David Tuller: Trial By Error: Professor Sharpe’s Pre-Hearing Briefing for Monaghan

Lucibee said:
Having now done it, I can totally agree with you! That was sooooo fun (not)!

https://lucibee.wordpress.com/2018/...for-the-21-june-2018-westminster-hall-debate/

[btw - I'm happy to edit to add links or comments or whatever - or if anyone wants to reblog it elsewhere - go for it!]
Click to expand...

Thanks a lot for that Lucibee. Made me feel even more irritated with Sharpe!

"This is demonstrated by the lack of any difference in the more objective outcomes used in the trial."

I thought I'd mention that there was a statistically significant difference between SMC and SMC+GET for the 6mwt.

That briefing from Sharpe makes it sound like he's really just not up to the job of engaging in a debate about the quality of his work.

Strange how Sharpe made a deal about TSC approval for the protocol deviations on their primary outcomes, but made no mention of this for the 'recovery' deviations.

Sharpe wrote: "Whist there is no agreed definition of recovery the figures we reported were approximately 20% with CBT and GET, and approximately 8% with the other conditions."

Why change from 22% for CBT/GET, 8% APT and 7% SMC?

Later on Sharpe said: "We published a secondary paper exploring different definitions of recovery and found about twice as many people could be considered recovered with CBT and GET (about 20%) than with APT and SMC (about 10%)."

Is he now trying to play down how impressive their claimed results were?

This is my favourite bit from Sharpe: "This released data has been used to “reanalyse” the trial results, with a claim that the published results were misleading. The reanalysis is flawed and misleading."

What's with the scare quotes around "reanalyse"? He can't even bear to use that term for the analysis which just followed his pre-specified protocol? Why won't he explain why this so-called reanalysis is flawed and misleading? Maybe he's just sure it is, but can't work out exactly how.

"Another reason is that patient experience of CBT and GET outside the trial is reported as often being unhelpful or even harmful. This is an interesting and worrying observation probably explained by the treatment being given to people dissimilar to those in the trial or being given incorrectly."

Interesting that he seems to here be suggesting that any positive results from PACE aren't actually that helpful for providing information to patients about the likely impacts of the treatments they might be considering outside of RCTs. This is another problem with a lot of rehab approaches - they can be applied in wildly different ways by different therapists.
 
from
@Lucibee blog
MS:
Another reason is that patient experience of CBT and GET outside the trial is reported as often being unhelpful or even harmful. This is an interesting and worrying observation probably explained by the treatment being given to people dissimilar to those in the trial or being given incorrectly.
L:Quite a few of such reports have come from patients who were treated at the same centres that were part of the trial.

the point about 'treatment being given incorrectly' is BS when talking about CFS clinics as the BPS brigade Chalder and co were the ones that wrote the manuals for PACE that were then adopted by the clinics and Chalder &co trained them up, wrote the book didnt she, so if they arent doing it right it is the faulty training by his friend Trudy
 
Esther12 said:
"Another reason is that patient experience of CBT and GET outside the trial is reported as often being unhelpful or even harmful. This is an interesting and worrying observation probably explained by the treatment being given to people dissimilar to those in the trial or being given incorrectly."

Interesting that he seems to here be suggesting that any positive results from PACE aren't actually that helpful for providing information to patients about the likely impacts of the treatments they might be considering outside of RCTs. This is another problem with a lot of rehab approaches - they can be applied in wildly different ways by different therapists.
Click to expand...
This is an interesting change of tack. Up till recently, the PACE authors have dismissed patient charity surveys on the grounds that they are inherently unreliable, because people who join these groups are socialised to believe they are ill. I think they realise that gives away too much about their attitude to patients and patient groups. So it's interesting to see them change tack, and go more for the "there must have been something different about those cases" line.
 
Last edited:
Esther12 said:
I thought I'd mention that there was a statistically significant difference between SMC and SMC+GET for the 6mwt.
Click to expand...

I think the data on this is invalid because such a large proportion of patients didn't do the 6 minute walk at the 12 month follow up. Given that refusal to do it is most likely to be deterioration of symptoms, and that more in the GET group didn't do it, any claims of the GET group doing better on the walk are invalidated.
 
Am I the only person appalled by Michael Sharpe's lack of attention to detail when it comes to grammar? I've noticed on twitter that when he gets petulant his spelling and grammar go out of the window, but this in a letter to an MP:

There has been many false allegation made
Click to expand...

(by not patients in general)
Click to expand...

I would frankly feel insulted to receive such a slovenly cobbled-together communication.

We are happy to supply further information to back up these statements if required.
Click to expand...

Ok, please do. Critics of the PACE trial have been waiting for you (singular or plural) to do precisely that for the last few years.

It is hard to see that if anyone had wanted to bias the results or commit fraud, how they would be able to do so.
Click to expand...
On the contrary, Michael Sharpe's problem is that it's very easy to see exactly what the authors did.

Most of these refusals were upheld, but on one occasion, a FOI tribunal ordered the release of some of the trial data.
Click to expand...
This is a complete misrepresentation of how the appeals process works. To try to imply that the tribunal decision was "one occasion" out of line with all the rest, when it was an appeal decision which examined and corrected the wrong decisions of a lower body, well, to try and get away with that when writing to an MP ...
 
Sean said:
The subjective-objective correlation test I most want to see is the 6MWT for the GET arm.

It is the only objective outcome that delivered a statistically significant result (though not clinical significance).
Click to expand...

Given that we have the data for SF-36, CFQ and the 6MWT, that could be done if anyone could be bothered. The question is, what do you do about all the missing data which invalidates the whole enterprise.
 
Trish said:
I think the data on this is invalid because such a large proportion of patients didn't do the 6 minute walk at the 12 month follow up. Given that refusal to do it is most likely to be deterioration of symptoms, and that more in the GET group didn't do it, any claims of the GET group doing better on the walk are invalidated.
Click to expand...
Yes, true. I think @Esther12 is just pointing out the need to choose our words with care.
 
Sean said:
The subjective-objective correlation test I most want to see is the 6MWT for the GET arm.

It is the only objective outcome that delivered a statistically significant result (though not clinical significance).
Click to expand...

Not sure whether this will upload, but I did a video of that a while ago... I have done 6mWT vs PF in GET arm separately, but I can't work out how to convert it into an uploadable format...

But here's the gif...(GET is bottom right) PACE_6mWTvsPF.gif

[update: I do have a better version of this gif, uploaded here]
 
Last edited:
Lucibee said:
Not sure whether this will upload, but I did a video of that a while ago... I have done 6mWT vs PF in GET arm separately, but I can't work out how to convert it into an uploadable format...
Click to expand...

Did you come to any conclusion as a result of this? I assume the initial scatter graphs are for the start of the trial with SF-36 PF on one axis and 6MWT distance on the other, then it moves to the end of 12 month figures.

From the initial graphs it looks like there's no correlation between PF and 6MWT, and at the end, a slight positive correlation that is no different between groups. Not sure what to conclude from that.
 
TiredSam said:
To try to imply that the tribunal decision was "one occasion" out of line with all the rest, when it was an appeal decision which examined and corrected the wrong decisions of a lower body, well, to try and get away with that when writing to an MP ...
Click to expand...

MS said:
Most of these refusals were upheld, but on one occasion, a FOI tribunal ordered the release of some of the trial data.
Click to expand...

The First-tier Tribunal upheld the Information Commissioner's decision, so it was QMUL's appeal (that was dismissed.) But the inference is that the Tribunal's decision was anomalous; given it was upholding the Decision Notice, it clearly wasn't.
 
You must log in or register to reply here.
Back
Top Bottom