ME/CFS Treatment Approaches Overview

[Trish]

Thank you for listening and taking down the ME/CFS section of your product.

I think if you intended this to be a serious money making business, you need to think again. We have shown you that the ME/CFS section is so flawed as to cause harm, but do you understand that the same argunents will apply to every other disease?

I am puzzled who you expect to be your target market. It's clearly not a suitable resource for patients, and can cause harm. And I can't imagine any good doctor, researcher or drug company relying on such a resource that has just used AI to trawl the internet, when they can do that themselves, and use more reliable research publication listings.

 

[Utsikt]

Thank you all for the feedback. To clarify my intent: I chose ME/CFS specifically because I thought patients dealing with severe fatigue would benefit most from not having to search hundreds of sources manually. I understand now that when the evidence base itself is empty, consolidating sources doesn't reduce the burden.

I’m puzzled by how you use the word «evidence». The evidence base for ME/CFS isn’t empty, it’s just that it only contains evidence against the efficacy of a given treatment, or evidence for the treatment that has too much uncertainty from e.g. a high risk of bias which makes it unsuitable for clinical decision making.

The ME/CFS report is permanently removed, and I've taken the site offline temporarily to rework both the evidence weighting and the target audience framing. We'll see if I can address all the issues raised here - there are more than a few.

I don’t see how you can solve the issue of judgement. Even a descriptive label like «trial» or «case series» might be wrong because we frequently observe that the authors outright lie when they label their own work. The only way to get around it would be to only list the sources, but where’s the value in that?

If someone are doing a systematic review or meta analysis it might be useful to have an additional search option called «asking ChatGPT to retrieve all publications on topic X», but they are not going to pay someone to do that for them.

For everyone else, the only thing you’ll achieve is to pull questionable research to the front and increase the likelihood that someone will try it.

And this isn’t limited to ME/CFS even though the field is particularly bad. 94 % of the interventions Cochrane assessed are not supported by high quality research, and that’s likely an overestimation of the quality because the Cochrane quality assessment tools are lenient. Harm was only measured for 37 % of the interventions and 8 % of those found harm. This is very likely a substantial underestimation of the harm done by interventions someone have thought might be useful.