Multiple large language models versus experienced physicians in diagnosing challenging cases with gastrointestinal symptoms

Faced with challenging cases, doctors are increasingly seeking diagnostic advice from large language models (LLMs). This study aims to compare the ability of LLMs and human physicians to diagnose challenging cases. An offline dataset of 67 challenging cases with primary gastrointestinal symptoms was used to solicit possible diagnoses from seven LLMs and 22 gastroenterologists. The diagnoses by Claude 3.5 Sonnet covered the highest proportion (95% confidence interval [CI]) of instructive diagnoses (76.1%, [70.6%-80.9%]), significantly surpassing all the gastroenterologists (p < 0.05 for all). Claude 3.5 Sonnet achieved a significantly higher coverage rate (95% CI) than that of the gastroenterologists using search engines or other traditional resource (76.1% [70.6%-80.9%] vs. 45.5% [40.7%-50.4%], p < 0.001). The study highlights that advanced LLMs may assist gastroenterologists with instructive, time-saving, and cost-effective diagnostic scopes in challenging cases.

Authors

Institutions

Publication Details

Journal
npj Digital Medicine
Published
2025-02-05
DOI
https://doi.org/10.1038/s41746-025-01486-5
Citations
22
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
article
Field-Weighted Citation Impact
14.40

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Multiple large language models versus experienced physicians in diagnosing challenging cases with gastrointestinal symptoms

Rongchun Zhang, Xintian Yang, Xiaoyu Kang, Shuhui Liang et al.
22 citations
npj Digital Medicine
Artificial Intelligence in Healthcare and Education
14.40
article

Multiple large language models versus experienced physicians in diagnosing challenging cases with gastrointestinal symptoms

Rongchun Zhang, Xintian Yang, Xiaoyu Kang, Shuhui Liang, Huihong Zhai, Rui Guo, Weijie Xue, Daiming Fan, Zhongyin Zhou, Zhi Ni, Yajun Li, Na Liu, Tiantian Zhang, Cheng Wang, Limin Wang, Ning Chen, Handong Wang, Yaling Liu, Yongzhan Nie, Jianghai Zhao, Lina Zhao, Xiaoning Yang, Tongxin Li, Hui Luo, Xiangping Wang, Yanglin Pan, Xiaoyin Zhang, Baiwen Li, Bo Ning, Linhui Zhang, Shanhong Tang, Fandong Meng, Yong Lv, Gui Ren
article en
22 citations

Abstract

Faced with challenging cases, doctors are increasingly seeking diagnostic advice from large language models (LLMs). This study aims to compare the ability of LLMs and human physicians to diagnose challenging cases. An offline dataset of 67 challenging cases with primary gastrointestinal symptoms was used to solicit possible diagnoses from seven LLMs and 22 gastroenterologists. The diagnoses by Claude 3.5 Sonnet covered the highest proportion (95% confidence interval [CI]) of instructive diagnoses (76.1%, [70.6%-80.9%]), significantly surpassing all the gastroenterologists (p < 0.05 for all). Claude 3.5 Sonnet achieved a significantly higher coverage rate (95% CI) than that of the gastroenterologists using search engines or other traditional resource (76.1% [70.6%-80.9%] vs. 45.5% [40.7%-50.4%], p < 0.001). The study highlights that advanced LLMs may assist gastroenterologists with instructive, time-saving, and cost-effective diagnostic scopes in challenging cases.

npj Digital MedicineVol. 8(1)
Fujian Medical University (CN), Union Hospital (HK), Shanxi Medical University (CN), Shanghai Jiao Tong University (CN), Capital Medical University (CN), Dalian Medical University (CN), Peking University (CN), Southern University of Science and Technology (CN), Wuhan University (CN), Henan University Huaihe Hospital and Huaihe Clinical Institute (CN), National Clinical Research Center for Digestive Diseases (CN), Beijing Shijingshan Hospital (CN), Second Affiliated Hospital of Chongqing Medical University (CN), Kumamoto University Hospital (JP), Peking University People's Hospital (CN), Zhongshan Hospital of Xiamen University (CN), Ningxia Medical University (CN), Hainan General Hospital (CN), Shanghai First People's Hospital (CN), Ningxia Medical University General Hospital (CN), Renmin Hospital of Wuhan University (CN), Beijing Friendship Hospital (CN), Xijing Hospital (CN), Shanxi Provincial People’s Hospital (CN), Huazhong University of Science and Technology (CN), Hainan Medical University (CN), Chongqing Medical University (CN), Air Force Medical University (CN)
National Natural Science Foundation of China, Huazhong University of Science and Technology
Openalex Percentile: Top 1%
Artificial Intelligence in Healthcare and Education
14.40
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.