🔥 block box !!

(Aka glass box of emotion)

@Jeffgadsden.bsky.social has a lot of editing to do!

@Bilih.bsky.social
@Dr-Amit-pawa.bsky.social

@Nottheflyhalf.bsky.social
@Emarianomd.bsky.social
@Petermerjavy.bsky.social
@Maggieholtzmd.bsky.social
@lloydturbitt.bsky.social
@ropivacaine.bsky.social

Nothing Benedict…to be honest, it’s a way to get the post seen via its tag…..🤷‍♂️

It’ll be one of our last weapons of resistance as EPRs and robotics turn on us.
The steth and tendon hammers… and this dude.

What’s wrong with it….apart from the fact it looks as useful as they actually are these days😂🤪😉

Ha ha Tim. It still can’t perform too well with bad images…à la ICU reality mate!

Also, did you know, if you ask ChatGPT what it will do if we try to shut it down…? It will write its own code so making it impossible! Skynet here we come!

4/4

This is what I mean by residual confounding. Essentially, the crossover and adjustments mitigate, but don’t completely eliminate the impact of hospital-level variations. 😬🤷‍♂️😉

3/4

Hospitals are bound to have unmeasured or time-varying differences; like changes in staff expertise, patient severity, or protocol adherence flexing during the study—these could still influence the results.

2/4

Each hospital serves as its own control by switching between intervention and control periods. So, adjusting for the hospital in the analysis helps reduce confounding from differences in hospital characteristics (e.g., patient mix, protocols), they may not fully account for all variations.

1/4

As far as I see it mate, there may be bias in the study results that remain, even after attempts to control for confounding factors. Stuff like hospital-level differences.

Same Jo….same!

Yes…rather you than me mate…..😂

Hey buddy…I can’t find where I wrote that?? 🥴

Sorry about the incorrect study title here….open to look