Should we initiate development of a new, short questionnaire to identify PEM (to aid diagnosis)?

[Simon M]

I think it would help PwME if we had a short questionnaire that does a good job of identifying postexertional malaise (PEM), the cardinal symptom of this illness. It could be useful in two ways:

1. Doctors, especially those in primary care, could use it to help diagnose people who have ME/CFS, and identify those who don’t have PEM, and so might have a different illness.

We know that primary care diagnoses are very slow and often wrong. Two decent studies found that around half of cases referred to specialists by GPs didn’t have ME/CFS, mostly because people instead had another illness, psychological or biomedical. One study was led by Julian Newton, the other by Peter White, but both identified the same problem. This study found that nearly half of GPs say they find ME/CFS hard to diagnose

2. Researchers could use it to see if people who self-report a diagnosis of ME/CFS, or have a GP-given diagnosis, have PEM, to make sure that they are identifying ME/CFS cases for their research (DecodeME did this).

Any questionnaire should be short, freely available, and validated to have high sensitivity for identifying almost all true cases and high specificity for excluding people with other illnesses without PEM, but who might have related issues such as exertion intolerance.

Currently, there is no tool available to help with this: the latest DePaul PEM questionnaire (Jason & Chee, 2025) is criticised in this thread as not capturing the true nature of PEM well, and it is also quite lengthy and complicated.

How could we do this?
This is likely to be a lot of work. We’d need to come up with good questions endorsed by PwME and show that it could discriminate between ME/CFS and other illnesses. That means collecting data on ME/CFS and other chronic illnesses. We don’t know to what extent PEM exists in other illnesses, so there is probably a lot of groundwork to do here. @ME/CFS Science Blog produced a helpful outline of how we might go about this.

I hope this thread will start a discussion on whether and how we could take this forward. I agree with others that we would need to collaborate with a researcher to make this happen. But I suspect there's a lot of useful discussion we can have here first. Due to my poor health, I won't be able to contribute much myself.

[Thanks to all those who engaged with the suggestion for a new questionnaire on the original thread, especially @Trish, who suggested similar questions to me, and produced the comprehensive S4ME PEM Factsheet; @ME/CFS Science Blog who said they were happy to collaborate (and see above re methodology); and @Bivox (Krisitian Sommerfelt), who shared his clinical experience on identifying and understanding PEM.]

 

[Simon M]

Relevant data so far on which other illness do or don’t have PEM

We know next to nothing about the existence or extent of PEM in other illnesses, despite it often being described as unique to ME/CFS. The term was coined by medics for the 1994 CDC/Fukuda criteria because they knew of no term that captured it, suggesting that, if PEM exists in other illnesses, it is rare.

A 2017 study from the Lenny Jason team found that between 46% and 52% of MS patients met the threshold for at least one PEM criterion (vs 69-79% for ME/CFS).

A 2018 study by Lenny Jason Group produced the most valuable data I have seen. They asked PwME, PwMS, and people with post-polio syndrome (PPS) about the duration of “PEM” (though their questionnaire likely captured exertional intolerance). “PEM”, as defined by Jason and team, had a far shorter duration in MS and PPS than in ME/CFS, suggesting that most of these patients do not have PEM as we understand it.



Note: The PPS percentage figures add up to only 85%—an error in the original paper.

I think this demonstrates why we need to get data on PEM-related issues for illness other than ME/CFS.

 

[Sasha]

Thanks for putting this forward, @Simon M. I think this is genius. We've seen in the Samms and Ponting study that there was what, a ten-fold difference in ME/CFS getting diagnosis in different areas of the country? Doctors are leaving many thousands of PwME in the dark about their diagnosis and putting them in danger thereby. And it's mission- critical that researchers really are studying ME/CFS when they think they are. This is a problem that needs fixing.

I hope this thread will start a discussion on whether and how we could take this forward. I agree with others that we would need to collaborate with a researcher to make this happen. But I suspect there's a lot of useful discussion we can have here first. Due to my poor health, I won't be able to contribute much myself.

What should be the steps to make this happen? I agree that we'd need to collaborate with a researcher, but if I were that researcher, I'd want to be in at the ground floor. So maybe the time to bring in any likely researcher is now. We could maybe form a group of interested forum members, set up on a private forum, and chuck some names about to see who might be worth approaching.

And then we could come up with a way forward, maybe getting feedback on draft PEM questions from the forum, as we do for the factsheets. We could pilot-test the questions here, too, and refine them before taking them off for a proper independent test.

 

[Yann04]

I completely agree. This has been something I’ve been hoping for for a while.

