NICE guideline for ME/CFS is unethical – Dr Diane O’Leary, Kennedy Institute of Ethics | 23 August 2

[Barry53]

I say this with some feeling because my wife is only alive today because her psychiatrists ignored explaining and got on and gave her ECT under section.

This is a perfect illustration of how medical professionals have no choice but to play god sometimes, it goes with the territory. That brings huge responsibility, and the need for tremendous integrity and ethical awareness etc. No one is perfect and no one can be expected to get it right all the time, but should be expected to always try their damnedest, no matter where in the clinical treatments chain.

If NICE are still recommending treatments based on evidence that is now much more questionable than when the guidance was originally issued, someone somewhere is playing god for the wrong reasons.

 

[Jonathan Edwards]

Hi Jonathon. There are patient surveys carried out by several national ME associations involving 1000s of participants, showing that GET/CBT are ineffective and even harmful. Surely NICE should take this evidence on board? I'm not aware of any counter surveys involving large patient numbers showing CBT/GET to be beneficial. And of course there's the PACE trial and we all know that on reanalysis the PACE recovery figures were found to be bogus. NICE is continuing to recommemd these treatments despite patients saying for over 20 years that these don't offer any benefits.

Patient surveys are no use to us. The fact that PACE is scientifically unsound rests on the likelihood of bias in the analysis. Patient surveys are also likely to be biased so they are just as bad. Patients may well believe they have been harmed but the problem is that people believe all sorts of things and that is why we have methods for eliminating bias. If NICE listened to the thousands of people who say vaccination is harmful then vaccination would stop and tens of thousands of children would die of measles and whooping cough. The only way to make a watertight case against NICE is to exclude anything that might be as weak as what they are relying on.

The PACE trial is uninterpretable so NICE should not base policy on it or the other similar studies. That has always been the case. The problem is simply getting NICE to follow its own guidelines about quality of evidence.

 

[Jonathan Edwards]

I appreciate that point, but something feels wrong to me. The notion of "new evidence", surely, should not just encompass additional supporting evidence, but also subsequently-negated previously used evidence. Else it's a mockery.

Surely evidence that is subsequently proven to be invalid, only becomes so by the introduction of further new evidence. The allowable new evidence must include any new evidence that calls previously used evidence into question.

I can see that NICE can reasonably feel, as for a court of law, that it has to be able to stick with a decision and not allow constant appeals. But since decisions of this sort are always based on opinion in the end there is always the risk that a bad decision stands. As with the law lobbying and publicity us sometimes needed to get a review of an old decision without new evidence. Since NICE can dictate its own rules it is ethically bound to do whatever is just and fair - which in this case is to review the decision.

The invalidity of PACE was apparent right from the outset. In fact it was apparent before the trial was completed. There is no new evidence as such. The claims of recovery have been shown to be bogus but the value of CBT and GET should not have been assessed on any claim for notional recovery. I assume that NICE would have based its decision on the 2011 paper. So the only way forward is to say to NICE, sorry, your rule of new evidence just is not fair and needs to be waived in this case. We did that for anti-TNF drugs, we did it for rituximab and for lots of other things.

NICE budge when the pressure is enough. But that will only work if they are taking advice from people without a pro-PACE bias. The problem is a person politics one. To my mind you can only win a political argument by showing that you are thinking more clearly than the other side. If thinking gets muddled with weak arguments of the sort the other side use then there is not much hope.

 

[Jonathan Edwards]

If NICE are still recommending treatments based on evidence that is now much more questionable than when the guidance was originally issued, someone somewhere is playing god for the wrong reasons.

Yes. As I said, the evidence was just as questionable at the start, but that has become more widely understood. Moreover, the circumstantial evidence for inappropriate motivation has increased. If authors go around comparing their critics to climate change deniers something is not right. Someone is playing God for the wrong reasons. We know who and we have a pretty clear idea what the wrong reasons are. But NICE should make its decision on evidence grounds and that is simple because the evidence was never there.

 

[trishrhymes]

@Jonathan Edwards I understand what you are saying about NICE only being interested in whether a treatment works, and works cost effectively. I assume this also includes taking into account whether a treatment has what they judge to be too many unacceptable side effects in large numbers of patients, or having the opposite to intended effect.

For medicines, I understand there is a reporting system so that if side effects show up only after the medicine is released into the community, a recommendation can be caveated with cautions to look out for that side effect, or recommendations made about particular groups who should not be given that drug. Or if bad enough the recommendation in NICE for that treatment can be withdrawn.

But how do we report side effects of GET and CBT if the doctors and researchers ignore reports of PEM after exercise and long lasting worsening of symptoms and reduced energy capacity? Is there a reporting system for side effects of psychological or exercise therapies?

I do understand the point about patient surveys not being accepted as evidence because they are self selecting and subjective reporting. I wonder whether there is any way around this apart from all patients undergoing these therapies wearing actometers and keeping symptom diaries to provide evidence of long term compliance and effects.

