'Recovery' from chronic fatigue syndrome after treatments given in the PACE trial

[kaffiend]

Have the means and standard deviations of the outcome measures (the selected ones) for each trial arm been reported anywhere? I don't fully grasp the use of the odds ratios analysis. A cynical reader would note that such a statistic allows for playing with the threshold for recovery. I'm actually surprised by the low levels of "recovery" given the broad entry criteria.

The paper does not seem to report on both of its primary outcome measures described in the PACE Trial Protocol.

5.2.1 Primary objectives
(1) Is APT and SSMC more effective than SSMC alone in reducing (i) fatigue, (ii) disability, or (iii) both?
(2) Is CBT and SSMC more effective than APT and SSMC in reducing (i) fatigue, (ii) disability or (iii) both?
(3) Is GET and SSMC more effective than APT and SSMC in reducing (i) fatigue, (ii) disability, or (iii) both?
(4) Are the active rehabilitation therapies (of either CBT or GET) more effective than the adaptive approach of APT when each is added to SSMC, in reducing fatigue, in reducing physical disability?
(5) What are the relative cost-effectiveness and cost-utility of these treatments?

 

[alex3619]

I just posted a blog on this "recovery" paper, The Bluk Factor.

http://forums.phoenixrising.me/index.php?entries/the-bluk-factor.1357/

In it I replace the word "recovery" with "bluk" and see where this leads me.

 

[Dolphin]

And just to add to Dolphin's list of interesting sites, there is the King's College CFS information for patients and practitioners. (The one for practitioners urges people performing CBT to try to look beyond their protocols and at the patient's actual situation, or 'what some analysts refer to as 'The Real'. A truly radical proposal.)

But even more exciting is the information contained in the advice to patients concerning 'The Physiology of CFS'. This begins with a list of the various physiological changes observed in people with CFS/ME, which leads the reader to think that, finally, they're acknowledging the physiological nature of this illness. But it turns out that every single one of these factors is caused by resting! (or may be, or probably is, or something very similar has been). It is particularly notable that when healthy subjects go to bed for three weeks, these symptoms appear, and when they get up again, these symptoms go away. The message is obvious: resting is a dangerous activity, while the curative potential of 'not resting' has yet to be fully explored. I don't know why this very important information, with its broad implications for both medical practice and everyday life, has been hidden away in an article about CFS.

It states that the article is 'based on the work of Pauline Powell'. I looked her up, and she has, indeed, written a number of articles making these assertions, but I didn't see any research papers.

Pauline Powell was involved in an earlier trial (Powell et al., 2001). This lead to a bigger trial, the FINE Trial (Wearden et al., 2010) which was basically a big flop. There's a thread on the manual here: http://forums.phoenixrising.me/index.php?threads/fine-trial-pragmatic-rehabilitation-manual.21415/ and a thread on the initial presentation here: http://forums.phoenixrising.me/inde...ation-initial-presentation-to-patients.21416/

Orla did an in depth analysis and critique of a similar manual by Pauline Powell in this thread: http://forums.phoenixrising.me/inde...liverpool-cf-cfs-clinic-patient-handout.3066/ with others commenting along the way.

 

[Dolphin]

Have the means and standard deviations of the outcome measures (the selected ones) for each trial arm been reported anywhere? I don't fully grasp the use of the odds ratios analysis. A cynical reader would note that such a statistic allows for playing with the threshold for recovery. I'm actually surprised by the low levels of "recovery" given the broad entry criteria.

The paper does not seem to report on both of its primary outcome measures described in the PACE Trial Protocol.

5.2.1 Primary objectives
(1) Is APT and SSMC more effective than SSMC alone in reducing (i) fatigue, (ii) disability, or (iii) both?
(2) Is CBT and SSMC more effective than APT and SSMC in reducing (i) fatigue, (ii) disability or (iii) both?
(3) Is GET and SSMC more effective than APT and SSMC in reducing (i) fatigue, (ii) disability, or (iii) both?
(4) Are the active rehabilitation therapies (of either CBT or GET) more effective than the adaptive approach of APT when each is added to SSMC, in reducing fatigue, in reducing physical disability?
(5) What are the relative cost-effectiveness and cost-utility of these treatments?

