1. Patients launch a $1.27 million crowdfunding campaign for ME/CFS gut microbiome study.
    Check out the website, Facebook and Twitter. Join in donate and spread the word!
No Longer Naive in the Ways of The Beast
After having lived for years with ME/CFS, Jody Smith learned there's more to this beast of an illness than she realized, and that what might help one person may not help others ...
Discuss the article on the Forums.

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

Discussion in 'Latest ME/CFS Research' started by Sam Carter, Jan 31, 2013.

  1. user9876

    user9876 Senior Member

    Messages:
    684
    Likes:
    1,549
    One of the things that Morris claims is that GET has been delivered safely and reliably within the context of a RCT but patient reports suggest the dangers of delivery by untrained people.

    However it should be noted that 15% (24 out of 160) people treated in the GET group received in adaquate treatment. Where as Morris concludes that GET should only be delivered by trained and supervised staff I think we could argue that given a very careful management of PACE was challenged to properly control delivery then even this would be a challenge. Another significant issue is we don't know how much people in the PACE trial increased their activities - it may be (judging from the 6mwt) that very little activity increase was encouraged so as to get a good safety record!

    Even if we were to believe the PACE trial scaling out such a service would be very hard.
    taniaaust1, Bob, biophile and 2 others like this.
  2. Valentijn

    Valentijn Activity Level: 3

    Messages:
    5,883
    Likes:
    7,964
    Amersfoort, Netherlands
    I think they're doing a very good job of outing themselves as people who read media summaries, or abstracts at best. They can't be bothered with reading the details of something before drawing conclusions about it and assuring everyone else that they know what they're talking about. It's also quite possible that as "authorities" themselves they have an increased likelihood of relying on other authority as a persuasive tactic.
    peggy-sue, Roy S, maryb and 1 other person like this.
  3. Sam Carter

    Sam Carter Guest

    Messages:
    296
    Likes:
    184
    And this is the same Richard Morris who worked on PACE's 'sister' trial, FINE? He seems to have forgotten all about that.
    Valentijn likes this.
  4. biophile

    biophile Places I'd rather be.

    Messages:
    1,350
    Likes:
    4,025
    PACE may have also been more careful than some random clinic to exclude those unsuited for GET. I'll guess that poorly applied GET is part of the problem in the wild, but I also think another problem lies with how harm is being defined. Patients' ideas of harm may be different than what has been defined for research.

    GET participants did not have to increase activity if they did not want to or could not tolerate it, there is no evidence to show they are increasing total activity levels, and as you said, the available data suggests they are not doing so to any major degree, so it would be questionable to imply that increasing total activity levels is therefore safe.

    Before PACE was published there was almost no published data from previous RCTs on adverse effects (the safety profile was based on dropout rates and group averages in scores).

    There has been a FOI request for the data on deterioration rates using the opposite threshold for improvement. I'm not sure if this will reveal much, as this is compared from baseline to 52 weeks (adverse events are more time constrained), and patients had 6 months after therapy to recover.

    Without going into all the definitions of the various adverse events, these were recorded by the research assistant at 12 weeks, 24 weeks, and 52 weeks. I'm not sure if patients were required to remember them or write them down before hand, but the form used specified a date start and a date end.

    The rates of "non-serious adverse events" were similar for GET vs SMC, but there were different degrees of severity for this outcome, mild/moderate/severe, which have all been lumped together when PACE reported on safety in 2011. It was possible to have a "severe" non-serious adverse event which was relatively incapacitating for up to 4 weeks.

    Next we have serious adverse events: "The increased rate of serious adverse events with GET compared with SMC is unlikely to be important because serious adverse events were not thought by the independent scrutinisers to be related to treatment." These were also rated as mild/moderate/severe.

    Once a serious adverse event was identified, the "independent scrutiniser" was unblinded to allocation to help determine whether it was a reaction to treatment. Who knows what effect that had, there were more SAEs for GET but when the scrutinisers found out it was for GET suddenly the difference disappeared.

