Do we need a long-term actimeter study of PwME outside of treatment?

[Sasha]

I know I'm only one person, but that wouldn't capture some of the important things about me. My fluctuations tend to play out over years, not months.

My thinking was that by snapshotting many people over 6 months (say), we'd capture how many people experience marked improvements over a six-month period. If each of us only has a marked improvement every ten years on average, for instance, you could capture someone having that improvement if you had twenty people followed for a half-year (again, on average, hence we'd need big numbers).

I agree that longer-term tracking would be good but again, I can't see many people wearing trackers consistently for years on end.

Also, there's often a complete mismatch between what I do and how I feel. Sometimes I'm still active when I feel terrible and shouldn't be doing anything (today would be a good example!) because I need to. Other times I feel as good as I ever do, and just sit on the sofa just listening to music or playing a few tunes. My step counts aren't much affected by my severity status; I can't walk any further on a good day than I can on the worst. They will sometimes show a small increase, but that may be down to necessity rather than feeling better.

I agree that day-to-day wellness won't be captured but bigger, longer-term changes would be, for the most part.

 

[bobbler]

My thinking was that by snapshotting many people over 6 months (say), we'd capture how many people experience marked improvements over a six-month period. If each of us only has a marked improvement every ten years on average, for instance, you could capture someone having that improvement if you had twenty people followed for a half-year (again, on average, hence we'd need big numbers).

I agree that longer-term tracking would be good but again, I can't see many people wearing trackers consistently for years on end.

I agree that day-to-day wellness won't be captured but bigger, longer-term changes would be, for the most part.

6months wouldn’t tell you anything I think

It’s the timeframe it would take for me to know if I was just in PEM from ‘that thing’ and must have had a cold as I haven’t got my feet under me or had actually got a lower threshold (need to plan in more recovery for ‘that thing’ which is likely to be eg an unavoidable medical appt) or just to have a sense of what my threshold is eg how much I can brush teeth shower things related to eating talking getting downstairs in a month.

So I’d say the measure of someone’s ‘threshold’ ie level rather than it being muddled up with what would be day-to-day busy weeks for healthy people but that’s months of it evening out (as we can only stop brushing teeth moving or other appts so much even when recovering from something ‘bigger’) is probably going to be taken over 6-12months to be accurate as an average of a sine wave rather than confusing a snapshot that might be peak-to-peak or peak-to-trough or vice versa

Which means you could say eg ‘that 12months was on average better than the 12 months after/before’ by doing a rolling basis and I’d feel approx like that would give a not misleading ‘sense’ of direction my illness level had gone. But I couldn’t compare this dec to June as it would probably say more about what cumulatively had been on my plate in the months leading up to them - it needs that longer timeframe to average out what are things with big impacts cumulating with smaller things.

It’s only at eg 3yrs in I realised I was a new severity and not just ‘this horrible time I’ve been thru means I’ll need a very long time to recover’ but assuming after that sleeping off the acute (knowing I needed a years worth of sleep type feeling) I’d have been optic the same ballpark severity it had just been a bad ongoing time. I guess that’s no coincidence to the ‘some people recover in the first 3yrs if they take it seriously’ recovery rates

 

[Kitty]

I agree that day-to-day wellness won't be captured but bigger, longer-term changes would be, for the most part.

It might depend on how long you've been ill, I guess? My last change was a disastrous crash in 1999 that took two years to come out of fully.

Everything else is relatively minor, a few percentage points. Often it reflects how well I'm managing ME/CFS (pacing was never my my strongest suit) rather than any real change in the underlying severity.

But of course you could decide to exclude boring people like me!

 

[Sean]

My answer is yes. Two actimeters. One on the dominant hand (wrist), and one on either ankle.

Maybe a third on the upper dominant arm, just below the shoulder. Though that may be too impractical for long-term measurement.

Using a range of bio-monitors on a large number of patients to measure long-term patterns of general activity and biological parameters will provide good data to inform both research and clinical management.

As with genetics, this kind of data is useful whatever the results.

As long as it is done properly.

 

[Sasha]

It might depend on how long you've been ill, I guess? My last change was a disastrous crash in 1999 that took two years to come out of fully.

Everything else is relatively minor, a few percentage points. Often it reflects how well I'm managing ME/CFS (pacing was never my my strongest suit) rather than any real change in the underlying severity.

But of course you could decide to exclude boring people like me!

I think what we're really looking for is the big shifts. I have similar shuffles in a few percentage points but larger shifts are rarer. I think we want to capture changes that are beyond the scale of what can be affected by management (i.e. pacing).

 

[Trish]

I am a fan of activity monitors, and find my fitbit steps data useful as a relection of my state of functioning. There are downsides to wrist worn monitoring as my daughter discovered over recent months as she has taken up knitting again. Each stitch registers as several steps, so her step count goes up from the usual few hundred to thousands.

