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

As always, the simple answer is: the rules are made-up and the details don't matter.

 

Perhaps that'll be so they can say (when there are no differences between treatment arms at long term follow up) "thats because the AM group all went on to have GET" ..... - like they did for PACE... (or was that PACE? perhaps it was one of the other studies iinto CBT/GET?)


@Beckster thanks so much for sharing, very best wishes to you and your daughter

ETA the word 'perhaps'

 

Results of the feasibility phase of the managed activity graded exercise in teenagers and pre-adolescents (MAGENTA) randomised controlled trial of treatments for chronic fatigue syndrome/myalgic encephalomyelitis
Pilot and Feasibility Studies volume 5, Article number: 151 (2019)

Published: 19 December 2019 (Open access)

https://pilotfeasibilitystudies.biomedcentral.com/articles/10.1186/s40814-019-0525-3

 

Having conducted multiple trials of the same thing, it's just bizarre to have a feasibility trial of something that Crawley herself has performed many times pretty much identically. It's been 4 years since initial recruitment. How is it even OK to call something a RCT when it is clearly not properly controlled. Do medical journals simply DGAF anymore?

And you have bizarre stuff like this:

That's quite the attrition rate:

57% did not even return a bunch of questionnaires at 6 months.

But they are totally capable of exercising regularly. Somehow this time the accelerometer was not dropped for being too "burdensome", though no data were provided.

I'm sure there will be a thorough evaluation of that. Absolutely. Very serious stuff.

What's randomization? Does it matter? About as much as proper controls, I guess.

As usual, no exploration of this obvious trade-off will be made.

Recording a few things on an app is too onerous but sure, they can exercise and, *checks note* watch television. Wow! Much improved! There really is no disease here!

Travel by car, as a passenger, is too much but, sure, they can totally exercise to the point of being able to watch TV!

Nevermind, they did drop the actimeters. Good grief, after having done this many times the review panel should be fired for gross incompetence for having approved of this. Zero accountability, ridiculous.

I'm at loss for words here.

This is serious research funded by serious research institutions approved by serious ethical approval people published in a serious medical journal by serious editors.

In a random sampling, therapists did not mix up rooms or patients. This is "fidelity of intervention". OK.

Too bad "activity management" is not actually an intervention, or a treatment, or anything, really.

Counterpoint: no.

 

I can't get excited about reading this stuff on Christmas Eve. But is there any sign that they realise that the feasibility trial has been found out, as a simple cheat? I guess one could argue that it is the proper thing to do to produce a published report even if the research is too poor to mean anything. At least someone might learn from people spending years on a phoney methodology.

Hopefully at least there will not be any more professors of feasibility trialling appointed. And while they about it no more professors of pragmatic trialling.

 

it will be interesting to look at what changes were implemented for the full trial based on the feasibility trial, because I'm assuming they folded in all the feasibility trial participants into the full trial, a la LP study. anyway it's only Christmas Eve early afternoon here.

 

Hmm...

 

Fitbits shouldn't raise questions. Kids love gadgets!

 

So the recruitment rate (±50%) was higher than expected (±25-30%).

The drop-outs were relatively low: 3 out of 40 (7,5%) in both the GET and activity management arm.

The vast majority of patients (155/161) were recruited at centre one, which I assume is Bath. The recruitment site at Cambridge was withdrawn from the study. The authors write that "clinicians were unable to deliver treatment, deliver recruitment calls or collect baseline data according to the protocol" without much information why this was the case.

34 patients were excluded because they were referred to psychological support (I assume this mostly means CBT) at assessment. So it seems that if patients at these recruitment sites are thought to be better off with CBT, that they will not be considered for MAGENTA, which creates selection bias.

There were some problems with the accelerometers not provided data for many participants (41.7% across the three time points) for example because patients did not wear them consistently.

The outcome measure EQ-5D-5 L has been changed to EQ-5D-Y because the latter is more suited for younger populations.

Adverse events will be monitored as follows: "For safety outcomes, we will prospectively collect serious and non-serious adverse events defined as any clinical change or illness reported at clinic or postal follow-up. We will define a serious deterioration in health as: a decrease ≥ 20 in SF-36-PFS or scores of “much” or “very much worse” on the Clinical Global Impression scale; clinician reported serious deterioration in health; or withdrawal from treatment because of feeling worse."

 

MAGENTA remains puzzling to me because the treatments are so much alike. It seems like one form of graded activity versus another that focuses on physical activity. So what exactly is being tested here: whether graded activity should focus on physical activity or not?

They also say they used a checklist of GET elements and activity management elements to check the fidelity of treatment. I would be interested in seeing that checklist, cause what exactly are the differences? The only one I notice is that GET is focused on physical activity rather than all forms of activity.

They mention that both approaches are recommended by NICE but poorly tested in adolescent ME/CFS patients. So why didn't they try to test them first?

 

They say they consulted with the University of Bristol CFS/ME Young Persons Advisory Group (YPAG). Does anyone know more about them?

 

If this can only be delivered according to protocol at a single site is the protocol not adequately defined/ are staff less engaged with the process ? I thi k more detail is needed for dropping a whole cohort. ( how much does this affect stats?)

I simply don't understand how any child with ME can have friends who are unaware re condition. That alone raises questions.

 

I don't understand why they didn't use both devices (FitBit and ActiveME app) for both groups. It would have been a great way to provide an internal consistency check of activity recording, and would also reduce expectation differences between the groups that might have affected outcomes ("Oh, but I only joined the GET group because I thought I would get a FitBit!" - actually, I bet that's why they found that participants didn't return them - they kept them because they liked them!).

Has anyone here used the ActiveME app? Might it be useful?

As some have said, the groups do seem awfully similar - GET vs GAT. What's next, I wonder? GIT vs GOT?

ISRCTN page seems to indicate that it was the Cambridge centre that dropped out, leaving Bath and Newcastle for the main trial. I expect we'll see the full results sometime next year.

 

And likely to cause significant harm to at minimum "some" children and adolescents (to my mind "some" is more than "a few" though even a few is too many). My guess would be over half unless they're using mis-diagnosed kids. They are so keen to "prove" that they are right, and have THE answer, that nothing would surprise me any more. Keep the funds flooding in for whatever ideas they have.

 

Anyone heard anything about MAGENTA? Said it was completed in 2019:

https://www.isrctn.com/ISRCTN23962803?q=&filters=conditionCategory:Nervous System Diseases&sort=&offset=6&totalResults=536&page=1&pageSize=10&searchType=basic-search

Presumably they'll be planning some PR push around NICE and I wondered if there was any sign MAGENTA would play a role.

 

published results here:
https://www.s4me.info/threads/esthe...-me-baseline-accelerometer-magenta-data.9141/

post with actual link to paper here:
https://www.s4me.info/threads/esthe...erometer-magenta-data.9141/page-2#post-177688