S4ME: Submission to the public review on common data elements for ME/CFS: Problems with the Chalder Fatigue Questionnaire

[Graham]

The PACE trial analyzed their step test in two ways: one effectively measured energy expended against increase in heart-rate, and the other used the Borg perceived scale of effort. Both clearly showed no difference between the groups with CBT or GET and the group without.

 

[Lucibee]

meandering over a boggy field

Like I felt reading this paper just now... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1739950/pdf/v057p00353.pdf

 

[Barry]

Like I felt reading this paper just now... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1739950/pdf/v057p00353.pdf

A quick look at the Conclusions and Discussion seems all they could identify is that it is an extremely blunt instrument, primarily able to discriminate between people who are fatigued and those who are not . And although the claim of evaluating change over time ...

According to Kirshner and Guyatt17 questionnaires can be used for three purposes: (a) discriminating among subjects, ( b) predicting prognosis, and ( c) evaluating change over time.

... there is no proof in there that the degree of change can be evaluated. e.g. If someone scores 60 on one day, and scores 30 six months later, there is no proper proof that halving the number equates to their fatigue actually being half what it was before; any such assumption is just that ... an assumption.

 

[Lucibee]

It was used as the *primary outcome* in the iCBT trial published a couple of days ago.

 

[Barry]

It was used as the *primary outcome* in the iCBT trial published a couple of days ago.

I think it comes down to: they know no better, have nothing better, but totally believe in their supreme knowledge and infallibility. Very narcissistic.

 

[Adrian]

It was used as the *primary outcome* in the iCBT trial published a couple of days ago.

It can't be worse than the CFQ can it?

I think certainly not suitable for a primary outcome.

 

[Invisible Woman]

.According to Kirshner and Guyatt17 questionnaires can be used for three purposes: (a) discriminating among subjects, ( b) predicting prognosis, and ( c) evaluating change over time.

Bolding mine: - first read that as discriminating AGAINST subjects......

 

[Woolie]

I would say yes. Your body is facing an additional burden, fighting off the infection. It has extra work to do, and feels the consequences, whether it's voluntary work or not.

Perhaps, for those at the very extreme end of severity are in the situation where the basic and minimum body function needed to stay alive is triggering PEM.

Gosh, that's a stretch of the concept of PEM though. That makes it considerably broader - in a way that might make it even harder to define.

(my symptoms are massively affected by infections too).

 

[Lucibee]

It can't be worse than the CFQ can it?

Oh I think it can. Take a look... (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1739950/pdf/v057p00353.pdf) and they can't even get the wording right in the instructions of how to fill the damn thing in!

The more questions you add, the more subtopics you cover, the more imprecise your 'scale' will become. There are multivariate statistical models that will capture the complexity that one probably needs to collect to model 'fatigue' and associated symptoms. But you lose all that information when you reduce it down to a single number. It becomes a proxy of wishy-washiness.

 

[Adrian]

Oh I think it can. Take a look... (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1739950/pdf/v057p00353.pdf) and they can't even get the wording right in the instructions of how to fill the damn thing in!

The more questions you add, the more subtopics you cover, the more imprecise your 'scale' will become. There are multivariate statistical models that will capture the complexity that one probably needs to collect to model 'fatigue' and associated symptoms. But you lose all that information when you reduce it down to a single number. It becomes a proxy of wishy-washiness.

Yes I don't get these scales - sorry the rest of this post is a rant about questionnaire scales in general!

I read some stuff on Likert scales and it was very clear ask the same question multiple ways in order to reduce answer error.

What some of these 'scales' ask is completely different questions in roughly the same area. So I feel they shouldn't be adding up the values and calling it a scale. I'm thinking there are issues of linearity and correlation between questions which means that a 1 point change in an answer could represent very different things. Also when one answer is changed from certain answer sets others are more likely to also change.

So for example, looking at the SF36 scale is the difference between having difficulty walking a 'block' and not having difficulty the same in terms of physical function of having difficulty walking a mile or not. Also when I looked at the ONS sf36 data there wasn't a good ordering with answers especially at the middle of the scale where there were many different answer sets giving the same score - to me that suggests that the answers to questions may be closely related in terms of physical function - hence the SF36 is a bad proxy for physical function. Or the notion of a scale for physical function is just not valid.

What surprised me when I looked there seem to be some work on this but not much and it was hard to find. What I didn't find was any form of sensitivity analysis that looked at different possible effects and how that may reflect on summary measures, significance checks or effect sizes.

I do see economists combining multiple things into a utility function (the eq5d is a good example) but here the utility function has some form of justification for example is surveys about how to weight different aspects. I think the eqd5 used country based surveys and then a linear model to compare results, the model wasn't a great fit which suggested to me that there should be a notion of accuracy below which differences had no real meaning (could be model errors rather than real differences).

 

[Lucibee]

It seems that some psychologists seem to be remarkably bad at paying attention to advances in their own field. I'm currently reading Daniel Kahneman's Thinking, Fast & Slow - and so many times I've thought, "why isn't anyone applying this?". One example is the difference between the 'experiencing' and 'remembering' self, which is crucial to distinguish between when doing surveys. All these surveys fail because they seem to hopelessly mix them up.

His TED talk is here. The talk is mainly about happiness, but it's equally applicable to pain or fatigue.

