Some of the “errors” that the authors mention are outright misleading.
1. Created a new definition of CFS/ME which automatically downgraded the certainty of trial evidence
I am puzzled by this argument. The
2007 NICE guideline (CG53) already required post-exertional fatigue or malaise with the typical 24h delay and prolonged recovery (§1.2.1.2). It also states that the diagnosis of CFS/ME should be reconsidered in the absence of PEF/PEM (§1.3.1.3). Therefore, had the PACE trial been published earlier and reviewed by NICE for the 2007 guideline, it should have been downgraded for indirectness because the Oxford criteria do not require PEF/PEM.
It seems important to note that the authors who were stakeholders in the 2007 guideline, in particular the first author and PACE trial lead investigator Peter White (via the St Barts CFS clinic), did not reject the mandatory presence of PEF/PEM. They cannot turn on their heels 15 years later.
Further, NICE merely modified the IOM criteria, which were one of the reasons why NICE started the process of developing a new guideline as mentioned in their
10 year surveillance report of the CG53 guideline.
Also, in one of the snippet above, the authors say that the “most widely used” CDC definition (Fukuda) doesn’t require PEM, but they omit to mention that the CDC now endorse the IOM criteria.
2. Omitted data from standard trial endpoints used to assess efficacy
Perhaps this refers to NICE’s decision to use the longest follow-up data available (144 weeks) that Simon Wessely
lobbied against by sending text messages to the former head of NICE.
If they want to use this argument, then surely one should have the right to invoke objective endpoints, which the authors tried to bury under the carpet by publishing them separately from the main results because they were null?
3. Discounted trial data when assessing treatment harm in favour of lower quality surveys and selective comments from qualitative sources
First, how should one interpret trial data on harms when participants were instructed to ignore and push through their symptoms, which they were told were the result of “false illness beliefs” and related maladaptive behaviour?
Second, it is known that reporting of harms in trials is often incomplete and this issue can only be compounded when the investigators have a strong bias in favour of the intervention they are trialing.
Third, 2274 patients responded to the
survey on GET/CBT that was analyzed by the Oxford Brookes University and submitted to NICE as evidence for the guideline development process. This is 3.5 times as many patients as the number of participants in the PACE trial (641). Why should such a strong signal be ignored even if it comes from non-RCT evidence? The
”First Do No Harm” report on harms from drugs and medical devices published in 2021 by the Independent Medicines and Medical Devices Safety Review was based on post-marketing (non-RCT) evidence, yet the harms were acknowledged.
Fourth, speaking of post-marketing evidence, there is no equivalent of the Yellow Card scheme for GET/CBT and ME/CFS clinics do not have a system to report harms.
https://journals.sagepub.com/doi/10.1177/1359105319854532
4. Downplayed the importance of fatigue as an outcome, even though it is the primary symptom of CFS/ME
Did they?
5. Failed to synthesise and GRADE trial evidence adequately
@MECFSSkeptic already responded to these allegations:
https://www.bmj.com/content/371/bmj.m4774/rr-9
With regards to synthesis, the appendices for the evidence review contain hundreds of pages of GRADE tables.
6. Misrepresented Graded Exercise Therapy as mandating fixed increments of change when the major trials defined it as collaborative, negotiated and symptom dependent
This is a blatant lie. NICE reviewed the trials carefully and summarized its findings on the definitions of graded exercise therapy in Box 5.
One can look at the
PACE GET manual for therapists. The “GET process” is summarized on pages 38 and detailed from pages 39 to 56. A fixed increment is mandated, first in duration (20%; p. 38, p. 49) then, once exercise can be sustained for 30 minutes, in heart rate by 10-20% (p. 49).
It is true that therapists are to collaborate with patients: “ALWAYS ensure the process is collaborative and that every stage is jointly negotiated” (p. 32). Collaboration is described as such:
PACE trial GET manual for therapists (p. 34) said:
Collaboration
Collaboration is an essential skill in working with people with CFS/ME. Up to the point of meeting you, many participants will not have been included in the management of their illness. They may not have been asked their opinion about what is wrong with them and may feel rather helpless and out of control. Collaborating throughout treatment will help participants to feel more involved in their treatment and will help them to regain some sense of control.
