Key Takeaways
- Hackensack Meridian School of Medicine students and faculty published a commentary on how medical schools can approach AI clinical summarization tools in physician training.
- The authors say medical students should have a voice in discussions about the ethical use of AI in clinical notetaking.
- They support a phased and supervised approach that lets students first build core clinical notetaking skills before using AI tools.
- Faculty supervision could help students learn how AI models work, recognize common errors and use the tools responsibly before entering medical practice.
- The commentary appears in The American Journal of Bioethics and includes perspectives from two HMSOM students and two faculty leaders.
AI generated.
Generative artificial intelligence (AI) is dramatically changing the workflows of many professions, and medicine is no different. The question of how AI clinical summarization tools can be introduced into training future physicians is digital territory still being explored at medical schools across the world.
Two medical students and two leaders at the Hackensack Meridian School of Medicine have now published a peer commentary, which said that trainees including medical students are a vital and missing stakeholder group in the ethical debate about the use of AI in clinical notetaking.
Excluding trainees from using AI as a tool as they learn to become doctors is a missed opportunity - and might do more harm than good, according to the commentary.
“The Missing Stakeholder and Their Values: Medical Students and the Integration of AI Clinical Summarization Tools” appears in The American Journal of Bioethics, and was authored by students Rachel Lozada and Ali Sarhan, along with faculty leaders Vice Dean for Academic Affairs Miriam Hoffman, M.D., and Charles E. Binkley, M.D., FACS, HEC-C, the director of AI Ethics and Quality for Hackensack Meridian Health, as well as associate professor of Surgery, and chair of the Faculty Assembly for the medical school.
Together, the four write in response to a proposed ethical assessment model recommending preventing medical trainees from using AI clinical summarization tools. That previous piece held that reliance on the automated tool would lead to “deskilling.”
The HMSOM authors contend instead that, while too much early reliance on the tool would be detrimental, phased and supervised integration into medical education would be a great help to doctors who need to incorporate these tools eventually into their profession.
Traditional note-taking is a critical function for clinicians, they write, since it provides the “cognitive scaffolding” and ability to construct a coherent clinical narrative in a variety of data inputs and observations, as well as logical conclusions. By learning such note taking, it also helps notice learners to avoid “automation bias” which the AI might adhere to absent the real-life experience of a human observing, and interacting with, patients.
Under faculty supervision, students might better learn how AI models operate, identify common errors the machines produce, and overall practice responsible usage of the tools before entering practice.
“With these tools becoming increasingly ubiquitous in clinical practice, it is not a matter of whether physicians will use them, but rather when,” the authors conclude. “There is no better time or place for students to apply the foundational clinical skills to AI summarization tools than during their education.”