Incremental / Adaptive Pacing Therapy". Another red flag in their attempt to muscle out pacing? I can imagine a future where patients go to a CFS clinic for "pacing" and get hoodwinked into simple or complex "Incremental Pacing" ie CBT/GET. The rationale for the name seems to be that it contains "elements of simple pacing to stabilise activity".

 

[WillowJ]

the cfids self-help site (which I'm not familiar with, other than recognizing the name as a site which exists) oddly states that all the therapies are somewhat like what they call pacing.

more interestingly, they say the therapies are all similar, but APT is new and written especially for the trial (which we knew; I wish they had a link or citation for this, but they do say they have it from the authors)

Adaptive Pacing

The steps for this approach are not laid out as clearly as those for GET and CBT. The researchers explain that the materials for this approach were written for the trial, while materials used with GET and CBT dated back to the 1990s. So one way to view the results of the trial would be to say that two mature systems for teaching pacing were found superior to third, less mature approach.

In any case, adaptive pacing therapy (APT) has many similarities to the other two.

my bolding and bold italics

http://www.cfidsselfhelp.org/library/pace-trial-shows-two-forms-pacing-more-effective-third-type

 

[Bob]

... but APT is new and written especially for the trial (which we knew; I wish they had a link or citation for this, but they do say they have it from the authors)

Yes, APT was an unsuccessful attempt at designing a novel 'therapy', and it was created specifically for the PACE Trial.

Here's a quote from the PACE Trial paper itself, but there is sure to be more about it in the protocols:

"Westcare and Action for ME helped in the design of the therapy and endorsed the final manuals."

 

[Esther12]

Yes, APT was an unsuccessful attempt at designing a novel 'therapy', and it was created specifically for the PACE Trial.

Here's a quote from the PACE Trial paper itself, but there is sure to be more about it in the protocols:

"Westcare and Action for ME helped in the design of the therapy and endorsed the final manuals."

It was always silly to view pacing as a 'therapy' imo.

No-one should be paid to provide 'pacing treatment' to CFS patients. If anything, it's an acknowledgement that it is not worth paying people to provide any particular treatment for CFS, and an attempt to provide some sort of medical legitimacy to the reduced capacity of patients. The results from PACE would seem to support those assumptions, and presumably the vast majority of patients who had not recovered after CBT/GET (90%? More? We still don't have the actual figures, and only know that 70%+ were nowhere near recovery) would now be pacing themselves.

 

[Bob]

It was always silly to view pacing as a 'therapy' imo.

No-one should be paid to provide 'pacing treatment' to CFS patients. If anything, it's an acknowledgement that it is not worth paying people to provide any particular treatment for CFS, and an attempt to provide some sort of medical legitimacy to the reduced capacity of patients. The results from PACE would seem to support those assumptions, and presumably the vast majority of patients who had not recovered after CBT/GET (90%? More? We still don't have the actual figures, and only know that 70%+ were nowhere near recovery) would now be pacing themselves.

All we know is that approx 13% (average for both CBT & GET in both primary outcome measures) of patients experienced at least a minimum improvement (not a recovery), after receiving CBT or GET.
If we extrapolate to the whole population (making an assumption that 25% are severely affected), then it goes down to 10%.
And that's not even taking into account all of the weaknesses of the methodology.

APT cannot be called a 'therapy' because it was invented for the PACE Trial, and failed to benefit patients. (i.e. if a proposed treatment doesn't work, then it can't be called a 'therapy'.)

 

[oceanblue]

I've posted some open questions about why APT had no apparent effect on fatigue in PACE controls/comparison thread. I realise this might be seen as provocative but that's not my intention at all: I'm merely trying to understand the results.

 

[biophile]

biophile wrote (post#1512): In a study of 51 community-dwelling patients (average age 71 5 years) with established chronic coronary heart disease, the average physical function score was 68 19 points, making them less disabled on average than all therapy groups post-treatment in the PACE Trial despite the latter being 30 years younger on average (http://archinte.ama-assn.org/cgi/reprint/159/19/2357.pdf).

Oops, wrong URL. I can't edit the post now but it should be : http://www.ncbi.nlm.nih.gov/pubmed/11773926 .

 

[biophile]

A patient charity is responsible for PACE dropping actigraphy?

I just noticed that PACE have released a comprehensive FAQ2 on their website on the 13th December 2011.

http://www.pacetrial.org/faq/faq2.html

Enjoy / have fun. Haven't looked at it all yet but noticed some of the usual spin, such as Q&A #46 (Some patients didnt improve: was it worth it?):

"Since approximately 6 out of 10 patients improved after either cognitive behavioural or graded exercise therapies, the results of this study are encouraging."

They fail to mention here that 45% of patients also "improved" in the SMC group, giving a NNT of 7 patients required for a single one to experience a small improvement, not all that "promising" when considering that the authors predicted a NNT of 2 in the protocol.

Also noticed Q&A #25 (Why did you omit actigraphy as an outcome measure?):

"Actigraphy is a measure of physical activity, measured by a wrist watch sized accelerometer, worn around the ankle continuously for a week. Before we started the trial, we were advised that the number and scope of the outcome measures were too great and that it might reduce the proportion of participants making it through to the end of the trial. Actigraphy was the obvious measure to reject because of its burden in time and effort required by participants. The patient charity advising us agreed that this would be sensible."

Would that be Action for ME? And so out of all the numerous measures, the "obvious measure to reject" just happened to be one of the few objective measures, one which associates from overseas have found to debunk part of the cognitive behavioural model being tested in the PACE Trial? Nothing suspicious about that, *cough wink*. A 6 minute walking test and a fitness test would have been more of a burden than a wrist watch sized accelerometer worn around the ankle for a week.

#41 is clever (asterisks added): "The *effectiveness* of cognitive behaviour therapy (CBT) as a treatment for CFS does not make assumptions about the nature of the illness." Then they mention CBT for heart disease etc. Plenty of inferred and implied assumptions were made about CFS in the CBT manuals.

 

[Valentijn]

9. Did the trial include patients with ME? All patients met the Oxford criteria for CFS. 67% also met the international (CDC) criteria for CFS, and 51% also met the London criteria for ME. We analysed our data separately for all these groups and found that the pattern of the effects of treatments were similar in all three groups.

Eh, maybe I'll give CBT and GET a try when I don't have ME anymore and I'm just feeling fatigued

But it's interesting to see the statistics, even from an untrustworthy source: half of people diagnosed by the Oxford Criteria don't actually have ME/CFS.

 

[Valentijn]

Would that be Action for ME?

Looks like it:

The trial was subject to independent scrutiny by both a Trial Steering Committee and a Data Monitoring and Ethics committee throughout, and was conducted to the highest standards. A patient charity, Action for ME, helped to design, implement and oversee the study throughout its existence.

Definitely not on MY list of charities to donate to. If I had money and stuff.

 

[Dolphin]

New thread: BACME conference '12-link to short report on Peter White's keynote speech

New thread: BACME conference '12-link to short report on Peter White's keynote speech:
http://forums.phoenixrising.me/show...-short-report-on-Peter-White-s-keynote-report
(discusses PACE Trial results)