Yes, a lot of the data was in this paper published Feb 2011: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(11)60096-2/fulltext

The other main paper that was previously published is here: http://www.plosone.org/article/info:doi/10.1371/journal.pone.0040808

They're both open access.

 

[Hope123]

The trial was randomized, but not controlled or blinded. Blinding wouldn't work because patients can tell which treatment they're getting. And without blinding patients, I don't see why they'd bother blinding the researchers or assistants involved in treatment. I'm not sure if SMC was provided by unaffiliated local clinics or the trial researchers, but the unblinded researchers are still the ones determining who is meeting the various CFS criteria.

1) In some studies, patients are reminded to not tell or suggest to the assessors (who are blinded) which treatment arm they are in. I've seen research offices that explicitly post such reminders all over the place.

2) Assessors can be blinded even if patients, the therapists, and physicians giving the medical care are not.

3) Different people can be used to select the subjects for the study so blindness is maintained.

4) That is, there is no excuse for not blinding if the original intent was to run a well-designed trial but maybe that's not the intent.

 

[Firestormm]

Firestormm.

Re: severely affected. Obviously not covered by published evidence and the FINE Trial was a failure, but IIRC Wessely has claimed that his clinic get good results in the severely affected too, and the PACE manuals claim that the GET model should apply equally or even more in severely affected patients. Wessely has also said about bedridden patients: "In that kind of disability psychological factors are important and I don’t care how unpopular that statement makes me." So, severely affected patients are only sometimes spared having CBT/GET pushed on them because of the lack of published evidence, not because CBT/GET proponents do not want to treat them or do not actually treat them in practice.

Re: targeting. CBT and GET only help a minority of Oxford criteria patients report modest improvements in some measures but not others. CBT/GET proponents admit to not knowing enough about what predicts and mediates the response, but there is still a wide application, all patients are encouraged to attempt it. Knowing more about what predicts and mediates the response may help to determine which patients need it and what aspects of therapy work best.

Thanks. I seem to recall that NICE talked about using the 'principles' of GET and CBT for the severe patient. Methinks that it is left to local deliverers of healthcare to decide and interpret what this might mean.

I am probably behind in my reading, but thought this was interesting, be grateful for comment:

Patients were classed as recovered if they no longer met several criteria for ill health which were initially used to define eligibility for the trial. These included not suffering from significant fatigue or physical disability, and no longer meeting diagnostic criteria for CFS.

Patients also had to rate themselves as being “much” or “very much better” in their overall health.

The findings showed that those who received CBT or GET, in addition to SMC, were three times more likely to meet the criteria for recovery than those receiving SMC alone or in combination with APT.

Overall 22 per cent of those who received either CBT or GET, in addition to SMC, met the criteria for recovery, compared to eight per cent after APT in addition to SMC, and seven per cent after SMC alone.

There were similar patterns of recovery however CFS was defined, including those who were diagnosed as having myalgic encephalomyelitis (ME), thought by some to be the same as CFS and by others as being different.

http://psychcentral.com/news/2013/02/01/recovery-from-chronic-fatigue-syndrome-possible/51060.html

Can you tell from the paper what the 'several criteria for ill health' were? Does it relate to that SPF-factor thingy? Thanks.

And what's with the ME? Several times I have heard the claim that ME was excluded from this Trial although I never really felt that what I think it was White said should have been taken literally. Any thoughts? Thanks.

 

[Valentijn]

The following paper that is referenced in the recovery paper is available for free at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3020067/

I haven't read it yet but I suspect the points White et al. make that reference this paper are very questionable

Yes, that cited paper is showing that randomly questioned UK residents (almost half of whom reported chronic conditions), have an average of 3-4 symptoms in the past two weeks. The relevant quote from the Recovery paper referencing it is:

In this context it is important to note that recovery does not mean being free of all symptoms; population studies show that the average person in the UK reports a mean of four symptoms in any 2-week period.

Actually it's 3.6 symptoms, so calling it 4 is taking some (significant?) liberties.

The three most common symptoms reported were fatigue, headache and joint pain; symptoms consistent with CFS.