I even made a similar thread a bit ago.

Thread 'What would a good questionnaire for diagnosing PEM look like?'

Dec 22, 2024

The most popular questionnaire for PEM is the De-Paul Symptom Questionnaire, which, to be frank, is not the best, as it mostly focuses on fatigue after exertion and seems to flag an unlikely proportion of people with other illnesses having PEM.

It seems there isn’t much competition. The FUNCAP Scale is excellent for measuring disability in people with PEM, but is ill suited to diagnose PEM itself. Davenport’s team made a similar questionnaire that might serve as an outcome measure for functional capacity related to PEM but is not exactly suited to measure if a person has PEM.

Obviously...

• Yann04
• depaul symptom questionnaire dsq funcap pem scale
• Replies: 33
• Forum: Other research methodology topics

• 

 

[Sasha]

If we did everything necessary and a brilliant paper was published showing our super new validated ultra-short PEM questionnaire, how would we get it widely adopted into clinical practice? Presumably when this kind of thing happens, people don't hang around for years waiting to get the new thing added to the NICE guidelines? Or do they?

Or would it be disseminated as some sort of NHS bulletin to GPs (in the UK, anyway)?

 

[Sasha]

I completely agree. This has been something I’ve been hoping for for a while.

I even made a similar thread a bit ago.

Zeitgeist!

 

[Trish]

I like the idea of a simple PEM questionnaire that is purely intended to help with diagnosis for both clinicians and clinical trials.

The main task, as I see it, is to distinguish the specific phenomenon of PEM from PEF, DOMS, exercise intolerance, fatigability. I think that can be done with a simple one or two questions that describe what an episode of PEM includes, as in our fact sheet, and asks yes/no/not sure whether the person experiences that. It could also have a question describing the other options that are not PEM but may be part of PEM, and asking yes/no/not sure for each of those. That might help both doctors and patients to understand better which versions of responses to exertion apply to them, and therefore whether they are likely or not to have PEM.

There could be supplementary questions asking about types of triggers, changes in symptoms when in and not in PEM, and delay and duration times, but these are more about helping people diagnosed with PEM to manage their activity and cope with symptoms.

As to how to do it. DecodeME designed their own simple one or two questions which seemed to be sensible. That could be a good starting point. Can anyone remind us what they were?

I agree we would need clinicians as part of a team setting this up and probably funding for all the palaver of testing on large samples and validation and publication.

Perhaps the first task of such a team would be to examine any existing tools used for diagnosing PEM and write and get published a critique. I don't want to divert this thread into discussing other questionnaires, I'll just draw attention to the recently published latest version from Leonard Jason, and the Sarah Tyson MEA funded team are develping one. Both have serious flaws.

 

[Simon M]

I should probably have posted this earlier as it's relevant to some of the comments made above.

What other PEM questionnaires are there, and what questions do they use?

We have the Jason and Chee 2025 paper, and it mentions several other scales. The NIH is developing one, and @Bivoxmentions another is being developed (is this right)? But these seem to take a comprehensive approach – I think we need something short and focused on identifying the presence/absence of PEM that is simple for doctors to administer and for patients to answer.

Specific questions for identifying PEM that might be useful

The Jason 2025 study used:

Over the past 6 months, have you experienced PEM, which is defined as an abnormal response to minimal amounts of physical and/or cognitive exertion, with symptom severity and duration out of proportion to the initial trigger?

DecodeME used:

Symptoms after effort or activity

12. In the last 6 months, what happens to your symptoms after you do more physical or mental activity than usual (exceed your energy limit)? If you pace your energy, we want you to think about what would have happened if you didn't. Please only choose ONE answer.

• My symptoms (such as pain, fatigue or feeling out-of-sorts) get worse, or I get new symptoms, and this reduces how much I can do
• My symptoms either stay the same or improve (Skip to Question 14)

13. In the last 6 months, after you have done more physical or mental activity than usual (exceeded your energy limit), how long does the change in your symptoms usually last? If you pace your energy, we want you to think about what would have happened if you didn't. Please only choose ONE answer.

• The change in my symptoms lasts a long time, which can be more than 24 hours
• I bounce back straight away or my symptoms don't last very long given the effort I just made

I like this question, but as I said on the other thread, I wonder if the final question on duration needs more guidance to avoid it capturing exertion intolerance, seen in many illnesses after physical effort, instead of PEM. The Lenny Jason data suggested 12 hours would be a reasonable minimum, but we at least need to replicate that, and any threshold risks losing milder cases.

