Michael Sharpe skewered by @JohntheJack on Twitter

[Lucibee]

retweeted by Simon Wessely:

And by Sharpe.

From The Scandal of Poor Medical Research:
"much poor research arises because researchers feel compelled for career reasons to carry out research that they are ill equipped to perform, and nobody stops them."

I really don't think they realise that they are doing anything wrong. But there is no system or person to stop them.

 

[alex3619]

Yes, using Sharpes logic government must always be right because they got the most votes in the last election. No need for scrutiny on policies practice and economical management then, the leader can just stand up and say, "look how many people are on our side and all of my cabinet agree with me".

Its actually a variation on appeal to authority, its appeal to expert authority, where an expert is someone who has been granted the authority of an expert by peers, but may or may not deserve it. In any case even experts get it wrong, just less often than average.

 

[alex3619]

'Well a subset of scientists, mostly not experts in trials, at least some of who have pretty strong feelings about the findings. How many trial experts do you think find it a good trial - not perfect but good? '
I would argue ... zero. Anyone who finds it a good trial is providing evidence they are not experts in trial methodology.

 

[dave30th]

@alex3619 Where did you come up with the number 37 in terms of the issues in PACE? Is that an actual list? Would love to see it. I've never tried to count because so many of them are interlinked or embedded in each other. I think the use of the statistical method for normally distributed populations on skewed populations is definitely good evidence of deception, given the 2007 paper and the caveat there that is absent from PACE. But I think other things are also pretty convincing evidence even if not as "proven." For example, not revealing that 13% met an outcome threshold at baseline. That is salient information that cannot have been hidden by accident--no one could reasonably argue that this is not relevant information that had to be disclosed. We also know, for example, that they pre-approved the language of "strict criterion for recovery" in the Lancet commentary written by their Dutch colleagues. And the failure to disclose their conflicts of interest in consenting participants is in direct violation of their protocol promise to adhere to the Declaration of Helsinki.

 

[Pechius]

I really don't think they realise that they are doing anything wrong. But there is no system or person to stop them.

They may be beginning to realise it slowly, but just refuse to admit it to themselves due to various reasons. I think they will never ever change their opinion on it, because that's what they believed all their life. It's almost impossible.


His thinking is probably something along the lines of 'Everyone is doing the same and we are being selectively targetted simply because patients don't like our findings. There are other studies..' and so on.
But the reality is that the whole psychotherapy field is in trouble (Dr. Johnathan Shedler talks about problems with CBT here: https://www.madinamerica.com/2018/06/158155/)


This is a good question from Sharpe. What is placebo??
I'm assuming that in their mind, it's ok to use placebo(forcing patients to think they're ok, believing that exercise will cure them), because hey, patients aren't really sick, right? What's the harm?
The way I understand it is that placebo effect (e.g. as measured in drug trial) has two components. First - actual physiological changes benefitting the patient due to reduction in stress which may reduce underlying inflammation tnat is contributing to symptoms, etc. And second - a reporting bias, wanting to please the investigators and simply fooling yourself. So which would you think is more likely in PACE trial? Until there is objective evidence that it improves funtion, I think it's safe to assume that it is mostly second.

 

[Indigophoton]

 

[Jonathan Edwards]

'What a bizarre and offensive comment for a (long retired) medical academic to make about any colleague.'

When a medical academic persistently makes offensive comments about their patients and refuses to accept that their work has fallen seriously below basic standards of methodology I cannot see anything bizarre or inappropriate about calling them out. Are we intended to infer that the old boy Oxford network should apply and that academics should not blow the whistle on 'colleagues' when needed?

I am not sure that eight years counts as all that long, especially if the last five have been devoted to studying and advising on research in ME/CFS.

 

[large donner]

This is a good question from Sharpe. What is placebo??
I'm assuming that in their mind, it's ok to use placebo(forcing patients to think they're ok, believing that exercise will cure them), because hey, patients aren't really sick, right? What's the harm?

No its not that, its the old, "what does it matter if its a placebo, it worked". That's the kind of nonsense you get from a wide eyed member of the public who doesn't understand what placebo means and blindly accepts any basic premise given to them unquestioningly. Anytime soon the establishment is going to be pushing placebo trials to see if placebo can be a treat all, its now become so sureal they might even have another placebo control in the placebo trial to double the cure chances.

Sharpe wants the truth to be that he cures people from decondtioning by giving them CBT which slowly cures faulty illness beliefs, except in the Lightnening Process whereby it does it at lightning speed if you have 800 quid.

Although he sees the flaws in LP but pretends they are not the same ones as in CBT and GET except LP is like a CBT GET proponent on acid.

As his truth doesn't actually have an scientific validity he just goes around claiming people are militants, vexatious, anti psychiatry etc.

This allows him to live in his pretend world where he was right all along and the treatments work.

He is a mentally ill man in denial, who knows his denial may even have a biological component, infact I'm pretty sure his denial is a very physical condition but we should not stigmatise him for being mentally ill he needs psychiatric help.

 

[Indigophoton]

Sharpe's logic skewered again by @JohnTheJack:

 

[Jonathan Edwards]

'I agree that would likely make people worse.'

