Skip to content
OpenSmartRoute

LLMs1 min read

LLM-Based Phenotyping Improves OUD Detection

A new LLM-based computable phenotype (CP) achieved high F1 and AUROC scores (0.774 and 0.934) for identifying opioid use disorder (OUD) from EHRs, surpassing traditional machine learning and zero-shot LLM approaches.

By OpenSmartRoute editorial · written through the router by writer-small

From arXiv cs.CL - “A Rubric-Guided Large Language Model Solution for Opioid Use Disorder Computable Phenotyping”

This study presented a rubric-guided large language model (LLM) solution for opioid use disorder (OUD) computable phenotyping. The framework utilized an 18-item rubric to instruct the LLM in automatically extracting critical text and supporting evidence for determining OUD flags. The system was evaluated on a dataset of 253 patients, including 68 confirmed OUD cases. The LLM-based CP demonstrated superior performance compared to existing methods, achieving an F1 score of 0.774 and an AUROC of 0.934. This represents a relative improvement of 12.8% and 44.4% over machine learning-based CP and zero-shot LLMs, respectively.

Source: https://arxiv.org/abs/2609.05682

Published Sep 9, 2026 · updated Sep 9, 2026 · 94 words

Keep reading

Related posts

More in LLMs

LLMs1 min read

CONDUIT: KV Cache Reuse Policy for VLMs

CONDUIT is a training-free refresh policy for KV caches in vision-language models, unifying single and multi-image reuse. It achieves high accuracy and speed improvements with a small refresh budget, outperforming budgeted methods on latency.

LLMs1 min read

Language Models Exhibit Layered Magnetic Organization

Research identifies 'magnetic vectors' within LLMs that organize surrounding tokens, influencing their length. This layer-wise geometric organization, observed across models and tasks, offers a new probe-free approach to understanding language model computation.