BMJ editorial: Updated NICE Guideline on chronic fatigue syndrome, 2020, Stokes and Wade

The editorial was linked to on the BMJ Facebook page in case anyone would like to comment:

 

The published version of your response @Michiel Tack reads very well, well done for getting this clear response out quickly.

Looking at the range of responses published so far all pointing out the major errors in the Turner-Stokes & Wade editorial one only hopes that everyone reading the editorial also reads the responses.

 

Agree. Will ask the moderators to merge this thread with the forum kitchen.

 

The Science Bit by professor Brian Hughes: Expert reaction to the BMJ editorial calling for the abandonment of standards

An invited, non-peer-reviewed guest editorial in the BMJ has claimed that behavioural interventions for “complex conditions” (such as ME or CFS) should not be judged using the customary criteria — and that the relevant studies should not be evaluated as though they were proper randomised controlled trials — because, among other things, “double blinding is impossible.”

I don’t think this makes sense. Not least because one of the editorial’s co-authors had previously argued that standard RCTs were, in fact, essential for this purpose.

 

And in a bizarre twist, the latest troll friend of the PACE gang was lauding the editorial and Stokes-Turner and Wade's vast experience at RCTs as the reason why they should be considered... I don't know, superior, I guess was his point?... and listened to no matter what...

In response to an editorial basically saying RCTs are overrated. Impressive cognitive dissonance, truly one for the ages.

 

Funny and spot-on article by Brian Hughes, as usual.

 

I think @Brian Hughes blog maybe should have also made the point, as made by @Jonathan Edwards at times, that where blinding truly is impossible, then reliability is still safeguarded provided robust objective primary outcomes are employed. It is the combination of lack of blinding with subjective outcomes that is the Achilles heel.

 

I especially love the selected past quotes, incredible hypocrisy. But again reminding me of the CODES trial insisting for years that number of seizures and only number of seizures is reliable, then doing a full 180 when it fails, basically pretending they never said that, even committing to praising secondary analysis, which is explicitly forbidden.

If only saying the exact opposite depending on what people personally benefits them were a problem. It generally is, it's just that it's exempted here because medicine hates us far more than they are committed to basic ethics, all violations are acceptable because after all we're not patients and therefore have no rights, those are for patients only.

 

I do love this bit ...

 

Trial by Error by David Tuller: BMJ Pushes Back Against New NICE Draft

The BMJ has published an online “editorial” slamming the new draft of clinical guidelines for ME/CFS from the National Institute for Health and Care Excellence. The position expressed is an interesting one: Non-pharmacological treatments for “complex conditions” cannot be adequately measured by randomized trials, according to the two authors. It is, of course, noteworthy that The BMJ is mounting this argument only after an authoritative review commissioned by NICE found the evidence for CBT and GET was mostly of “very low” quality– although some was just of “low” quality.

 

Trial By Error: A Letter to BMJ From Prof Hughes, Prof Racaniello and Me

https://www.virology.ws/2020/12/21/...-bmj-from-prof-hughes-prof-racaniello-and-me/

 

Re-reading the Turner-Stokes and Wade Editorial I was struck that they gave two reasons for rejecting the reports of harm arising from GET as it is “based on qualitative evidence provided by a small number of service users”.

I would regard the ‘small number of service users’ as a downright misrepresentation. Hopefully later on today I will be up to counting the number of respondents involved in the various surveys.

But what is most bizarre is their rejection of this evidence as qualitative when the central tenet of their account is that their preferred behavioural intervention can only be adequately evaluated by qualitative means because of its complex nature and its real life context, and in this situation this is perfectly acceptable, despite the fact that the only such evidence is that some people do get better from ME/CFS and from Long Covid, despite the fact that there is no evidence at all for such recovery being reliably associated with their preferred intervention.

So when qualitative evidence is counter to their preferred intervention(s) it is unreliable but where there is a logically possibility that it might support their view it is of high quality.

 

Yes, I thought that was rich - so I described their claims in my rapid response as coming from a 'small number of service providers'! Very much smaller.

They contradict themselves right the way through.

 

Also published on Brian Hughes' blog The Science Bit - Letter to the BMJ

 

The thing that really gets me is that NICE has been using the GRADE system since 2009. If as they argue it unfairly downgrades the evidence base of rehabilitative treatments then why did they not raise the issue earlier before over a decades worth of NICE reviews were produced using 'poor' methods?

 

Turns out there may be consequences to people constantly lowering the bar to make it easier to fabricate evidence for their beliefs, such as the fact that the bar is the same for everyone and people will naturally follow the path of least resistance.

It's as if centuries of hard scientific lessons about the need to eliminate all sources of bias never happened. That lesson is too far back, it's too abstract by now, some people just want to make it easier for themselves, the ends justify the means.

I'm actually surprised at how little the alternative medicine industry is capitalizing on this. The bar has been set so low they could pass anything, zero difference between Reiki for diabetes or CBT for ME. Maybe the LP will provide the model for this.

https://twitter.com/user/status/1341205977127923715

 

This makes me curious about the recent rewrite of the Cochrane guidelines, in which they marked some very low quality ones as being of high quality. Doesn't Cochrane use GRADE? How did they rate the same studies so differently using the same tools? Given that some of those studies could not possibly be more biased yet were rated as having low risk of bias, it probably came down to who was doing the grading, then who signed off on an obviously fantastic interpretation of the rules.

With Cochrane and with the 2007 NICE guidelines, no one in the room actually enforced a proper grading system (not sure which they used in the 2007 NICE review but the same problem seems to have happened, rules only mean something if people respect them). Makes sense, as usual this entire thing is simply exempt from rule enforcement.

 

Simple, because GRADE allows assessors to downgrade at their own whim. The system is garbage. The whole idea of starting from the premise that randomised trials are brilliant and then you can downgrade them is completely bogus. You don't start anywhere and change. You assess the study. If it is no good that is all you need to say.