You will be demonstrating a collaborative style at your first meeting when you individualise the GET model to their illness. By this we mean drawing a model together, examining factors they think have been responsible for triggering as well as maintaining the illness. Agreeing an agenda for each treatment session, asking for their input in making suggestions for their activity programme and evaluating previous sessions will help participants to feel valued and included in the treatment process.
However, the extent of collaboration is very much limited by the GET model and process.
First, collaboration mostly revolves around negotiating goals before starting the therapy and a baseline of exercise (p. 38). But even then, it is stated that:
- “Goals for GET should have relevance to exercise, physical activity or physical functioning” (p. 45) and not be “a reduction or absence of symptoms” (p. 42), which may differ from the patients’ preferred goals (and also highlights that GET is not meant to be a cure for ME/CFS)
- “The baseline should be renegotiated if, after subjective assessment or review of the Physical activity/exercise diary, the therapist feels this level is inappropriate” (p. 48), so the therapist has the upper hand on deciding the baseline activity level.
Then, the process insists on maintaining activity during crashes, which are “normal, and likely” (p. 51), even when the patient is reluctant to do so, because it is presumed that “hurt does not equal harm” and, otherwise, “the boom/bust cycle continues, and the body is not able to desensitise to the increase in activity”:
PACE trial GET manual for therapists (p. 51) said:
A central concept of GET is to MAINTAIN exercise as much as possible during a CFS/ME setback. This is to reduce the many negative consequences of rest, and to allow the body to habituate to the increase in activity. If activity and exercise is reduced at this time, the boom/bust cycle continues, and the body is not able to desensitise to the increase in activity: which is, of course, an essential component of a graded increase in exercise and activity.
Although it can be difficult to encourage maintenance of exercise despite an increase in symptoms, participants usually are able to understand the reasoning behind this and are often pleased they were able to maintain activity during this time. It is important to explain that although they have an increase in difficult symptoms, ‘hurt does not equal harm’ (as you would do with somebody with chronic low back pain).
Some participants may be resistant to this approach, and will wish to reduce both activity and exercise during this time. If they cannot be encouraged to maintain their previous level of exercise, then encourage them maintain as much as they are able to, and work towards building up the activity/exercise as soon as possible. Additional support may be required at this time.
It is helpful to explain the theory behind maintaining exercise during a setback to participants BEFORE they have a setback if possible, as while they are symptomatic it may be more difficult to encourage. It is therefore useful to discuss and review a written setback plan, outlining useful information to follow in the event of a setback.
Therefore, going back to the authors’ claims about GET not having fixed increments and being collaborative, negotiated and symptom dependent:
- The increments are fixed (10-20%), whether in duration or in exercise intensity;
- The GET process leaves little room for collaboration before starting the program and almost none when it is ongoing — the patient is expected to comply with their exercise regimen and to maintain it even when they are experiencing a flare;
- How negotiation during the program should be done and what it entails is not specified; in particular, there is no mention of negotiating increments;
- To say that GET is symptom dependent while encouraging the patient to continue exercising even while they are crashing is disingenuous.
7. Deviated from NICE recommendations of rehabilitation for related conditions such as chronic primary pain
Here, the authors contradict themselves with regards to error 5.
NICE followed the same methodology for reviewing the evidence for the chronic pain and ME/CFS guidelines, so they synthesized and graded the evidence the exact same way. It makes no sense to criticize NICE’s protocol for the ME/CFS guideline but not for the chronic pain one.
Also, NICE’s evidence review of non-pharmacological interventions for chronic pain rated many of them as low quality (some moderate), just as in the ME/CFS evidence review. The decision to recommend exercise and psychotherapy in the chronic pain guideline was based on the committee’s clinical experience, as is clearly stated in the “Rationale and impact” section.
8. Recommended an energy management approach in the absence of supportive research evidence.
____________________________
GET was recommended in the 2007 guideline without definitive evidence of its efficacy. It should also be noted that this energy management approach can hardly be considered a treatment.
