Sasha
Senior Member (Voting Rights)
That's ironic, because every time I use voice recognition software, it turns rude words into polite ones.
-
New Forum Software Installed (Still a few issues but reopening the forum) More details.
2025: The 2019/24 Cochrane Larun review Exercise Therapy for CFS - including IAG, campaign, petition, comments and articles
- Thread starter S4ME News
- Start date
Utsikt
Senior Member (Voting Rights)
Thank you for dealing with all of my questions!
Wouldn’t the likelihood depend on the selection method, the number of participants, and the distribution of humans in the pool of animals that the participants were selected from? I.e. doesn’t the the number and ratio of humans affect the likelihood that the results apply to that subset?
You've just described in your last paragraph exactly my 6 week experience of CBT and Activity management in a clinic, diagnosed under the 2007 NICE guideline and PACE trial results in 2011. My form filling was worse on fatigue but I felt better on everything else. I had been given tools to help me manage my illness going forward, or so I thought. A week after the end I collapsed and been severe ever since. A six month following SF36 was sent to me, but I wasn't well enough to complete it. That was it, lost to follow up, but hey, on the subjective forms they held I was a success and have been for the last 13 years!
Added: The review is causing harm at treatment level globally.
What I haven't seen in the review, unless I've missed it, is the outcome measures reviews which Crawley and Crawley and White did in 2013 on 5 and 6 treatment centres respectively in the UK, showing deterioration in a proportion of patients.
Hence the new PROMS?
Also, regarding criteria, I noticed a couple of months back that the Review had been cited in at least two US trials concerning GWI. What's included in that criteria? Because in the UK, last time I checked, that illness was on 'watch and wait'.
Been trying to catch up with the last 2 days of post on here and failing miserably, so apologies if this is a replication.
Last edited:
Yes, it could be useful to study "different sorts of asthmatics..." The key there is they all have asthma.
But we are talking about a group of "different sorts of chronic fatigue". IMO, that is no longer - and I think never was - an appropriate group to stuff ME/CFS into. Especially with what we know about PEM. Continuing to claim that studies with Oxford for instance can appropriately be used to
I completely agree that a poor trial design with respect to outcomes and controls is a disaster.
BUT a poor trial design with regards to patient selection is also a disaster when a trial selects one group of patients who have not been demonstrated to have the condition for which the intervention is being evaluated don't actually have that condition.
Yes, it may be legitimate to generalize from a wider cohort to a narrower one. But when the wider cohort selection criteria are so broad as to be effectively meaningless from a clinical perspective - e.g. nonspecific medically unexplained chronic fatigue of any kind - then one has to question the appropriateness of generalizing from that wide cohort to ALL the clinical populations that might be stuffed into it.
Utsikt
Senior Member (Voting Rights)
Just to clarify, this is what I’ve been trying to say. Medfeb said it a lot better. @Jonathan Edwards
I don’t know if it’s correct, but it’s my current understanding.
Jonathan Edwards
Senior Member (Voting Rights)
There are a whole lot of different likelihoods to consider here so it gets complicated. The first thing is the likelihood that the results of a trial can be taken as showing that it is more probable than not that at least some people treated in the future will benefit. Subsets do not come into that much. There is a similar estimation of likelihood of harm.
But a likelihood of 95% that at least some will benefit may go along with a likelihood of benefit for an individual of only 30%, or even less. That likelihood may well be affected by features of the individual that put them in a subset - like PEM. But the PACE trial selection criteria are justified on the basis that if it gave a positive result with 95% confidence that would be a starting point for treating anyone with the criteria if there were no obvious eta features.
The problem then is what features one thinks can legitimately be used to change one's assessment of likelihood of benefit. Say a trial shows that a particular dose of methotrexate shows a good risk/benefit ratio in treating rheumatoid arthritis. Should one then say that this cannot be applied to people with red hair or who are Japanese because there was only one person in the study with red hair and no Japanese.
The answer is quite interesting because we have no reason to think people with red hair would be different and after decades of usage we still don't, but we had no reason to think Japanese people would be different but after decades of use we have discovered that the drug is much more likely to be toxic because Japanese people have different liver enzymes.
If we come to exercise intolerance we might say that (if PACE had had a positive result) an intelligent doctor would be less confident of a good result in someone with PEM from GET, since they are exercise intolerant. That still leaves the PACE trial as legitimate but allows for intelligent interpretation, which is always going to be needed. (You tend not to use injection therapies for people with needle phobias.)
But, interestingly, if you follow the PACE authors' logic, people with PEM ought to be more likely to benefit, because the idea was to desensitise them to exercise intolerance and a reconditioning cycle. People without exercise intolerance shouldn't benefit.
The bottom line is that the choice of entry criteria for PACE was entirely legitimate. Whether PACE could be used as evidence for likely benefit for people with PEM is more debatable but it is and unknown and as such you cannot throw out PACE on that basis. But what you can do is throw out PACE because there were no eggs in the box after all.
