In this study led by Yusuke Tsugawa and Atsushi Miyawaki with Anupam Jena, we found better outcomes for hospitalized patients treated by female physicians, with greater benefit for female patients. Annals of Int Med

Higher-performing LLMs can better self-assess confidence relative to accuracy. Using the Dunning Kruger effect to improve reliability of LLMs. NEJM AI ai.nejm.org/doi/full/10.10…

Outstanding documentation goes hand-in-hand with exceptional clinical care NYU Langone Health.

What makes a great note? And how can we assess note quality across our system?

Check out our 5Cs in NEJM Catalyst:
✅ Complete
✅ Clinical Reasoning
✅ Contingency
✅ Concise
✅ Correct

A warm welcome to an amazing new team of Associate Editors at NEJM AI including Enrico Coiera Roxana Daneshjou MD/PhD Sam Finlayson Venk Murthy MD PhD Roy Perlis Adam Rodman Effy Vayena Gabriel Brat

ai.nejm.org/about/editors-…

New paper from Verity Schaye, MD, MHPE and NYU Grossman School of Medicine on AI in #MedEd : 'The generative artificial intelligence revolution: How hospitalists can lead the transformation of medical education' In Society of Hospital Medicine . View here: …mpublications.onlinelibrary.wiley.com/doi/10.1002/jh…

Was a bit concerned when 1st heard American Board of Surgery was nixing ABSITE %ile reporting.
Have since learned its a solid evidence-based move w preserves %correct & standard score w are⬆️ accurate for formative & predictive uses while removing the most misused number: the dreaded percentile.

Our new study in JAMA Internal Medicine looking at the reasoning abilities of GPT-4 compared with human physicians just came out.

Big picture: AI displays (much) better reasoning than humans, makes diagnoses similarly, but hallucinates considerably more.

A 🧵to put in context ⬇️

There is considerable confusion in terms used to describe #troponin , including terms that are imprecise (e.g., “troponin leak”) or deprecated (e.g., “type 2 NSTEMI”) (Lang, 2022). What's the difference between 🫀myocardial injury, ischemia, & infarction❓ 🧐 #MedEd 🧵

We know LLMs can ace multiple-choice exams. Taking us deeper, an important new study led by Steph Cabral and Adam Rodman conducts a nuanced evaluation of the clinical reasoning abilities of GPT-4 wrt physicians. Guess who wins?

Need more of these!
jamanetwork.com/journals/jamai…

New in JAMA Network Open 🚨 we show that initiation of meds for alcohol use disorder on hospital discharge is associated with ⬇️ 30-day ED visits and readmissions! 1/7 jamanetwork.com/journals/jaman…

Now on the #ACGME Blog! Read a summary of the #ACGME 2024 Featured Plenary 'The Generative AI Revolution: Innovations and Opportunities in Medical Education,' with panelists Bob Wachter, Eric Holmboe, Jesse Burk-Rafel, and Eva Aagaard. #MeaningInMedicine #MedEd #AI acgme.org/newsroom/blog/…

The Family Health Centers at NYU Langone Health created a novel clinical decision support tool that alerts primary health care providers to patients' gaps in care and

4/ Two helpful references from this session were

🔥 Academic Medicine last page on teaching when time is limited on rounds
🔥 behaviors of effective general wards teachers by Houchens Sanjay Saint et al

Simulation can inform interventions for MDs to address racial #implicitbias in clinical encounters. #MedEd Congrats to Cristina Gonzalez & colleagues on JAMA Network Open paper: jamanetwork.com/journals/jaman…
Sandy Zabar NYU Grossman School of Medicine

Congratulations to Robin Ortiz, MD, MS, FAAP for being named one of NMQF 40 Under 40 Leaders in Minority Health. Annually 40 health leaders from minoritized populations under the age of 40 are selected for making a positive impact. Learn more: bit.ly/43uFdtF #womenshistorymonth

Grateful to introduce our new paper “A Toolbox for Surfacing Health Equity Harms and Biases in Large Language Models”. Adversarial datasets and human assessment rubrics to evaluate LLMs for potential biases and equity-related harms: arxiv.org/abs/2403.12025 🧵:

Why are some non-antibiotic drugs toxic to bacteria 💊☠️🦠? Do they work like standard #antibiotics (ABX)? Our latest work by Mariana Noto published Science Magazine addresses this fundamental question (1/n)
science.org/doi/10.1126/sc…

Peer observation of teaching is an effective & efficient model for #FacultyDevelopment . James Nixon recommends grounding peer observation in 3️⃣ learning theories/frameworks:

·🤝Social learning theory
·🌱Growth mindset
·🎯Deliberate practice

🔗:bit.ly/3P59ZUi

Kudos to team NYU Hospital Medicine helping us understand evolving capabilities and limitations of LLMs in #HospitalMedicine . NYU Department of Medicine

“What is uniquely human in your field?” asks Jesse Burk-Rafel about the future of AI in medicine. He suggests that which is will persist and that which is not will likely drive changes. #ACGME2024 #MeaningInMedicine