Thread by @WhistlingDixie4

Are biomarkers helpful? #LIVES2022

(That’s the question)

Jokes, it isn’t

“Are biomarkers helpful in characterising inflammation V infection?”

The goal is to prevent both under AND over treatment of infection.

As clinicians we have bias
- Tend towards ‘action’
- Over estimate risk of infection
- Over estimate/misattribute improvement (outcome bias)

A meta-analysis has show that value had highest diagnostic calorie for sepsis in adults

pubmed.ncbi.nlm.nih.gov/33902476/

At this point the man beside me is clicking all of his upper limb joints progressively in a step wise manner and so I am distracted and missed something about PCT and the surviving sepsis guideline

Now we’re on to this which proposed a candidate to rapidly diagnose CAP on ICU: FAIM3:PLAC8
pubmed.ncbi.nlm.nih.gov/26121490/

Speaker then says there are two tests published on genomic expression

pubmed.ncbi.nlm.nih.gov/33297498/

And

pubmed.ncbi.nlm.nih.gov/29624409/

What do we need to manage the uncertainty? The biomarker needs to be able to correct the clinician’s decision

She points to @msh_manu study and that CRP and PCT were not significant

pubmed.ncbi.nlm.nih.gov/30291379/

The cracky knuckle man has now made me get up to let him out

And we’ve moved on to this pubmed.ncbi.nlm.nih.gov/35831640/

And it’s been about fifteen seconds and we’re on to machine learning and bio-markers

pubmed.ncbi.nlm.nih.gov/33786999/

And ~ten seconds later, we are already concluding with a call for new study designs that
-repeat tests at a different time point
-use combinations
-or use machine learning

I’m going to close with a plea to #LIVES2022 to give people the time they need to do their talk…or coach them craft it that way

Such wonderful points to be made by picking them up in a meaningful way is often difficult…

End