Discriminating Myalgic Encephalomyelitis/Chronic Fatigue Syndrome and comorbid conditions using metabolomics in UK Biobank, 2024, Huang et al

[MelbME]

I still don’t understand how this tests is supposed to add value. I’ll repeat my question from above:

Can you please give a high level outline of the diagnostic process with and without the test and explain which activities that are different?

Patients currently go to a GP and first present with fatigue or tiredness ongoing. They might details recent events like infection or so that could explain it.

Now I guess long COVID comes in to play but fatigue can be explained by many conditions. So they likely go through this fatigue pathway to rule out explanations for the fatigue. Early on they will run standard blood tests. The consistency of this is helpful.

After they get blood results back the doctor can input into our algorithm (or algorithm is input to an existing interface from the path lab) and they are given report with % chance breakdown of all the possible explanations for fatigue. At this point it may say ME/CFS 60% probable and the top hit, we'd be using this as a point to provide education to the GP on the diagnosis process for ME/CFS.

 

[Utsikt]

Patients currently go to a GP and first present with fatigue or tiredness ongoing. They might details recent events like infection or so that could explain it.

Now I guess long COVID comes in to play but fatigue can be explained by many conditions. So they likely go through this fatigue pathway to rule out explanations for the fatigue. Early on they will run standard blood tests. The consistency of this is helpful.

After they get blood results back the doctor can input into our algorithm (or algorithm is input to an existing interface from the path lab) and they are given report with % chance breakdown of all the possible explanations for fatigue. At this point it may say ME/CFS 60% probable and the top hit, we'd be using this as a point to provide education to the GP on the diagnosis process for ME/CFS.

Thank you, that’s much more useful information for understanding what you’re trying to achieve.

I have some follow up questions:

How would you be able to predict the likelihood of all the possible explanations for fatigue?

How would this algorithm save time for the doctors when they have to check the results manually anyways?

How would it deal with actual comorbidity where you might have asthma, diabetes and ME/CFS?

How would you provide education on how to diagnose ME/CFS?

And if you’re already planning on providing education, what’s the value added by the algorithm? It seems to me like just doing the education part would achieve the same outcome.