Discussion in 'Latest ME/CFS Research' started by Dolphin, May 12, 2010.
Only when they have been thoroughly indoctrinated first about what answers they ought to give.
And used the questionnaires before in smaller trials.
I completely agree, I don't understand why the majority of the criticisms of the trial do not mention this.
For a trial that is measured on subjective patient reports there seems an outrageous amount of coaching of the patients in exactly what effects they should expect to report.
Describing one treatment as
in previous research studies most people with CFS/ME felt either 'much better' or 'very much better'
and another as
can improve coping and provide the conditions for natural recovery but in itself does not fundamentally change the course of the disease.
would seem to be enough to explain most of the difference in scores.
One of the practical problems with quoting bits in Letters to the Lancet was quotes took up quite a bit of words. 250 words can go rather quickly - one can't develop many points with it. Also a lot of people haven't read the manuals. But I agree that it is a good point worth raising and I hope one or more letters that went in did make this point.
I think it could be ethically challenged on that claim alone.
Do you have a link to these quotes? I'm assuming they're from the manual but haven't ploughed through them myself.
Yes, straight from the manuals
in previous research studies most people with CFS/ME felt either 'much better' or 'very much better' - in bold on page 28 of the GET patients manual
can improve coping and provide the conditions for natural recovery but in itself does not fundamentally change the course of the disease - page 5 of the APT patients manual
Thanks for that
Now that I've looked through both therapist and participant manuals, I think this has been a bit overblown, though as Esther12 says, there are ambiguities. First, that 'is there a cure?" quote in full, from the FAQ section for therapists:
I still think they should not have given such a blunt answer but it's only used if the participant asks (nb not anywhere in the participants manual, as far as I can see). Overall the approach elsewhere seems more balanced and generally positive. Here are the relevant quotes on recovery from the participants manuals.
As in the FAQ, they seem to be saying there is no cure (as in some external treatment like a magic bullet), but you can nevertheless recover.
I do think that PACE's APT is fundamentally flawed* but I'm not sure it's due to a strong nocebo effect. 'There's no cure but natural recovery is possible' strikes me as a pretty reasonable position, partiucarly for a group of patients who have only been ill for a few years.
*it's the sticking to a '70% limit' of perceived energy that dooms APT to failure, in my view.
Now suppose you were told this at the beginning of a twelve month period, and at the end of it, while you may have avoided crashes, you have sensed no "natural recovery." Wouldn't you feel that you have been sold a bill of goods?
(Always remembering that a sizable portion of this cohort in fact probably had primary depression, for whom this particular line of treatment would have been disastrous.)
And recalling that this study was designed to "test" (=discredit) patient organizations' claims that pacing is the best intervention they have found to reduce crashes and symptom levels. Do patients among themselves tell each other that pacing will bring about "natural recovery"?
Do the designers of the PACE protocol themselves *believe* in this natural recovery? Of course not, because they believe that ME is a cycle of fear of activity and deconditioning, and that pacing will only perpetuate the disease. Where did they get this idea about "natural recovery" anyway?
I think one of the problems with APT is that they really jerk the patients around in terms of expectations. (and I agree with Oceanblue that, while there's a difference between what patients are told in the CBT/GET groups and the APT group in terms of recovery, the gap isn't a total gap; the APT group is told they will recover--they just are given inconsistent messages)
They first tell the patient (wwll, that isn't in the patient manual, but it is in the therapist manual and I assume they are to communicate this to the patient because they are continually told to remind the patient of the theory of pacing) that the theory of pacing is that CFS/ME is caused by a completely mysterious disease process which doctors are unable to test for (translation: it's all in your pretty little head, but we aren't going to be so rude as to tell you that directly OR it's martian-level science which won't be solved in your lifetime... up to you which you want to believe, BUT either way, don't expect any help from a physician).
Next they hand out a nifty pyramid diagram which basically implies that the cause of debility is the boom-bust cycle (it's not a mysterious disease process--it's your bad coping skills, but we aren't going to tell you that either because then we'd be impinging on CBT territory), and if you can regularize your activity (now they're impinging on GET) and avoid overactivity, the result WILL be empowerment and recovery. However they define recovery as a reduction in symptoms.
we can avoid disability? wow
Next they are told, rather than a mysterious biological cause, stress and overbreathing actually cause some of the symtoms of CFS/ME. (Again they impinge in CBT.) Next lack of sleep hygeine.
Other sorts of contradictions:
If the goal here is to conserve energy, why is a stool better than sitting in a chair?
I don't think there's an actual ergonomic reason for not reclining, so why is it there?
This article is in print form now in the April issue of The Psychologist. I'm going to try to write a letter in response. They publish quite long letters so it's worth a go.
Also, it would be worth thinking about submitting an academic article to this journal. Articles can be shorter than those in most journals. The Psychologist has a very large readership compared to most academic publications - over 40,000.
