New results on Graded Exercise Therapy from recent AfME survey report

[Tom Kindlon]

Results for Graded Exercise Therapy (GET) in Action for ME's 2014 survey report, "Time to Deliver":

23% of the sample reported having tried it. The total sample size for the survey was 2018 so I presume it's 23% of that which gives an approximate sample size of 464.

A little or very helpful: 35%

No change: 18%

Made a bit or much worse: 45%

"Time to deliver" http://www.actionforme.org.uk/Resources/Action for ME/Documents/get-involved/me-time-to-deliver.pdf

The document comments:

"Patients also reported, overall, finding GET less helpful than in 2008 (45% then compared to 35% now). This may be because people with M.E./CFS are self-prescribing GET, rather than working with a specialist therapist (see p 16). We continue to hear positive and negative experiences of GET from people with M.E./CFS and recognise that further investigation is needed to find why it may benefit some people and not others."

My comment:
Page 16 doesn't give any quantitative data on it.

Unless they have data to show this, they shouldn't say the bit I have marked with asterisks.

Their own survey in 2008 found that 31.27% of those who had done GET under an NHS specialist had been made worse with it. This was not statistically different from the overall percentage (31.92% were made worse by GET). The "other" category which would include self-prescribed GET actually had a marginally lower rate of people saying it made them worse (28.93%). (see slide 9 at: http://afme.wordpress.com/5-treatments-and-symptoms/)

When I posted the results on Twitter, some people said that the results categories were overly broad.

Here are the more nuanced results of a survey the ME Association reported on in 2010

(n=906)

Graded Exercise Therapy
Greatly improved: 3.4%
Improved: 18.7%
No change: 21.4%
Somewhat worse: 23.4%
A lot worse: 33.1%

 

[SOC]

I see no point in quibbling details of the report when we don't know the actual illness or condition from which the 2018 patients were suffering. Unless they used CCC or ICC (which I seriously doubt) to diagnose the patients, the sample set could very easily be more than 50% people with the symptom chronic fatigue, which can be caused by a number of conditions exercise would be expected to improve.

To my mind, there's only one response to this kind of research -- Garbage In, Garbage Out.

 

[Tom Kindlon]

I see no point in quibbling details of the report when we don't know the actual illness or condition from which the 2018 patients were suffering. Unless they used CCC or ICC (which I seriously doubt) to diagnose the patients, the sample set could very easily be more than 50% people with the symptom chronic fatigue, which can be caused by a number of conditions exercise would be expected to improve.

To my mind, there's only one response to this kind of research -- Garbage In, Garbage Out.

Yes, but I think other forms of chronic fatigue would be less likely to report adverse effects from exercise programmes so I think such data can be useful to show exercise programs can cause problems rather than being perfectly safe, as some people like to present them.

 

[SOC]

Yes, but I think other forms of chronic fatigue would be less likely to report adverse effects from exercise programmes so I think such data can be useful to show exercise programs can cause problems rather than being perfectly safe, as some people like to present them.

Good point. That fact that there is a significant number of patients reporting adverse affects from exercise does indeed suggest that there is a problem with exercise for some fatiguing conditions. We might be able to use that information IF we can make it clear that the original sample set is not made up of patients with ME/CFS, but patients with chronic fatigue. Then the question arises, what is different about those patients who have an adverse effect from exercise? I can guess -- that was the subset that has ME rather than some other fatiguing condition.

 

[Tom Kindlon]

Research33 funded by Action for M.E. and the National Institute for Health Research estimates lost earnings of more than £102 million a year to the UK economy as a result of M.E./CFS.

33 Crawley, E. et al. (2011) The impact of CFS/M.E. on employment and productivity in the UK: a crosssectional study based on the CFS/M.E. National Outcomes Database. BMC Health Services Research, 11, p 217

This is not worded well to put it mildly.

This is the paper being referred to:

The impact of CFS/ME on employment and productivity in the UK: a cross-sectional study based on the CFS/ME national outcomes database.
Collin SM, Crawley E, May MT, Sterne JA, Hollingworth W; UK CFS/ME National Outcomes Database.
BMC Health Serv Res. 2011 Sep 15;11:217. doi: 10.1186/1472-6963-11-217.

The estimate of £102 million is based on 4424 adults:

Total productivity costs among the 2,170 patients due to discontinuation of employment in the years preceding assessment by a specialist CFS/ME service (median duration of illness = 36 months) were £49.2 million. Our sample was equivalent to 4,424 UK adults accessing specialist services each year, representing productivity costs to the UK economy of £102.2 million. Sensitivity analyses suggested a range between £75.5-£128.9 million.