    PACE used the CGI but collapsed the scores. The "very much worse" and "much worse" scores were merged for "negative change". However the "a little worse" score is merged with "no change" and "a little better".

    Definitions for serious deterioration were fairly strict.
    taniaaust1, Svenja, user9876 and 3 others like this.
  5. Bob

    Bob

    Messages:
    7,755
    Likes:
    9,287
    England, UK
    I'm very late coming to this thread. I've read through it, but might have missed or forgotten a lot of discussion.

    The composite/cumulative "trial recovery" criteria is outlined in the top half of section 'b' of Table 1: "Cumulative criteria for trial recovery."
    It uses the Oxford criteria (not the CDC or London criteria) and I think these are the results that are quoted in the abstract.

    This composite trial recovery criteria consists of:
    Both CFQ and SF-36-PF in normal range
    And Oxford criteria not met
    And CGI 1 or 2 (95% CI)


    I want to comment on the SF-36 PF scores for the composite "trial recovery" criteria.

    This is how I interpret it:

    The 'normal range' individual recovery criteria requires an SF-36 PF score of 60, and above, for 'recovery'.

    The 'Oxford criteria not met' individual recovery criteria requires an SF-36 PF score of 70 of more. (For a diagnosis of CFS using the Oxford criteria, in terms of assessing 'recovery' for this paper, participants had to have a score of 65 or less. And the SF-36 PF scores go up in increments of '5', so the next possible score above '65', for an individual, is a score of '70'.) But, importantly, a score of 70, and above, is only required for a 'recovery' (for 'Oxford criteria not met' individual recovery criteria) if all of the other criteria for the Oxford criteria are met. For example, if a patient fails to meet the 'Oxford criteria' for a reason other than the SF-36 PF scores, then the patient will be considered 'recovered', whatever their SF-36 PF scores. So, for example, if 'fatigue' is no longer the main symptom, then a participant will fail the 'Oxford criteria not met' recovery criteria (i.e. they will be considered 'recovered') whatever their SF-36 PF score.
    As another example, if a patient has a SF-36 PF score of 50, but 'fatigue' is not their main symptom (i.e. 'post exertional malaise' or 'pain' is now the main symptom), then they will not be eligible for a diagnosis using the Oxford criteria, and will be considered 'recovered' for the 'Oxford criteria not met' individual criteria. Thus a patient could have a SF-36 PF score of 50, and still be considered 'recovered' for the individual 'Oxford criteria not met' recovery criteria.

    But, to be considered as 'recovered' for the composite recovery criteria, all participants must pass the 'normal range' test, with a score of 60.

    So, my interpretation is that a score of 60 and above is required for the composite trial recovery criteria.
    (This will be the case except for specific scenarios with regards to the 'Oxford criteria not met' recovery criteria, for which, in some cases, participants will need a score of 70 or more.)
    But generally speaking, patients considered 'recovered', using the main composite trial recovery criteria (as reported in the abstract) must all have a SF-36 PF score of 60 and above, but do not necessarily have to have a SF-36 PF score above 60. In other words, 'recovered' (composite trial recovery) patients only need a score of 60, and do not necessarily have to have a score of 70 or above.

    As such, a 'recovery' could theoretically be a 'deterioration', in terms of physical function, after treatment with CBT or GET.


    Any thoughts? Have I got this right, or have I missed something vital?


    Edit: The paper says that a research assistant assessed patients for the Oxford criteria, and the authors acknowledge that this is a subjective process, making bias possible:

    "The assessments of caseness (CDC, London and Oxford criteria) relied on a mixture of self-ratings and research assistant assessments, making some observer bias possible."

