UK:ME Association funds research for a new clinical assessment toolkit in NHS ME/CFS specialist services, 2023

[Adrian]

Yes Adrian some of us have flagged up that use of trackers should be considered and there are symptom tracking apps like visible (which already incorporates the Norwegian FUNCAP questionnaire as a monthly review)

unfortunately these comments have fallen on deaf ears and 20th century models are presented as gold standard

My point though is more that the use of trackers. AI really enables very different interfaces and ways to interact so its not just say an accelorometer or a form on a phone. Someone could speak/type a comment into their phone of what is on their mind about symptoms and health this can now be transcribed, interpreted, classified and advice given - using LLMs (Transformer networks).

I'm starting to think much more radical recording and interaction approaches can be enabled now.

 

[Jonathan Edwards]

I'm starting to think much more radical recording and interaction approaches can be enabled now.

Detailed recording ought to be useful for research but the issue here is how people should be assessed in routine practice. Until we have evidence of benefit of one action in one situation and another in another I see no point. People should see a physician who can listen to and advise on any specific problems they raise, check alternative diagnoses, activate OT support etc. All the things physicians did forty years ago and presumably still do. A questionnaire just gets int he way of focusing on what matters at the time. So would a long actometer read out, I suspect.

 

[Adrian]

Detailed recording ought to be useful for research but the issue here is how people should be assessed in routine practice. Until we have evidence of benefit of one action in one situation and another in another I see no point. People should see a physician who can listen to and advise on any specific problems they raise, check alternative diagnoses, activate OT support etc. All the things physicians did forty years ago and presumably still do. A questionnaire just gets int he way of focusing on what matters at the time. So would a long actometer read out, I suspect.

How about a summary of how someones symptoms change with activities written in a short paragraph and derived from data recorded at the time. The problem is that people forget especially with cognitive issues or represent the most severe issues rather than the complete picture. So why not have some one record their situation as they see it at the time and automate the summarization for clinical situations (in a way that a doctor could interact with). And I'm not talking about a long actometer trace. More people saying what their symptoms are as they are of concern or just after and then using AI methods to transcribe, interpret and summarize (the technologies are now there to enable that now with the latest language models). So I'm thinking we could look at how recording can be transformed over the next year or so in a way that could deal with issues with ME. I'm thinking in terms of monitoring (detecting change and treatable symptoms) rather than diagnosis.

It just feels like technology is bringing radical new possibilities and healthcare will at some point catch up.

 

[Trish]

Crossposted with Adrian.

I think Adrian's idea is worth exploring. It wouldn't necessarily produce a long readout, rather, if I understand it correctly it would also include analysis of features such as connections between activity, medications, symptoms and signpost patterns the pwME might not have spotted in their symptom contingent pacing.

The aim for pwME medical care should be to empower the pwME to manage their own activity and symptoms as much as possible with the help of tech. Then any time spent with a doctor can focus on diagnosis, new or troublesome symptoms, prescriptions, referrals etc.

If tech is designed well it should be able to provide the most relevant data in summary form, not pages of printouts, that I would love to have to back up what I tell my GP so she doesn't just think I'm tired.

 

[Trish]

People with other chronic conditions such as asthma and diabetes are self monitoring and recording data on apps which help both with the patient managing their symptoms, and their data is accessible to their doctor or specialist nurse in real time. That should be possible for pwME too.

 

[Jonathan Edwards]

People with other chronic conditions such as asthma and diabetes are self monitoring and recording data on apps which help both with the patient managing their symptoms, and their data is accessible to their doctor or specialist nurse in real time. That should be possible for pwME too.

The monitoring should be possible, but what would you do with it?

For asthma and diabetes we have well documented treatments that provide evidence for doing this in this situation and that in that situation - more bronchodilator or more insulin maybe.

For ME/CFS we don't have that. We don't even have any data showing that pacing according to monitoring of activity is helpful. Patients develop an idea of how to pace but who knows whether some numbers on an App are more use than how you feel today?

