MAGENTA (Managed Activity Graded Exercise iN Teenagers and pre-Adolescents) - Esther Crawley

Yes because our MAGENTA inspired #stopGET petitions were written when we didn’t have PACE data. Though when we wrote to the ethics committee a bit later with our concerns they justified MAGENTA based on Cochrane rather than PACE (ie criticism of PACE is irrelevant).

PACE is a big contributer to the Cochrane data in the meta analysis and one of the few with harms data.

Also much of the criticism of pace applies to other data in the Cochrane meta analysis in terms of the subjective outcomes in open label trials and the outcome switching (in the Cochrane meta analysis).
 
PACE is a big contributer to the Cochrane data in the meta analysis and one of the few with harms data.

I agree but they seemed to be under the impression that this field would still stand without PACE. In a sense it’s a repercussion of our focus on PACE methodology problems- PACE is disregarded as potentially bad science, rather than CBT and GET being disregarded as bad treatment.

Cochrane uses a similar GRADE approach to NICE. A general critique has been that “the Cochrane Collaboration may cause harm by giving credibility to biased studies of vested interests through otherwise respected systematic review” (Ioannidis, 2016). This problem is potentially relevant to any systematic review of ME and CFS studies.

https://www.sciencedirect.com/science/article/pii/S0895435616001475

An issue is studies perceived to be well designed RCTs have high status, but they don’t assess individual studies they focus on overall evidence for a treatment/therapy. I’m guessing this means that they might be very quick in looking at individual trial methodology or perhaps even just the abstract?!
 
Technically the success of an intervention isn’t relevant to whether it is feasible. But in this case they must have recorded the Full Trial outcomes in the feasibility stage. If accelerometers have been now dropped, can we FOI the feasibility trial accelerometer results? Or if the children faked them, is this unhelpful for us to know?


Just dropping the measure when they are continuing the trial seems dodgy to me. This should be seen as a mid way assessment and change in outcomes measured rather than a feasibility study and then a new full trial. They are pulling data from the feasibility study into the full trial (as they did with smile).
 
I wouldn't assume that they've dropped accelerometers. They might be keeping them in their back pocket as a moderator, or measure of compliance, rather than having them registered as an outcome. That way they can release a positive result with much fan-fare, a null result can be downplayed or ignored.
 
I agree but they seemed to be under the impression that this field would still stand without PACE. In a sense it’s a repercussion of our focus on PACE methodology problems- PACE is disregarded as potentially bad science, rather than CBT and GET being disregarded as bad treatment.

Cochrane uses a similar GRADE approach to NICE. A general critique has been that “the Cochrane Collaboration may cause harm by giving credibility to biased studies of vested interests through otherwise respected systematic review” (Ioannidis, 2016). This problem is potentially relevant to any systematic review of ME and CFS studies.

https://www.sciencedirect.com/science/article/pii/S0895435616001475

An issue is studies perceived to be well designed RCTs have high status, but they don’t assess individual studies they focus on overall evidence for a treatment/therapy. I’m guessing this means that they might be very quick in looking at individual trial methodology or perhaps even just the abstract?!

Have you seen the post PACE reanalysis letter that bristol sent to the ethics committee and then the ethics committee gave them a clean bill of health. I think they ethics committee were mislead by bristol. The Cochrane review is bad as Bob and Tom pointed out to them. But they appear to have no quality control. The review itself pulls in the PACE data so it is heavily reliant on PACE - I'm pretty sure its the biggest trial and they mark its quality as high.
 
Here’s a thought. Bear with me. . .

My background is biology, and as a student, when surveying plants in the wild, we always used quadrats that were placed in very precise predetermined locations. So if surveying a transition from wood to grass land for example the quadrat positions could not be selected once you were in the field. As young students we always wanted the best results, and we’re always tempted to reposition our quadrats to include the plants that fitted our notion of what we thought we should see.

Of course we had it drummed into us, that biasing results like this by reselecting where our sample boundaries lay was NOT science, no matter how much tidier it made the results appear.

So, could these date switches be means a means of shifting the sample boundaries (a bit like repositioning a quadrat) to give a sample that gave the “best” (in their view) results?

It all seems a little weird
Certainly a possibility. This seems to be the default approach to 'science' by these folks. The most scary thing is they seem to think it is the right way to do it! If the answer does not solve the problem, then restate the problem so it does! @Brian Hughes book covers this lots, and is both fascinating and scary. It's "pantomime science" - would be hilarious if it wasn't for real.
 
I can’t remember but it sounds like we (people at #MEAction) got a very similar response @Adrian

The committee we wrote to is Frenchay based (also in Bristol the city).

