Exploring the validity of the chalder fatigue scale in chronic fatigue syndrome - 1998 Morriss et al

Interesting. Do you have a link to the trial?

 

I do and meant to include it! https://academic.oup.com/rheumatology/article/42/9/1050/1772059

It's Tench et al 2003 Fatigue in systemic lupus erythematosus: a randomized controlled trial of exercise

I've put the link in my original post too.

 

Thank you!

 

So they're saying two measures agreed (and gave a null result) and one differed so therefore the one that differed must be correct and the other two are wrong for reasons? You couldn't make this stuff up!

 

we were also discussing it at the other place in 2017
https://forums.phoenixrising.me/thr...medicine-again-by-d-tuller-et-al.50158/page-5

the original paper on the development of the Fatigue scale
http://simonwessely.com/Downloads/Publications/CFS/32.pdf

 

old thread: S4ME: Submission to the public review on common data elements for ME/CFS: Problems with the Chalder Fatigue Questionnaire
https://www.s4me.info/threads/s4me-...-with-the-chalder-fatigue-questionnaire.2065/

 

Why didn't Graham tell us he had it all sorted?
At least we may have come to the same conclusions independently!

 

I think the Chalder fatigue scale is misleading because in chronic conditions, patients cannot accurately do what the scale asks them to do, which is to compare their current state with their pre-illness state.

I don't have a memory of how being healthy felt like. What I think happens is that the brain, when it is asked for memories it doesn't have, simply makes the best guess. Maybe this best guess is further or particularly influenced by expectations (patients having been told that the treatment will improve their fatigue etc)

 

How yer doing? (visual analogue) is more relevant and understandable than the Chalder Fatigue Scale.

 

It is important to point out that there is a big difference between correlations at a static point in time and correlations in an unblinded prospective clinical trial which is inherently subject to various biases.

 

I looked up other publications by the author of this article (Craig Jackson) and found this one:

Variation and Interactional Non-Standardization in Neuropsychological Tests: The Case of the Addenbrooke's Cognitive Examination.

https://www.ncbi.nlm.nih.gov/pubmed/31550997
https://sci-hub.tw/10.1177/1049732319873052

interesting that did not apply any of this to the interpretation/administration of the CFQ. (eg instructions on how to interpret 'as usual')

(I think this is him [there's an email contact address]
https://www.bcu.ac.uk/social-sciences/about-us/staff/psychology/craig-jackson )

would be interesting to see what he thinks about @Lucibee @Michiel Tack and others assessment of the CFQ on the latest Cochrane review thread.

eta; and of course the comments on this thread

 

Agree with both.

I was coming from an other perspective.

There seems to be a consensus that in trials on therapist-delivered treatments where blinding is not possible, objecive measures are needed.

There also seems to be a certain consensus that for trials on certain illnesses like ME, because there are no objectively measurable signs and symptoms specific to that illness, you have to mostly rely on subjective measures to assess any treatment's effect on symptoms / illness severity.

Cochrane and other institutions assessing evidence of treatment effects seem to be rather confident with this situation; it might be considered as just one factor among others in the 'tools' they use to assess the quality of evidence. Which seems at odds with the standards used for trials investigating other illnesses (with 'hard' signs and symptoms).

Coming from that perspective, I found it was interesting that people tested the correlation of some items on the Chalder Scale with objective measures. Even if those are very unsprecific meausures, don't they invaildate the otherwise used argument that subjective symtoms like fatigue can't be objectivly assessed per se?

 

Subjectivity has several meanings but the important one here is that a measure can be skewed enough in one direction or another by knowing what result you want to provide a statistically significant effect. That is entirely consistent with there still being some degree of correlation with an objective measure. Scoring in ice dancing is subjective but the results tend to correlate quite well with an absence of technical errors like falling over.

As we have debated in the past, the answer is to have multiple threshold measures that require both a significant change in the desired subjective measure AND a significant change in some objective measure that corroborates the subjective measure. There is still room for getting spurious results but much less.

 

I think @Jonathan Edwards a while back said something along the lines of if you wanted to know if the patient felt better you could just ask the patient.

Even where there are objective measurements the definition recovery & harms for any trial should be agreed with the patient population.

In a healthcare system where there are never enough resources the point is not to treat people regardless of outcome, it is to improve the outcome for the patient.

The BPSer have long held that patients with ME, & now those with MUS, shouldn't undergo unnecessary tests, even though you can't be 100% sure a test isn't necessary until you have done the test and seen the result. Yet at the same time feel it's appropriate for patients to undergo treatments that do nothing to improve the outcome for the patient, except perhaps to make the majority worse. As long as the treatment is provided by their own speciality or cronies of course.

The CFQ is simply a means to create noise. A distraction to stop people noticing that there was no meaningful improvement in function or quality of life.