The total prevalence of "CFS/ME" in the UK is many times higher than 4424.

 

[Valentijn]

If their data is similar to the earlier survey from MEA, then a very small percentage are finding GET "very helpful" while about 10 times as many are "much worse" as a result of it. Since AFME isn't showing more detailed data, and are pretty pro-establishment, I'm betting the sub-categories would look even bleaker than the "helpful" and "worse" combined categories.

I think the presumed increase in the "a little helpful" group might be due to GET now typically incorporating pacing as a preliminary step prior to starting the graded exercise. Some patients will stop after learning pacing, or that will be all that's taught by a clinic even though they still call it "GET". Hence they'll never get to the graded exercise stage, and will thereby avoid the harm it usually causes.

 

[A.B.]

More likely, people who report getting better with GET simply don't have the same type of problem seen in ME/CFS patients diagnosed with stricter criteria. The evidence for impaired recovery from exertion is there.

If the Wessely school's goal was obfuscation of the truth, they could hardly do a better job. Looking at how many of their key figures have ties to insurance companies, that may actually be their goal.

Psychiatry has a history of broadening diagnostic criteria to an unreasonable degree in other contexts, and presumably for different reasons. Overmedication due to lax diagnostic criteria is a known problem. Most people that take psychotropic medication don't actually need or benefit from them.

All of these are symptoms of diagnostic methods that are dysfunctional because they rely on questionnaires and lists of vague symptoms open to interpretation, rather than on biomarkers. Labeling a symptom, such as depression, as illness itself is the wrong approach as well because the assumption that there's one underlying cause is most likely completely wrong, so searching for it cannot succeed. This is similar to what has been done with CFS and focusing on the subjective symptom of fatigue. Fever was also once seen as illness, rather than a symptom of illness, and goal of treatment was to reduce fever which probably often made things worse. What's needed is a divide and conquer algorithm (http://en.wikipedia.org/wiki/Divide_and_conquer_algorithms) which in practice would be continuing to divide patients into subgroups until clear patterns emerge. Note how the Wessely school has done the opposite.

That isn't to say paying attention to subjective symptoms has no place. It does, but diagnoses should be supported by solid biomedical research.

 

[Tom Kindlon]

Sonya Chowdhury, CEO of Action for ME, has taken the time to reply to comments of mine that I posted on various forums.

As you can see, she says:

"I am very happy for you to share any or all of this email should you wish to."

So that is what I'm doing.

I've post my response underneath

Hi Tom - I thought it would be worth providing a little more detail regarding the Time to Deliver report in response to your posts/tweets.

I have attached a copy of the initial survey. As you will see, question 32 asks people to rate 'very helpful' and 'helped a little'. The reference in our report combined these two responses. This was then correlated with the data from people about whether they had received input from a specialist service in the last 5 years alongside additional comments from people which provided the basis for the statement about self-prescription of GET.

We of course recognise that the scaling used is subjective and not standardised. We have not, in any way, intended to spin the results and I would be very happy to respond to any further questions or points that you have. I am copying in my colleague, Clare (who is on holiday this week) as she has been leading this work and can also respond to any queries. We will be producing more detailed reports once we have had a chance to analyse the data further and have input from experienced researchers and students (who are being overseen by Julia Newton). This will obviously take some time but we are happy to discuss any aspects of the report at any stage.

I hope this helps clarify the concerns that you have vocalised and I am very happy for you to share any or all of this email should you wish to.

Warmest wishes -----

My response to Sonya:

Dear Sonya,

Thank you for taking the time to write.

The report says: "Patients also reported, overall, finding GET less helpful than in 2008 (45% then compared to 35% now). This may be because people with M.E./CFS are self-prescribing GET, rather than working with a specialist therapist."

I will await to see the further analysis but I would be very surprised if none of the 45% who reported GET made them worse were under a specialist therapist which is what the wording of your report suggests.

I can't remember all the details of your 2011 survey but certainly neither your 2008 survey didn't find this: It found that 31.27% of those who had done GET under an NHS specialist had been made worse with it. This was not statistically different from the overall percentage (31.92% were made worse by GET). The "other" category which would include self-prescribed GET actually had a marginally lower rate of people saying it made them worse (28.93%). (see slide 9 at: http://afme.wordpress.com/5-treatments-and-symptoms/)

The 2003 survey used very small numbers but the quantitative data didn't suggest that the people who got worse had solely not done it under a professional, yet that was the spin Chris Clark put on it.