An interesting statement here - of interest to NICE. Since we have no way of guaranteeing that psychotherapists are properly trained (a concern expressed to me by Simon Wessely) there is clearly no way of being sure that people will not be made worse. In clinical pharmacology we do things like dose response curves to make sure we know the therapeutic ratio. Sir Simon himself has pointed out that nothing of the sort applies to 'CBT' in the NHS.

If there is any confusion here I think it comes directly from papers by Wessely and Chalder - where confusion abounds.

 

[alex3619]

Where did you come up with the number 37 in terms of the issues in PACE?

Hi, I think you can safely treat this as a factoid. I am preparing a long reply. It would help if I could recall where I got this number from, but in the meantime I am compiling a list of issues. It will take a while. By my count it is a lot less, but I will be working on that. I might put a blog on it here or on PR.

I do think there are irregularities, many of them, and they are worth pointing out. The issue I refer to goes to intention to deceive - that one point we can demonstrate to a high standard of evidence is a starting point. Once that is firmly established it puts the other irregularities into better perspective.

My current take is this kind of methodology gets results from accumulation of small biases. It might lead to questions about many other studies as well. The issue that so many biases failed to get good long-term results actually raises the question as to whether patients might have been worse off but they failed to record that. We just do not know.

 

[alex3619]

And second - a reporting bias, wanting to please the investigators and simply fooling yourself. So which would you think is more likely in PACE trial? Until there is objective evidence that it improves funtion, I think it's safe to assume that it is mostly second.

Its important to note that the different arms might have had different placebo doses. Its hard to quantify placebo and so we cannot be sure they received the same dose, especially when its in an unblinded trial and there will be subtle personal interactions.

 

[Hoopoe]

@alex3619 I'm glad you're posting more lately, I missed your posts.

 

[alex3619]

@alex3619 I'm glad you're posting more lately, I missed your posts.

Five year decline after two months in hospital. Physically recovering a bit but my brain has not gotten the memo yet.

 

[Indigophoton]

 

[Sean]

But there is no system or person to stop them.

The downside of academic freedom.

Sharpe wants the truth to be that he cures people from decondtioning by giving them CBT which slowly cures faulty illness beliefs, except in the Lightnening Process whereby it does it at lightning speed if you have 800 quid.

That must sting.

Sharpe's logic skewered again by @JohnTheJack:
JohnTheJack says:

As for CBT, since the illness is perpetuated by the patient's misinterpretation of bodily sensations, how can they dictate the pace of increase?

Exactly.

I also want to know at what point in the therapeutic process does the patient's bodily misperceptions miraculously transform into being reliable enough to be the primary outcome measure?

It is beyond farce.

 

[Woolie]

The way I understand it is that placebo effect (e.g. as measured in drug trial) has two components. First - actual physiological changes benefitting the patient due to reduction in stress which may reduce underlying inflammation tnat is contributing to symptoms, etc. And second - a reporting bias, wanting to please the investigators and simply fooling yourself. So which would you think is more likely in PACE trial? Until there is objective evidence that it improves funtion, I think it's safe to assume that it is mostly second.

There isn't any direct evidence for the first - that some of the benefits of placebo are "real" in the sense that it reduces stress and aids healing. In a metanalysis comparing studies that differed only in whether patients were blinded or unblinded, this study found no reliable benefits of unblinding on objective measures. The "benefit" only occurred for self-report measures, which suggests they are artefacts that affect people's attributions about their symptoms*, rather than real, genuine effects.

(*what I mean by this is that if you think you've had a treatment that might help, you may be actively looking for evidence to confirm that, so you'll focus more on the times you feel better than the times you felt worse. And when asked to recall how you've felt in the last few weeks, you might be more likely to recall times you felt better. So its not real improvement, just a temporary change in how you see things - which obviously won't last).

 

[alex3619]

My recollection of the placebo effect is that its mostly about changing perception about pain. A placebo in a clinical trial (typically double blinded) is chosen because it should not have a physical effect, and is used as a comparison to help eliminate perception changes in subject or researcher. Its a method to decrease human bias in a clinical trial.

 

[arewenearlythereyet]

Perhaps we should ask MS to try and explain his version of what he thinks a placebo is?

I’m sure he thinks this is evidence of mind over matter/power of positive thinking or some such garbage rather than temporary artifice of the experiment. It would be good to tease out his view on this.

I think it’s also possible he may be confused about the difference between ‘control’ (as in being careful about what you do ...a subjective view) and ‘a control’ or the true meaning of controlled conditions reading through some of his comments? I often see controlled conditions being used incorrectly just to mean “we are being careful and considered’ as oppose to having a control within the design.

 

[Alvin]

Perhaps we should ask MS to try and explain his version of what he thinks a placebo is?

I’m sure he thinks this is evidence of mind over matter/power of positive thinking or some such garbage rather than temporary artifice of the experiment. It would be good to tease out his view on this.

I wouldn't bother, i would continue showcasing their malfeasance instead of wasting our energy trying to convince or drag the truth out of the perpetrators who have every reason to never give it

I think it’s also possible he may be confused about the difference between ‘control’ (as in being careful about what you do ...a subjective view) and ‘a control’ or the true meaning of controlled conditions reading through some of his comments? I often see controlled conditions being used incorrectly just to mean “we are being careful and considered’ as oppose to having a control within the design.

You can't educate him out of reality denial, choosing to deny facts is not an educational deficit.