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

[ladycatlover]

So a 1 point improvement on a mental fatigue and a 1 point worsening on physical fatigue will lead to a (potential) worsening of the CFQ but the other way around an improvement.

Could the mental fatigue and physical fatigue be "measured" using separate questionnaires? Or does that make things over-complicated? The Chalder thing is a load of carp anyway, so why not make something better for both fatigue components. Oh, and when it comes to having problems with starting things, I have no problems with that. Now finishing them is a different matter of course!

 

[arewenearlythereyet]

I'm not sure about your point 4, If the scale is intended to act with a threshold to help discriminate a group then I would agree although tests on different groups would be needed. But if the scale is intended to be a linear proxy that measures fatigue accurately then more needs to be done to validate that scores are equi-distant and have a meaningful representation in reality. This is how the scale is being used in a trial where it is basically being used to measure change.

Yes I was thinking more about the scores we use in food rating. There are a number of different ones depending upon what you plan to measure (discrimination tests would be used to measure difference, acceptance tests for approval rating and just right for determining if things are too little or too much). For most of these the test methodology originated out of NASA or the USA army and were tested ad infinitum in the 50's and 60's. Most marketing survey companies adopt some of these principles although I've seen quite a few howlers though that did need to feel the point of my pencil.

It does seem difficult to fathom what TC had in mind when she designed this, but I seem to recall people have used it as a screening tool subsequently (that might be brain fog of course). Either way its not very good for either job (I'm being polite and nearly used one of @Woolie descriptive pictures but refrained)

 

[Trish]

The Chalder thing is a load of carp

Sorry, couldn't resist

 

[Adrian]

Could the mental fatigue and physical fatigue be "measured" using separate questionnaires? Or does that make things over-complicated? The Chalder thing is a load of carp anyway, so why not make something better for both fatigue components. Oh, and when it comes to having problems with starting things, I have no problems with that. Now finishing them is a different matter of course!

Originally they talked about quoting the two scores separately and as a combined score.

 

[Barry]

The stats (quoting mean and sd) assume that the scale is linear otherwise if it is ordered (ordinal scale) then the median and MAD can be used. If not (as I think is the case for the CFQ) only the mode should be given.

When comparing improvements you cannot do that if an improvement of 1 point in real fatigue is measured as an improvement of 3 points on your proxy scale (the CFQ) when at position X but only 1 point if at position Y. So the proxy scale needs to map well onto the actual fatigue values (if such a thing really exists!). Problems such as possible answer set for the scale not equally representing all the set of values but say cutting off at bad (and not representing very bad or changes from that) have an obvious effect.

But so do issues about question content. If three questions are very related then a fairly small real change may lead to large amounts of reported change if the person is on the scale at a point where the answer to those questions will flip. Hence it is very important to understand how each question fits onto a real measure of fatigue. I would argue that to be a proper scale these questions would been to be ordered and the order would be valid for most people.

There is a further issue in terms of ordering which is if mental and physical fatigue can vary independently (or sometimes do) then the CFQ is actually the sum of at least two different basic things to measure with different weightings given different numbers of questions for each. So a 1 point improvement on a mental fatigue and a 1 point worsening on physical fatigue will lead to a (potential) worsening of the CFQ but the other way around an improvement. This is not based on a 'fatigue utility' i.e. looking at how people value the different components but is arbitary in the make up of questions. It also implies non-monotonic orderings.

Many thanks for that Adrian, very helpful once I got my head round it.

This part of your post was what I mainly had in mind:

When comparing improvements you cannot do that if an improvement of 1 point in real fatigue is measured as an improvement of 3 points on your proxy scale (the CFQ) when at position X but only 1 point if at position Y. So the proxy scale needs to map well onto the actual fatigue values (if such a thing really exists!). Problems such as possible answer set for the scale not equally representing all the set of values but say cutting off at bad (and not representing very bad or changes from that) have an obvious effect.

The problem reminds me of an engineering example, where high temperatures are typically measured using thermocouples; temperature changes result in small voltage changes, but the transfer characteristic is often very non-linear, the voltage change varying considerably for a given temperature change, depending what the temperature actually is. The big advantage in this case is that the characteristic is very well known, and so the appropriate compensation can be applied, so what you end up with is a linear relationship between volts (mV actually) and calculated temperature, via the characterisation step.

