Auditable Emergency Triage for Maternal and Newborn Care in India
- Source
- arxiv.org
- Author
- Shobhit Jagga, Aman Dalmia, Niharika Priyadarshini, Neelima Devadas, Amrita K Prasen, Nikhil Nalin, Santhosh SJ, Sreeram Nurani Ramasubramanian, Muhammed Afeer K, Anubhav Arora
- Date
Context
Noora Health, whose nurses answer caregivers' medical questions on a WhatsApp-based service, made its emergency-triage assistant easier to inspect by separating message interpretation from clinical decisions. A extracts symptoms and patient ; clinician-written rules decide which messages need urgent . This lets staff distinguish an extraction mistake from a missing rule and correct the relevant component.
On 427 held-out messages, the system identified 81% of emergencies, compared with 56.5% for its predecessor. These are detection results under the team's internal labeling protocols. Nurses and doctors retain clinical authority, and effects on patient outcomes remain unmeasured. The paper establishes improved emergency detection and clearer error diagnosis within this service.
- LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- attention — The mechanism that lets a model weigh which earlier words matter for the word it's currently processing — the core operation of a transformer.
“we decomposed triage into two steps: an LLM extracts canonical symptoms and patient context from the query using a clinician-authored vocabulary, and a deterministic rule engine captures the scenarios that indicate an emergency”
“the new system has triaged 152,421 patient queries and flagged 28,535 (18.7%) as emergencies”
“Clinicians have also added 48 new rules since deployment, evidence of the faster correction loop we set out to build”
articleInstitution-Specific LLM Prompting Recovers PHI That De-identification Systems and Their Gold Standards Both MissDaniel Palacios, Matthew Brady Neeley, Angel Adetomike Otto, Shalini Dhamodharan, John P. Woodhouse, Chi-fan Lin, Mark Zobeck, Zhandong Liu, Hyun-Hwan Jeong
articleBeyond the Traceback: Using LLMs for Adaptive Explanations of Programming ErrorsAlexandru-Radu Moraru, Shreyan Biswas, Ujwal Gadiraju
articleTowards a Deterministic Math Solver for Clinical Language ModelsFelipe Ocampo Osorio, Sebasti\'an Andr\'es Cajas Ordo\~nez, Maximin Lange, Rafi Al Attrach, Sahil Kapadia, Zakaria Laouabdia Sellami, Angelo Antonio Talio, Leo Anthony Celi
Checking sign-in…
Loading comments…