Steven Ilko πŸ₯ πŸšπŸ’¨

@zentensivistilko.bsky.social

EM/ACCM/E-AEC Zen-tensivist. Humanist. Philosophizer/Belly-Button Starer. Med interests: MCS, HEMS-CC, de-prescribing, and communication. My opinions are my own, not of my employer(s).

Proud to share the published results of an international collaboration assessing the outcomes of various ED-ICU care models. (I read 2000 abstracts so you didn't have to) πŸ“‰Lower ICU admission rate πŸ“‰Shorter hospital LOS (0.5d) πŸ“ˆLonger ICU LOS (0.5d) πŸ™…No difference 28d mortality rdcu.be/fqNAI #EMIMCC

Critical care delivery models in emergency departments: a systematic review of the literature and meta-analysis of related outcome effects

rdcu.be

If I have a spill in my kitchen. And you provide me with 5 different ways to categorize and stratify my spill. But then tell me I'm to use my own judgement on how I clean it. You've done nothing. You've created a distinction without a difference. Waste of time, energy and money.

JAMA@jama.com Β· 3mo ago

Med News: A new framework for defining and diagnosing #obesity has sparked debate among experts. Find out more about how it could affect patient care:

For anyone paying attention... Of course this is the solution. The tech is sold to admin as a way to allow me to see more patients in less time. Not for my wellness, nor the benefit of the patient. But for higher productivity. Their $$ benefit. My liability. Checks out.

brookswalsh@brookswalsh.bsky.social Β· 3mo ago

Problem: Medical AI scribe hallucinates Solution: Attestation turns these into *physician* errors. "The auditor recommended that the province implement IT controls in the systems to enforce an attestation from doctors, to confirm they reviewed the notes." www.cbc.ca/news/canada/...

More poorly designed "AI Trials" with even worse reporting. The frequency with which ppl are using AI to alter care/decision making is rising. AI queries can be a helpful part of the conversation to generate questions... But I'm seeing cases of people using AI as the sole mode of decision making.

brookswalsh@brookswalsh.bsky.social Β· 3mo ago

As Dr @kmpanthagani.bsky.social points out, the AI did not obtain the history. It used the history obtained by the ER doc. And getting that history is the *highest*-level cognitive task ER doctors perform. youcanknowthings.substack.com/p/did-ai-rea...