Guardian Article on NICE pause

Discussion in '2020 UK NICE ME/CFS Guideline' started by Jonathan Edwards, Aug 30, 2021.

  1. Hoopoe

    Hoopoe Senior Member (Voting Rights)

    Messages:
    5,424
    If ME was a psychological or behavioural problem, it would be liberating for patients to receive the right treatment. They would feel understood and be supportive and grateful of the research and therapy.

    Somehow this thought does not occur to many people. Belief in a Freudian view of psychosomatic illness (patient categorically in denial) is widespread.
     
    Sean, Michelle, alktipping and 20 others like this.
  2. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,563
    Location:
    UK
    I wish people would challenge them when they make such claims to ask them what evidence they have to back this up. Given what we know of how the clinics work (with lack of follow up) along with trial results it seems to be a very empty claim.
     
    Michelle, alktipping, Amw66 and 21 others like this.
  3. Peter Trewhitt

    Peter Trewhitt Senior Member (Voting Rights)

    Messages:
    4,081
    Also given they are now admitting GET/CBT can cause harm and have as of yet presented not indication as to how to distinguish those that will be harmed from those that will not, they are asking the patient to play Russian roulette. But also Russian roulette where they can not give any meaningful probabilities for the different outcomes.

    At times medicine involves risk, and it is acceptable to ask the patient to take risk, if they are informed of the consequence and can balance the likelihood of benefit versus harm, however given the GET advocates and current specialist service do not adequately record harm and up till now have refused to even acknowledge the possibility of harm, we are not seeing informed choice.
     
    Wits_End, Chezboo, Michelle and 16 others like this.
  4. Sean

    Sean Moderator Staff Member

    Messages:
    8,064
    Location:
    Australia
    Patients who get better tended not to be as vocal as those who had had bad experiences, she said.

    That would be the patients "who get better", who for some mysterious reason, also failed to show up in the results of your own "definitive" study?
     
    Last edited: Sep 13, 2021
  5. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,563
    Location:
    UK
    Its interesting that so many claim to have an interest in the 'lived experience' and yet seem to use this as a reason not to listen to what patients are saying but would rather listen to psycologists (or whatever Chadler is) and what they say about 'lived experience' even when they have no evidence.
     
  6. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,563
    Location:
    UK
    When I first read PACE (without much knowledge about anything ME or about trials) my first thought was that some would probably benefit and others not. I thought this would lead to a bimodal disrtibution over improvements (i.e. one part of the population improving and one not) such that it would look like the sum of two Gausians. This also suggests that using mean differences wouldn't be great. But from what I remember of the data it doesn't take this form at all but rather is fairly smooth (normalish) in terms of the distribution.

    @Jonathan Edwards do you recall trials where the intervention helped a subset of people and how that was represented in terms of the distribution on outcome measures? Would it show say a clear separation on a scatter plot?
     
  7. Barry

    Barry Senior Member (Voting Rights)

    Messages:
    8,420
    The crucial insight for people that often gets lost along the way: Treatments are given to pwME for a condition they do not suffer from. Full stop.

    Their ME/CFS is not due to unhelpful beliefs preventing them from overcoming deconditioning, yet that is what they are being treated for! If you treat any patient for an illness they do not have, there will always be the potential for deterioration of their condition, either through lack of the correct treatment, or because the prescribed treatment directly impairs their real condition.

    So as I see it we need to cut to the chase. Just highlight the blindingly obvious, but which people don't realise until someone cuts through all the cr@p for them, and spells it out succinctly and clearly: GET and CBT are intended to treat an illness that pwME do not sufferer from. So akin to being prescribed ibuprofen for a kidney condition that might be worsened by ibuprofen, without the clinical wisdom or evidence to know if it would be safe for the patient.

    GET and CBT are intended to treat an illness that pwME do not sufferer from. pwME suffer from a physiological illness; CBT and GET are prescribed to overcome a psychological condition.

    Misapplication of a treatment - THAT is the root of what pwME object to. Explanations more complex than that just provide breeding ground for conflation by the BPS crew to sow confusion and misinformation.

    Also, when MS and Co "patiently explain" that they know there is a physical component to ME/CFS, it comes across to the uninitiated that pwME's assertion of a physical illness is being acknowledged by MS and Co. But of course that is absolutely not so, and is just part of the confusion MS and Co deliberately spread; they are talking about deconditoning as the physical component.
     
    Last edited: Aug 30, 2021
  8. Sean

    Sean Moderator Staff Member

    Messages:
    8,064
    Location:
    Australia
    @Adrian

    Far as I know, there is literally nothing in the PACE data (or in FINE, GETSET, etc) that even justifies further research into the psycho-behavioural approach, let alone its application in the clinic and medico-legal settings.

    If PACE, et al, show anything it is that this is not a productive line of inquiry.
     
  9. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,563
    Location:
    UK
    There isn't but I'm wondering if there is a simple analysis of the distribution of the values for the outcome variables that would be able to test their hypothesis. My feeling is if there is a real effect on a reasonable proportion of the trial population then that would show in the distribution whereas if you have a common effect (such as reporting bias) then that would lead to a different distribution of values.
     
  10. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    Fluge and Mella produced pharmacodynamic profiles for each patient showing progress over time. Although we now know that the responses were spurious (placebo or expectation bias of whatever sort) you can see very clearly that some patients spiralled up to reporting feeling normal and others did not change. For the phase II unblinded extension my memory is that the situation was fairly clearly bimodal, although messily so.

    When I was doing rituximab studies I similarly plotted individual patient progress. For some nothing happened. For others there was remission.

    Basically the PACE authors shod have presented their data the way F and M did and it would have been even clearer that the treatments were no unwinding some psychosocial perpetuating factor..
     
