GETSET (white) in Lancet 22/06/17

[Jonathan Edwards]

I am travelling with limited internet but from what I can see so far this is just as a non-starter as PACE.

Why was the trial called 'pragmatic'? That usually means 'we know it's bad but it is the best we could do'. It is not usefully controlled. It is supposed to be randomised but that is not enough to make it any use.

 

[user9876]

The mean is not necessarily an accurate figure as it can be reduced by very low scores; the median is more useful in such a situation. Also in the general population are a lot of elderly people as well as some with significant ongoing disabilities and illnesses who would be excluded from the study. The mean for the general population is 84. Getting close to that doesn't mean you are recovered.

Also the general population is not a normal functioning population. It sounds to me that they used the mean to mislead the ethics committee to get changes in the primary outcomes.

 

[user9876]

So we are given information on physiotherapist-rated data on adherence but as far as I can see not information on how the patients rated their adherence to the protocol.

What would be interesting is a sensitivity analysis to see if adherence made a difference to the results.

 

[user9876]

In the GES group, a similar proportion of participants improved by a clinically meaningful amount on both primary outcomes (34%) and scored themselves in the high range of physical activity (IPAQ; 30%) at follow-up, which provides some support for these thresholds.

I don't see how they can claim this helps support their arbitary threshold for clinically meaningful. IPAQ measures subjective reports of physical activity so get patients doing more then they will report this but that does not necessarily correlate with a health improvement. It could even be to keep therapists happy by saying yes of course I did some exercise today.

 

[Dolphin]

Two (2%) patients in the control group and ten (9%) patients in the GES group were lost to follow up (p=0·03; figure). Two (2%) participants in the GES groups and none from the control group actively withdrew from the intervention; this was not significantly associated with study group (p=0·50).

For anyone who doesn't know, the usual threshold for statistical significance is p <0.05. So the difference in terms of lost to follow-up a statistically significant for graded exercise self-help.

I see they do describe this in words later:

Significantly more participants were lost to follow-up in the GES group than in the control group; a sensitivity analysis, assuming participants who withdrew would report worsening, found no significant difference in worsening between study groups. Two participants actively withdrew from GES, but neither reported their withdrawal as being due to the intervention causing them harm.

 

[Dolphin]

Serious adverse events were uncommon (a participant attended Accident and Emergency [A&E] department after falling and damaging an arm; no fracture was found, and they were discharged; a participant attended A&E after twisting a knee, a damaged cartilage was diagnosed in the knee, and they were discharged; and a participant was admitted to hospital overnight for numbness in the right arm and leg, a neurologist assessed them and they were discharged the next day), and no serious adverse reactions were reported in either group.

Only one of these 3 were in the GES group, so not that interesting.

But it is a bit disappointing that in general we are not told which is in which group. It means the data can't be collated from a number of studies to look for any patterns. I doubt it would be acceptable with most drug trials.

 

[user9876]

In terms of the SAR I would be concerned that some experts looked to see if they felt anything could be attributed to the treatment. A much safer way to do it is to ask if an event could clearly be attributed to an independent event and otherwise say it may be due to treatment. Randomisation should even things out a bit but it is harder with low probability events.

 

[Daisymay]

AfME & the MEA don't really have much influence. They go along to the CMRC and other meetings and nothing changes.

AfME take the minutes to avoid having to voice an opinion. So what does the MEA do?

On the entry criteria they used NICE but they did an assessment for Oxford and CDC. What seems a little strange is that only 83 and 87% that met the NICE guidelines met the Oxford ones.

Could that be because Oxford doesn't require PEM but NICE does?

 

[Dolphin]

severity of disability according to the Short-Form 36 physical function subscale (SF-36 PF, ≤40 and ≥45).17

Participants in the GES group with poor baseline physical functioning were more likely to report improvement at follow-up (eight [20%] reported improvement in CGI-CFS and nine [23%] in CGI-health) whereas this was not observed in their counterparts in the control group (zero patients for both CGI-CFS and CGIhealth).

Based on these figures, around 40 were in the low functioning group at baseline (and so around 57 or in the higher functioning group).

After adjustment for covariates, patients in the GES group who had worse baseline physical functioning (ie, low SF-36 PF score) had a higher mean score at follow-up (56·9, SE 3·3) than their counterparts in the control group (44·2, 3·3).

This is interesting because the overall average was only 55.7.

This means that the group that had a baseline SF-36 physical functioning scale of 45 or more actually ended up with a very slightly lower average score (54.9) than those who who started with scores of 40 or less (56.9). Anyone can score much more than the mid-fifties on the SF 36 physical functioning subscale so it's not really believable that this is to do with any sort of ceiling effect when they claim that patients can get back to normal functioning.

 

[Dolphin]

Our finding that GES was more useful in those with worse physical functioning is reassuring and has been reported previously,11 but further exploration is necessary because it might be related to a ceiling effect in those with good physical functioning at baseline. This ceiling effect might also explain the relatively smaller difference in the effect size for physical function, which would reduce the overall difference between study groups.

Just to reiterate my last post with this particular quote.

Those who started off with an SF-36 physical functioning score of 45 or more only ended up at an average of 54.9. This is nothing close to a ceiling score. Healthy people of working age tend to score 95 or 100.

Based on the following

However, during recruitment we noticed that a significant minority of participants scored close to the mean of the general population (ie, normal physical function) so could be considered recovered even before any intervention [43].