I agree there is potentially significant value in large samples of pwME monitoring over years of whaever can be done practically with wearables. I think this needs to be done alongside regular recording of overall symptom burden, for example filling in once a day on an app an overall symptom burden score from 0 to 10 with 0 being well and 10 being PEM/very severe, and perhaps filling in FUNCAP once every 3 months and ticking a box for ME/CFS severity level at the same time. I think there should also be some simple way to include any treatment being tested, meds and supplements, and major new health changes with unrelated illnesses, and major life events that impact health.

 

[Sasha]

I am a fan of activity monitors, and find my fitbit steps data useful as a relection of my state of functioning. There are downsides to wrist worn monitoring as my daughter discovered over recent months as she has taken up knitting again. Each stitch registers as several steps, so her step count goes up from the usual few hundred to thousands.

Good point!

I agree there is potentially significant value in large samples of pwME monitoring over years of whaever can be done practically with wearables. I think this needs to be done alongside regular recording of overall symptom burden, for example filling in once a day on an app an overall symptom burden score from 0 to 10 with 0 being well and 10 being PEM/very severe, and perhaps filling in FUNCAP once every 3 months and ticking a box for ME/CFS severity level at the same time. I think there should also be some simple way to include any treatment being tested, meds and supplements, and major new health changes with unrelated illnesses, and major life events that impact health.

If people are allowed to test treatments during monitoring, doesn't that negate the purpose of the study - to be sure that we're looking at natural variation? This is partly why I suggested a six-month monitoring period, because that seems a reasonable period for people to agree not to try new treatments. Longer than that, and people might not want to take part.

 

[Trish]

Good point!

If people are allowed to test treatments during monitoring, doesn't that negate the purpose of the study - to be sure that we're looking at natural variation? This is partly why I suggested a six-month monitoring period, because that seems a reasonable period for people to agree not to try new treatments. Longer than that, and people might not want to take part.

If people are to be asked to provide longitudinal data over years to get a really good picture of things like effects of long term pacing or not pacing, prevalence of big fluctuations, relapses and remissions as they occur in real life, then the it would not be ethical to ask them to stick to the no medication use or changes in use over years. Allowing them to act normally, including treatments for other conditions such as cancer might actually lead to new discoveries of potentially useful treatment avenues, as happened with rituximab.

 

[Sean]

If people are to be asked to provide longitudinal data over years to get a really good picture of things like effects of long term pacing or not pacing, prevalence of big fluctuations, relapses and remissions as they occur in real life, then the it would not be ethical to ask them to stick to the no medication use or changes in use over years. Allowing them to act normally, including treatments for other conditions such as cancer might actually lead to new discoveries of potentially useful treatment avenues, as happened with rituximab.

I agree, as long as there are good records of what the patients do so it can be factored in. Should help reveal what helps and what doesn't, among other things.

The key is having a large sample size, and doing it for long enough.

 

[Peter T]

I think the advantage of such a study would be more than just providing an overview of variation in ME/CFS but it would add to the information on using activity levels and hopefully work towards actimeter use becoming a standard outcome measure in ME/CFS trials.

 

[Sasha]

If people are to be asked to provide longitudinal data over years to get a really good picture of things like effects of long term pacing or not pacing, prevalence of big fluctuations, relapses and remissions as they occur in real life, then the it would not be ethical to ask them to stick to the no medication use or changes in use over years.

Agreed, which is why a years-long study might not be suitable for answering the question of what the natural course is of ME/CFS when you don't intervene.

Allowing them to act normally, including treatments for other conditions such as cancer

It would be unethical to expect that people should not be treated for other conditions, especially cancer, during the observation period - but those people might need to be analysed separately from the others

might actually lead to new discoveries of potentially useful treatment avenues, as happened with rituximab.

Aren't we then simply back into the world of anecdote, but with Ns too small to determine much of anything? Couldn't we argue that every case of big improvement following some random therapy could lead to new discoveries of useful treatments? DecodeME's second questionnaire included asking about interventions and I'd expect we'd get better data from that.

 

[EndME]

Agreed, which is why a years-long study might not be suitable for answering the question of what the natural course is of ME/CFS when you don't intervene.

It would be unethical to expect that people should not be treated for other conditions, especially cancer, during the observation period - but those people might need to be analysed separately from the others

I don't think one needs to complicate things. I can't see any problem of doing certain monitoring things long-term. If someone develops a condition and is treated for that you simply record that in your data, as one would for other treatments and do an appropriate analysis as one does for other conditions. People can live their life as they should want to without being part of any study (but have to deal with filling out certain questionnaires). The only restriction I see to that is that you might forbid access to activity tracking devices during a certain period of the study during which you use tracking devices that don't feed any information back to participants.

 

[Utsikt]

Just an annual FUNCAP in a national patient register could provide some value. It might be easy enough that you could get the improvers to stay on as well.

Of course using it requires that you acknowledge that PEM is a thing..