IiME letter to Mark Baker (NICE) re: CBT & GET as recommended treatments

Thanks.
I posted them to this thread:
http://forums.phoenixrising.me/inde...ed-as-primary-treatments-for-me.37782/page-15

Also here:
ME/CFS: Graded Exercise Therapy. Experiences & Comments from 2015 (UK) MEA CBT/GET/Pacing Survey
https://www.pinterest.co.uk/tomkindlon/mecfs-graded-exercise-therapy-experiences-comments/

ME/CFS: Cognitive Behavioural Therapy. Experiences & Comments from 2015 MEA CBT/GET/Pacing Survey
https://www.pinterest.co.uk/tomkindlon/mecfs-cognitive-behavioural-therapy-experiences-co/

 

Its going to be a real challenge for NICE to review issues with the PACE trial and continue with the current recommendations but at the same time by admitting PACE is flawed they will be admitting peer reviewed published work in the Lancet funded by the DWP is garbage.

As NICE have already replied that one trial doesn't disqualify all others the issue of the criteria used could be crucial in how we win this argument.

 

Today a psychologist said to my husband (both in acadamia), that they pick data so as to give "good results" (I guess p < 0.05), and if they published raw data (as required for data archiving) people would see that they manipulated. That's why they cannot publish their raw data.

 

Perhaps this paper by Nacul et al helps in arguing against the Oxford criteria,

How have selection bias and disease misclassification undermined the validity of myalgic encephalomyelitis/chronic fatigue syndrome studies?

The seems like an argument for not using any evidence that relies solely upon the Oxford criteria, assuming one can argue that the CCC are useful.

Based on this, it does appear that if we can get the Oxford criteria thrown out, that would effectively knock out CBT and GET.

 

Nacul's argument is an important one.. The other thing that strikes me is that if it is agreed that PEM is central to the category of ME/CFS, CFS/ME or ME+CFS that the guidelines will be for then there is an immediate problem with studies that voluntarily recruit patients to possibly receiving GET. It is highly unlikely that People with PEM will want to be included. And Oxford allows all the people without PEM through, to have the effect of GET on them studied. This is a consistent problem with all the Oxford studies that include GET or GET and CBT.

So Oxford is no good if trial design biases out all (or even most of) the Canadians. So although Oxford is reasonable to use on a purely set theory basis, if we know that the probability of the right subset being included is seriously biased against we have a positive reason to reject an Oxford sample.

It might also be that people with PEM would not sign up for CBT since they would be the ones, in the BPS words, most convinced they have a physical illness (because it feels like it) and therefore sceptical about CBT.

I wonder if there is any chance of discovering the PEM status of the recruited and not recruited patients in any of these studies. I suspect not.

Edit: I realise that some of this is probably what Large Donner has been trying to persuade me of, but I needed to find a way of thinking it that required no need to prejudge ill intent.

 

Would you be willing to explain this?

Also, I get confused. Didn't I understand you correctly that from a scientifc view it is correct to use Oxford criteria and transport the results to ME? I.e. didn't you mean that Oxford-CFS is a superset of ME? If so, I realize I didn't understand why that is correct?

 

Yes, despite Valentijn's reasonable protestations I think Oxford is roughly a superset of ME. So you can apply inferences drawn from an Oxford population to ME if you have no reason to think there is an important confounding factor that skews the probabilities. But if such a confounding factor is staring us in the face, and in fact we have debated it several times before, then the same inferences are not reasonable to apply.

 

I think for a mathematician this is hard to understand. But I hope I start to understand your way of thinking nonetheless. Thanks for explaining.

 

I think this is where I have been confused as I haven't considered Oxford as a superset but as a large, loosely defined set where there are some common members with other separate sets such as ME.

As far as I can tell I would have been excluded from Oxford, even though there is little doubt about the diagnosis of ME /CFS.

 

Doesn't Oxford criteria exclude those with neurological signs? This sounds like a permission to exclude many ME/CS patients, because I'm pretty sure that a determined doctor could easily find some neurological abnormalities. I can't walk straight with closed eyes for example (tandem gait test it's called I believe).