I'm frustrated by this as sometimes when I have been discussing in journals the adverse reactions that people have reported with GET in surveys, I have had to deal with people not accepting this view quoting Action for ME statements and survey analysis on the issue.

Graded Exercise Therapy as defined by Peter White is about breaking the link between symptoms and activity levels: you don't choose how much you do based on your symptoms. So if you increase your activity levels, and suffer increased symptoms, you are supposed to maintain the same level of activity levels for a week or two. I have no problem believing that such an approach can be dangerous/cause relapses. So I don't accept that GET under a professional is necessarily safe if properly applied. I think a lot of people don't stick to the rules, either on their own or under a professional, and hence there are fewer people than there might be having problems i.e. in effect, not sticking to the rules and rationale for GET is safer I believe.

Feel free to forward this on to whomever you want.

Best regards,

Tom

 

[Esther12]

Thanks.

Good that she's happy to have a public discussion about it... will be interesting to see where it goes.

 

[Firestormm]

The assumption is ever that GET under 'specialists' is the same as that prescribed in PACE. And it ain't. That is to say, 'we', have no evidence to say that it is - not for those who answer these surveys. I am never really sure where these discussions are meant to be taking us - they seem to be circular.

Are we trying to say that 'specialists' are all as bad as the manual for GET and the results from PACE? Or are we attempting to say, that more care needs to be imposed by 'specialists' in terms of personalising their approach and recognising when a patient cannot 'push through' with any incremental exercise plan? Or simply that GET - a la Peter White - should be abandoned for everyone with our diagnosis and that these surveys can be used to support his argument?

NICE recommends a management strategy based on personalisation: how - in practice - can a personalised approach ever be measured collectively: especially in any 'patient survey'?

I was taught GET, it wasn't helpful or appropriate, so we switched to a programme of Graded Activity, which I try and follow to this day and in my own mind is not dissimilar to pacing.

If I answered a survey, I would say I had found GET 'unhelpful', but would not have then been able to indicate why, or that I had found something that is structured and is helpful. And my answer would have no bearing on PACE or Peter White.

The GET I was taught, was also personalised, and whilst it did suggest increasing walking over weeks, it was only if I could manage to do so. If I was experiencing a setback, I would begin again. There was no 'pushing through'. And that was from (two) NHS specialists as was the Graded Activity Plan.

To my mind, it isn't exercise or activity per se that's 'bad' for people with ME, it is the means by which any plan is structured and then enforced that can lead to hardship - but only if there is no built in mechanism for flexibility.

If Peter White - and anyone else - believes that GET pursued vigorously and without regard for the patients state of health is the way to recovery: then I can't see how you can disprove his/their theory from these patient surveys - or even imply that GET a la Peter White is reflected by these results let alone has any bearing on PACE.

I am also not convinced that comparison to previous patient surveys is especially helpful in this regard. I think these kind of surveys need to be more inciteful, more detailed, and ask more about what it was that people found helpful or not helpful. Then we might be able to turn round to NICE and say, 'This does not appear to be as appropriate or as useful for people with this diagnosis...'

But that then still leaves us with the big hole: What else might be? And how can we demonstrate it? No ME charity would be willing (neither would it's supporters), to invest funds in a trial to prove something akin to the various 'pacing' prescriptions in favour, are 'more effective' than those therapies in PACE.

What might work, is for an ME charity, to review it's results and work with a specialist NHS centre, to produce a more acceptable management strategy: one that can be measured locally or something and then be studied and published. It might then influence NICE or other NHS provisions etc. If not an NHS centre then one of the private ones.

These surveys are great. But where do they get us? Need something to challenge to demonstrate an alternative more so than we have at present...

 

[user9876]

The issue is that the quantitative information is not given in the survey but they make a sweeping statement that the change is down to people self prescribing GET. Sonya's comment suggests that there is some correlation between different questions but I think they need to give the numbers here to back up the point.

What they are saying however, is that people are reporting that they find GET less helpful than they did in a previous survey so as well as looking at the correlations in the current survey they need to compare this to the previous survey. I did once do some work on a (business) survey that was carried out for many years and they took a great deal of effort to keep most of the questions the same so they could do such comparisons.

I think Tom would really like a different question answered which is around the risks associated with get. Given they are able to correlate the data the question here becomes one of how many people who had GET from a specialist reported being made much worse and how many people being treated by a specialist reported being made a little worse. This is important information since it looks at the risks - as I understand it doctors and other medical professionals have a duty to discuss risks and benefits with patients.