    I've underlined the criteria that I think were assessed subjectively by a research assistant, to assess for 'recovery':

    "Research assessors judged whether participants still met Oxford criteria for CFS at 52 weeks; specifically they determined if : (1)fatigue was the main symptom, (2) it was of definite onset
    and not lifelong, (3) fatigue was severe, disabling and affected physical and mental function, and (4) fatigue had persisted for 6 months or more and was present 50% of the time (Sharpe et al. 1991)."
    taniaaust1, Valentijn, Roy S and 2 others like this.
  6. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    Yes, I think that's a good summary.
    Small point: You switched between "research assistant" and "research assessor" - I don't think the former is appropriate language as I am presuming they are a doctor (the SMC doctor)? (but I haven't checked to answer this)
    ETA: biophile checked the striked through bit - two posts down

    I think many people could read it that everything has to be outside the baseline criteria (or at least both the CFQ and SF-36 PF have to be). I think the record of the thread showed I was confused at one stage (I think I crossed that out) and I've seen various people show signs of confusion here and elsewhere. And these are people who are familiar with the other PACE Trial papers and the protocol, while many people will just read this paper on its own and probably a bigger number will just read the abstract.
    Bob likes this.
  7. biophile

    biophile Places I'd rather be.

    Messages:
    1,350
    Likes:
    4,025
    Bob.

    I'm not convinced the CFQ and SF-36/PF entry requirements mentioned along side the Oxford criteria are actually part of the criteria for recovery, and even if they were, as you said they did not both have to be satisfied, many reasons could lead to no longer meeting Oxford criteria despite ongoing illness, not just from the participants (eg, fatigue not main symptom, etc) but also from those assessing CFS caseness (eg, post-trial optimism), knocking them out of diagnosis. I agree with the argument that a proportion of those diagnosed with Oxford criteria at the start of the trial may have already been close to no longer meeting Oxford criteria and being within normal range for fatigue and physical function.

    As you say, someone could theoretically have "Oxford criteria not met" and a SF-36/PF score of 50. However, this alone was not classified as a recovery, the normal range in both fatigue and physical function was the first base criteria in the composite definition for recovery. To be classified as recovered, the participant first had to meet this normal range first, which would exclude those with SF-36/PF scores of <60. If scoring 65 at entry, 60 would still be a slight deterioration.

    No longer meeting Oxford criteria does not seem to be that robust when comparing to the other outcomes:

    No longer meeting Oxford criteria was 41% for SMC and 54% for CBT (100% at baseline for both).
    Normal CFQ and SF-36/PF was 15% for SMC and 30% for CBT.
    Normal CFQ and SF-36/PF, and, no longer meeting Oxford criteria, was 14% for SMC and 28% for CBT.

    End-point thresholds overlap with the entry criteria for disabling fatigue, but almost all participants within normal range for both measures at the end of the trial did not meet Oxford criteria, yet a large proportion of participants no longer meeting Oxford criteria, roughly half but better for CBT, still had abnormal levels of fatigue and/or physical function.

    Bolting on the CGI criterion to the above tightened it up, but again, not itself a recovery and not necessarily much better compared to if they had been in the SMC group ("much better" was compared to their own baseline, not compared to SMC). It reduced the figures down to 7% for SMC and 22% for CBT, which shows some subjects declared normal and without CFS did not in fact get "much better" but "a little better" at best, as the next choice down the list on the CGI is "a little better". Perhaps those participants who would have scored "a little better" if in the SMC group were helped slightly by CBT and were more likely to report feeling "much better"?

    As others have pointed out, there should have been additional criteria to compare with. Dr Charles Shepherd recently noted that sickness benefits are incompatible with recovery, and pointed out that the rate of receipt even increased slightly in the CBT and GET groups (without significant difference between any group in the trial), despite all this "recovery" going on.
    Bob, Valentijn and Dolphin like this.
  8. biophile

    biophile Places I'd rather be.

