Impact of an Evidence-Based Large Language Model (LLM) Diagnostic Decision Support System: A Randomised Controlled Trial

Sangah Ahn, Joongheum Park, Sujeong Hur, Kwang Yul Jung, Jae Ho Lee, Sae Won Choi, Meong Hi Son, Min Jeoung Kang, Youn Jung Kim, Hanna Park, Won Chul Cha, Junsang Yoo

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Generative artificial intelligence (AI) influences clinical decision-making in healthcare by analyzing medical data and proposing personalized treatment options based on patient records. However, generative AI limited due to its inability to provide accurate evidence. Therefore, this study aims that the influence of AI-generated diagnostic suggestions on emergency healthcare providers' diagnostic patterns and decision-making, and evaluates the correlation between clinicians' adoption and diagnosis accuracy.

Original languageEnglish
Title of host publicationMEDINFO 2025 - Healthcare Smart x Medicine Deep
Subtitle of host publicationProceedings of the 20th World Congress on Medical and Health Informatics
EditorsMowafa S. Househ, Mowafa S. Househ, Zain Ul Abideen Tariq, Mahmood Al-Zubaidi, Uzair Shah, Elaine Huesing
PublisherIOS Press BV
Pages273-277
Number of pages5
ISBN (Electronic)9781643686080
DOIs
StatePublished - 7 Aug 2025
Event20th World Congress on Medical and Health Informatics, MEDINFO 2025 - Taipei, Taiwan, Province of China
Duration: 9 Aug 202513 Aug 2025

Publication series

NameStudies in Health Technology and Informatics
Volume329
ISSN (Print)0926-9630
ISSN (Electronic)1879-8365

Conference

Conference20th World Congress on Medical and Health Informatics, MEDINFO 2025
Country/TerritoryTaiwan, Province of China
CityTaipei
Period9/08/2513/08/25

Keywords

  • Clinical Decision Support System (CDSS)
  • Emergency Medicine
  • Large Language Model (LLM)

Fingerprint

Dive into the research topics of 'Impact of an Evidence-Based Large Language Model (LLM) Diagnostic Decision Support System: A Randomised Controlled Trial'. Together they form a unique fingerprint.

Cite this