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

Sly Saint

Senior Member (Voting Rights)
(Mods, do what you will with this.)

This paper from 1998 is regularly cited in various papers as showing that the CFQ is a reliable measure for CFS research.

But read beyond the abstract..........
Abstract

The Chalder fatigue scale is widely used to measure physical and mental fatigue in chronic fatigue syndrome patients, but the constructs of the scale have not been examined in this patient sample. We examined the constructs of the 14-item fatigue scale in a sample of 136 chronic fatigue syndrome patients through principal components analysis, followed by correlations with measures of subjective and objective cognitive performance, physiological measures of strength and functional work capacity, depression, anxiety, and subjective sleep difficulties. There were four factors of fatigue explaining 67% of the total variance. Factor 1 was correlated with subjective everyday cognitive difficulties, concentration difficulties, and a deficit in paired associate learning. Factor 2 was correlated with difficulties in maintaining sleep. Factor 3 was inversely correlated with grip strength, peak VO2, peak heart rate, and peak functional work capacity. Factor 4 was correlated with interview and self-rated measures of depression. The results support the validity of mental and physical fatigue subscales and the dropping of the “loss of interest” item in the 11-item version of the fatigue scale.

https://www.sciencedirect.com/science/article/abs/pii/S0022399998000221

Inclusion criteria specified that patients had to meet operationalized “Oxford” research criteria for CFS [2

The distribution of the 14 items of the fatigue scale in the 136 CFS patients was examined. Scores on eight items were normally distributed, but six items (“tiredness,” “resting more,” “lacking energy,” “feeling weak,” “feeling sleepyor drowsy,” and “starts things without difficulty but gets weaker as goes on”) were highly skewed with the majority of patients reaching the maximum score.
RESULTS
Two hundred twenty-seven patients were assessed for eligibility into a treatment
trial of graduated aerobic exercise and fluoxetine

so they were recruiting for a trial
Sixty-two did not fulfill criteria for CFS. Of the remaining 165, 25 did not consent
to the treatment trial, 3 could not come off antidepressants, and 1 required other
medical treatment first. The 136 CFS patients not taking psychotropic medication
who started the treatment trial completed the assessments examined in the current
study. Physiological assessments were missed for four patients: three because they
dropped out during the initial assessment, and one with no meaningful measure-
ments. Nine patients failed to complete their sleep assessments.
The mean (SD) age of the total sample was 39 (11) years, median (interquartile
range) duration of CFS symptoms was 28 (40) months, and median (interquartile
range) education was 11 (1) years. There were 97 (71%) females, 83 (61%) were
married or cohabiting, and 29 (21%) were currently in paid employment. Sixty-two
patients (46%) fulfilled DSM-III-R criteria [6] for a psychiatric diagnosis: 14 (10%)
had major depression; 32 (24%) either dysthymia or depressive disorder not other-
wise specified; 14 (10%) various anxiety disorders; and 2 (2%) somatization disor-
der.
The 50 patients who took part in the cognitive testing were older than the rest of the sample
Near-maximal scoring on six physical fatigue scale items from the total of 14
items constituting the Chalder fatigue scale supports the validity of scoring the
physical fatigue scale on a two-point scale (presence or absence) rather than the
four-point Likert scoring. Most of the variance in the total fatigue scale is thus ac-
counted for by the mental fatigue score.
There are a number of methodological issues that need to be addressed. The sub-
jects in our study were volunteers for a randomized, controlled trial and differing
constructs might apply to a sample of all attenders with CFS at a medical clinic

http://sci-hub.tw/https://doi.org/10.1016/S0022-3999(98)00022-1#
 
Write up looking at the CFQ in Occupational Medicine (2014)
The CFQ 11 allows the user to differentiate between fatigue ‘cases’ and ‘non-cases’—responses in the two left-hand columns are scored with 0, while responses in the two right-hand columns receive 1. The sub-scales of physical and psychological fatigue are not used here, but rather the respondent receives a global binary fatigue score ranging from 0 to 11. The authors established that a global binary fatigue score of 3 or less represents scores of those who are not fatigued, with scores of 4 or more equating to ‘severe fatigue’. The scoring systems used in the CFQ 11 are virtually identical to the concepts of ‘caseness’ used in General Health Questionnaire and will be apparent to anyone familiar with that scale [4].

Reliability coefficients for the CFQ 11 have been high in studies of CFS patients [5] as well as occupational and general population research, ranging from 0.90 for the Likert scoring method and 0.83 for the binary scoring method [3].
https://academic.oup.com/occmed/article/65/1/86/1433061
 
I cannot see how this scale could EVER be taken seriously.

Look at the instructions, and think like a patient.

To what should the patient be comparing their symptoms “in the last month”?

Should it be compared to “when last well” or to “usual”? The header says to “usual”.

So, a patient who has just undergone treatment will quite possibly think that their 2nd taking of this questionnaire is to compare how they feel after treatment, to their “usual” state before treatment.

Hence a treated patient might answer that they have a symptom “no more than usual” - ie they think they are reporting that the treatment hasn’t changed their “usual” symptoms! (& Where is the “no LESS than usual” option?)

BUT those analysing the questionnaires will read a “no more than usual” response to mean, NO MORE THAN WHEN LAST WELL! - ie they will record the patient as recovered!!!

Think about this.... when first taking a this questionnaire, the patient might have no difficulty thinking that “usual” means “when well”, because they just have two states to consider. Being well, and being ill.

However when the patient takes the questionnaire the 2nd time after a treatment (perhaps months later) they now have THREE health states to consider: Being well, being ill, and the state they are in after treatment.

Now see how easy it is to create confusion for the responses. “Than Usual” in the header of each response means it is a mental leap every time to recognise that “usual” actually means “when well”.

What a mess. Whey not just put “when well” in the headers?

