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!
9th Invest in ME International ME Conference, 2014 - Part 1: Autoimmunity and ME
Mark Berry begins a series of articles on the 9th Invest in ME International ME Conference in London, with a look at three presentations on autoimmunity
Discuss the article on the Forums.

PACE Trial and PACE Trial Protocol

Discussion in 'Latest ME/CFS Research' started by Dolphin, May 12, 2010.

  1. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Thanks oceanblue

    Great spot. :Sign Good Job: (one tiny point: it's slightly awkward as they changed the entry criteria part way through from <=60 to <=65. But I think they would let this through ok and it would take up words to cover both).
    :Retro smile:
    :worried:
  2. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    I think the actual evidence from PACE, that CBT & GET have a pretty limited impact, can be powerful tools to expose the limitations of the biopsychosocial model. However, it's also clear from the way that the results were presented in the paper, spun by the authors and reported in the media that as things stand the PACE trial is pretty unhelpful to a better understanding of this illness.

    The question is, how do we use the evidence in PACE to turn things around? Currently I'm not sure, though letter to the Lancet will be a start. So I still think it's a big opportunity but i'm not at all sure how we make the most of it.
  3. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Need full paper: Chalder fatigue normative data from Norway

    Can anyone get the full paper?
    It'd be good to see more of the data e.g. SDs

    Remember they're claiming a score of 18 or less and one has a normal level of fatigue.

  4. Hope123

    Hope123 Senior Member

    Messages:
    1,141
    Likes:
    531
    Very slowly reading here but quick comments:

    1. PM me if you want the full article.

    2. 6 minute walk test:

    Optimal reference equations from healthy population-based samples using standardized 6MWT methods are not yet available. In one study, the median 6MWD was approximately 580 m for 117 healthy men and 500 m for 173 healthy women (48). A mean 6MWD of 630 m was reported by another study of 51 healthy older adults (53). Differences in the population sampled, type and frequency of encouragement, corridor length, and number of practice tests may account for reported differences in mean 6MWD in healthy persons. Age, height, weight, and sex independently affect the 6MWD in healthy adults; therefore, these factors should be taken into consideration when interpreting the results of single measurements made to determine functional status. We encourage investigators to publish reference equations for healthy persons using the previously mentioned standardized procedures.

    Article from 2002 from American Thoracic Society guidelines -- you can spring off of this to find a more recent article.

    http://ajrccm.atsjournals.org/cgi/content/full/166/1/111
    3. Have heard that CAA will be analyzing PACE and publishing a statement. I've asked them to consider widely circulating it as much as possible.
  5. Cort

    Cort Phoenix Rising Founder

    Messages:
    7,025
    Likes:
    428
    Raleigh, NC
  6. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    A look at what happened to the outcome measures that had been promised

    Registering one's trial and outcome measures is very important now - some journals won't accept trials that haven't been registered.
    The authors went one step further and published their protocols in a journal.

    http://www.biomedcentral.com/1471-2377/7/6

    Here's a look at what happened to them
    [Aside: The analyses they promised would need to be looked at separately (I'm not promising I'll do it)]

    Not given per protocol.
    Likert scoring used (not necessarily a bad thing but a change)
    Normal fatigue no longer given as a primary outcome measure
    (when it is given later, not in the abstract, the 50% threshold is not used and a score of 18 is seen as acceptable which, most people would say, does not equate to 3 or less on bimodal scale (out of 11); a score of 18 is a minimum of 4 (out of 11) on the bimodal scoring and a maximum of 9).

    Not given per protocol.
    Normal physical functioning no longer given as a primary outcome measure
    (when it is given later, not in the abstract, the 50% increase is not used and a score of 60 or more is suddenly seen as acceptable despite the fact that a score of 65 or less lets one into the trial).


    Not given per protocol.
    Data is given, but not as a primary outcome measure, but it uses these new, lower thresholds, mentioned above

    Has become one of the primary outcome measures.

    Per protocol.

    Neither of these are given.

    Not given.
    The commentary by Knoop & Bleijenberg talks about the percentage in "recovery" but doesn't use anything close to this definition (just to clarify, they don't give extra data).