It looks like they are referring to people with scores of 75 or 80 perhaps (domain for the general population is 84) who still have scope to improve.

And there probably wasn't very many of them given the overall mean and standard deviation.

Also at baseline only 3% in the graded exercise therapy group had a high physical activity level so it looks like they had scope to improve.

There shouldn't be that many if any people with physical function scores of 95 or 100 (i.e. with no scope for improvement) diagnosed with CFS and waiting for therapy.

 

[Dolphin]

Our finding that GES was more useful in those with worse physical functioning is reassuring and has been reported previously,11 but further exploration is necessary because it might be related to a ceiling effect in those with good physical functioning at baseline. This ceiling effect might also explain the relatively smaller difference in the effect size for physical function, which would reduce the overall difference between study groups.

A freedom of information act request might yield interesting information on the baseline scores. Ideally you would want all the baseline scores but even the number at baseline over a high threshold e.g. 85 or 90 would be interesting.

 

[Dolphin]

All outcomes were selfrated, which might lead to bias by expectation, although the effects of such self-rating are uncertain because of the mixed perception of GET.30

Bias due to expectation is only one of the possible problems that can be caused by relying on self-report measures alone. Undertaking the therapy itself with all the time invested might cause one to hope it was worth it and give more positive results. Also due to the contact with a therapist, you might want to want to say you are feeling better to please them or because you wanted to claim you tried hard (e.g. social desirability bias) (though technically they do not ask the questions for the outcome measures from what I can see).

 

[Dolphin]

On the entry criteria they used NICE but they did an assessment for Oxford and CDC. What seems a little strange is that only 83 and 87% that met the NICE guidelines met the Oxford ones.

One difference is that the Oxford criteria require 6 months of symptoms while it's only 4 months for the NICE guidelines.

 

[Dolphin]

We did not measure any objective outcomes, such as actigraphy, which might have tested the validity of our self-rated measures of physical activity.

They mentioned this now but why didn't they include them as part of the design.

 

[Sea]

So 1 in 5 patients can have a small improvement by doing GET. Not exactly a very positive result. It can probably be explained by some sufferers under-estimating how much exercise they are capable of as when your energy goes down a lot because of this illness you have to estimate your new activity limits. So there probably isn't ANY real physical improvement going on, even in those who benefited. They are simply realizing they can do slightly more than they thought and then report that as an improvement when their illness hasn't actually improved at all.

e.g. Before the GET they may have thought they could walk 10 minutes a day. Then by the end of the GET they find they can actually walk 20 minutes a day. They report that they can now walk twice as far and thus it seems like they have improved. Whereas in reality they could walk 20 minutes a day the whole time and just estimated how far they could walk wrong in the first place because we can only guess/estimate how far we can walk without triggering a post exertion malaise reaction. So it's really merely changing their perspective on their illness rather than actually improving it.

What is even more likely for some is that they did less of other things they were doing before the trial and substituted exercise making it appear they were doing more. Without objective measures we cannot know.

 

[NelliePledge]

Sorry for my ignorance folks but even if you ignore all the issues over the way it's set up if you take it on face value is this really only saying that the "benefit" of GET over standard treatment is only that participants scored 4% lower on the fatigue scale at the end of 12 weeks. Does that really even register as significant.

Secondly does this actually help back up the reanalysis of the PACE data which showed a similar very small effect.
In which case - despite the media coverage around publication - maybe this actually can help with debunking PACE. Or am I just being exceptionally naive.

I noticed there was a very short comment in the SMC experts which seemed surprisingly disparaging.

So in NiCE terms can the cost of GET programmes continue to be justified for such a minimal effect on the minority who perceive a benefit. And given the negative impact report by a significant number with the consequences for individuals and cost to the economy.

It isn't as if it's an actual treatment like some cancer drugs that might not cure but having a small impact is still of value by prolonging the lives of those who benefit.

 

[RogerBlack]

A bit odd that this group was chosen given this is not a study on young people in particular. I wonder did they ask anyone else and they refused or alternatively did they immediately pick what they saw as the weakest group.

Weakest? Naah. Strongest. Esther Crawley is an adviser for them after all.
I would not be truly shocked if EC was on the steering committee.

However, it seems according to the following post I was incorrect.

 

[Dolphin]

Weakest? Naah. Strongest. Esther Crawley is an adviser for them after all.
I would not be truly shocked if EC was on the steering committee.

The members of our Trial Steering Committee were Christopher Williams (Chair), a patient charity representative, Alastair Miller, Rona Moss-Morris, and a patient representative. The members of our Data Monitoring and Ethics Committee were Astrid Fletcher (Chair), Charlotte Feinmann, and Irwin Nazareth. Hiroko Akagi and Vikki McKeever scrutinised our adverse event data.

 

[NelliePledge]

Moss Morriss is the big proponent of MUS I believe - does that make her the chief MUPPET

And isn't Alistair Millar the ex medical adviser to AFME?? If so yet again they are making weasel statements with a veneer of criticism to paper over their integral inviolvement in this bullshit

 

[Sea]

Those who started off with an SF-36 physical functioning score of 45 or more only ended up at an average of 54.9. This is nothing close to a ceiling score. Healthy people of working age tend to score 95 or 100.

I read it as a ceiling of benefit of GET rather than reaching the ceiling of normality. So for example if the maximum gain that could be had was to say 60, a person at 45 who improved to 60 did better than a person at 55 who improved to 60. I'd have to reread it to see whether I understood them properly