(Couple minor edits for clarity)

C312F2DF-474B-40ED-AA89-73B75C196A25.jpeg
 
Last edited:
Does anyone know if they have changed it from usual nowadays?

I searched online to find the questionnaire, and looked at several places where it is available. All use “usual”.

I think were that changed, it would need to be described as a modified version of the CFQ?

I really don’t understand how this questionnaire ever passed the first glance from some-one other than the author. The confusion would be so simply cleared up by putting “when last well” in each of the headers instead of the word “usual”.

So it really begs the question, WHY did that not happen?


And another thing - there is no option for “No change”, or “no less than usual” - the closest option for the patient using the questionnaire and wanting to declare no change, being the “No worse than usual” option.

Thus the phrasing for the patients duped into ticking this box (thinking about how they are comparing their current state to their “usual” illness state) will read exactly the way that the researchers want.

That then means they can record “recovery” from a treatment, just because the box ticked is interpreted by them as meaning “No worse than when last well”. (No matter what the patient thought they were recording!)


In other words, the use of this “no more than usual” phrasing as the “no change” option means these patients are easily reported as RECOVERED in the outcomes from the trial. This is also a sleight of hand in my view!
 
The lack of rigour in drafting something like this sums the whole BPS lot up really.

Some of the questions asking 'more' or 'less', when the answers are more/less options anyway.

If "you have been feeling tired for a long while", and your brain is fried as part of your symptoms, how well are you going to remember what it felt like when you were last well anyway? I suspect the therapist's advice might be "Oh, just do the best you can" ... !
 
Does anyone know if they have changed it from usual nowadays?
It’s definitely usual in the survey from Liverpool that’s just been put out - on separate thread
Yes I was going to say the same thing. I've seen the chalder scale before but it pulled me up short again when I did that survey - just because it's such an utterly stupid way of describing - leading to all the issues you describe @Unable . Since she is a professor it's difficult to believe she is lacking in IQ department, therefore one has to wonder if Chalder (et al) are just being deliberately obtuse, or if it is, as suggested, intentional.
 
It is impossible to fathom why they thought this was a good scale.

a global binary fatigue score of 3 or less represents scores of those who are not fatigued, with scores of 4 or more equating to ‘severe fatigue’.
Why is there not a section for mild fatigue or moderate fatigue?

They seem to think it is a good thing that so many people cluster at the high end.

Scores on eight items were normally distributed, but six items (“tiredness,” “resting more,” “lacking energy,” “feeling weak,” “feeling sleepyor drowsy,” and “starts things without difficulty but gets weaker as goes on”) were highly skewed with the majority of patients reaching the maximum score.

How can a scale be good if everyone scores the maximum? Why not extend it?

Then using "usual". Anyone reading measurements using this scale will assume it is like a pain scale, a measure of how fatigued a person feels. They will assume the measure is against no fatigue or normal everyday fatigue. Any pain scale which just measures how much pain you are in compared with usual would find cancer patients scored near the low end while someone with a splinter will score high.

It couldn't be that they deliberately want a scale to back up their theories
 
I wouldn't make any assumptions about that. I mean in general, not this individual in particular.
Yes, I guess it doesnt always follow. I just meant i expect a certain level of intelligence if someone has made it through to that stage in their academic life. However I think this issue is more to do with common sense, & thinking it through, than anything else so it may not follow...
 
I got sick in my early 20s. I am now in my mid-50s.

I have been sick for much longer than I was healthy.

How am I supposed to remember accurately what it felt like to be healthy?

Like you I have been ill so long I can't remember but they are not specifically asking you to compare with what you were like when you were well, we are just assuming that. They are asking as usual which means the way things usually are - I can't think of another way to say "usual" it means what it means.

So you have to try to work out what they are asking before you can answer and who knows what anyone is thinking. It is unbelievable!
 
Like you I have been ill so long I can't remember but they are not specifically asking you to compare with what you were like when you were well, we are just assuming that. They are asking as usual which means the way things usually are - I can't think of another way to say "usual" it means what it means.

So you have to try to work out what they are asking before you can answer and who knows what anyone is thinking. It is unbelievable!
And it's not as if there were countless studies showing how unreliable memory is. Might as well ask people to guesstimate how much they think they weigh in mushrooms as part of a weight loss trial. What mushrooms? Dry? Freshly picked? Cut? Washed? Exactly.

A more relevant question would be how accurate it is. The answer to that is obvious but then that's not the point of this questionnaire at all.
 
but they are not specifically asking you to compare with what you were like when you were well, we are just assuming that. They are asking as usual which means the way things usually are - I can't think of another way to say "usual" it means what it means.

So you have to try to work out what they are asking before you can answer and who knows what anyone is thinking. It is unbelievable!
To be fair, in the Liv Uni study mentioned above (where the Chalder scale is included) it did say at the beginning - sorry I cant remember the exact words - but words to the effect of 'if you've been ill a long time answer compared to when you last felt well'.

But I don't think, IIRC, that that clarification is always given, which makes it even sillier & less interpretable between studies.
 
Im miffed with the Liverpool questionnaire because I didn’t spot wording at the beginning that said you should fill it in with reference to when you were last well and in the questionnaire itself it definitely said by reference to a month ago because I was filling it in on that basis and put no change. I got bounced out anyway. I don’t know if I can be bothered to go back to it. I think whoever it was on that thread who said they interpret your responses in the way they want anyway is probably right.
 
I think the general wording of the scale is of interest in that all the questions are really psychological questions - what you feel, what you need, what your problems are. The assumption is that 'fatigue' is a self-value judgment rather than a symptom as such. For arthritis we do not ask 'Do you have problems with pain?'. We have a pain level score and the patient creates their own baseline that they can refer to.
 
Back
Top Bottom