I feel like this photo of masked, armed men pepper spraying a pastor protecting his community is going to be a defining picture of this moment in America for a long, long time.

Should we (specifically in pediatrics) be using linezolid more often for complicated community-acquired pneumonia? 🤔 #pedsID #IDsky #Medsky

Very cool #dataVisualization of train travel in Europe. I don't know if this is a Shiny App, but sure looks a lot like one.

As we continue to enhance interoperability, bolstering security will be vital. Updating standards related to HIPAA would be a good step.

This will be a huge challenge ➡️ “We need an ecosystem where health systems themselves take on the responsibility for ongoing validation.” FDA Comish Califf

Bias in medical AI can compound throughout the development pipeline. Assessment and mitigation of bias needs to be a part of healthcare AI strategies/implementations #AIinMedicine

“Why is it so complicated to be an academic physician?” Seriously, between MOC, credentialing, licenses, COI, academic work tracking, grant paperwork, mandatory education, etc, it feels like I spend half my time tracking/updating things. 😡🤬 #medsky #IDsky

Non-inferior! Will have to dig into the subanalyses and nuances ➡️ "Antibiotic Treatment for 7 versus 14 Days in Patients with Bloodstream Infections" #IDsky #USAAW

#AntibioticAwarenessWeek "Fighting AMR takes all us" Everyone has a role in fighting AMR. Working together, we must improve abx prescribing to benefit human health, animal health, plant health and the environment.

Rise of the 'CAIO'? Highlights the ever growing need for more informatics trained physicians! #clinicalInformatics

Here's a good NY Times article on the LLM study for aiding diagnostic reasoning I linked to today #AIinMedicine

This study on LLMs aiding diagnostic reasoning buries the lede (https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2825395 ). The LLM alone outperformed the humans. Make sure to read the accompanying commentary (https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2825399 )

Implications for EHRs based on ancient code 👀 "...AI has significantly increased the real cost of carrying tech debt...generative AI dramatically widens the gap in velocity between ‘low-debt’ coding and ‘high-debt’ coding."

This is one of the wildest applications of LLMs I've seen. Incredible initial effect and then apparently durable over time (!) but wide variation in response #artificialIntelligence

I'm skeptical but deeply interested... #AIinMedicine #artificialIntelligence https://buff.ly/3CtGTuh

Fierce Healthcare and Sermo surveyed physicians, revealing 76% use general-purpose LLMs for clinical decisions. HUGE caveats: (1) respondents were LLM users, (2) 60% were in academia, (3) Sermo users tend to be tech-savvy. Interesting nonetheless #artificialIntelligence