So basically this study is being used to "prove" that having 4 symptoms from a list is normal, even though it's a different list of symptoms. I'm not sure if they're talking about "4 symptoms = normal" to excuse their poor results, or if they're saying that for trial or clinical purposes, patients were considered normal/recovered if they had 4 or fewer symptoms of any criteria. The context is quite odd.

 

[Valentijn]

And what's with the ME? Several times I have heard the claim that ME was excluded from this Trial although I never really felt that what I think it was White said should have been taken literally. Any thoughts? Thanks.

The first condition for being included for the trial was meeting the Oxford criteria, where "fatigue is the main symptom". If PEM or OI or pain is the main symptom, ME patients would be excluded for failing to meet Oxford criteria.

And since PEM is the "main symptom" of the CCC and ICC definitions, it's likely most CCC/ICC ME patients would be excluded. Though if patients aren't familiar with the various criteria, and doctors keep telling they have this chronic fatigue stuff, they might think that their ME symptoms are basically just fatigue.

 

[Purple]

The first condition for being included for the trial was meeting the Oxford criteria, where "fatigue is the main symptom". If PEM or OI or pain is the main symptom, ME patients would be excluded for failing to meet Oxford criteria.

And since PEM is the "main symptom" of the CCC and ICC definitions, it's likely most CCC/ICC ME patients would be excluded. Though if patients aren't familiar with the various criteria, and doctors keep telling they have this chronic fatigue stuff, they might think that their ME symptoms are basically just fatigue.

Not sure if this is how this works in practice. I suspect many people in the trial did have ME. Many patients may not be aware of the different criteria etc.

 

[biophile]

Can you tell from the paper what the 'several criteria for ill health' were? Does it relate to that SPF-factor thingy? Thanks.

And what's with the ME? Several times I have heard the claim that ME was excluded from this Trial although I never really felt that what I think it was White said should have been taken literally. Any thoughts? Thanks.

The several criteria for ill health and recovery are described in the text but the results are in Table 1:

1) A normal range in fatigue and physical function. Despite clarifying in 2011 that this is not recovery, this is now the base or mandatory criteria for recovery (the remaining are optional add-ons). Yes, it is based on the CFQ score for fatigue (<=18 points out of 33, Likert scoring) and SF-36 score for physical function (>=60 points out of 100). Both thresholds here overlap with the entry criteria for "disabling fatigue": >=6 points out of 11 for fatigue in bimodal scoring (18 points Likert scoring can be anywhere between 4-9 points in bimodal score), and <=65 points in physical function.

2) No longer meeting Oxford criteria. The Oxford criteria is ambiguous on what is significant fatigue and disability, so PACE added additional criteria to it here, which is the same threshold used to define "disabling fatigue" at trial entry as mentioned above. In other words, <=5 points in fatigue (bimodal), and >=70 points in physical function is now regarded as a recovery. Note that in the original protocol, <=3 points in fatigue (bimodal, so 4 and above was regarded as abnormal) and >=85 points in physical function, these were the threshold for recovery. Adding this to the normal range criteria barely changes the figures of recovery at all, which seems odd to me. It also leaves open the possibility that without the additional criteria for fatigue and physical function, people who are "recovered" from CFS in PACE could still be diagnosed with Oxford criteria CFS if they went elsewhere and felt that their fatigue and physical function were abnormal for them (e.g. 2/3 of the general population score 95 or 100 points in physical function and would not think 70 is a full recovery).

3) A CGI (clinical global impression) score of 1 i.e. "very much better" or 2 i.e. "much better". Adding this onto the previous two criteria for recovery does significantly lower the recovery rates, which suggests that a significant proportion of the "recovered" participants only felt slightly better from baseline. Also note that in the original definition of recovery, only 1 or "very much better" was regarded as the CGI criterion for recovery.

As for the CDC criteria for CFS and London criteria for ME, it is unsurprising that adding these on does little to the recovery rates since the Oxford criteria is broad and everyone who meets the CDC criteria or London criteria but not the Oxford criteria were already weeded out before the trial began. Also, by changing the symptom requirement from 6 months to one week, they do not appear to have used the CDC criteria properly anyway.