@Firestormm I think the issues of personalization and variation in delivery are different but play into the overall risk and benefit equation. I'm a big believer in having a strong process for delivery of a service (medical protocol) so that you get consistency but that process should be responsive and contain feedback loops (personalization?). Where the treatment is basically left up to the skill of an individual therapist it becomes impossible to generalize risks and benefits as we don't know what is happening.

 

[WillowJ]

there is a qualitative survey on exercise programmes:
http://forums.phoenixrising.me/inde...fs-probed-via-online-survey-pheby-2013.23613/

Positive experiences of rehabilitation therapies include supportive communication with a therapist, treatment which included routines and goals, and value attached to baselines and controlled pacing.

By contrast, factors leading to negative experiences include poor communication and support, conflict in beliefs about CFS/ME and rehabilitation, pressure to comply with treatment, worsening of symptoms, baselines experienced as unsustainable, and feeling blamed for rehabilitation not working.

 

[WillowJ]

Jason did do a head-to-head study of CBT (with exercise component), cognitive therapy (coping skills and pacing), anaerobic exercise (a la Workwell, Pacific Fatigue Labs at the time), and relaxation (originally designed as a control group, but they found it helped with pain & other symptoms).
http://www.cfids-cab.org/cfs-inform/CFStreatment/jason.etal07.pdf

The coping skills and pacing outperformed the other interventions in the study.

Jason has other studies on envelope theory showing that most patients are operating outside of their "energy envelope" and that they improve when they are able to cut back on activity to stay within illness-imposed limits.

 

[Esther12]

The assumption is ever that GET under 'specialists' is the same as that prescribed in PACE. And it ain't. That is to say, 'we', have no evidence to say that it is - not for those who answer these surveys. I am never really sure where these discussions are meant to be taking us - they seem to be circular.

Are we trying to say that 'specialists' are all as bad as the manual for GET and the results from PACE? Or are we attempting to say, that more care needs to be imposed by 'specialists' in terms of personalising their approach and recognising when a patient cannot 'push through' with any incremental exercise plan? Or simply that GET - a la Peter White - should be abandoned for everyone with our diagnosis and that these surveys can be used to support his argument?

Lots of people try to claim that properly conducted GET cannot harm patients. Lots of people also make misleading claims about the efficacy of GET.

In order for patients to be able to give informed consent for treatments, and make informed decisions about their healthcare, they need to be aware that many patients do report being made worse by GET and have access to more reliable information from trials. If AfME is not pushing for the release of PACE's protocol defined outcomes, and is attempting to unreasonably downplay patient reports of being harmed by GET, then this is a bad thing.

I think that this should just be about allowing patients access to honestly presented data, and should not require any promotion of an alternative management plan. Why assume patients should be managed? I don't want to be, and never would have agreed to being so if I had been allowed access to honest information about GET and it's efficacy when I first fell ill.

 

[Tom Kindlon]

Continuing correspondence with Sonya Chowdhury (CEO, Action for ME) on interpretation of GET survey results
---------
For anyone interested, Sonya Chowdhury (CEO of Action for ME) has replied to my last message. She said she would be happy for me to share her response.

Extracts from my last message have ">" in front of them.

----------------
>Dear Sonya,

>Thank you for taking the time to write.

>The report says:
>"Patients also reported, overall, finding GET less helpful than in 2008 (45% then compared to 35% now). This may be because people with M.E./CFS are self-prescribing GET, rather than working with a specialist therapist.”

>I will await to see the further analysis but I would be very surprised if none of the 45% who reported GET made them worse were under a specialist therapist which is what the wording of your report suggests.

This wasn’t the intention with the wording and we know that GET (under a specialist therapist) makes some people worse. The report was read by a four people before publication, within the ‘patient’ community, and this wasn’t picked up as us being misrepresentative with the wording. I do see your point and will look to get the wording changed.

>I can’t remember all the details of your 2011 survey but certainly neither your 2008 survey didn’t find this:
It found that 31.27% of those who had done GET under an NHS specialist had been made worse with it. This was not statistically different from the overall percentage (31.92% were made worse by GET). The "other" category which would include self-prescribed GET actually had a marginally lower rate of people saying it made them worse (28.93%). (see slide 9 at:
http://afme.wordpress.com/5-treatments-and-symptoms/)

>The 2003 survey used very small numbers but the quantitative data didn’t suggest that the people who got worse had solely not done it under a professional, yet that was the spin Chris Clark put on it.