But a likelihood of 95% that at least some will benefit may go along with a likelihood of benefit for an individual of only 30%, or even less. That likelihood may well be affected by features of the individual that put them in a subset - like PEM. But the PACE trial selection criteria are justified on the basis that if it gave a positive result with 95% confidence that would be a starting point for treating anyone with the criteria if there were no obvious eta features.
The problem then is what features one thinks can legitimately be used to change one's assessment of likelihood of benefit. Say a trial shows that a particular dose of methotrexate shows a good risk/benefit ratio in treating rheumatoid arthritis. Should one then say that this cannot be applied to people with red hair or who are Japanese because there was only one person in the study with red hair and no Japanese.
The answer is quite interesting because we have no reason to think people with red hair would be different and after decades of usage we still don't, but we had no reason to think Japanese people would be different but after decades of use we have discovered that the drug is much more likely to be toxic because Japanese people have different liver enzymes.
If we come to exercise intolerance we might say that (if PACE had had a positive result) an intelligent doctor would be less confident of a good result in someone with PEM from GET, since they are exercise intolerant. That still leaves the PACE trial as legitimate but allows for intelligent interpretation, which is always going to be needed. (You tend not to use injection therapies for people with needle phobias.)
But, interestingly, if you follow the PACE authors' logic, people with PEM ought to be more likely to benefit, because the idea was to desensitise them to exercise intolerance and a reconditioning cycle. People without exercise intolerance shouldn't benefit.
The bottom line is that the choice of entry criteria for PACE was entirely legitimate. Whether PACE could be used as evidence for likely benefit for people with PEM is more debatable but it is and unknown and as such you cannot throw out PACE on that basis. But what you can do is throw out PACE because there were no eggs in the box after all.
Jonathan Edwards
Senior Member (Voting Rights)
I don't buy it. Remember that Simon Wessely said that yes there are very ill disabled people who he loved to look after with chronic fatigue syndrome but they are different from people who think they have ME because having ME is just thinking you have ME. Admittedly, the logic starts to fall apart if you treat people with chronic fatigue syndrome as if they have unhelpful beliefs and a reconditioning cycle, but from listening to Peter White and other psychiatrists talk I am fairly sure that most of their patients had ME/CFS.
Most people who are bad enough to need to see a specialist - unable to work, housebound or whatever - with long term fatigue type symptoms probably have ME/CFS I suspect. I am very suspicious of the suggestion circulating in the ME/CFS advocacy community that there are masses of people around with just chronic fatigue who are likely to do fine on GET and shouldn't gum up research into ME (usually without the CFS). I see no evidence for that.
And if you are going to argue that people with ME/CFS do not fall under the wide umbrella of Oxford or Fukuda it needs to be clear why. As far as I know most people with ME/CFS do fall under Oxford and Fukuda. The do not fall under criteria for asthma or rheumatoid arthritis, so we can't say these criteria are meaningless. They do not overlap with any other medical conditions much. (The other conditions require other specific features.)
For all their faults I suspect that people like Peter Denton White spent years doing their best to look after people with ME/CFS. It is just that they did not apply clear reasoning and got stuck with doing things that don't work. It doesn't help if ME/CFS advocates try to counter this with equally poor reasoning.
The problem with your hypothetical example is that you are assuming that the researchers could easily tell the difference between a human and a rat, and that that distinction makes a difference with respect to their response to treatment. An alien from an advanced civilisation might be interested to know if all of these organisms require oxygen to breathe. Their experiment might find that, yes, they do. They have learned something useful about this group of creatures that they can generalise to all of the creatures that they group together.
As Jonathan said, asthma is a construct, just as chronic fatigue is. Your opinion is that the chronic fatigue construct is not particularly helpful, that it is not appropriate to put ME/CFS into it, is only an opinion, even if a reasonable one. Other people think that ME/CFS can be put in that group, and that it can also include people with post-viral fatigue symptoms with no obvious PEM. Given how difficult it is to know if someone has PEM, that is not a completely unreasonable stance to take. It's okay to ask 'does this treatment (which might be a drug) fix people with, for example, post-viral fatigue symptoms?'.
If you do your trial well, you will get a result that gives you some information, maybe no, it helps no one, maybe it helps the majority, maybe there is one sub-group that it might be helping and one sub-group that seems to get worse.
It is your opinion that an Oxford criteria is so broad as to be meaningless from a clinical perspective. Even with the knowledge we have today, it may not be so. If you collect up people who meet the Oxford criteria, with persisting fatigue, with their illness onset immediately after a specific infection, maybe SARS-CoV-2, you might find that a treatment helped nearly all of them. It could be a reasonable thing to try, especially if you record a range of characteristics about them such as whether they report exertion-related crashes and do some post hoc analyses.