Guidelines for submitting are here:
I think you have to be a psychologist to submit a paper.
If anyone would like to collaborate with me to submit something, let me know. Before I had to give up work I was an academic psychologist and journal editor.
4th letter published this week.
I wonder will we top the 5 they published in reply to a piece last year.
Reply to thread post range 601-630
ancientdaze, I think you may be making some good points, but I'm afraid I lack the experience to fully comprehend them and put them to practical use on the PACE trial.
Some good points there. Again, I agree the phrasing of CBT and GET vs APT is a significant enough problem for the PACE trial, since being a psychological study using subjective questionnaires and aimed at changing patients' perceptions and beliefs.
If PACE really wanted to adequately test the "pathological model of ME" (as they claim they do in the APT arm), they shouldn't have used such a highly selective process that went to great lengths to exclude anyone with potential pathology or symptoms from the CCC which White simply doesn't like. 80% of candidates were excluded from the trial (admittedly, some of those refused). What they were really testing was the psychological effects of the "pathological model of ME" on people they believe have "abnormal illness beliefs", certainly not a CCC-like "pathological model of ME". White et al have never agreed with pacing, and APT itself was a classic strawman argument used to discredit a position that the ME/CFS community doesn't hold.
Indeed. On average the CBT and GET groups scored about the same as the 75-84yrs age group in Bowling et al. How they defined and calculated "normal" physical functioning was one of the more severe and obvious blunders made by the authors. Except a blunder is defined as "an embarrassing mistake" and this was not a mistake and the authors aren't going to be embarrassed when they are receiving funding and praise. For anyone willing to look beneath the gloss as we have been doing, it's bad enough to cast doubt on them as researchers, and no one except the ME/CFS community has noticed or cared.
What this shows us is that it's OK in mainstream academia to publish inappropriate definitions of "normal" which are clearly abnormal. As far as I am concerned, the reputation of the Lancet is in question over letting that through, along with the editorial confusing the original protocol's definitions of "positive outcome" and "recovery" with the relaxed definitions of "improvements" and "normal". And that's what happens when the Lancet doesn't bother to peer-review its editorials, factual errors escape through into print for the world to read.
So it seems that the only plausible explanations are: B&K didn't even bother to read the PACE Lancet paper close enough to notice PACE pulling the old switcheroo, or, basically they lied. That's pretty damning either way. It encourages me to distrust them, and the Lancet.
I think they are forced to be more "honest" in established diseases. CFS is more of a free for all. Anything that is "subjective" is more susceptible to spin and post-modernist interpretations.
My impression was that the GET group had a "safety net" from severe exacerbations because they were very cautious and while participants were encouraged to increase activity on a regular basis if possible they were not forced to do so, so it was possible to remain in the GET group at 52 weeks without any actual increases in activity whatsoever since we are not told how many patients managed to increase their activity levels overall. As you and others have implied, the data on adverse effects has also been obscured and may be worth investigating in more detail.
Thanks, that will be useful.
Thanks for that.
That sounds about right, 58 or 60/100 points was roughly the 15th percentile for the general population in Bowling et al, which includes the diseased and elderly. Wouldn't be surprised if it is more like 5% percentile for a healthy age-matched control group. And because of the skewed distribution we have the odd situation where roughly 75% of the population are scoring above the average! 60 may be the 15th percentile but 85 is still only (roughly) the 25th percentile, a figure White et al used to help define a full recovery but omitted from the PACE trial results.
Reply to thread post range 631-660
Yes, my mistake, I must have confused the outcomes.
Good point. Now I wonder if adverse effects from the 6MWT would have been ignored even at 0 weeks.
Researchers who assume CFS is largely a psychosocial deviancy involving a faulty perception of functional symptoms, I don't think they would be concerned much about response bias and "social desirability effects" in self-reports, along the lines of "as long as we can get these patients believing and behaving as if they were healthy then they effectively are so why worry".
Was interesting to see how they derive "normal" values for CFS vs other conditions, the suspicion of spin keeps growing when we see such obvious double standards. I was surprised to see that CBT seems to be a lot more effective at reducing fatigue in MS than CFS. Apparently so was Marco: "Doesn't this tell you everything you need to know about how CBT 'cures' fatigue? [...] Unless CBT miraculously produced physiological changes making participants healthier than healthy controls?". What it tells me is response bias is a major problem in psychological research and that the "fatigue" in CFS is different than in MS, which of course involves other disabling signs and symptoms.
I agree with your argument but when I looked at Figure 1 of PACE I arrived at n=251 for PF/SF-36 >65 exclusions.
But according to White, treatment shouldn't take years to work!
Reply to thread post range 661-690
This was good.
Great idea! The authors defended the 6MWT as "objective", and now they should live with the result. As oceanblue points out, GET compared to SMC does not even meet their own definition of a "clinically meaningful" improvement (+0.5SD). No group average hit the 400m mark while most comparable healthy people will score about 600-650m on average. The 6MWD values for all the PACE groups including the GET group at 52 weeks are similar to a wide range of serious medical conditions.