As to whether ME patients were studied, this is more complicated. For the purposes of PACE, ME was not regarded as an "organic brain disease". However, the medical assessment in PACE may have excluded some clinical characteristics which are allowed in ME criteria. Also, PACE only included candidates if their only main symptom was fatigue, which is not a requirement in other definitions. ME focuses on other symptoms more, symptoms which PACE relegate. I would guess some ME patients did get in though, but whether or not the London ME criteria could locate them is another matter.

I think it is misleading for White et al to claim that their report on recovery is both "comprehensive" and "conservative". They go as far as erroneously claiming it was more conservative than the paper by Knoop et al (2007) when it seems to be the opposite, particularly for the physical function score anyway. The peer-reviewers should have at least demanded that the original definition of recovery be added onto the top of the hierarchy for comparison. Why did White et al abandon them? Their claim that "we made the changes before analysis and to more accurately reflect recovery" is unconvincing.

 

[Valentijn]

1) In some studies, patients are reminded to not tell or suggest to the assessors (who are blinded) which treatment arm they are in. I've seen research offices that explicitly post such reminders all over the place.

2) Assessors can be blinded even if patients, the therapists, and physicians giving the medical care are not.

3) Different people can be used to select the subjects for the study so blindness is maintained.

4) That is, there is no excuse for not blinding if the original intent was to run a well-designed trial but maybe that's not the intent.

Based on reading the PACE protocol, the people rating the patients' improvement (SMC doctors) are the same people providing information about the patients adhering to the trial protocols - so they'd have to know what was expected of individual patients.

All 600 randomised participants will receive SSMC. This will be provided by a doctor with specialist experience in CFS/ME.

I wonder if there's a list of the CFS/ME specialist doctors involved in providing the SSMC treatment to trial participants.

The therapists (GET/CBT/etc) also gave multiple CGI ratings after the 14th session (page 54 of protocol).

 

[user9876]

Yes, a lot of the data was in this paper published Feb 2011: http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(11)60096-2/fulltext

The other main paper that was previously published is here: http://www.plosone.org/article/info:doi/10.1371/journal.pone.0040808

They're both open access.

I think there is more detail in the appendix for the lancet paper so it is worth looking at this as well as the main paper

 

[user9876]

2) No longer meeting Oxford criteria. The Oxford criteria is ambiguous on what is significant fatigue and disability, so PACE added additional criteria to it here, which is the same threshold used to define "disabling fatigue" at trial entry as mentioned above. In other words, <=5 points in fatigue (bimodal), and >=70 points in physical function is now regarded as a recovery. Note that in the original protocol, <=3 points in fatigue (bimodal, so 4 and above was regarded as abnormal) and >=85 points in physical function, these were the threshold for recovery. Adding this to the normal range criteria barely changes the figures of recovery at all, which seems odd to me. It also leaves open the possibility that without the additional criteria for fatigue and physical function, people who are "recovered" from CFS in PACE could still be diagnosed with Oxford criteria CFS if they went elsewhere and felt that their fatigue and physical function were abnormal for them (e.g. 1/3 of the general population score 100 points in physical function and would not think 70 is a full recovery).

Under their definition to be diagnosed according the the oxford criteria you need a sf-36 score <= 65 and a a fatigue score of >= 6. But there are also other conditions which are assessed by the research assessors who I believe were unblinded. Hence to not meet the Oxford criteria you could still have a low sf-pf and fatigue score but the research assessor could decide that fatigue is not your main symptom or that you are fatigued less that 50% of the time. Perhaps because you have been told to have a positive attitude or you will never recover. Or it might be that since they included fibromyalga patients they can persuade some that pain and not fatigue is the main symptom.

Hence a patient could enter the trial with a sf-36 score of 65 worsen to 60 and still meet the recovery targets or at least their trial defined ones (just not meeting the oxford criteria).

Basically the only new information in this paper is that they had some unblinded research assessors do an assessment on various criteria and it reports these results the rest was in the lancet paper.

I've realised that the CGI numbers are new as well.

The other piece of information is perhaps that they have no shame in what they define as recovered.

 

[user9876]

The first condition for being included for the trial was meeting the Oxford criteria, where "fatigue is the main symptom". If PEM or OI or pain is the main symptom, ME patients would be excluded for failing to meet Oxford criteria.

And since PEM is the "main symptom" of the CCC and ICC definitions, it's likely most CCC/ICC ME patients would be excluded. Though if patients aren't familiar with the various criteria, and doctors keep telling they have this chronic fatigue stuff, they might think that their ME symptoms are basically just fatigue.

But when the therapists then tell them that they are better and no longer fatigued and just need to have positive thoughts or exercise then fatigue no longer becomes their main symptom again, the no longer meet the oxford criteria and hence they can be labelled as recovered if they were not too severly affected.

 

[Research 1st]

Undoubtedly recovery from CFS without medical treatment is possible if someone really does have a chronic fatigue syndrome that is unexplained by medical science. (This excludes ME CFS sufferers with demonstrable activate immune systems, autonomic disturbance and so on). But it doesn't. Because other still have a CFS you don't have if this is you. And here lies the problem.

The medical profession will be skeptical of any claims of an effective psychological treatment for CFS because their recovery from CFS is based on self reported replenishment of energy levels that cannot be measured scientifically (PACE). So there is no evidence people with unexplained chronic fatigue were organically ill to begin with, before treatment. Any treatment physical or mental.

PACE should have used a treadmill to measure VO2 max rather than a Chalder fatigue questionnaire as then there would be a scientific basis for recoveries demonstrated that are robust. A heart rate monitor could also have been used and a step counter to count how far the patients walked and if their strength increased cardiovascular. In science, there is a fundamental problem of researchers citing a recovery in patients with no explanation how this is possible.

Chalder fatigue questionnaires are not useful for an organic medical disease because the questions can be steered by situations the patient is forced into from the consequences of disease disability, but, the answers (the tick boxes penciled in by patients) still reveal to the data collector what they think the answer reveal but may not.

To give an example a person with ME CFS may tick on a Chalder fatigue sheet they ''don't enjoy things as much as before'' as they did before they got sick. Thus the patient is effectively telling others they are depressed, as they tick more and more boxes and the score increases.

If someone with heart failure is bed ridden, not depressed but isolated by their physical disability they can get a high score on the Chalder fatigue questionnaire too, yet not be depressed. If this person wasn't able to have a diagnostic test for heart failure they may then be misdiagnosed with a depressive disorder and not heart failure and be ushered into psychotherapy.

Precisely what happens to people bed ridden with ME and severe CFS. I notice, that severely affected CFS were excluded from the PACE trial and indeed any studies performed by the Wessely School since time began.

 

[joshualevy]

Lets compare the medical therapy performance of the hugely expensive UK's Wessley School PACE trial vs a biomedical CFS drug trial:
PACE trial = 22% improvement.
Chemotherapy drug trial for CFS = 67% improvement. *

This comparison is completely wrong.
The PACE trial 22% was measuring "recovery" (based on their ideas on recovery), and the Rituximab study 67% was just reporting improvement. Did Fuge and coworkers ever even report on "recovery"? I didn't think so.

Joshua Levy

 

[Valentijn]

This comparison is completely wrong.
The PACE trial 22% was measuring "recovery" (based on their ideas on recovery), and the Rituximab study 67% was just reporting improvement. Did Fuge and coworkers ever even report on "recovery"? I didn't think so.

Well, they did report on SF-36 scores. According to the math done at http://forums.phoenixrising.me/inde...g-the-real-size-of-the-rituximab-effect.1043/ , Rituximab patients on average improved 22 points more than placebo on physical functioning.

In PACE, 11-12% more CBT/GET patients than those in the half-assed control group met SF-36 "trial recovery" standards, which might mean they didn't improve at all, or ever scored lower on the SF-36. We don't really know what the average change was, because the non-scientists running the study won't tell us

 

[Research 1st]

Hello Joshua L nice to meet you. I am not wrong in my view as I am approaching this in an unorthodox way. The Rituximab paper percentages I quoted were indeed based on the patients ideas (beliefs) because the patients after the drug was administered self reported to be less fatigued. This is a psychological measurement of improvement (the way we feel, fatigue is a perception even if organic/caused by cellular processes) about a physical barrier.