What I'm trying to get across is that using a selection criteria that does not require PEM is not a flaw in itself and it could still tell you something useful about ME/CFS.
Last edited:
Utsikt
Senior Member (Voting Rights)
I still don’t understand how you can use a CF study to say something in general about ME/CFS unless you’ve 1) proven that you have a representative sample of the CF population, and 2) the ME/CFS subset is large enough to have the required statistical power.
The fact that I don’t understand doesn’t mean that I disagree. I don’t have an opinion yet, I’m just struggling to understand. Maybe we should just leave it here.
That might say more about the suitability of GET for CF than the ratio of disabling non-ME/CFS CF to ME/CFS.
For all I know, the X-factor(s) for ME/CFS might also be responsible for some non-PEM CF.
The fact that I don’t understand doesn’t mean that I disagree. I don’t have an opinion yet, I’m just struggling to understand. Maybe we should just leave it here.
That might say more about the suitability of GET for CF than the ratio of disabling non-ME/CFS CF to ME/CFS.
For all I know, the X-factor(s) for ME/CFS might also be responsible for some non-PEM CF.
Jonathan Edwards
Senior Member (Voting Rights)
Statistical power does not come into the extrapolation from the wider set to the narrower one because there is nothing to do statistics on in this situation. Other than what the PACE authors did, which was a post-hoc sub analysis that showed no difference between people with PEM and others. But if only one person in PACE had red hair there are no statistics that tell you how legitimate or otherwise it is to apply the overall result to people with red hair.
You aren't looking for confidence limits. You are looking for balance of probabilities and that is a different analysis.
rvallee
Senior Member (Voting Rights)
Notable that since this follows, this has been Garner's most recent promotional tour. Out of some trial on LC that doesn't actually say that, but he's been making this claim loudly, that those with PEM benefit the most, for a few months.
And I think this must be why decision makers in Cochrane don't think that there are any grounds for withdrawal of the review. They just don't believe that there really is any risk of harm and they still believe, regardless of the lack of evidence, that exercise helps.
Perhaps they really do believe that we are all a bunch of hypochondriac whiners and that stories of deterioration after GET are just made up. Perhaps they do believe that certain ME/CFS charities have whipped up this global myth that exercise is harmful. Perhaps they believe that telling someone that they could be well if they wanted to be is not psychologically damaging, because they really believe that we could be well if we wanted to be.
Why MEAction is not be pushing for withdrawal, I do not know. Perhaps MEAction believes that it is impossible to make the case that GET might cause harm sufficiently well?
@Medfeb, do you know why people seem to be settling for a note on the review rather pushing for withdrawal?
This is not just my opinion. The IOM and NICE both require PEM for a diagnosis of ME/CFS.
Yes, it may be useful to compare people with chronic fatigue (as in the example you gave) to people with ME/CFS. But the problem is when all of those people, regardless of what they have, are labelled ME/CFS. That creates a body of evidence that can't be interpreted.
We are just going to have to agree to disagree.
Edited to put the end quote in the right place
Last edited:
ME/CFS Skeptic
Senior Member (Voting Rights)
There are a couple of studies that looked at people referred to specialist centres with a suspicion of ME/CFS and around 50% did not meet ME/CFS criteria (it did not seem like solely a problem of the criteria). The non-ME/CFS group included conditions like depression, sleep disorders, overtraining syndrome, burnout etc.
Newton JL, Mabillard H, Scott A, Hoad A, Spickett G. The Newcastle NHS Chronic Fatigue Syndrome Service: not all fatigue is the same. J R Coll Physicians Edinb. 2010 Dec;40(4):304-7.
Devasahayam A, Lawn T, Murphy M, White PD. Alternative diagnoses to chronic fatigue syndrome in referrals to a specialist service: service evaluation survey. JRSM Short Rep. 2012 Jan;3(1):4.
Darbishire L, Ridsdale L, Seed PT. Distinguishing patients with chronic fatigue from those with chronic fatigue syndrome: a diagnostic study in UK primary care. Br J Gen Pract. 2003 Jun; 53(491): 441–445.
Mariman A, Delesie L, Tobback E, Hanoulle I, Sermijn E, Vermeir P, Pevernagie D, Vogelaers D. Undiagnosed and comorbid disorders in patients with presumed chronic fatigue syndrome. J Psychosom Res. 2013 Nov;75(5):491-6.
Johnston SC, Staines DR, Marshall-Gradisnik SM. Epidemiological characteristics of chronic fatigue syndrome/myalgic encephalomyelitis in Australian patients. Clin Epidemiol. 2016 May 17;8:97-107.
and the obvious conclusion to that being that when we die, we really did bring it upon ourselves through behaviour and chronic deconditioning, as the current 2024 (re-dated 2019 review) implies, because the therapy causes no harms at all, and that's why there'd be no need for a reporting mechanism.