I wonder how many of the 15% of SMC participants and 28% of GET participants who reported "normal" fatigue and physical functioning (which we now know were still abnormal) also scored a more normal 6WMD of 600-650m? Instead we get this: "6-min walking distances were greater after GET than they were APT and SMC, but were no different after CBT compared with APT and SMC. [...] The objective walking test favoured GET over CBT, whereas CBT provided the largest reduction in depression."
So now we have a possibility that White et al used the wrong data? Not surprising when they didn't bother to use the appropriate age group either.
Excellent points. It is quite revealing that the PACE authors didn't bother to take this into consideration. Incompetence, or spin? Again, pretty damning either way.
I'd guess that some patients simply didn't want to get involved. However, being a relatively new bunch it is hard to say what they knew about CFS at the time of their invitation and we are not given reasons for refusal, "new" patients may be more "open" to biopsychosocial explanations, I doubt many long-term patients would sign up to CBT/GET research after having already been scorned by the biopsychosocial approach.
Good points. If they modified the London ME criteria and then claimed their results apply to all ME, I wouldn't be surprised if they did something similar to the "international (CFS) criteria" to claim their results apply to all CFS around the world.
Good question. I'd guess that the difference of suggestion is between "this WILL happen QUICKLY" vs "this MAY happen GRADUALLY".
The authors can always argue that CBT/GET aren't very effective because the patient has not managed to successfully challenge their abnormal illness beliefs. Afterall, CBT is not a silver bullet for any psychiatric diagnosis.
true that CBT is not a silver bullet for any psychiatric diagnosis. however the working model of the wessely school crowd is that cfs is not even equivalent to a psychiatric illness. an abnormal illness belief is a delusion, not an actual illness like a standard psychiatric illness.
of course they have so much doublespeak they are likely to claim in the self-same paper that (a) cfs is hysteria and (b) it would be awful for anyone to say cfs is somehow not real, but even whilst arguing the latter point the best they can do is compare cfs to the disordered self-assessment of anorexia nervosa. [this is not to suggest that anorexia nervosa is somehow not real, not serious, or not devastating, because it is all of those, but it does include an element of incorrect self-assessment, which is simply not the case in me/cfs (except for the part where the PWME is likely to overestimate her or his ability), but the point here is that cfs is not an inability to determine when one's self is truly ill and in fact disabled, which is what the psychobabblers are suggesting]
For a hysteria, claim (a), CBT should work just fine and ought to be a silver bullet
the fact that it is not a silver bullet calls their model into serious question, doublespeak and all--inability to determine when self is ill and disabled should also be readily correctable through CBT, supposing social support is provided which was ostensibly done in this particular trial, at least at a surface level (my contention is that this support is hypocritical since it involves actually disbelieving the patient while providing verbal support).
They cannot claim that the CBT model works, unless they have objective measures of behavioural changes, specifically actometer measurements.
Otherwise the change is merely 'cognitive', and potentially due to inflated self-efficacy and potentially due to 'response bias'. You could say that the patients didn't want to feel that they had wasted their time, as well as the time of those who provided the treatment and thus were likely to report subjective improvements on questionnaires. Hence the minimal effect shown on 5 year follow-ups.
Could you clarify how you did this and maybe give links to your online calculator. I used a pooled SD (combining SD of both SMC and GET) to get my figure of 0.34 but I'm not sure if I did this the right way. [ps the SF36>65 figure should indeed by 251, no idea where I got my figure from.]
CBT is very effective for some psychological disorders eg it has a large effect on generalised anxiety, PTSD and Depression (see this review of meta-analyses). Given this, the fact that they have clearly identified 'flawed thoughts', CBT is supposed to be good a tackling such flawed thinking and they've had 20 years to optimise their treatment I don't think they can realistically blame the patient.
Also, the high levels of patient statisfaction (82%) and strong therapeutic alliance (independently rated as 6.5/7) for PACE CBT suggests the therapy was implemented effectively; it's failure indicates a problem with the underlying biopsychosocial model.
As I recall both Prins and Chalder have made that assertion, and no doubt others have too in one form or other.
Of course, such an explanation is totally unfalsifiable. No matter how much CBT doesn't work, they can just keep saying you are not trying hard enough, which is a particularly nasty assertion as it can never be proved or disproved, leaving patients at the mercy of the subjective opinion of the vested interest ridden 'expert'.
The fact that they have to resort to such blatantly self-serving pseudo-science to defend their model is powerful ammunition against it, and they should be questioned vigourously about that whenever possible, without mercy.
It is just the establishment version of the same poisonous double-bind loaded drivel found in that New Age fraud The Secret.
Exactly. Quite the opposite, as I am sure you will agree.
You can also try a Google Site Search
Separate names with a comma.