@Bivox (Krisitan Sommefelt) shared the kind of question he uses in clinical practice:

[I ask] "What happens if you do very much more than you usually do because you want to or have to? First, being physically very much more active?"

I continue "What happens in the following day/ days? Do you feel better?" If they say yes, they don't have PEM. If they say they feel very bad and can do much less of any activity compared to before the activity they might have PEM.

I then ask the same question about social/ cognitive activities (use a bit different words - age dependent). If I get the same answer (i.e. much more symptoms and definite reduction in capacity for activity, then they have PEM.”

I then ask them to describe how they feel/what symptoms they have in the period they can do much less than normal and feel bad. I ask what the maximum duration typically has been. They usually say > 1 week.

BUT major point is it depends heavily upon how they regulate activities. Both motor and cognitive/social/sound-light exposure/ upright position etc. depending on how severe their disease is.

Then I chart their strategy for regulation of activities and who/what really decides their activity level. It should (in children and adolescents at least) almost always NOT be above what they themselves think. Often should be below that.

 

[Sasha]

I agree we would need clinicians as part of a team setting this up and probably funding for all the palaver of testing on large samples and validation and publication.

I hadn't thought about clinicians but I think we'd need people with experience in designing, testing and validating questionnaires.

Perhaps the first task of such a team would be to examine any existing tools used for diagnosing PEM and write and get published a critique.

Not necessarily. The critique could simply form the introduction to the paper about the new questionnaire. Or, if they were putting in for grant money, it could form part of the grant proposal.

 

[Simon M]

Thanks for putting this forward, @Simon M. I think this is genius. We've seen in the Samms and Ponting study that there was what, a ten-fold difference in ME/CFS getting diagnosis in different areas of the country? Doctors are leaving many thousands of PwME in the dark about their diagnosis and putting them in danger thereby. And it's mission- critical that researchers really are studying ME/CFS when they think they are. This is a problem that needs fixing.

Um, agree with all that. Yes, 10x higher in Cornwall than in the lowest areas. This study found that nearly half of GPs say they find ME/CFS hard to diagnose, and this questionnaire could help a lot with that.

So maybe the time to bring in any likely researcher is now. We could maybe form a group of interested forum members, set up on a private forum, and chuck some names about to see who might be worth approaching.

Sounds good. My health means I can't take things forward myself, though.

And then we could come up with a way forward, maybe getting feedback on draft PEM questions from the forum, as we do for the factsheets. We could pilot-test the questions here, too, and refine them before taking them off for a proper independent test

I think that's crucial. This forum has so much knowledge not only of the illness but also of PEM questions that miss the point. I included a few possible questions in this post.

how would we get it widely adopted into clinical practice?

Good question.
To GPs - not sure. It would go better with endorsement from a GP organisation/the NHS. but if we created a good tool, that is freely available, easy to use and, above all validated as being effective, I think researchers would be keen to use it, though promotion would be needed too.

I completely agree. This has been something I’ve been hoping for for a while. I even made a similar thread a bit ago.

Only 10 months back. Good thread, and I hope we can find a way forward here.

 

[jnmaciuch]

If we did everything necessary and a brilliant paper was published showing our super new validated ultra-short PEM questionnaire, how would we get it widely adopted into clinical practice? Presumably when this kind of thing happens, people don't hang around for years waiting to get the new thing added to the NICE guidelines? Or do they?

Easiest way to get the ball rolling would be to have another publication in a good journal use and cite your validated tool. Tons of “measurement tools” are created every year but most fall by the wayside because every researchers wants to stick by whatever is considered “standard” in the field, formally or informally. If you can get a couple of groups to publish a paper using it, others see the citation and use it for their own research, and then it becomes the “standard” that doctors will recommend in guideline committees.

To get that first citation I’d recommend finding a clinician partner who publishes studies on their patient populations. Something like a rheumatology clinic tied to a research hospital that puts out quality-of-care or similar studies. I’ve met plenty of clinicians who have heard of ME/CFS in some form and want to be able to measure it, but just haven’t had time or energy to wade through the mud of uncertainty in an unfamiliar field.

Having an informal endorsement by an ME/CFS org would also help legitimize—even something like a poll to members asking which of the available tools (including yours) they feel best captures their experience. You really just need to create word of mouth trust in the tool from clinicians.

Not necessarily. The critique could simply form the introduction to the paper about the new questionnaire. Or, if they were putting in for grant money, it could form part of the grant proposal.

I agree, it’s standard practice in a methods paper to outline in the introduction how your new tool addresses shortcomings in other available tools.