I would think it very unlikely there is going to be a neat linear relationship between actual fatigue (as you say, whatever that is), and people's rendering of it onto something like the CFQ. But unlike my thermocouple example, the characteristic is far from well known I image, or even well recognised. Yet without understanding that relationship, it could be that someone being interpreted as twice as fatigued as someone else, might for example actually be four times more fatigued, or whatever. Presuming a neat cosy relationship would be a bad idea, unless already proven to be neat and cosy, which I very much doubt.

Have any studies ever been done I wonder to try and better understand the characteristic that relates a person's actual "fatigueness", with their reporting of it into something like the CFQ? Very tricky I imagine, even if the will is there.

 

[Adrian]

I would think it very unlikely there is going to be a neat linear relationship between actual fatigue (as you say, whatever that is), and people's rendering of it onto something like the CFQ. But unlike my thermocouple example, the characteristic is far from well known I image, or even well recognised. Yet without understanding that relationship, it could be that someone being interpreted as twice as fatigued as someone else, might for example actually be four times more fatigued, or whatever. Presuming a neat cosy relationship would be a bad idea, unless already proven to be neat and cosy, which I very much doubt.

Sounds right to me. I do come at it from more of an engineering or maths background hence perhaps I worry about the properties.

Have any studies ever been done I wonder to try and better understand the characteristic that relates a person's actual "fatigueness", with their reporting of it into something like the CFQ? Very tricky I imagine, even if the will is there.

Not that I know of. It could be argued that there is nothing as simple as fatigueness that can be measured (not as a simple measure such as temperature). But if there was something it can not be observed directly only through reports or perhaps observations of activity. Thus it is very difficult to measure. I think Jason did a paper on comparing fatigue scales (but the CFQ didn't have the necessary properties to make it in to the comparison).

I think the inability to observe fatigue is why I would prefer activity measures to be used.

 

[Barry]

There is a closer analogy when considering the decibel scale, and noise perceptions. This link ...

http://www.noisehelp.com/decibel-scale.html

... sums it up nicely with:

The relative loudness that we perceive is a subjective psychological phenomenon, not something that can be objectively measured. Most of us perceive one sound to be twice as loud as another one when they are about 10 dB apart; for instance, a 60-dB air conditioner will sound twice as loud as a 50-dB refrigerator. Yet that 10-dB difference represents a tenfold increase in intensity. A 70-dB dishwasher will sound about four times as loud as the 50-dB refrigerator, but in terms of acoustic intensity, the sound it makes is 100 times as powerful.

Somewhere in the past someone must have done some experiments to identify this relationship. Interesting that they will have been using an objectively measurable value (acoustic power) to then assess people's percepetions of loudness. In fairness, I imagine the physical characteristics of the ear play a significant part in things.

 

[Lucibee]

I'm just going to throw my twitterings into the ring...

Good critique by Carolyn and Graham. There are so many problems with CFQ that it's difficult to know where to start!

The key one for me was the use of different time periods in the instructions. "In the last month" is underlined on the scale commonly used. And although the instructions tell you to compare yourself with "when you last felt well", this then poses a conflict if you repeat the questionnaire several months later, particularly if the intervention has fiddled about with your definition of what is "usual" (particularly as the first few weeks of the intervention will be about establishing new baselines).

It also strikes me that it is terrible choice of instrument if you want to capture "improvement", because it is so heavily weighted towards getting "worse".

Hence my summary:

CFQ should only ever be used as a diagnostic tool (and probably only to diagnose Chalder Fatigue Syndrome). It is NOT a repeated measures instrument. It does NOT measure absolute fatigue. It is NOT suitable for use as a trial outcome measure.

My bloggings are here (in case anyone is interested):
https://lucibee.wordpress.com/2016/04/02/fatigued-by-scales-as-outcome-measures/
https://lucibee.wordpress.com/2017/03/30/more-on-fatigue/
https://lucibee.wordpress.com/2017/03/31/even-more-on-fatigue-getting-really-tired-now/

 

[deboruth]

I didn't say that my activity is primarily limited by my primary feelings of fatigue, but that doesn't meen that I don't experience feelings of fatigue, and the things this questionnaire covers.

That's a very big extrapolation to jump to, and it's completely untrue. I assure you, I know my symptoms far better than you know them.

