UC San Diego Health highlighted a new clinical study showing how artificial intelligence can make sepsis quality assessment faster, more efficient, and more useful for physicians.

Published in JAMA Network Open, the study evaluated the use of large language models to automatically review complex SEP-1 quality measures and provide targeted, near-real-time feedback to emergency department care teams.

Traditional SEP-1 review can involve a 63-step evaluation of extensive medical records and may require months of effort. The study found that AI can scan hundreds of patient charts, identify important clinical context in seconds, and provide feedback while patients may still be receiving care.

This approach helps move sepsis quality reporting beyond delayed, retrospective chart review and toward timely, actionable guidance that can support improved compliance with national sepsis care measures.

Read the full UC San Diego Health press release

‍