I absolutely sure you there was no attempt to spin any of the results; in fact we worked hard to look at even statement from all angles to prevent this and to use the statistics. Clearly, we did not succeed on this point. I would be very happy to engage with you directly and discuss elements of our report. We were very clear about our purpose and are not, despite the opinion of some, out to purely prove or support the statements/theories/beliefs of one or two individuals. Our purpose was to undertake a survey (not ‘scientific’ we recognise) to help highlight people’s experiences which then provides an evidence base for us to use in out informing and influencing work alongside highlighting any positive changes (of which we know there is very little) or good practice. I will also listen to the criticism levied; that is how we learn and develop what we do. What is hard is when people do not take the time, as you are doing now, to highlight specifically what the issue is so that we can address it. With the best intentions and will, we will sometimes miss things or see things from a different perspective. I appreciate you taking your time to highlight this. We have a team of volunteers working with us to do the same – you are always welcome to input!

> I’m frustrated by this as sometimes when I have been discussing in journals the adverse reactions that people have reported with GET in surveys, I have had to deal with people not accepting this view quoting Action for ME statements and survey analysis on the issue.

And I regularly hear of people who have adverse reactions as well as those who say it has helped (albeit sometimes not implemented in the way it was in the PACE trial) so I wouldn’t in any way want to suggest otherwise.

> Graded Exercise Therapy as defined by Peter White is about breaking the link between symptoms and activity levels: you don’t choose how much you do based on your symptoms. So if you increase your activity levels, and suffer increased symptoms, you are supposed to maintain the same level of activity levels for a week or two. I have no problem believing that such an approach can be dangerous/cause relapses.

We also know that this approach is not the GET applied in all clinics whereby if symptoms increase, the individual is told to immediately reduce activity, not persist - I think is a very interesting difference

> So I don’t accept that GET under a professional is necessarily safe if properly applied. I think a lot of people don’t stick to the rules, either on their own or under a professional, and hence there are fewer people than there might be having problems i.e. in effect, not sticking to the rules and rationale for GET is safer I believe.

> Feel free to forward this on to whomever you want.

> Best regards,

> Tom
---------

I think I will leave this for now. I have had another letter accepted for publication on this issue - hopefully it will be out soon (it was accepted for publication last October!)

 

[Esther12]

They need your proof reading!

I do worry that even if there is the start of an attempt to turn things around, they are so full of people that don't see the problems with their approach that not much can really change. I still don't get any sense from them that they really understand what went wrong with PACE.

PS: would be good if they did change the part you pointed out Tom.

 

[Firestormm]

Action for ME Facebook:

5 June 2014

"We've published an updated version of M.E. Time to deliver, our new report based on a survey of more than 2,000 people with M.E.

This is because we were asked to clarify our statement regarding our survey findings on graded exercise therapy (GET).

“I would like to absolutely assure anyone with concerns that there was no attempt to spin any of the results,” says our CEO Sonya Chowdhury. “Our survey found that around one in five people with M.E/CFS had tried GET, and of these a third said they found it helpful or very helpful.

“Conversely, nearly half said it made them a bit or much worse. We acknowledge that this can occur even when GET is delivered by a trained and experienced professional.

“We continue to hear positive and negative experiences of GET from people with M.E./CFS. We recognise that further investigation is needed to find why it may benefit some people and not others.”

We are grateful to patient advocate and M.E. campaigner Tom Kindlon for his involvement in updating our survey report, and for all the other comments we have received about M.E. Time to deliver. We welcome feedback from people affected by M.E. and others, which offers us a great opportunity to ensure the transparency of our work.

You can find the updated version at http://www.actionforme.org.uk/get-i...the-facts-spread-the-word/me-time-to-deliver/

 

[Min]

There seems to be quite a number of people who have been made so very ill by GET that they are unable to fill out polls. They have, as usual, not been considered.

 

[Valentijn]

They're still screwing around with the numbers. Only a small percentage found GET "very helpful", with a rather large percentage finding it "very harmful". Lumping together "helpful" and "very helpful" disguises that most of the helpfulness was marginal, just as combining "harmful" and "very harmful" disguises that most found it "very harmful", not "harmful".

 

[Tom Kindlon]

They're still screwing around with the numbers. Only a small percentage found GET "very helpful", with a rather large percentage finding it "very harmful". Lumping together "helpful" and "very helpful" disguises that most of the helpfulness was marginal, just as combining "harmful" and "very harmful" disguises that most found it "very harmful", not "harmful".

Do we know that? I don't recall seeing a breakdown for this survey - are you thinking of another survey?