    Messages:
    1,350
    Likes:
    4,025
    I just had a look. On page 53 of the complete protocol, it was the research nurse/assistant that did the Oxford criteria assessments. It also looks like the 6MWT was done at 24 weeks too.
    Simon and Dolphin like this.
  9. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    (This point may been referred to before - apologies if it has) If people are doing letters, one possible useful angle would be to quote any long-term follow-up studies where benefit deteriorated in the long-term.
  10. Firestormm

    Firestormm Senior Member

    Messages:
    5,819
    Likes:
    5,941
    Cornwall England
  11. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    Charles Shepherd has posted this on the MEA FB page today:
    Firestormm likes this.
  12. alex3619

    alex3619 Senior Member

    Messages:
    6,806
    Likes:
    10,342
    Logan, Queensland, Australia
  13. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    It most likely has been pointed out that Knoop used a different definition of recovery (incl. SF36 PF score >=80).

    I'm not sure it has been pointed out that the post-treatment mean in the Knoop et al. study was 76.3 (23.0), which is a lot different from the CBT: 58.2 (24.1) and GET: 57.7 (26.5) results in the PACE Trial.
    biophile, Simon and Bob like this.
  14. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    (Apologies if this point has been made before)
    In
    (which involved one of the PIs), they used this approach:

    Ref 27 is:
    Has anyone read it? I think it could contain useful arguments against some of what they did, but I don't want to pay for it if it is of no use. Thanks.
  15. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    If anyone had the time, it'd be interesting to find studies out there on other groups who were said to have fatigue but we know that some of these fatigued group had Chalder fatigue questionnaire (CFQ) Likert scores of 18 or less.

    I'm thinking of other illnesses, where CFQ would be lower than in ME/CFS

    This one is a possibility but not perfect:
    At baseline, in two groups with MS fatigue:


    Scores where Interquartile ranges were given would be better as that would likely confirm for many studies that 25%+ had scores of <=18. Some might even have means of <=18 which means there are definitely one or more patients with scores of <=18*.

    *given the max score is 33, one can probably say with certainty there were more under 18, depending on the sample size of an individual study
  16. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    A wiki was mentioned earlier. I suggested people should concentrate on letters for the moment. But I know some people are not going to send in a letter so if you're not going to do a letter, or you've already sent one in, if you could do up a wiki it would be great. It's hard to keep track of the points that were made - it could be a useful resource for anyone considering doing a letter.
  17. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    Has this finding been mentioned in the thread? Probably worth repeating even if it has been:

    http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(11)60688-0/fulltext


    Valentijn, biophile, Bob and 2 others like this.
  18. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    From Knoop et al. (2007) full recovery paper
    I was intrigued by what these might say so read them.
    However, I don't believe they are of much use.

    Here's the abstract for the second one - it sums up what it is about

    The other article is just an editorial that talks about this paper. Neither really are about recovery - they're just about how "positive well being" (my wording) is not necessarily the opposite of "mental health problems" (my wording).

    I noticed before with this journal that odd references of articles from the journal would appear in papers. And a high percentage of publications had articles from the journal. My guess is what happens is the editor or others from the editorial team review articles and suggest certain references be mentioned; researchers just want to get their papers published so go along with them.
  19. user9876

    user9876 Senior Member

    Messages:
    684
    Likes:
    1,549
    I've come across economists who as reviewers insist that their papers are cited even when they are not strictly relevant so it could happen in this journal.
    Dolphin likes this.
  20. Dolphin

    Dolphin Senior Member

    Messages:
    6,519
    Likes:
    4,900
    Reminder for anybody who wants to know about the Chalder Fatigue Sclae and its scoring:

    The 11 questions are on page 162
    http://evaluatingpace.phoenixrising.me/PACE_Protocol.pdf

    With likert scoring, the different answers reading from the left are
    0, 1, 2, 3 for each question (i.e. range of 0-33).
    With bimodal scoring, the different answers reading from the left are
    0, 0, 1, 1 for each question (i.e. range of 0-11).

See more popular forum discussions.

Share This Page