  11. adambeyoncelowe

    adambeyoncelowe Senior Member (Voting Rights)

    Messages:
    2,736
    Since journalists often ask the two parties to contribute quotes in isolation, there's never really a chance for one side to interrogate the other. In a news article, where the journo won't want to (overtly) editorialise, they end up not challenging this themselves in case it seems biased.

    It's the nature of the beast. Perhaps what we really need is media training for our charity reps, so that they can preempt the usual statements. It doesn't take precognition to know what they will say.
     
  12. Jonathan Edwards

    Jonathan Edwards Senior Member (Voting Rights)

    Messages:
    15,175
    Location:
    London, UK
    Except that in this case I had spent the best part of an hour on the phone with Grover going through the arguments and counter-arguments. She specifically rang me to ask me to explain why the two sides differed so much.
     
    Sean, FMMM1, Peter Trewhitt and 28 others like this.
  13. JemPD

    JemPD Senior Member (Voting Rights)

    Messages:
    4,500
    Well i dont see why not :rolleyes::D

    couldnt agree more.
     
  14. adambeyoncelowe

    adambeyoncelowe Senior Member (Voting Rights)

    Messages:
    2,736
    It seems the article has been updated, because your comment about believing patients when they worsen as well as when they recover is there, as well as your comments about the step counters.

    However, the Forward ME press release for the 18th August had an unhelpful focus on the dualism argument, so that's what I was referring to more than your comments.
     
    bobbler, ukxmrv, Cheshire and 18 others like this.
  15. JemPD

    JemPD Senior Member (Voting Rights)

    Messages:
    4,500
    So disingenuous. As if huge numbers of patients dont say the treatment made them worse. "if you believe the patient you believe the patient" Indeed. But you have to believe them when they're not saying what you want them to say because they havent been asked leading questions/been manipulated into saying they feel better, as well, not only when they say what you want them to.
    It's diabolical the way things are being spun. Very clever.

    Edit Cross posted with adambeyoncelowe above, it seems the article has been updated to adress this now
     
  16. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,563
    Location:
    UK
    I had something different in mind than looking at individual patient progress.

    Basically if we have a trial population T which is made up of two groups A and B (as defined by some function s) then one group (say A) will have an average improvement effect on the outcome at a given point of i. And compare that to a trial population T' which is made up of a group of patients C with an average improvement i' such that the overall trial improvement for improvement (T') == improvement(T).

    Then if we look at the distributions for improvement over the patents in both trials then will they appear to have been drawn from different distibutions (i.e say a significance test to see if two sample distributions are the same) and could I tell which is which?

    Underlying this is the thought that say we had 10 patients and 5 improved then if we looked at a scatter plot of the improvements then we would see 5 in a cluster with improvement and 5 in a cluster around 0. But if we have all 10 improving a bit (to give the same overall result) then we would see one cluster on the scatter plot (and probably a higher degree of spread).

    My feeling is (thinking a bit about it) that there would be too much noise to reliably pick out a signal (unless the subgroup improving was around 50% of the population and the improvement significant). I guess I'm just wondering if there is any theory around this.
     
  17. cassava7

    cassava7 Senior Member (Voting Rights)

    Messages:
    1,051
    The article has a number of issues (the title and the headline first and foremost) and errors from interviewees left uncorrected.

    The "best available criteria" at the time the PACE trial was being designed were the Fukuda criteria, not Oxford, and dropping actigraphy measurement was the sole decision of the trial steering committee, not that of medical bodies.

    However, it seems to me that, from an external (non-ME patient) point of view, the arguments from critics of the PACE trial and ME charities effectively rebuke those of Sharpe and Chalder. The latter give the impression of being weak justifications against well-founded criticism, and they become a confused muddle.

    One such instance is believing patients about their self-reported symptoms:
    While the article unhelpfully spirals down into the "psychological vs physiological" debate, the very reasonable position of ME charities on the matter offsets Chalder's comment on not getting embroiled in "dualistic thinking". Clare Ogden (Action for ME) and Peter White (Forward-ME) both recognize without question that any chronic illness can have a significant psychological impact; the argument is not about denying it, but that ME patients could have benefited from more biomedical research rather than funding being endlessly granted to psychological research.

    ETA: a somewhat exquisite example of survivorship bias in Chalder's response:
     
    Last edited: Aug 30, 2021
  18. JohnTheJack

    JohnTheJack Moderator Staff Member

    Messages:
    4,789
    Jonathan, thanks for all your work with these journalists.
     
  19. Tia

    Tia Senior Member (Voting Rights)

    Messages:
    519
    Indeed, thank you @Jonathan Edwards
     
  20. Barry

    Barry Senior Member (Voting Rights)

    Messages:
    8,420
    [my bold]

    As always, obfuscating and playing down the grim fact hidden within those words "some patients do not benefit from these therapies". It is a sleight of words designed (yes I do mean 'designed') to give the impression that those who do not benefit, simply remain unchanged, and are no worse off. That is the generally accepted interpretation of such language.

    But the use of such language here is intentionally different I'm sure. It is intended to convey exactly that impression, whilst it also embraces a more sinister, but equally valid interpretation: That within the group of patients who do not benefit, there will be a subgroup who deteriorate.

    So when these folk very ambiguously state that not everyone benefits from GET/CBT for ME/CFS, they really should be challenged to be more specific regarding those who do not benefit - what about those whose mode of "not benefitting" is to in fact deteriorate?

    Ambiguity is second nature to these people, knowing that people will assume the most obvious interpretation, and thereby completely miss the more incisive interpretation they want people to miss.
     
    Last edited: Aug 30, 2021

Share This Page