I wonder whether you have any suggestion of how we could overcome the problem of compiling evidence of harms acceptable to NICE.

On a separate point, you say there is no further evidence beyond what was known in 2007. I take the point that since PACE was a non scientific trial, its evidence is worthless. However, I do think it provides useful evidence - though not of the sort White and co wanted.

Along with the FINE trial, PACE clearly shows evidence that however hard the researchers tried to bend the trial to get the outcomes they wanted, they failed to do so. In other words, they showed no long term benefits for either treatment.

For FINE this was evident at 1 year, for PACE it was evident at the 2 year follow up. Even without the reanalysis using the protocol specified outcomes, the PACE and FINE authors themselves showed the treatments don't work. Surely this should be enough to get the recommendations withdrawn.

 

[Sasha]

The premise under which CBT & GET are offered is that patients have false illness beliefs.

That I think is a very fair point for CBT. I am not sure it is quite true for GET since that is in part based on the idea of deconditioning and fear of exercise.

From the 2011 Lancet paper:

GET was done on the basis of deconditioning and exercise intolerance theories of chronic fatigue syndrome. These theories assume that the syndrome is perpetuated by reversible physiological changes of deconditioning and avoidance of activity. These changes result in the deconditioning being maintained and an increased perception of effort, leading to further inactivity. The aim of treatment was to help the participant gradually return to appropriate physical activities, reverse the deconditioning, and thereby reduce fatigue and disability.

I think it's unavoidably implicit here that GET is also based on 'false illness beliefs' - specifically, the patient's belief that they have an organic disease and that their feeling of getting worse after activity is due to that disease rather than to (reversible) deconditioning.

 

[Jonathan Edwards]

For FINE this was evident at 1 year, for PACE it was evident at the 2 year follow up. Even without the reanalysis using the protocol specified outcomes, the PACE and FINE authors themselves showed the treatments don't work. Surely this should be enough to get the recommendations withdrawn.

I think it is a good point that the PACE follow up paper fails to show long term benefit.

 

[user9876]

No new evidence has appeared since 2007 as far as I am aware.

That in itself is quite a statement - I think with PACE there is more evidence that CBT/GET don't work or maybe just more evidence that the proponents of these treatments can't run trials good enough to give solid evidence.

We should be reflecting that back to groups like the MRC. 10 years on there is no more evidence to support potential treatments of a disease with no real treatments. That seems quite damning of the funding authorities.

 

[Sean]

I think it is a good point that the PACE follow up paper fails to show long term benefit.

For patients with chronic conditions the long term outcome is the primary endpoint, so no long term benefit is the central empirical result from PACE (and GETSET).

 

[docsimsim]

Patient surveys are no use to us. The fact that PACE is scientifically unsound rests on the likelihood of bias in the analysis. Patient surveys are also likely to be biased so they are just as bad. Patients may well believe they have been harmed but the problem is that people believe all sorts of things and that is why we have methods for eliminating bias. If NICE listened to the thousands of people who say vaccination is harmful then vaccination would stop and tens of thousands of children would die of measles and whooping cough. The only way to make a watertight case against NICE is to exclude anything that might be as weak as what they are relying on.

The PACE trial is uninterpretable so NICE should not base policy on it or the other similar studies. That has always been the case. The problem is simply getting NICE to follow its own guidelines about quality of evidence.

Thank you. So we need a placebo controlled double blind large-ish study to test the efficacy of CBT/GET. Et voilà, we already have that in the form of NICE, and the MRC/DWP funded it themselves. And it showd absence of efficacy or failed to prove presence of efficacy because it had hugely flawed methodology. NICE are just taking the easy option and procrastinating. And now they have the added excuse of waiting for NHS FITNET results, which aren't expected till atleast 2022!

 

[joshualevy]

No new evidence has appeared since 2007 as far as I am aware.

That's not true. Just off the top of my head there was Sandler and Worm-Smeitink in 2016, Wiborg in 2015, Janse, Schroder, and Qure all published in 2017: That's six studies in the last three years. The problem is they all reinforce the NICE guidelines. They all show various forms of GET and CBT work, and don't cause serious side effects. There is plenty of new evidence. The problem is, it all supports the NICE guidelines as they exist now.

As long as studies supporting GET and CBT are published at a rate of 2 per year, and studies showing GET and CBT are dangerous and/or are not the best treatment are published at a rate of never, there is no reason for NICE to change anything.

The real question (and it is an important question) is this: Why has there not been a single controlled, intervention study published in the last few years that found that GET or CBT did not work? There have been six that show it does work.

 

[Hutan]

The real question (and it is an important question) is this: Why has there not been a single controlled, intervention study published in the last few years that found that GET or CBT did not work?

Umm, because anyone with any sense and/or not determined to prove that GET and CBT work at all costs has better things to do and more respect for people with ME than to do yet another study of GET and CBT?

Especially given that both PACE and FINE proved that CBT and GET don't bring any significant benefit.