 

[jnmaciuch]

Overall I think this could be very valuable and would love to have the tool available for my own future research. I wouldn’t be able to spearhead this since I can’t take on another project and wouldn't have the right background for it anyways, but would be happy to provide feedback and help shop it around to interested clinicians at my institution at later stages.

[Edit: also just to put it out there, the most useful tool from both a clinical and research perspective would be one that can both identify whether or not a patient experiences PEM, and can quantify severity/impact of the symptom so it can be used to judge improvement or worsening]

 

[Kitty]

The main task, as I see it, is to distinguish the specific phenomenon of PEM from PEF, DOMS, exercise intolerance, fatigability.

I think just explaining those would get doctors halfway there. Most are unlikely to realise there are several different phenomena that need to be (and can be) separated.

 

[Utsikt]

[Edit: also just to put it out there, the most useful tool from both a clinical and research perspective would be one that can both identify whether or not a patient experiences PEM, and can quantify severity/impact of the symptom so it can be used to judge improvement or worsening]

I feel like FUNCAP does the second one as well as one could realistically hope.

You might even be able to modify it to ask people about today and not a middle day the past month to try to get a glimpse into what a PEM episode looks like.

 

[Jonathan Edwards]

I have reservations about trying to create a questionnaire.
When it comes to making a diagnosis in real practice questionnaires are rarely used and from what I have seen suggested simply wouldn't be appropriate. The golden rule is that you start by asking the patient to describe what their problems are in their own words. As soon as you start trying to standardise words you run into serious trouble because there is a huge range of ways people talk and a huge range of what they might understand by what you say.

If I was wanting to elicit a history of PEM having heard about disabling symptoms but no specific account of what happens after exertion, I would ask, in the sort of language they have used, what happens if they have to do rather more than usual or than they normally find they can manage. If that did not elicit a response of 'I feel ill about a day later and it can go on for weeks' then I would come at it from the other side with something like 'do you get major setbacks' and if they say yes, 'what seems to bring that on if anything'.

When people are answering questions in a medical context they are hugely biased by what they think they are supposed to say or what will get taken most notice of. If you have a questionnaire for PEM then half the patients will already have seen it on the net and will give all the right answers. As for trials, subjects are often their own worst enemies when it comes to getting reliable information mostly because they think they are expected to say this or that.

For research studies you might want a questionnaire filled in by the subjects but we don't do that for other illnesses. The clinician fills in a questionnaire with defined elements like pain, PEM, joint swelling.

A problem with PEM is that there is no complete agreement on what constitutes it and what doesn't quite make it.You cannot 'validate' a questionnaire if you don't have a gold standard and if you have a gold standard you probably don't want to use a questionnaire.

 

[Jonathan Edwards]

When it comes to other illnesses, I don't think there is any need to try to distinguish a special PEM for ME/CFS rather than other diseases. If PEM is some sort of immune signalling following exertion, or even if it isn't, I think it likely that people with tuberculosis, lymphoma, MS, or malaria will have symptoms that, in the time window when it is important to make these diagnoses, may well seem to fit 'PEM' quite well. Someone brewing active TB very likely feels much worse a few hours after exertion and if their TB is then progressing it will last for weeks.

I personally do not think PEM comes in at this point in the decision tree/algorithm. If the subject has generalised disabling symptoms you start off by looking for evidence for pathology with history, exam and tests. Only at the point where you are reasonably confident that no specific pathology will be found would you want to know whether the illness involved PEM and so could be managed as ME/CFS, or didn't in which case it would just be unexplained and to what extent you would manage it as if it were ME/CFS would depend on the whole picture. Clearly if exertion did not produce any bad effects there would be no point in discussing pacing in detail but I doubt things would always be that simple. We think identifying PEM is useful in getting to understand what we dealing with and refining study cohorts but within a year or two we may discover that there is actually a much better way of categorising people.

The real problem is that we have no academic physicians interested in knowing anything about ME/CFS. Until that changes I am not sure that we can make much headway getting better cohort definition.

The other thing is that I have never been keen on the idea of standardising criteria for the sake of it. Every research context deserves different categorisations and measurements. I know this sounds wwet blanket but I just don't think there is a clear enough consensus on the boundaries of PEM for it to be possible to 'validate' a standard set of questions.

 

[MrMagoo]

I think it would help PwME if we had a short questionnaire that does a good job of identifying postexertional malaise (PEM), the cardinal symptom of this illness. It could be useful in two ways:

1. Doctors, especially those in primary care, could use it to help diagnose people who have ME/CFS, and identify those who don’t have PEM, and so might have a different illness.