    Given (both)

    Given (both)

    Not given

    Given

    Not given

    Not given

    Not given

    Not given per protocol.
    We are just give a composite yes/no score for the 8 CDC symptoms combined and a percentage present/absent score for two of the symptoms.
    Likert scoring is the opposite of yes/no (or present/abscent) scoring
    What they actually asked that patients at different stages about each of the symptoms was the following:
    but we don't get this data.

    Not given

    Given

    Not given per protocol
    A lot of people would miss this but they redefined it to be “any two consecutive assessment interviews” [Quote: “Serious deterioration in health was defined as any of the following outcomes: a short form-36 physical function score decrease of 20 or more between baseline and any two consecutive assessment interviews”] Comment: it is already hard for a patient to drop 20 points on the SF-36.
  7. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Web Appendix Table C: Description of Serious Adverse Reactions

    To me, although it's a small number, the two experienced by people doing GET could be important. Maybe hard to bring up in a letter of 250 words. But something to be used perhaps when safety is being claimed.
  8. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    More great digging, dolphin, thank you. The PACE group have a lot of questions to answer. Frankly, so does the MRC Trial Steering Group which is supposed to independently oversee the trial and keep the authors honest. It seems to have fallen asleep on the job.
  9. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    Dolphin, I think I love you! :D

    Sorry, getting carried away here because there are some gems in that paper that expose the flaws in the PACE choice of thresholds.

    This study is based on completed questionnaires form 9,332 people of working age in central England.

    The survey gives SF-36 PF scores for the whole sample, for people who reported a longstanding illness and, crucially, people who did not report a longstanding illness. This last group might be the best estimation of 'healthy'. Here are the mean SF-36 scores with SDs in brackets and the threshold that would result from using the PACE formula of "mean minus 1 SD":

    'Healthy': 92.5 (13.4) = 79.1
    'Chronically ill': 78.3 (23.2) = 55.1
    'Population*': 89 (16) = 73
    *oh damn, they don't seem to give this separately, this is my guesstimation from looking at the data they do give

    Since the PF scale only scores in 5 point intervals (e.g. 60,65,70) these translate as PF threshold scores as:
    Healthy = 80, population = 70 or 75. PACE used 60.

    slightly more complex point
    As a bonus, they provided SF-36 PF scores for people who had consulted a doctor in the 2 weeks prior to completing the questionnaire. This is a pretty close approximation of the 'GP attenders' used to establish norm data for the fatigue scale. The scores are
    81.6 (23) = 58.6

    This shows that not only are the GP attenders substantially less well than 'healthy' people (81.6 vs 89-ish), they also have a much bigger SD, which has the effect of lowering the threshold even further (60 vs 70 or 75). Obviously this is for PF scores not fatigue scores, but it does illustrate how GP attenders differ from the normal population​
  10. Esther12

    Esther12 Senior Member

    Messages:
    5,087
    Likes:
    4,856
    Thanks for that OB... this does seem to be the key point so far.
  11. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Some people had problems with the RapidShare link.
    I've placed it in another place now:
    https://www.yousendit.com/download/T2pGd0VBaFI4NVh2Wmc9PQ
    or http://bit.ly/htzc1Y

    I can also E-mail to you if your account can cope with an 11MB file (gmail accounts can).

    My granny (91) is here at the moment so won't be able to contribute for a little while.
  12. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    Thanks for pointing that out. I'll keep it simple and just compare the PF score of 65=disabling fatigue with 60 = 'normal'.
  13. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    Re: reference used to justify major changes in reporting of primary outcomes

    Sorry, my mistake, it was ref 30 not 31, and this was publisehd in 2009, after their protocol
    Measurement in clinical trials: a neglected issue for statisticians?

    I'm afraid that tackling this paper is way above my pay grade.
  14. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    Thanks, Esther12.
    OB (Obi?!)
  15. Cort

    Cort Phoenix Rising Founder

    Messages:
    7,025
    Likes:
    428
    Raleigh, NC
    Dolphin are you saying that all the "not-givens' per protocol are things they said they would and did not make it into the study?

    If so do we have any idea if they actually measured any of those; ie did they measure them and decide not to report on them or did they just not do them?

