Michael Sharpe skewered by @JohntheJack on Twitter

For those not on Twitter, or not following:
https://twitter.com/user/status/1008361250953285632

https://twitter.com/user/status/1008366585462054912

https://twitter.com/user/status/1008342750201303041

https://twitter.com/user/status/1008364695311585282

 

@Esther12 What design do you think would have? It always seemed to me that if they had published honest results, it would have been like the reanalysis paper--end of story. CBT/GET don't work except to produce transient subjective reports of improvement. The trial would have served its proper purpose.

 

https://twitter.com/user/status/1008381718703886336

 

That may be true but there's also alot of solid research out there across all other medical fields.

Can anyone point to the DSM and show any solid/falsifiable research for any single "diagnostic" label in it?

 

MS: "I don't want to be rude, but..."

https://twitter.com/user/status/1008381718703886336


I so want MG to say: "Do you know who I am?" [in a booming voice]

 

That they dropped actometers seemed really worrying to me at the time (and TSC minutes now show that these were dropped once they became aware of the null results from earlier trials). There was an outcome that (if used over a reasonable length of time, and after treatment had been completed) had the potential to avoid the problems with bias/distorted behaviour that CBT/GET could induce even if ineffective.

I think that I expected GET to lead to improvements in their fitness outcome, even if ineffective, just because participants would have been more likely to prioritise 'exercise' over other activity. It's pretty amazing that this null result led to so little concern from the PACE authors, and was only released long after their initial results in their mediators paper.

There also seemed to be problems with SMC vs SMC+GET vs SMC+APT vs SMC+CBT when APT looked pretty rubbish (and from participant testimony it has since seemed that SMC was an even more minimal intervention than it sounded). At the time I thought that the use of a sham intervention designed to induce to sort of positive expectations and bias one would expect from CBT/GET would have been useful.

I also expected them to get better results for their primary outcomes, although again the potential for problems with bias would have made it difficult for anyone to know whether that reflected any genuine improvement in participants health or not. [edit: Even with their poor results for these outcomes, Sharpe is still banging on about them as if they show what a success CBT and GET are, and how all the criticisms of PACE are just an irrelevant distraction].

I'd been re-reading some early PACE discussions following Bob's death, and was amused by seeing myself first learning about the problems with PACE, but it's hard for me to remember exactly what I was thinking back then as I've leant so much since.

 

@Esther12 I agree with all those points, especially about the objective measures. I wasn't aware of anything beforehand and never heard of PACE until the day it came out and I had to write a news story about it. I just meant that had they actually reported their original findings, that would have been the end of it, whatever the pluses and minuses of their study design.

 

Oh my gosh, when I said a couple pages back that I wondered if Sharpe would even tell someone of Godwin's stature to 'read the paper', I was just joking around. But now he has actually told Godwin exactly that! And even more amusingly, the link Sharpe gave doesn't work, it leads to a 'This page does not exist' page on The Lancet...

 

Well at least Mike Godwin gets his "read it again" prefixed with an "I don't want to be rude but ...". I suppose that's what passes for respect from Michael Sharpe. Godwin should be honoured.

 

You have too little faith in their ability to delude themselves! When PACE did release a partial analysis of their primary outcomes, after they lost the Information Tribunal, they still tried to present this as supporting their claims about the efficacy of CBT/GET.

If they'd released their prespecified outcomes, and all their objective outcomes, in their initial paper it's pretty difficult to see how they could have sold that as a big success. But then the Lancet's claims about a 'strict criterion for recovery' are pretty shamelessly detached from reality.

 

Presumably the only way to understand is the way MS (and others like him) understand it... everyone else who reads the paper and comes to the conclusion that GET/CBT are not the wonder treatment/cure for ME presumably 'does not understand'.

 

Bad tactics by Sharpe I would say. There is a chance that Godwin has not actually read the paper yet. If he does then things can only get much worse.

 

And remember that the PACE authors read that commentary by their Dutch buddies and vetted it before publication. So they were very well aware of the words that were used to describe their outcomes. They have no possible claim of ignorance on how the commentary was used to get the word "recovery" into the conversation. The press coverage was all about "recovery" and "getting back to normal." They did nothing to "correct" the press coverage based on these two clearly false claims.

 

https://twitter.com/user/status/1008399982494396416

 

Indeed, and because it is a very complex analysis it is therefore much too difficult for mere mortals to understand. Therefore everyone should defer to the authority of the PACE authors on this.. they are the only ones who can fully understand all the nuances of such a large trial.

The fact they can’t seem to comprehend the criticisms of the whole fiasco tells me only that they were so certain of their end result, that any suggestion that their assumption was in error, leads them to redefine their outcomes rather than reassess their assumptions.

 

I say, this is hotting up nicely.

 

Godwin: 'That seems possible. There's a lot of rubbish research out there, produced primarily by people who aren't actively trying to produce rubbish. I'd say it's bigger than psychiatry, quite frankly.'

It's bigger than psychiatry but liaison psychiatry/psychosomatics seems to be very different in its standards from the medicine I have known for 40 years. It is interesting to contrast Dr Sharpe's comments with those of Dr Oystein Fluge, oncologist:

'I hope we have not misled you with our results [of immunological treatment of ME/CFS], Jo.'

Fluge then goes on to show that his theory was wrong, with a properly designed experiment.

 

Brilliant! If Sharpe keeps this up it'll be like an extra high profile advocate on our side