Medical Incident Causal Factors and Preventive Measures Generation Using Tag-based Example Selection in Few-shot Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Haseyama, Yuna, Ito, Tomoki, Sakaji, Hiroki, Noda, Itsuki
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913110573449216
author Haseyama, Yuna
Ito, Tomoki
Sakaji, Hiroki
Noda, Itsuki
author_facet Haseyama, Yuna
Ito, Tomoki
Sakaji, Hiroki
Noda, Itsuki
contents In high-stakes domains such as healthcare, the reliability of Large Language Models (LLMs) is critical, particularly when generating clinical insights from incident reports. This study proposes a tag-based few-shot example selection method for prompting LLMs to generate background/causal factors and preventive measures from details of the medical incidents. For our experiments, we use the Japanese Medical Incident Dataset (JMID), a structured dataset of 3,884 real-world medical accident and near-miss reports. These reports are variably annotated with a wide range of tags--some include descriptive information (e.g., "medications," "blood transfusion therapy"). We compare three few-shot example selection strategies--random sampling, cosine similarity-based selection, and our proposed tag-based method--using GPT-4o and LLaMA 3.3. Results show that the tag-based approach achieves the highest precision and most stable generation behavior, while similarity-based selection often leads to unintended outputs and safety filter activation. These findings suggest that selecting examples based on human-interpretable dataset tags can improve generation precision and stability in clinical LLM applications.
format Preprint
id arxiv_https___arxiv_org_abs_2605_10025
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Medical Incident Causal Factors and Preventive Measures Generation Using Tag-based Example Selection in Few-shot Learning
Haseyama, Yuna
Ito, Tomoki
Sakaji, Hiroki
Noda, Itsuki
Computation and Language
Artificial Intelligence
In high-stakes domains such as healthcare, the reliability of Large Language Models (LLMs) is critical, particularly when generating clinical insights from incident reports. This study proposes a tag-based few-shot example selection method for prompting LLMs to generate background/causal factors and preventive measures from details of the medical incidents. For our experiments, we use the Japanese Medical Incident Dataset (JMID), a structured dataset of 3,884 real-world medical accident and near-miss reports. These reports are variably annotated with a wide range of tags--some include descriptive information (e.g., "medications," "blood transfusion therapy"). We compare three few-shot example selection strategies--random sampling, cosine similarity-based selection, and our proposed tag-based method--using GPT-4o and LLaMA 3.3. Results show that the tag-based approach achieves the highest precision and most stable generation behavior, while similarity-based selection often leads to unintended outputs and safety filter activation. These findings suggest that selecting examples based on human-interpretable dataset tags can improve generation precision and stability in clinical LLM applications.
title Medical Incident Causal Factors and Preventive Measures Generation Using Tag-based Example Selection in Few-shot Learning
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2605.10025