    Can we show there's a pattern to the non-used tests; ie that they tend to better indicators of functionality?
  16. Snow Leopard

    Snow Leopard Senior Member

    Messages:
    2,203
    Likes:
    1,535
    Australia
    They definitely measured them. I bet they will claim that they will publish the metrics like the EQ-5D and the hours worked in a later paper where they can spin the unimpressive data by saying, 'but look, the treatment will still save a few dollars'. Compared to the billions in economic costs each year....
  17. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Like Snow Leopard said, I think there is a 99.99% chance they were measured. The protocol paper was published in 2007 after the trial had long started. The questionnaires can be seen at: https://www.yousendit.com/download/T2pGd0VBaFI4NVh2Wmc9PQ . So "not given" means they measured them but didn't publish them.

    I'm not sure. The whole point about registering primary and secondary outcome measures is that one can't then not publish them and come up with new ones that suit so it's a bad thing by itself. Such "rules" are mainly designed for pharmaceutical companies but really apply for anyone. They may be part of the CONSORT guidelines.
  18. Dolphin

    Dolphin Senior Member

    Messages:
    6,439
    Likes:
    4,663
    Yes. Or alternatively they may be buried. In the FINE Trial, the authors were challenged that they didn't publish the step test (the only objective measure they used) - they hadn't mentioned in the initial paper they hadn't used it.
    Their reply was a weak one:
    This sort of behaviour is watched like a hawk when it is done in trials of pharmaceuticals. I'm no expert on who does it - the FDA would be one (and similar agencies in other places). We need to do it for non-pharmacological trials although it probably will get mentioned in some reviews like Cochrane.
  19. Bob

    Bob

    Messages:
    7,432
    Likes:
    8,569
    England, UK
    I might be wrong about this... So please let me know if I am...
    But I think I've just come across an almost glaringly obvious bit of spin in the reported results...

    All the patients in the trial received specialist medical care (SMC), and then the CBT and GET was given along side this.
    I think that the improvement scale measurements (or at least some of them - I haven't checked them all) have been reported as compared to the SMC alone group. So any measurements are reported over and above the improvements that the SMC alone group reported.

    But I think that the percentage of patients who reported improvement has been reported as an absolute figure, and not as a comparison with the SMC along figures.

    In table 5, 25% of patients reported an improvement in their symptoms when using SMC alone, and 41% of patients reported an improvement with GET (which i think means GET+SMC).
    In which case, only 16% of patients reported an improvement using GET over and above the findings of the SMC alone group.
    So GET and CBT actually only helped 16% of the patients, as compared to the SMC alone group, and so 84% of patients were not helped by CBT or GET.

    Am I right in thinking that the SMC-alone group was used as the control group in this study? If so then why haven't the figures been reported as such? i.e. why haven't the GET and CBT figures been reported as a comparison with the control group?

    Basically, only 16% of patients reported an improvement using GET or CBT, over and above the control group who were receiving ordinary medical care.

    Am I right in this, or have i missed something here?


    16% is an appallingly low figure for a 5m government funded study. What a bloody waste of money.

    If this is the case, then I think we should all be focusing on the 16% figure rather than the 41% figure, and bring this to the attention of the media and the ME organisations.

    Like I said in an earlier post, it would be hard for the NHS to justify giving group courses of GET when 60% of the patients would not be helped.
    But if only 16% of patients are helped by CBT or GET, and 84% are not helped, then this makes is almost totally unjustifiable.
  20. oceanblue

    oceanblue Senior Member

    Messages:
    1,174
    Likes:
    342
    UK
    You're right that CBT/GET basically helped a net 16% extra patients over and above SMC, which is not very impressive. However, I'm not sure the researchers were bound to report this figure as a difference, as opposed to reporting figures for both CBT and SMC. I don't know enough about stats to know what is strictly 'correct'. And it looks like you're quoting the CGI figures for patients who say they are much/very much improved.

    But basically when Peter White said in the media that CBT helped 6/10 patients (using their slightly odd definition of 'improved', rather than CGI scores) he was missing out the crucial info that in this case 4.5/10 patients were helped by nothing at all (SMC).

See more popular forum discussions.

Share This Page