I can't imagine enjoying the luxury Wooly suggests -- of basing my movements on some intellectual prediction of how they will make me feel the next day. I simply find myself unable to move further. Nor do I feel "fatigue." I feel that I have no further capacity. My body fails or my brain fails or both. "Exhaustion" would be a more accurate word than "fatigue." With "fatigue" one feels poorly but can carry on. With ME one finds oneself -- often many times in a day, or continuously -- simply unable to carry on. The essence of "PEM" in ME is the inability to move, the inability to recover and the inability to resume in respect of normal human activities.

 

[Woolie]

I can't imagine enjoying the luxury Wooly suggests -- of basing my movements on some intellectual prediction of how they will make me feel the next day. I simply find myself unable to move further. Nor do I feel "fatigue." I feel that I have no further capacity. My body fails or my brain fails or both. "Exhaustion" would be a more accurate word than "fatigue." With "fatigue" one feels poorly but can carry on. With ME one finds oneself -- often many times in a day, or continuously -- simply unable to carry on. The essence of "PEM" in ME is the inability to move, the inability to recover and the inability to resume in respect of normal human activities.

Since I spend about 80% of my time bedbound (physically, I'm on the more severe end of the spectrum), I feel a little offended at the idea that there's anything "luxurious" about my situation. But I'm figuring you didn't mean it that way, @deboruth.

What you say does point to the idea that different people might have different illness mechanisms.

So here's my thought. There's a group of us where the primary problem is to do with inflammation. This inflammation is stimulated by exertion. When stimulated we feel awfully sick, flu-like, with lots of burning and aching and headaches. These effects are often delayed (for example, they may peak the day after the exertion). So our main goal is to keep these symptoms to a minimum - hence, avoidance of PEM. I'm almost certainly in this group.

Then there's another group where the problem is primarily to do with energy production. People in this group talk in terms of a battery - once the energy's used up, that's it. @deboruth, your description suggests you might be in this group.

I do occasionally get what you describe as collapse if I push too hard, but its more like burning, screaming pain than a feeling that I've run out of energy. More often though, I push through with no immediate symptoms, not till the next day. But then the price may be severe malasie, burning, headaches, so bad I might not be able to move my head for days. Its all very episodic so its easy to send yourself to bed for a fortnight, then feel a little better one day, do a few small things and then start the cycle all over again.

 

[Barry]

i suspect that the Chalder scale has had no proving done on it since it looks like it's been crafted fairly crudely, and seems to be missing an awful lot of basic best practice. I doubt that there was any validation in terms of elliminating bias from the scales used or the rather dubious wording of the questions.

I suspect it has been "validated by opinion".

 

[chrisb]

I suspect that we are here getting to the nub of a part of the problem. There is a variety of types of fatigue expressed in different ways, over different time scales. Some people may suffer two or more types of "fatigue", all of which may be different to what others suffer.

If people, who shall be nameless, had been doing there jobs, this might have been noticed by them. But we learned from "Reviewer 2" of the attention to detail displayed in the collection of histories.

 

[Barry]

Since I spend about 80% of my time bedbound (physically, I'm on the more severe end of the spectrum), I feel a little offended at the idea that there's anything "luxurious" about my situation. But I'm figuring you didn't mean it that way, @deboruth.

What you say does point to the idea that different people might have different illness mechanisms.

So here's my thought. There's a group of us where the primary problem is to do with inflammation. This inflammation is stimulated by exertion. When stimulated we feel awfully sick, flu-like, with lots of burning and aching and headaches. These effects are often delayed (for example, they may peak the day after the exertion). So our main goal is to keep these symptoms to a minimum - hence, avoidance of PEM. I'm almost certainly in this group.

Then there's another group where the problem is primarily to do with energy production. People in this group talk in terms of a battery - once the energy's used up, that's it. @deboruth, your description suggests you might be in this group.

I do occasionally get what you describe as collapse if I push too hard, but its more like burning, screaming pain than a feeling that I've run out of energy. More often though, I push through with no immediate symptoms, not till the next day. But then the price may be severe malasie, burning, headaches, so bad I might not be able to move my head for days. Its all very episodic so its easy to send yourself to bed for a fortnight, then feel a little better one day, do a few small things and then start the cycle all over again.

I suspect there may be a strong commonality in underlying problem, and the difference may be more to do with manifestation.

