OpenNotes

🧠 Can large language models (LLMs) like ChatGPT help people understand their open notes?

A new study shows they might. Expert patients Dana Lewis and Liz Salmi worked together on this research to find out how well AI can answer real questions from real patients. The questions were based on a medical note from a visit with a brain cancer specialist. (The note belonged to Liz, and she gave permission to use it.)

Both Liz (the patient) and her doctor looked at the answers from different AI tools—but they didn’t know which tool gave which answer. They found that when the AI was told to “act like a doctor,” it gave better answers.

Why does this matter? Many patients read their notes at home and have questions. LLMs could help explain medical language and support people between visits with their care team.

The main takeaway: Generative AI has potential to help patients—but how we ask the AI questions makes a big difference. It’s important to teach both patients and clinicians how to use it safely and clearly.

📖 Read the full open access article here: https://lnkd.in/gShAW6iP

LinkedIn This link will take you to a page that’s not on LinkedIn

Did you know that the way clinicians write their notes has a direct impact on the care the patient receives?

Notes have historically been written with the goal of communicating to other clinicians. You've probably seen stigmatizing language in your own medical notes like "non-compliant."

Dr. Anita Vanka and our team at are teaching students at Beth Israel Deaconess Medical Center to write with the patient in mind.

And the stakes are high according to a 2018 study in PubMed.

"What they found was that the group of trainees who read the note with the stigmatizing language was less likely to treat the patient’s pain and had a more negative perception of the patient.

So it not only affects our relationship with the patients or how we may view them but can impact care as well," said Vanka.

Have you ever encountered stigmatizing words in your medical notes? What would you want the next generation of medical students to know about communicating with patients?

Teaching Doctors to Write for Patients All Articles Teaching Doctors to Write for Patients March 2025 Teaching Doctors to Write for Patients How patient access to medical notes is changing clinical documentation Winter 2025 by Amos Esty 7 min Interview Over just five years, from 2017 to 2022, the percentage of Americans who accessed thei...

“Patients and clinicians agree on the importance of SDOH data, but significant concerns remain about how this information is used and documented,” said Cait DesRoches, DrPH MSc, Executive Director of .�

Many people worry about how sharing personal information—such as their income, education level, or living conditions—will be used in healthcare settings.��

Read the rest of our pivotal study in JAMA Network Open in collaboration with the Assistant Secretary for Technology Policy and Office of the National Coordinator for Health Information Technology (ASTP/ONC): https://lnkd.in/eNiP3Hw3

Meet Chethan!

Chethan Sarabu, MD, FAAP, FAMIA, is Director of Clinical Innovation for the Health Tech Hub at Cornell Tech, where he focuses on broadly connecting the dots across academic health innovation programs and the broader health innovation industry.

In addition, he is a Clinical Assistant Professor of Pediatrics at Stanford Medicine, where his work centers around the practical application of open notes in pediatric settings, with a particular emphasis on health equity and patient privacy considerations.

Chethan has been involved in the utilization of Language Models (LLMs) and AI in healthcare, bringing a nuanced perspective to their integration. He consults for OpenNotes as an AI and Informatics Strategist with the Lab.

Read more: https://www.opennotes.org/family/chethan-sarabu-md-famia/