Search results

And very little has happened since, Decode happened due to a lot of lobbying and was funded through the strategy board and working with the MRC to get something they were happy with. It wasn't just someone thought it was a good idea it was a result of massive amounts of work. There have been...

I believe lots of LC clinics are closing (funding gone) or merging with ME clinics (that don't really offer anything).

I agree - the paper didn't look like a good paper to me. It should also talk about ways to avoid and test for bias.

It is worth making the point that this is something that researchers and charities are pushing for, at the moment I can't see it happening without lots of pressure. I was going to comment that we have seen money ringfenced before and it hasn't lead to more research - hence I would argue that...

Given the history of ME research I think that is very naïve.

By focusing on the important questions - set out by the JLA exercise. Wise spending of research money will always be an issue but there can be a huge cost to not spending, not building research capacity and ending up with no research and hence no treatments. We've seen too many years of no...

But similar results to recruitment through clinics.

The numbers are uncertain and my guess would be higher. The 50% does come from what few studies there are. The 400k for ME comes from Chris Ponting's latest paper (I think published today). But for years we have talked about a vague 250k in the UK. The important thing is numbers give an idea of...

Lots of research money is wasted. But the notion of having some joined up strategy to build capacity and research platforms should tend to create useful artifacts and people with expertise to continue. No one is suggesting that peer review should not happen. But it needs to be acknowledged...

Thread split from [UK] A proposal for an ME/CFS, Long Covid, and Post-Infectious Disease research platform about a proposal that mentioned that around 50% of people with Long Covid have ME/CFS. There have been a number of studies Jason did one Estimating ME/CFS prevalence in individuals with...

There have been a number of studies Jason did one Estimating ME/CFS prevalence in individuals with long COVID. I think there was one study which had smaller numbers. The point about including long covid is there are probably common mechanisms (given ME criteria are met for a significant number...

As you know ME research is at a very very low level in the UK and across the world - there are very few researchers who know anything about it or have any interest. On odd occasions when someone expresses interest they are quickly put off by the difficulties (things like bad review processes)...

I've always read bias in AI as being between the model and the real world. (data is just used to train the model). There is work on fairness testing which if I remember correctly groups data according to meta data and checks that the overall results are the same with different splits of the data...

There need not be privacy issues. If a standard LLM is used then this can be run on a system that is not connected elsewhere and data can be kept secure. If a model needed fine tuning then it would need data but controls could be put on the resulting model to keep it secure.

I've wondered if better stats could be created using an LLM to read and classify medical records of people to look for patterns/symptoms that could be associated with ME and this could be done as scale. It may be necessary to do some fine tuning but that can be relatively cheap.

I don't think they have ME stats the delivery report states 280,000 which is scaled from the assumed value of 250,000 a few years ago taking account of population growth,

Were they paid I thought influencers basically made money by being paid to promote stuff.

If you are experimenting on humans then you need to show value from those experiments (i.e. show that the risk of carrying out the experiment is worth it). But if your methodology is too weak to get any value from the experiment then you shouldn't be doing it as there is only risk and very...

I suspect the whole mental health research industry has issues with measurement (too much subjective stuff and also some really badly done questionnaires that shouldn't be used as a measurement scale). But perhaps that just says we need better research into measurement approaches - and with the...

With the step test they published a bar chart with values but they turned down a FoI request from Graham earlier and @JohnTheJack may have got numbers in an FoI. I remember Graham trying to estimate numbers from their plot.