I would be very surprised if significantly restricted energy availability isn't core to what all PwME experience. But it may be that for one group, their normal safety mechanism of fatigue kicks in to warn them their energy supply is dropping through the floor, and they need to back off; albeit for PwME it typically kicks in much too late (probably because their energy availability drops off so rapidly, maybe after a time delay), hence the need for pacing. My wife would be in this category. But by the sound of it there is perhaps another group where maybe the normal fatigue safety mechanism just doesn't kick in at all, and the first the person knows about it is when their energy availability drops though the floor.

 

[Woolie]

I suspect there may be a strong commonality in underlying problem, and the difference may be more to do with manifestation.

Yes, I wondered that too, @Barry, since inflammation seems to interfere with cellular energy production.

But by the sound of it there is perhaps another group where maybe the normal fatigue safety mechanism just doesn't kick in at all, and the first the person knows about it is when their energy availability drops though the floor.

Yes, I think that's the whole reason people are encouraged to learn how to pace. For most of us, our bodies don't always tell us to stop early enough. If it did, then we wouldn't need to learn to pace.

 

[JemPD]

My brain is mush so i wont comment on the details of whats being said here but will say that i've agreed with all the comments so far. The CFQ is a joke. It's a pity its consequences are far from funny.

Well done everyone for all your comments & submission, much appreciated

 

[Amw66]

There is a closer analogy when considering the decibel scale, and noise perceptions. This link ...

http://www.noisehelp.com/decibel-scale.html

... sums it up nicely with:

Somewhere in the past someone must have done some experiments to identify this relationship. Interesting that they will have been using an objectively measurable value (acoustic power) to then assess people's percepetions of loudness. In fairness, I imagine the physical characteristics of the ear play a significant part in things.

Yep. Sound level is logarithmic scale. Noise and distance. Db levels are adjusted ( weighted) for human hearing range too

 

[Barry]

Yep. Sound level is logarithmic scale. Noise and distance. Db levels are adjusted ( weighted) for human hearing range too

I don't think distance has anything to do with it directly. You could sit someone in front a speaker at some distance and leave them there, and have some form of acoustic power meter at the same position. Then you set the acoustic power at say 1W, and the person has some sense of its loudness. Then wind the amplifier up until the acoustic power meter shows 10W, but the person will perceive the loudness to have only increased by a factor of 2, not 10. At 100W (ear drums permitting) the person will sense the loudness to be 4 times what it was at 1W.

But coming back on track a bit, the relationship between objectively measured sound power, versus subjectively measure loudness, is a very well understood relationship that fits a nice neat mathematical transfer function. For fatigue perception versus "real fatigue", no such neat relationship is known, nor is it even acknowledged as needing to be understood. It's just a case of "Trust me ... I'm a psychiatrist".

 

[Amw66]

I don't think distance has anything to do with it directly. You could sit someone in front a speaker at some distance and leave them there, and have some form of acoustic power meter at the same position. Then you set the acoustic power at say 1W, and the person has some sense of its loudness. Then wind the amplifier up until the acoustic power meter shows 10W, but the person will perceive the loudness to have only increased by a factor of 2, not 10. At 100W (ear drums permitting) the person will sense the loudness to be 4 times what it was at 1W.

But coming back on track a bit, the relationship between objectively measured sound power, versus subjectively measure loudness, is a very well understood relationship that fits a nice neat mathematical transfer function. For fatigue perception versus "real fatigue", no such neat relationship is known, nor is it even acknowledged as needing to be understood. It's just a case of "Trust me ... I'm a psychiatrist".

Shame the objective/ subjective wasn' t researched as well as in acoustics.

Thought i was going a bit mad, as i was taught the relationship between sound ( amplitude) and distance. There is a relationship ( phew! Brain still there at times- i did acoustics a long time ago)
Sound power is independent of distance- sound pressure is a distance dependent effect. Db relationship to sound pressure ( amplitude), and intensity ( energy) - for these there is a drop of around 6dB for every doubling of distance. Just in case a pub quiz question comes up ....

 

[Barry]

Thought i was going a bit mad, as i was taught the relationship between sound ( amplitude) and distance.

Yes I think that must also be right. Inverse square law presumably, similar to light intensity.

 

[Woolie]

I'm less worried about whether the underlying relationship between ratings and fatigue is linear or logarithmic or whatever. I just want there to be a relationship.

We can worry about scale when we've got that first part right.