I wouldn't assume that they've dropped accelerometers. They might be keeping them in their back pocket as a moderator, or measure of compliance, rather than having them registered as an outcome. That way they can release a positive result with much fan-fare, a null result can be downplayed or ignored.

Of course we’re just doing detective work at this point, but the latest list of outcomes in this thread had dropped all objective measures except school attendance as a secondary outcome. They shouldn’t really be allowed to reclaim accelerometers just because they throw up a significant result.
 
Of course we’re just doing detective work at this point, but the latest list of outcomes in this thread had dropped all objective measures except school attendance as a secondary outcome. They shouldn’t really be allowed to reclaim accelerometers just because they throw up a significant result.

Oh, sorry. Were they ever listed as an outcome? I thought they'd just not been listed as an outcome, rather than listed and then dropped
 
Were they ever listed as an outcome?
"
Feasibility analysis will include: the number of
young people eligible, approached and consented to the trial; attrition rate and
treatment adherence; questionnaire and accelerometer completion rates."

"
Quantitative data will include: the number of children who were eligible, approached,
consented and retained in the study; the completeness of questionnaire data at
baseline assessment and follow:up; the percentage of children providing usable
accelerometer data;"

"
We will calculate the proportion of children who wear the accelerometer and provide
usable data. We will assume periods of sixty minutes or more with zero readings as
“non:wear” time. Participant’s data will be included if they provide two or more
weekdays of data with at least 500 minutes of data between 6am and 11pm."

"
In addition to questionnaire measures,
participants in both trial arms will be asked to
wear an accelerometer (GT3X+) to measure physical activity for seven days within
one month of randomisation and at 3 and 6 months follow:up. Accelerometers will be
posted to participants with instructions. Participants will be asked to complete a log
of wear time (time worn and time taken off). Accelerometers are small, match box
sized devices that measure physical activity. They have been shown to provide
reliable indicators of physical activity among children and adults
23
. The accelerometer data will be processed to identify mean minutes of sedentary, light
and moderate to vigorous intensity physical activity per day using established
accelerometer cut off points and protocols
16 24
. The mean accelerometer counts per minute, which provides an indication of the volume of physical activity in which the participant engages, will also be calculated using established methods."

"We will interview children and adolescents and their parents about their use of the accelerometer, whether it is an acceptable device to wear and whether there are particular issues we need to consider in this patient group for the full trial."
 
"
Feasibility analysis will include: the number of
young people eligible, approached and consented to the trial; attrition rate and
treatment adherence; questionnaire and accelerometer completion rates."

"
Quantitative data will include: the number of children who were eligible, approached,
consented and retained in the study; the completeness of questionnaire data at
baseline assessment and follow:up; the percentage of children providing usable
accelerometer data;"

"
We will calculate the proportion of children who wear the accelerometer and provide
usable data. We will assume periods of sixty minutes or more with zero readings as
“non:wear” time. Participant’s data will be included if they provide two or more
weekdays of data with at least 500 minutes of data between 6am and 11pm."

"
In addition to questionnaire measures,
participants in both trial arms will be asked to
wear an accelerometer (GT3X+) to measure physical activity for seven days within
one month of randomisation and at 3 and 6 months follow:up. Accelerometers will be
posted to participants with instructions. Participants will be asked to complete a log
of wear time (time worn and time taken off). Accelerometers are small, match box
sized devices that measure physical activity. They have been shown to provide
reliable indicators of physical activity among children and adults
23
. The accelerometer data will be processed to identify mean minutes of sedentary, light
and moderate to vigorous intensity physical activity per day using established
accelerometer cut off points and protocols
16 24
. The mean accelerometer counts per minute, which provides an indication of the volume of physical activity in which the participant engages, will also be calculated using established methods."

"We will interview children and adolescents and their parents about their use of the accelerometer, whether it is an acceptable device to wear and whether there are particular issues we need to consider in this patient group for the full trial."

Thanks, even if they didn't have them as outcome measures in their registration, I'd say that listing them in the 'Clinical Outcome measures' section of the protocol counts: https://bmjopen.bmj.com/content/6/7/e011255.long
 
We will calculate the proportion of children who wear the accelerometer and provide usable data. We will assume periods of sixty minutes or more with zero readings as “non:wear” time. Participant’s data will be included if they provide two or more
weekdays of data with at least 500 minutes of data between 6am and 11pm."

Is this meaningful in relation to children with ME? There are the problems of the disrupted sleep patterns , such you can even see reversals of normal daily rhythms (ie active at night and sleeping during the day)? It could be that only using data between 6am and 11pm is misleading, especially if there is any intervention that impacts on sleep patterns (either directly or indirectly).

