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

Barry said:
Just re-read the original submission again, and I'm as always appalled by the naive incompetence so clearly evident in those who came up with / advocate the CFQ as a general instrument for measuring fatigue. Their obvious inability to think something like this through, the inbuilt heavily biased presumptions that inevitably load the results, and failure to comprehend its laughably blatant failings, casts massive doubt on their competence to do anything requiring clear uncluttered reasoning ... like being scientists for instance. Then add into the mix how these faux scientists have held sway over the health - and indeed safety! - of so many sufferers over so many years, it just beggars belief. It's medieval. And to me at least it feels downright criminal.
Click to expand...
As Simon Wessely joked about, if they didn't do this, no one would "recover" according to their own criteria.

So funny haha. Medical research fraud is hi-la-rious.
 
Lucibee said:
I've finally got round to writing this up... https://lucibee.wordpress.com/2023/...se-assumptions-plague-the-medical-literature/

In doing so, it's revealed some things I hadn't previously realised (or had forgotten) about the CFQ. That it is based on Goldberg's General Health Questionnaire, and that inception and validation is a bit dodge. I touch on that in the next post: https://lucibee.wordpress.com/2023/08/08/problems-with-pawlikowska-et-al-bmj-1994/ but I probably should try to go into more detail. Basically, as long as you use the right (complicated and flashy) tools and get broadly expected results, no-one will look too closely and reveal that all your assumptions are pants.

And it pains me that one of the most recent citations of Pawlikowska is in the development of a Turkish version of the CFQ. :aargh:
Click to expand...

I just wanted to link through to somewhere that discusses the CFQ and its problems, because I had another, closer, look at the original "Development of a fatigue scale" paper by Chalder et al., and spotted something I hadn't realised before. And it's important, because I'm sure that whenever any criticism is made of it, the response will be "But it's been validated. It's been published in a peer-reviewed journal. It's been used in thousands of studies without issue." etc etc.

So let's look at the validation.

Normally, that would be done against some "gold standard" - in this case, a standard (hopefully, something actually *objective*) used to diagnose "fatigue".

So what did they do? They took one question [item] out of Lewis (and Pelosi's) Clinical Interview Schedule (Revised) questionnaire - which was devised to screen for psychiatric conditions in general populations, and used *that* as the "gold standard" after turning it into a series of questions [correction - this "item" already consisted of a series of questions]:
1. “Have you noticed that you’ve been getting tired recently?"
2. “Have you felt that you’ve been lacking in energy?”

If a positive answer was elicited [to either or both is not clear], four supplementary questions were asked:

1. On how many days have you felt tired or lacking in energy during the past week?
(Score 1 for greater than or equal to 4 days.)
2. Have you felt tired/lacking in energy for more than 3 hr on any day in the past week?
(Score 1 for greater than 3 hr.)
3. Have you felt so tired that you’ve had to push yourself to get things done during the past week?
(Score 1 for yes.)
4. Have you felt tired or lacking in energy when doing things you enjoy during the last week?
(Score 1 for yes.)

That give a total fatigue score from 0-4. Scores of 2 or above are regarded as fatigued.
Click to expand...

And that's their "gold standard" pseudo-objective measure of fatigue.

It sort of begs the question, why didn't they just use that as their instrument?

Anyway. Any validation process involving ROC curves and PCA with two scales that use similar questions is going to end up with the scales agreeing by default. It doesn't take any kind of complex statistical analysis to see that. The analysis is merely a way of going through the motions.

It's also clear that they used Goldberg's General Health Questionnaire as a template - hence the reason it uses that unbalanced Likert scale, and measures change rather than absolute values (intensity).

What shocks me more is that you don't need to read the whole paper to get an indication of this. It's in the abstract.
...100 consecutive attenders to a general practice completed the fatigue scale and the fatigue item of the revised Clinical Interview Schedule (CIS-R). These were compared by the application of Relative Operating Characteristic (ROC) analysis...
Click to expand...

And now CFQ is being used to assess folks with Long Covid.

What can we do about it? How can I help???
 
Last edited:
Apart from all the points highlighted by the excellent analysis here, I've always been intensely irritated by the fact that Question 11 - How is your memory? is a different type of question from the others. It invites a narrative response, but is supposed to be scored as a value. "More than usual" could be interpretated as "I'm having more problems with my memory than usual" OR "my memory is functioning better than usual".

Patients must query it all the time, but the powers that be seem to have never noticed this anomaly.
 
tornandfrayed said:
Apart from all the points highlighted by the excellent analysis here, I've always been intensely irritated by the fact that Question 11 - How is your memory? is a different type of question from the others. It invites a narrative response, but is supposed to be scored as a value. "More than usual" could be interpretated as "I'm having more problems with my memory than usual" OR "my memory is functioning better than usual".

Patients must query it all the time, but the powers that be seem to have never noticed this anomaly.
Click to expand...

If that's how Q11 is displayed, they are using an incorrect copy of the CFQ. It should also have the options "Better than usual" >>> "Much worse than usual" for that question.

Wait till you see the GHQ!!!
 
You must log in or register to reply here.
Back
Top Bottom