Everyone wants to measure things and everyone wants to find reasons why this happens or that, but measuring things and providing reasons why this happens are exactly the things that are wrong with current rehabilitative treatment paradigms. Recommending measuring things when we don't know it is any use is what we are up against.

But I would agree that what are desperately needed are research studies that fin out if measuring things is useful.

 

[JellyBabyKid]

I'd say predictably those comments have fallen on deaf ears. Nothing has moved on from 'the therapist knows best' and 'the rehab professional knows best' - and as demonstrated right here on this forum, ME patients who disagree with therapists/rehab professionals will be told "Don't be silly" (the quote from Sarah Tyson)..

This really gets to me. Paternalism at its worst.

Michael J. Fox says “Parkinson's patients are the experts on what we have. We have a responsibility as patients to share our experience — what works for us, what we respond to, what we can contribute to research.” But then he had to set up The Michael J. Fox Foundation for Parkinson's Research to make any progress

You can read about apples all you want, but until you eat one, you haven't got a clue what it tastes like. Which is why researchers need to listen to patients as there is a chasm in understanding between what we experience and what they think we experience.

This all feels like rearranging deckchairs when we should be building life rafts.

We don't need new ways of measuring - we have so many ways - we need curiosity about why this is is happening and how to fix it - or at least not make it worse and not adding reams of complex admin.

 

[Trish]

The monitoring should be possible, but what would you do with it?

I think I would have found it very helpful when I was newly learning to cope with symptoms and crashes and still working part time. It might have been possible to have objective data both to demonstrate to my disbelieving employers and doctors (and even some family and friends) that my activity levels were leading to crashes and I needed adjustments at work and home. I think it would also have helped me with pacing especially if it could warn me I need to stop and rest, and not exert more that day. Similarly when applying for benefits, if I we can show what activity levels lead to crashes and how long they last etc, then we have much more objective evidence that what we say is true.

 

[JemPD]

This really gets to me. Paternalism at its worst.

Michael J. Fox says “Parkinson's patients are the experts on what we have. We have a responsibility as patients to share our experience — what works for us, what we respond to, what we can contribute to research.” But then he had to set up The Michael J. Fox Foundation for Parkinson's Research to make any progress

You can read about apples all you want, but until you eat one, you haven't got a clue what it tastes like. Which is why researchers need to listen to patients as there is a chasm in understanding between what we experience and what they think we experience.

This all feels like rearranging deckchairs when we should be building life rafts.

We don't need new ways of measuring - we have so many ways - we need curiosity about why this is is happening and how to fix it - or at least not make it worse and not adding reams of complex admin.

I completely agree, but the challenge is that Sarah Tyson... IS a patient. And the mother of another patient. Russell Fleming is a patient. Keith Geraghty is a patient.

Clearly it takes more than simply being a patient, to understand why PROMS arent the ideal thing

Just to clarify... not intending anything personal to any of them with that comment, just wanted to illustrate that its not as simple as patients vs researchers/rehab professionals

 

[Jonathan Edwards]

I think it would also have helped me with pacing especially if it could warn me I need to stop and rest, and not exert more that day. Similarly when applying for benefits, if I we can show what activity levels lead to crashes and how long they last etc, then we have much more objective evidence that what we say is true.

Yes I agree absolutely on this, but it need to be done as a research exercise to see what predicts what.

 

[Trish]

Yes I agree absolutely on this, but it need to be done as a research exercise to see what predicts what.

Every pwME's efforts at pacing is a research exercise carried out by the pwME as we navigate every day's activity and rest.

All I'm suggesting here is adding more information to that research exercise to enable us to learn what we find works best given our individual disease and life situation.

To take a simple example, even after decades of attempting to navigate a safe route for me between what I needed to do and what I could do without crashing, it wasn't until I started noting my step counts each day using a fitbit that a pattern emerged and I was able to find a ceiling of number of steps I needed to stay below, and was able to significantly reduce frequency and severity of crashes.