Here is what the Norwegian group concluded:

'' Major or moderate overall response, defined as lasting improvements in self-reported Fatigue score during follow-up, was seen in 10 out of 15 patients (67%) in the Rituximab group and in two out of 15 patients (13%) in the Placebo group (p = 0.003). ''

67% of the CFS patients in Fluge et al 2011 reported the drug had a beneficial effect, they felt it did. Those are feelings. No inflammatory profile or any organic biomarker was provided to demonstrate the self reported improvement which is identical to the PACE trial which is also based on patient's behavioural responses that made them perform more activity.

Filling in a questionnaire with a pen about how you feel after a drug is an assessment of beliefs. Doing more because someone coerces you to do more with CBT is also a belief, because the belief alters the behavior = The whole premise of CBT. (You challenge faulty illness beliefs, you do more, the activity increases = PACE).

This doesn't remove the fact that Rituximab likely caused a biological response in a cohort with neuroimmune disease without psychological chronic fatigue, but no evidence was presented in the research paper by the researchers to demonstrate this. It will no doubt be in the future.

PACE thus has a lot in common with other therapies for CFS which may be medical (drug based) but only use fatigue as a biomarker. Hence I proposed this when CFS researchers and oncologists alike are both using fatigue as a biomarker. The results show PACE is inferior to Rituximab. That's all I have to say as that's just my observation and my opinion on the matter. Thank you.

 

[joshualevy]

Well, they did report on SF-36 scores. According to the math done at http://forums.phoenixrising.me/inde...g-the-real-size-of-the-rituximab-effect.1043/ , Rituximab patients on average improved 22 points more than placebo on physical functioning.

You're comparing average peak maximums in the Rituximab study average at end of study for PACE. That's not a fair comparison. You either need to do peak max for both studies or end of study for both.

In PACE, 11-12% more CBT/GET patients than those in the half-assed control group met SF-36 "trial recovery" standards, which might mean they didn't improve at all, or ever scored lower on the SF-36. We don't really know what the average change was, because the non-scientists running the study won't tell us

There is a very nice diagram in the study which shows exactly the average change:
http://www.thelancet.com/journals/l...r2&sectionType=red&hasDownloadImagesLink=true
And even better slide might be here:
http://www.thelancet.com/journals/l...r3&sectionType=red&hasDownloadImagesLink=true

Joshua Levy

 

[Simon]

Got to admire the Chutzpah

Many people have waited a long time to see the recovery stats for the PACE Trial, particularly as Recovery had been defined in an unambigous and very reasonable way; judging by the other results the recovery outcomes would be poor. As recovery is a pretty important outcome it was surpring that recovery rates were not included in the orignal 2011 Lancet paper. Even more surprising, the authors argued they left them out due to lack of space, yet found the space for a raft of post-hoc 'normal' measures.

Scroll forward nearly 2 years and the recovery rates are finally published. Yet they do not use the original recovery measure they clearly specified in their 2007 protocol. In fact, reading the paper, you might assume that they hadn't even considered how to measure recovery until after the Lancet paper was published.

Just check out some of these quotes from the paper [my bolding]

Although the PACE trial found that many patients improved with CBT and GET, the question of how many patients recovered remains unanswered

[because they failed to publish recovery data along with the other trial outcomes in 2011]

...Before we can determine the proportions recovered we need an operational definition of recovery itself.

[They don't: they had published one in 2007]

...The aims of this study were to : (a) define operationalized criteria for recovery on relevant domains

[already defined in the 2007 protocol]

The main limitation of this analysis is the absence of a generally agreed measure of recovery

[The authors agreed and defined one in the 2007 protocol]

We changed some of the thresholds for measuring recovery from those of the original protocols (White et al. 2007); we made the changes before analysis and to more accurately reflect recovery.

[more accurately than the 2007 protocol definition? So here, right at the end of the paper, the authors do mention they changed the protocol Recovery measures; must have slipped their minds before]