 

[Adrian]

It seems that some psychologists seem to be remarkably bad at paying attention to advances in their own field. I'm currently reading Daniel Kahneman's Thinking, Fast & Slow - and so many times I've thought, "why isn't anyone applying this?". One example is the difference between the 'experiencing' and 'remembering' self, which is crucial to distinguish between when doing surveys. All these surveys fail because they seem to hopelessly mix them up.

His TED talk is here. The talk is mainly about happiness, but it's equally applicable to pain or fatigue.

I've not read this book but I read some of his papers a few years ago and thought they were good. I have a feeling his work was one of the few pieces that one of the replication projects managed to reproduce.

 

[Barry]

Perhaps, for those at the very extreme end of severity are in the situation where the basic and minimum body function needed to stay alive is triggering PEM.

I think this is very likely true. The more severe a persons's ME is, the less energy is available to their whole body, including running their basic life support systems. If their energy availability drops so low, that just to stay alive they are going beyond PEM threshold, then bang - just staying alive maybe pushes such a person into PEM, which would be the most appalling vicious circle. I think of this life support energy requirement as akin to the power an engine needs just to tick over; if the energy flows slower than that, the engine starts to run poorly, and at some point as power reduces even more, the engine 'dies'.

 

[Suffolkres]

Another consideration. CCGs?/NHS ME and CFS Services have to PAY to use the Chandler Scale I am told as it is a "commercial product". Patients find it lengthy, onerous and challenging- and for what outcomes....?

It should be discontinued on lack of cost effectiveness and for dissipating the few precious £ spent on any ME services. We have tried (unsuccessfully) to challenge local commissioning asking them to address this within the service specification and contractual arrangements. CCGS have "patient participation" within their target for the year locally but, watch this space- they are not really taking our participation seriously. We have challenged the use of this scale within the business model for the 7 CCGs across Norfolk & Suffolk.

 

[Trish]

I'm astonished that a list of 11 back of the envelope statements with tick boxes for less, the same, more, much more than usual when well, could be considered a commercial product and charged for, especially as they are freely available on line. That is outrageous:

1. Do you have problems with tiredness?

2. Do you need to rest more?

3. Do you feel sleepy or drowsy?

4. Do you have problems starting things?

5. Do you lack energy?

6. Do you have less strength in your muscles?

7. Do you feel weak?

8. Do you have difficulties concentrating?

9. Do you make slips of the tongue when speaking?

10. Do you find it more difficult to find the right word?

11. How is your memory?

At first I wondered, how could filling that in be onerous? Then I realised it's not because it's a long list with lots to read, it's because it's nonsensical. How can you have a symptom less than when fully well? What is the difference between more and a lot more? If I say, 'a lot more' now, what happens if my symptoms get worse? And how do you answer 'How is your memory' with any of the 4 options? I have no idea how to fill it in.

 

[Sasha]

I'm having trouble reading at the mo and can't tell whether the NIH have now dropped the Chalder scale (they've just given an update on what they're doing but I can't read it).

 

[BruceInOz]

I have no idea how to fill it in.

Yes, that "How is your memory" question makes the whole thing seem ludicrous! How are you supposed to answer it? What does "My memory is more than usual" mean? To fit with the other questions, answering more than usual should imply more impairment but it sounds more like saying your memory is better than usual!

 

[Luther Blissett]

If I may butt in here with an observation that may or may not help. Haven't been able to read this thread as it got too big before I noticed it.

The whole problem of the CFQ in my opinion, is that it is not measuring what other people think (not most with the illness) it is because the assumptions are not mentioned explicitly.

• It is a measure of the subjective cognitions of a phobic person.
• It measures the subjective cognitions because in a phobic person, the cognitions are the target of treatment.

For that reason, it doesn't have to be objective.

If you were trying to cure a person of a fear of heights for example, moving them on from refusing to stand near a suspension bridge to agreeing to walk, however far, onto the bridge is a success.

Any improvement in a subjective scale from 'not thinking about heights' to 'confidence to walk halfway across and admire the view' would be viewed as a success. Without the improvement in the scale, the actual physical (objective) measures that follow are useless and not sustainable.

The basic point is that the CFQ and scale are useless to us because we are not activity phobic. The BPS does not care if there are no objective measures in a study, because they believe that the subjective cognitions are the hard part of the treatment, and objective results will follow on from an improvement in subjective measures.

This post is probably redundant, but I get frustrated by people pondering why it is a popular measure and used so often.

 

[Valentijn]

I'm having trouble reading at the mo and can't tell whether the NIH have now dropped the Chalder scale (they've just given an update on what they're doing but I can't read it).

It's not listed by the NIH in their data elements for ME/CFS. The report from the fatigue subgroup doesn't mention it, even as "exploratory", despite that two other fatigue scales only previously used in research with crappy criteria were listed as exploratory:

In a few instruments, reported thresholds for ME/CFS were based on studies using case definitions that have since been downgraded by the NIH Pathways to Prevention report or the 2015 Institute of Medicine report because those definitions could encompass patients with other conditions. Those instruments are both rated exploratory and a note added about the issue.

The message may be that the CFQ is too bizarre and substandard to consider it as potentially being useful, even if validated in real ME/CFS patients.

 

[Sasha]

The message may be that the CFQ is too bizarre and substandard to consider it as potentially being useful, even if validated in real ME/CFS patients.

Thanks! I wish the NIH would be clearer about it, if only to send a message that they're actually paying attention to and acting on feedback, which would make them look good and show that they welcome high-quality submissions from patients.