The idea would be to replace laborious hour by hour activity and symptom diaries commonly used by therapists with more objective and less onerous data collection and analysis.

Some people find heart rate also provides immediate warnings of need to slow down.

 

[JellyBabyKid]

I completely agree, but the challenge is that Sarah Tyson... IS a patient. And the mother of another patient. Russell Fleming is a patient. Keith Geraghty is a patient.

Clearly it takes more than simply being a patient, to understand why PROMS arent the ideal thing

Oh. Oh heck.

Stupid question, but has the S4ME letter been sent to Keith and Russell as well as the trustees..? Keith seems to be pretty switched on, so I understood, so am surprised by this.

Russell and MEA seem to be focused on patients on the mild scale, but surely even at that level a 98 question form would be a challenge?

 

[Trish]

Stupid question, but has the S4ME letter been sent to Keith and Russell as well as the trustees..? Keith seems to be pretty switched on, so I understood, so am surprised by this.

They surely will have seen our letters and are free to read this thread.

 

[JemPD]

I am right that Keith is a patient arent I? just suddenly had a moment of doubt but dont i remember that Wessely etc discounted his research because of that? or am i getting confused with someone else?

 

[Kitty]

I think I would have found it very helpful when I was newly learning to cope with symptoms and crashes and still working part time.

To me, there might also be an opportunity to work on something sooner rather than later, iron out as many of the creases as we can, and then it's available as part of post-treatment monitoring when we do have potentially useful interventions being trialled.

If somebody who likes partnership working is running a trial, a tool like this might even look good on a grant application as an experimental part of it, run in parallel with whatever conventional monitoring the PI plans to do.

 

[Jonathan Edwards]

Every pwME's efforts at pacing is a research exercise carried out by the pwME as we navigate every day's activity and rest.

In a sense maybe, but what we normally call research is investigation that reliably transfers to a general situation.

And when it comes to the individual we have to face the question as to just how reliable these conclusions are about what helps pacing. How can we be sure that the change in tactics relating to concluding x causes y led to a better patch because x causes y. Maybe actually p causes y and x causes z but there is enough overlap for it to seem x causes y.

The point being that patients accounts of discovering what works have to be treated on a par with therapists accounts that they have discovered what works - GET and CBT and LP - unless we have reliable evidence that can be generalised.

I agree that if I had ME/CFS then I would probably get an actometer and try to work out how it might help me pace. But the exercise here is devising a monitoring format for general use. That needs a reliable evidence base.

 

[MrMagoo]

I've not been following this debate but it occurs to me that developing a new PROM could be the wrong approach. A PROM is basically a questionnaire that gets people to fill in symptoms/abilities/etc retrospectively (which isn't great when there are cognitive issues). Equally ME has a huge range of symptoms many of which could be relevant in a clinical context where symptoms could be treated (or help given).

We live in a world of rapidly changing technology and could we have new approaches to patient recording rather than a occasional PROM (its all very 20th century!)

Many have smart phones (with interactive screens, accelerometers, microphones)and there is readily available cloud compute for storage and processing along with AI for interpreting what people say.

Hence I have a question: If you were able to record daily records of symptoms + some notion of underlying symptoms what would work? Some sort of diary recording (say allowing someone to say or write 'I overdid it yesterday so I'm feeling crap today' or 'feeling more dizzy when I stand up' etc or even 'excited about family visiting tomorrow'). - We have transformer networks that do transcription and classification/interpretation of text which could then feed into a patient symptom record which could be augmented with accelerometer or phone/laptop usage data,

Or would an app asking for worst symptoms of the day work.

I feel we could come up with a much better interaction model that could help capture a patients experiences in a way that could be summarized and used for clinical decision making.

We certainly could, but it falls on deaf ears!