Does 60 minutes of zero readings necessarily equate to non wear time? When I have severe migraines or other extreme pain episodes I find that lying well supported with pillows as motionless as possible is the only thing I can do. I have no idea what would then show up on an accelerometer.

If electronic activity measurement was restricted to only sample periods of the day, it does does not allow for the possibility/probability that participants merely adjust what they do to allow for the increasing demands of a target activity without altering their total activity over a whole day or a whole week.

Having said that, this is irrelevant if they have completely dropped the accelerometer data. However it would seem that they had not really thought it through before the feasibility study.

If children were deliberately fixing their accelerometer results what does this tell us about the children's motives for involvement, their understanding of the experiment and their relationship with the researchers. How had they established that deliberate fixing occurred and what reasons did the children give for it?

If an objective measure was subject to fixing by the subjects, what grounds do we have for assuming that any subjective measures were immune to the same 'cheating'.

It does feel all this is rather a mess.
 
I thought the whole point of getting pwME to wear actometers in clinical trials was to gauge their entire activity over at least a week, preferably throughout the trial to look at both fluctuations and increasing or decreasing activity over time.
This nonsense about just wearing them for sample periods of the day demonstrates to me that they are not serious about getting anything interpretable from them.
 
I look forward to someone actually doing some robust research for paediatric ME. Surely it is not beyond the wit of man?

Since we obtained a FITBIT type device, we have gained a valuable insight into fluctuations and exacerbations. Although I am not convinced re absolute accuracy, it provides a relative comparison.

To me a trial that does not include continuous monitoring taking advantage of such devices is stuck in the 20th century.

Perceptions of things are not always accurate- we now know that there is little deep sleep for my daughter, that heart rate can spike at particular times/ with particular activities. Having begun to get the gist of it, thanks to @Keela Too 's blogposts, we are looking forward to aiming for a bit more stability - it is both surprising and fascinating finding out what actually affects heartrate as it is different for everyone.

To not have accelerometer data as an outcome, is for the reasons @Trish highlights, dodgy.
.
 
Outcome measures:
"
Primary outcome measure
Feasibility trial:
Feasibility and acceptability of investigating GET in a randomised controlled trial measured after 1 year.

Full trial:
Physical function is measured with the 36-Item Short Form Health Survey (SF36, physical function sub scale), collected at the 6 month time point.

Secondary outcome measures
Feasibility trial:
No secondary outcome measures.

Full trial:
1. School attendance is measured as percentage attendance of expected sessions
2. Fatigue is measured using the Chalder Fatigue score
3. Pain is measured using the visual analogue scale
4. Depression and anxiety are measured using the Spence Children’s Anxiety Scale (SCAS) and the Hospital Anxiety and Depression Scale (HADS, if they are 12-17 years old)
5. Health related quality of life is measured using the EQ-5D-Y

All of the above outcomes will be measured via child self-completed questionnaires at baseline, 6 and 12 months as well as a measure of physical function the SF36-PFS at 12 months."

http://www.isrctn.com/ISRCTN23962803?q=&filters=conditionCategory:Nervous System Diseases&sort=&offset=6&totalResults=536&page=1&pageSize=10&searchType=basic-search
Notice that, according to this, school attendance data will be obtained "via child self-completed questionnaires". So, once again not using actual school records?
 
Notice that, according to this, school attendance data will be obtained "via child self-completed questionnaires". So, once again not using actual school records?
Yes, that's an important point.

There's also the period over which the school attendance data are collected. If it's just a two week period, then a particular effort might be made to get to school for that time, with the young person collapsing afterwards. Of course, if it's longer than two weeks, a self-reporting approach means that the data are very very suspect.

Also, the approach taken to counting part-days. Does just getting to school count as a day attended, even if it's only for one class? Or does the young person have to attend for the whole day for it to count as 1.0? Are all the respondents taking a consistent approach to this?

But, no proper control treatment ... All the other flaws are insignificant compared to that one, when there is so much debate over whether these type of interventions are better than what would occur with just the passage of time and some sympathetic support.
 
Yes, that's an important point.

There's also the period over which the school attendance data are collected. If it's just a two week period, then a particular effort might be made to get to school for that time, with the young person collapsing afterwards. Of course, if it's longer than two weeks, a self-reporting approach means that the data are very very suspect.

Also, the approach taken to counting part-days. Does just getting to school count as a day attended, even if it's only for one class? Or does the young person have to attend for the whole day for it to count as 1.0? Are all the respondents taking a consistent approach to this?

But, no proper control treatment ... All the other flaws are insignificant compared to that one, when there is so much debate over whether these type of interventions are better than what would occur with just the passage of time and some sympathetic support.
But given that there is also the 80% recover " naturally" when looking at essentially " fatigue", that perhaps would undermine a whole lot more? ....
 
Back
Top