Stephen Ma

CI alums! @JenniferLeeLee1 ACMIO Nationwide; @medtechstef C informaticist Suki; @chethanr Dir Innovation at Cornell Tech; @StephenMaMDPhD Dir analytics Stanford. Many great career options after CI fellowship. #cic25 #whatisclinicalinformatics

🩺 + 🤖 Introducing MedAgentBench: EHR Environment to Benchmark Medical LLM Agents ✅ 300 agent tasks (beyond QA) from 10 physician-written categories ✅ 100 realistic patient profiles with 700,000+ data elements ✅ FHIR-compliant interactive environment for translation

3) A huge thank you to all the clinicians who participated in the pilot and our co-authors Margaret Smith, Jeannie Yejin Jeong, Anna Devon-Sand, Trevor Crowell, Clarissa Delahaie, Caroline Hsia, Steven Lin, Tait Shanafelt, Michael Pfeffer, @tophersharp

2) Despite the overall success, there were significant variations in utilization and time savings across individuals, highlighting the heterogeneity of clinical practice and the need to better support personalization

Sharing work led with @PGarciaMD, Shreya Shah, & @StanfordCIF extraordinaire @AprilLiangMD! Our pilot @StanfordMed showed that AI scribes save time and improve the clinician experience, published in @JAMIA_Journal: doi.org/10.1093/jamia/… doi.org/10.1093/jamia/… Some takeaways:

How a blend of magic & music helped #StanDOM's @jonc101x navigate the toughest year of his life during the pandemic --> stanford.io/4cFj8w9

Very excited to share these promising results that can further our ability to protect adolescent patient privacy (and beyond)! Huge thank you to coauthors @ShivamVedakMD @adussaq @donghanyaoMD @wui_ip @KMorseMD @StanfordKidCMIO, and to @StanfordCIF for bringing us together!

@StanfordCIF fellows @NymishaC closing out @AMIAinformatics #CIC24 with her great work on expanding patient portal access for Spanish-speaking families, and @drjulielee on her work on understanding impact of EHR-integration with the CA drug monitoring program!

@StephenMaMDPhD leading discussions on AI-generated patient message response drafts at @AMIAinformatics #CIC24!

@AprilLiangMD did a great job discussing how to leverage LLMs to enhance adolescent confidentiality in patient portals at @AMIAinformatics #CIC24!

Standing room only at the AI prompt engineering workshop at @AMIAinformatics #cic24 led by @StanfordCIF fellows @ShivamVedakMD and @donghanyaoMD!

🚀Want an image classifier within minutes? Just prompt latest models like GPT-4o and Gemini 1.5 with a bunch of demo examples (you can include thousands of them now) and ask multiple queries in one go! Work with @AndrewYNg at @StanfordAILab and @jonc101x . Inspired by @JeffDean

Our new publication @StephenMaMDPhD @jonc101x Thanks @JHospMedicine! The promises and limitations of AI for quality improvement, patient safety, and research in hospital medicine - …mpublications.onlinelibrary.wiley.com/doi/10.1002/jh… @StanfordHospMed @SocietyHospMed

5) Huge thanks to co-authors Shreya Shah, Margaret Smith, Yejin Jeong, Anna Devon-Sand, @mtaiseale, Kevin Takazawa, Danyelle Clutter, Kyle Vogt, Carlene Lugtu, Matthew Rojo, Steven Lin, Tait Shanafelt, Michael Pfeffer and my fellowship program @StanfordCIF

4) Despite these limitations, clinicians are excited about the use of these tools, and even early iterations are intuitive to use. There is enormous potential for improvement as these tools continue to rapidly evolve

Grateful for the chance to lead this work at @StanfordMed with @PGarciaMD with the support of @tophersharp and our many co-authors. A thread with some of my personal takeaways: