When patients and their families leave the emergency room, they’re usually clutching a sheaf of papers: Called discharge notes, they contain important instructions about how to care for themselves and when to follow up with a doctor. If patients don’t speak English, though, they may leave with directions they can’t understand.
At most hospitals, it takes between hours and days to translate discharge notes for outpatients who have limited English proficiency. But many health systems are considering automated machine translation to speed up the process, driven by the proliferation of powerful large language models — along with a looming deadline in federal nondiscrimination rules.
Now Children’s Hospital Los Angeles is taking the next step, beginning to pilot a program that will use generative AI to translate discharge notes for Spanish-speaking patients in seconds.
This article is exclusive to STAT+ subscribers
Unlock this article — plus in-depth analysis, newsletters, premium events, and news alerts.
Already have an account? Log in