If you find us a CBT or GET study that does not suffer from the following errors, then let's all have a good look at it:

• Overly inclusive participant criteria
• No long term outcomes (this is a fluctuating disease)
• Inadequate controls (this is a fluctuating disease)
• Inadequate objective measures/ reliance on patient self-reporting
• Inadequate reporting of the data, especially to do with reporting of patients who did not complete the trial

 

[GreyOwl]

If you find us a CBT or GET study that does not suffer from the following errors, then let's all have a good look at it:

• Overly inclusive participant criteria
• No long term outcomes (this is a fluctuating disease)
• Inadequate controls (this is a fluctuating disease)
• Inadequate objective measures/ reliance on patient self-reporting
• Inadequate reporting of the data, especially to do with reporting of patients who did not complete the trial

I'd just like a plausible model for how CBT/GET can influence objective immune dysfunction.

 

[A.B.]

The real question (and it is an important question) is this: Why has there not been a single controlled, intervention study published in the last few years that found that GET or CBT did not work? There have been six that show it does work.

Because CBT/GET supporters are all following the same study design recipe which allows plenty of bias to influence the results. In an unblinded study with subjective outcome measures (and an inadequate control) something like this should be expected to happen. Asthma patients say they feel better even with a placebo, but if you measure how well they can breathe only the medication is effective. All the CBT/GET studies are designed to exploit this effect (the authors doing them may just be copying previous studies without realizing what they are doing). The objective measures of health and function in CBT/GET studies show no or disappointingly small results, and even the subjective improvements are temporary. CBT/GET are a placebo.

They also lie about their results. See for example the FITNET trial http://www.mdpi.com/2076-328X/7/3/52

 

[Barry53]

There is plenty of new evidence.

But is any of this supposed "evidence" based on objective outcomes? Presumably not. Just the usual smoke and mirrors subjective guff that convinces those so desperately eager to be convinced.

 

[joshualevy]

If you find us a CBT or GET study that does not suffer from the following errors, then let's all have a good look at it:

Actually, I think it would be a great idea to compare all the studies which show that GET/CBT is the best treatment, to all the studies which show that some other treatment is better (any other treatment). I've already listed six which found that GET/CBT was better, why don't you post a list all all the studies which show that some other treatment (like "within the envelop pacing" or Rituximab or anything else) is a better treatment. We can then compare which clinical trials use better methodology, more patients, better end points, etc.

 

[A.B.]

Actually, I think it would be a great idea to compare all the studies which show that GET/CBT is the best treatment, to all the studies which show that some other treatment is better (any other treatment). I've already listed six which found that GET/CBT was better, why don't you post a list all all the studies which show that some other treatment (like "within the envelop pacing" or Rituximab or anything else) is a better treatment. We can then compare which clinical trials use better methodology, more patients, better end points, etc.

First, please acknowledge the methodological problems in CBT/GET studies or explain why you disagree. There is no point comparing clinical trials with a person that doesn't understand how a reliable clinical trial looks like.

 

[joshualevy]

But is any of this supposed "evidence" based on objective outcomes? Presumably not. Just the usual smoke and mirrors subjective guff that convinces those so desperately eager to be convinced.

Are you talking about GET/CBT or Rituximab?
For Rituximab the answer is none. No studies published to date show objective improvements compared to a control group.
For GET/CBT, the answer is yes. I found two (so far) published studies which reported objective improvements compared to a control group.
For "within the envelop pacing" the answer is none. No studies published to date show objective improvements compared to control groups (or GET/CBT groups).

Why don't you publish the objective studies show that pacing or Rituximab are better, if measured objectively?

 

[Esther12]

Actually, I think it would be a great idea to compare all the studies which show that GET/CBT is the best treatment, to all the studies which show that some other treatment is better (any other treatment). I've already listed six which found that GET/CBT was better, why don't you post a list all all the studies which show that some other treatment (like "within the envelop pacing" or Rituximab or anything else) is a better treatment. We can then compare which clinical trials use better methodology, more patients, better end points, etc.

What would be the point of that? Most patients recognise that there is no good evidence for an effective treatment for CFS. NICE should not be recommending any treatment at the moment (if the multi-centre rituximab trial is positive it could be worth re-examning the guidelines, but I don't think anyone is aguing NICE should recommend rituximab in its guidelines now).

The problem is that a tollerance for junk-science is allowing some quacks to make money from unfounded claims of expertise and treatment efficacy, telling patient how it is that they should respond o their ill health when they do not have good evidence to support their claims. That should be stopped regardless of whether or not there is some alternative treatment with a strong evidence base.

 

[A.B.]

@joshualevy it's the combination of lack of blinding and subjective outcomes that is problematic, especially if there is an inadequate control group that would be expected to produce weaker nonspecific effects. Under these conditions it's very easy to get a positive result for almost any intervention, even homeopathy. Since virtually all CBT/GET studies are designed like this, it doesn't matter how many of them report that CBT/GET work better than some inadequate
control intervention.

With effective blinding, subjective outcomes are perfectly fine.