Has there been a review of ME/CFS treatment randomised trials since the NICE 2021 review?

Sasha

Sasha

Senior Member (Voting Rights)
Not sure if I'm posting this question in the right place, but has there been a review of treatment RCTs in ME/CFS since the comprehensive review done for the NICE guidelines?

Is there an easy way to find out what trials have been published since?
 
Sasha said:
Not sure if I'm posting this question in the right place, but has there been a review of treatment RCTs in ME/CFS since the comprehensive review done for the NICE guidelines?
Click to expand...

I am not aware of any competent trials since 2021. What matters more is that they are properly controlled and if necessary blinded. Randomisation is necessary but not enough to make a trial of any use.
 
Jonathan Edwards said:
I am not aware of any competent trials since 2021. What matters more is that they are properly controlled and if necessary blinded. Randomisation is necessary but not enough to make a trial of any use.
Click to expand...
I agree, and on a quick skim I'm a bit surprised to see this review taking pre-2021 studies as showing efficacy when NICE bombed them on quality grounds. It's surprisingly big on Ampligen.

It's a pity that there's no handy up-to-date review that takes quality properly into account.
 
Utsikt said:
Not a review but an overview from last year:
Click to expand...

Sasha said:
I'm a bit surprised to see this review taking pre-2021 studies as showing efficacy when NICE bombed them on quality grounds. It's surprisingly big on Ampligen.
Click to expand...

That reads like a very long student essay, with very little insight into the real problems. Worth remembering that most reviews now are written by students wanting to expand their publication record.
 
Does anyone here routinely do the same sort of PubMEd/whoever search that NICE did, just to check out what's new?

I'm wondering if we even know how many trials have been done since then. That review paper above is over a year old now.
 
Sasha said:
Does anyone here routinely do the same sort of PubMEd/whoever search that NICE did, just to check out what's new?
Click to expand...

Maybe not in the same way but the team here are very impressive in their coverage of relevant papers. I think we may miss one or two things but they usually surface a bit later. I think it very unlikely that we have missed anything important. Despite all the other things he does Andy is still up there almost every day with new papers. So are half a dozen others. The pity is that up until now so much of the stuff has been hardly worth the effort, but I think that will change once we have leads from the basic science.
 
From what I recall the "german version of the nice review" the IQWiG for ME/CFS pretty much went with NICE when it came to assessing RCTs and just took data from NICE probably because there's nothing noteworthy to add on that end. I suspect there might be other countries doing something similar, it's possible that some of those also induced evidence from newer trials, but I doubt it is worth bothering as they largely lack sufficient control. The whole IQWiG report seems to be missing the point a bit and discusses at length whether or not people might have had PEM but seems to miss that the results for the trials look unimpressive independently of whether PEM is present or not.
 
Jonathan Edwards said:
Randomisation is necessary but not enough to make a trial of any use.
Click to expand...
Even then, and I might be wrong in my understanding of what randomized means in the context of clinical trials, but I don't see how a single trial on ME/CFS, or any variation of it, has actually been truly randomized. Randomization has two elements: selection and allocation. They both have to be met, it's not optional. This is like an AND circuit, both bits have to be 1 for the answer to be 1. Otherwise it's 0.

By definition, a randomized trial can't rely on subjective selection criteria, or then it just wouldn't be truly randomized, it would be selected then randomized. So for a cancer trial, once there is enough objective evidence that a person meets the criteria, something that can be blinded entirely, patients will be recruited consecutively until they meet their threshold. This way the selection is random, whereas what typically happens in so-called pragmatic trial is that the selection is made by clinicians with subjective requirements, sometimes even that list is subject to a subjective selection effect, which isn't actually random, especially so in the hands of biased ideologues.

Selecting people at an astrology conference and randomly assigning them to treatment arms in a trial of astrology isn't random, because the people who were selected were selected using a biased sample. They aren't randomized, even if the arm allocation is done using a truly random method, such as a ball draw. They are selected, then randomized. Which isn't randomized, not even pseudo-random.

What I'm not sure is why people pretend this is so. Even trials that clearly aren't controlled are always referred to as randomized, so this is clearly accepted. And yet it doesn't add up to being randomized in a mathematical or information theory sense, only partly, and I can't imagine this actually applies in, let's go for 'real', clinical trials. This has lead to absurd levels of bias such as pretending that the heavy filtering selection done for quackery like the LP still means the participants were randomly selected, when in fact they were explicitly selected for being open to manipulation through a subjective interview process. Which is a lot like a lottery where you know which prizes are low and can simply remove them from the draw. Not random, not a lottery. In fact this would be considered fraudulent by any gaming laws and regulations.

For sure by the real definition of randomized, no trial of ME/CFS has ever been truly randomized. Including the better ones, because we don't have objective criteria for selection, and so a selection effect is unavoidable. But the standard in the clinical trial industry has been that if they can't meet a standard, then they just don't bother, and pretend like it's just as good as meeting that standard anyway. This is where things like GRADE become useful, in whitewashing what is basically a popular lie that everyone buys into.
 
rvallee said:
Even then, and I might be wrong in my understanding of what randomized means in the context of clinical trials, but I don't see how a single trial on ME/CFS, or any variation of it, has actually been truly randomized. Randomization has two elements: selection and allocation. They both have to be met, it's not optional.
Click to expand...

I don't think they do have to be met for something to be considered a randomised trial. It's understood that you're simply randomising the selected population who are taking part in your trial.
 
I don't think anyone is suggesting that randomised trials are randomised in the sense that you are meaning, @rvallee - ensuring that they are a statistically representative sample of a wider population that you might want to recommend the treatment for. That is a very important issue but it is not what is being referred to. It relates purely to the randomisation of arms once patients are recruited.
 
Yeah I figured that much. So: randomized, does not mean randomized. Controlled, does not mean controlled either. As for trial...

Words, and their meaning, so complicated, so convoluted in their application.
 
rvallee said:
So: randomized, does not mean randomized.
Click to expand...

Actually it does mean randomised. What it doesn't mean is ensured to be representative by matching sampling to variation within a target population. You might theoretically be able to achieve that matching by random sampling but in practice you almost certainly never can, so randomisation isn't really a relevant term for it.
 
Jonathan Edwards said:
Actually it does mean randomised. What it doesn't mean is ensured to be representative by matching sampling to variation within a target population. You might theoretically be able to achieve that matching by random sampling but in practice you almost certainly never can, so randomisation isn't really a relevant term for it.
Click to expand...
Yup, as I said, it doesn't mean random. Random has a meaning. This isn't it.
 
You must log in or register to reply here.
Back
Top Bottom