In the same way that patients are not allowed to see their mental health records, so there's no requirement to publicly explain all that. All one needs to do is republish with the date 16 December 2024 and stay silent.
Utsikt
Senior Member (Voting Rights)
Iirc, the studies that have looked self-reported ME/CFS status in combination with a symptoms survey, about 50 % of the self reported patients do not fulfill the criteria.
The main causes are probably a poor understanding of what ME/CFS actually is, from either the patients or their doctors.
But, we are not talking about what the best definition of ME/CFS is. That's a different question. We are talking about whether the BPS researchers did something fundamentally unscientific by doing a trial using a selection criteria that did not require PEM. And whether a review can be reasonably made of the studies that used a selection criteria that did not require PEM.
And the thing is, doing that is not unscientific, and that is why making the main criticism of those trials and the review about diagnostic criteria is wrong.
What is unscientific is using a poor trial design that does not have the capacity to adequately assess the hypothesis that exercise helps the group of people you are looking at. What is also unscientific is doing a review of such trials and not making it abundantly clear that the trials are basically worthless. The BPS researchers and Cochrane deserve criticism for that, first and foremost.
If you don't call that out that behaviour, then people will go on using that recipe to prove that basically anything is a somewhat effective treatment, in just the same way as the sham treatments in the asthma study appeared to be effective when only subjective outcomes were used.
I hope you don't give up on this conversation at this point @Medfeb. We have precious few sincere and influential advocates for ME/CFS, and we need you all to understand this point when you advocate on our behalf. It's really important.
Last edited:
I completely agree that poor trial design is a huge problem. But for me, patient selection is a critical part of trial design - that's exactly what defines "the group of people you are looking at."
At this point, PEM is mandatory for a diagnosis of ME/CFS and using selection criteria that do not require PEM to study people who are claimed to have ME/CFS is unscientific. Any review that does not grapple with that fact - in addition to addressing the issues with subjective outcomes in unblinded trials, failure to evaluate PEM as a hard, failure to address the impact of outcome switching, etc - can not be relied on as the basis of recommendations specifically for people with ME/CFS.
IMO, its not just any one of those issues. It's all of them. I accept you feel differently
Thanks @Medfeb
The Larun et al review refers to chronic fatigue syndrome, not ME/CFS. It's a moot point whether PEM is required for a diagnosis of CFS. They were answering the question they set out to answer.
The lack of use of a diagnostic criteria including PEM is an argument for there needing to be a new review relevant to ME/CFS. But it's not a reason for saying that the Larun et al review was unscientific. (Although of course there are many other reasons why it was unsound.)
An example of a study with a wide criteria that is useful
The recently published MAGENTA study says the NICE 2007 diagnostic criteria were used, and, as far as I can tell, PEM was not mandatory in that criteria.
https://pmc.ncbi.nlm.nih.gov/articles/PMC1962830/
But, I would absolutely want any future review of exercise therapy in ME/CFS to consider the MAGENTA study; it does tell us something useful about GET and about the harm that can be done by telling people that they can be better if they put the work in. It included objective outcome measures, accelerometer data, that show that the young people returning accelerometer data had overall reduced activity after 6 months, and particularly reduced activity in the moderate and vigorous activity categories. It made an attempt at including a control treatment that participants might have regarded as equally valuable, a therapised version of pacing.
The MAGENTA study doesn't use a diagnostic criteria where PEM is mandatory, and yet, against the odds e.g. Crawley's involvement, the 2024 report does manage to be useful.
Medfeb, would you reject any consideration of the MAGENTA study when thinking about treatments for ME/CFS on the grounds that PEM was not mandatory? Do you think it tells us nothing useful about ME/CFS? Would you toss it in the rubbish bin and say that we need a new study of GET in young people who have been carefully determined to all have PEM?
If we had a CFS/ME review made up of exercise therapy studies that used the NICE 2007 diagnostic criteria (PEM not mandatory) and objective outcomes (and that were otherwise sound), would you say that that review would have nothing useful to contribute to decisions about treatments for ME/CFS? Would you say that it was unscientific?
The NICE 2007 diagnostic criteria is an example of a wide criteria, but I think studies that used it potentially could tell us something about how to look after people diagnosed with a narrower ME/CFS criteria that requires PEM (if they were not otherwise flawed).
Last edited:
Of course all of the issues are relevant, but some have wider reaching impacts than others. And some aren't evidence of unscientific practice.
So far, the people involved in the new Cochrane review process who have spoken seem to have focussed on one single issue, claiming that the issue is evidence of unscientific research, when it unequivocally is not. They have ignored both a trial design issue that makes any reported benefits unreliable, and, I think I am right in this, the fact that the finding (that there was no clinically useful benefit) was not properly presented.