 

Reply from Mark Baker and response to it from IiME - http://www.investinme.org/IIMER-Newslet-1801-01.shtml#IiMER-reply18-1

"protect what is good" I've no idea what he's talking about here, I've yet to experience anything good from the NHS while I have progressed from mild to the severe side of moderate.

 

This is my whole point.

To declare you diagnosed with CFS to dump you they ignore all the observable neurological signs from the examination and fail to document them in your medical files hence it adds to the argument that you are medically unexplained and disqualifies you from testing.

Yet when they want to run a CBT GET trial to "prove" their treatments they wouldn't let you within a million miles of the trial because as White says in his own words.....

CFS defined simply as a principal complaint of fatigue that is disabling, having lasted six months, with no alternative medical explanation.

He couldn't make it any clearer that's what he means and its a built in disclaimer.

 

That's why I agree with @large donner and others that ME is not a subset of Oxford-CFS.
If I understood correctly, PEM is excluded in Oxford-CFS, but it's an ICC symptom. So, ICC-ME is no subset.

The more I think about that - and having in mind Mr. Bakers words - the more I think that is correct. And in a way, it fits what the psychologist said about picking data to get nice results.

 

Perhaps this is why there are recent efforts to trace and check status of PACE participants? And NOD data?

 

Then could we simply argue that Oxford is invalid because it does not require PEM?

Proving the need to use PEM in any criteria might be relatively straightforward, as there seems to be an increasing consensus on that.

For example, The BMJ Best Practice Review from just last month states that "PEM is the hallmark of CFS", and that the Oxford criteria are inappropriate:

http://bestpractice.bmj.com/topics/en-us/277

(My bold).

If PEM is agreed, and demonstrable as, essential to ME/CFS, and therefore to any trial criteria, then Oxford fails, the trials that rely on Oxford fail, and CBT and GET are gone.

So the question might be, is there any evidence to suggest that PEM is not a requirement in ME/CFS. If there is none, while there is plenty for it having a core role, or key place, in the condition, then Oxford doesn't have a leg to stand on.

 

Looking at Figure 1 on the original PACE paper, after the first set of exclusions were carried out, it left 1460 patients. It seems as though 533 of those failed the consent part (mostly the patients refused, but 46 were on doctors' advice). We don't know why, but it's a substantial proportion (over a third). A further 29 were dropped for unrecorded reasons.

On the second run through, where the 1460 - 533 - 29 = 898 were further reduced to 722, a further 69 failed the consent part (mostly the patients refused to take part, but 2 were withdrawn on research assessors' advice). Together with a further 12 that were unrecorded, this brings the total down to the 641.

I'm sure there's a lot of interesting data in here, but we will never be able to get hold of it.

 

It does seem as if we are getting a message through to him though.

This sounds like the something is better than nothing approach.

He doesn't seem to realise the ME services he is concerned about withdrawing may be the very ones offering CBT and GET.

 

PACE authors claimed that about half of their Oxford patients also met London criteria, which includes a vague form of exercise intolerance:

I think it comes down to the disease being defined. NICE purports to describe diagnosis and treatment for ME/CFS, and sort-of requires PEM as a symptom. It's ridiculous for NICE to then use criteria for a different condition to determine treatments for ME/CFS. Oxford fatigue studies are applicable to patients with Oxford fatigue, whereas studies honestly using NICE criteria or similar (PEM or at least exercise intolerance sort-of required) would be applicable to patients who can be diagnosed by the NICE criteria.

I think part of the problem is that NICE is trying to combine chronic fatigue along with ME/CFS into one illness. On one side are patients who are principally fatigued, and on the other hand are patients who have a specific form of exercise intolerance. Both groups need descriptions, diagnostic criteria, and treatments, but mashing them together has resulted in bizarre recommendations that are unlikely to be particularly helpful for either group.

Perhaps what is needed is to show that Oxford and PEM patients are definitively distinct - different symptoms, different impact, different biochemistry, different needs. I think there has been research comparing symptoms and severity of different groups, but not biomedical or other measurements. However comparing different studies using different criteria regarding biochemistry, etc, might also be productive.

 

Also if they didn't report the results or provide the data separately for the two claimed different definitions why did they bother using two definitions.