 

[MrMagoo]

My point though is more that the use of trackers. AI really enables very different interfaces and ways to interact so its not just say an accelorometer or a form on a phone. Someone could speak/type a comment into their phone of what is on their mind about symptoms and health this can now be transcribed, interpreted, classified and advice given - using LLMs (Transformer networks).

I'm starting to think much more radical recording and interaction approaches can be enabled now.

I’ve used free trackers like Bearable in the past to track symptoms. I often can’t speak so an app is preferable.

There’s an issue with your idea regarding “advice given” as what advice would the NHS give to me?!Manage your symptoms? Try pacing?

 

[MrMagoo]

How about a summary of how someones symptoms change with activities written in a short paragraph and derived from data recorded at the time. The problem is that people forget especially with cognitive issues or represent the most severe issues rather than the complete picture. So why not have some one record their situation as they see it at the time and automate the summarization for clinical situations (in a way that a doctor could interact with). And I'm not talking about a long actometer trace. More people saying what their symptoms are as they are of concern or just after and then using AI methods to transcribe, interpret and summarize (the technologies are now there to enable that now with the latest language models). So I'm thinking we could look at how recording can be transformed over the next year or so in a way that could deal with issues with ME. I'm thinking in terms of monitoring (detecting change and treatable symptoms) rather than diagnosis.

It just feels like technology is bringing radical new possibilities and healthcare will at some point catch up.

I’m happy with the reports which visible spits out, and I can request further CSV symptom data extracts if I wish. I could do that and crunch the data. Absolutely no GP or Consultant would be interested.

I think Tyson, Gladwell, want to do what they’ve always done, and we’ll get what we’ve always got. And they want my fundraising cash (on top of an NHS salary) to do it. Shame on the Me Association

 

[bobbler]

I've not been following this debate but it occurs to me that developing a new PROM could be the wrong approach. A PROM is basically a questionnaire that gets people to fill in symptoms/abilities/etc retrospectively (which isn't great when there are cognitive issues). Equally ME has a huge range of symptoms many of which could be relevant in a clinical context where symptoms could be treated (or help given).

We live in a world of rapidly changing technology and could we have new approaches to patient recording rather than a occasional PROM (its all very 20th century!)

Many have smart phones (with interactive screens, accelerometers, microphones)and there is readily available cloud compute for storage and processing along with AI for interpreting what people say.

Hence I have a question: If you were able to record daily records of symptoms + some notion of underlying symptoms what would work? Some sort of diary recording (say allowing someone to say or write 'I overdid it yesterday so I'm feeling crap today' or 'feeling more dizzy when I stand up' etc or even 'excited about family visiting tomorrow'). - We have transformer networks that do transcription and classification/interpretation of text which could then feed into a patient symptom record which could be augmented with accelerometer or phone/laptop usage data,

Or would an app asking for worst symptoms of the day work.

I feel we could come up with a much better interaction model that could help capture a patients experiences in a way that could be summarized and used for clinical decision making.

The issue no one talks of is coercion and all of the forms that come under this (social pressure , culture, all these links to schools, benefits, employers everything on which someone’s ability to live relies)

And these measures not striving to protect safety of those who will be handing over what will be treated as exam results as much as medical exam results. All activity logged against their name as a record inferring all sorts. If someone will lose their job if they aren’t seen to improve to start getting back to work that’s going to coerce results. Do you get forced to say you managed to increase something and it isn’t finishing you off so you don’t lose your house and that gets logged as ‘treatment worked’? Are they solving people getting pushed into increasing - whatever it does to their body- with no responsibility from the HCP if that leaves them more disabled 12months down the road?

These things shouldn’t, and wouldn’t in any normal clinical context of other illnesses, be being fed off all of the place.

along with no objective measures that show the actual health and physical function - just either short term ‘can’ or subjective self-report ‘did’ behind which many life sustaining pressures (or eg ‘we will remove your freedom’) might be levied