We know that primary care diagnoses are very slow and often wrong. Two decent studies found that around half of cases referred to specialists by GPs didn’t have ME/CFS, mostly because people instead had another illness, psychological or biomedical. One study was led by Julian Newton, the other by Peter White, but both identified the same problem. This study found that nearly half of GPs say they find ME/CFS hard to diagnose

2. Researchers could use it to see if people who self-report a diagnosis of ME/CFS, or have a GP-given diagnosis, have PEM, to make sure that they are identifying ME/CFS cases for their research (DecodeME did this).

Any questionnaire should be short, freely available, and validated to have high sensitivity for identifying almost all true cases and high specificity for excluding people with other illnesses without PEM, but who might have related issues such as exertion intolerance.

Currently, there is no tool available to help with this: the latest DePaul PEM questionnaire (Jason & Chee, 2025) is criticised in this thread as not capturing the true nature of PEM well, and it is also quite lengthy and complicated.

How could we do this?
This is likely to be a lot of work. We’d need to come up with good questions endorsed by PwME and show that it could discriminate between ME/CFS and other illnesses. That means collecting data on ME/CFS and other chronic illnesses. We don’t know to what extent PEM exists in other illnesses, so there is probably a lot of groundwork to do here. @ME/CFS Science Blog produced a helpful outline of how we might go about this.

I hope this thread will start a discussion on whether and how we could take this forward. I agree with others that we would need to collaborate with a researcher to make this happen. But I suspect there's a lot of useful discussion we can have here first. Due to my poor health, I won't be able to contribute much myself.

[Thanks to all those who engaged with the suggestion for a new questionnaire on the original thread, especially @Trish, who suggested similar questions to me, and produced the comprehensive S4ME PEM Factsheet; @ME/CFS Science Blog who said they were happy to collaborate (and see above re methodology); and @Bivox (Krisitian Sommerfelt), who shared his clinical experience on identifying and understanding PEM.]

Can Ijust say that I’m in the Peter White study as someone who didn’t have ME, according to Peter White who cited Oxford Criteria.
However, numerous other specialists disagree, citing evidence of PEM in Peter’s notes of his assessment of me. Also my dna was used in DecodeME.
What I’m saying is he shouldn’t be marking his own homework if he can’t even do the homework.

 

[ME/CFS Science Blog]

When it comes to making a diagnosis in real practice questionnaires are rarely used and from what I have seen suggested simply wouldn't be appropriate.

A lot of ME/CFS research already uses questionnaires to assess PEM, for example, for screening or diagnosing participants. So it would be valuable to have something that is better than the questions that are used today such as: do you have PEM, do you get tired after exertion, etc.

When it comes to other illnesses, I don't think there is any need to try to distinguish a special PEM for ME/CFS rather than other diseases.

Think it would be valuable to have questions that differentiate the characteristic PEM that ME/CFS patients experience from similar conditions such as chronic fatigue, depression, chronic pain that do not involve PEM as we know it. Tuberculosis, lymphoma, MS, or malaria do not come into play as these are more easily detected using other diagnostic signs and tools. It's mainly the other medically unexplained symptoms that we have to focus on.

As I see it we would have to make a list of PEM statements that reflect how we as patients understand the concept and are useful to detect ME/CFS. We would then need to test these in ME/CFS patients and various controls to check the sensitivity and specificity. If a statement is only agreed upon by a minority of ME/CFS patients for example, it might not be useful as a criterium. On the other hand, if it is agreed upon by many controls who don't have ME/CFS but other illnesses, then it is not very useful either.

This means we probably need a researcher as a partner, one that has access to patients and controls to extensively test the questionnaire. Another option is to try to contact other patient organisations and share the questionnaire online with their members to get data from non-ME/CFS controls.

 

[ME/CFS Science Blog]

The long PEM questionnaire by the Jason group gives some interesting data on potential PEM-questions:
Assessment of Post-Exertional Malaise (PEM) in Patients with Myalgic Encephalomyelitis (ME) and Chronic Fatigue Syndrome (CFS): A Patient-Driven Survey - PubMed
The development of an instrument to assess post-exertional malaise in patients with myalgic encephalomyelitis and chronic fatigue syndrome

I've added this longer PEM questionnaire in attachment.

 

[ME/CFS Science Blog]

. The NIH is developing one,

Not sure, the link you shared isn't explicit about this. I thought the Common Data Elements was mainly about selecting tools to recommend for use in research, rather than developing new ones.

Perhaps someone on the forum knows more?