Accepted for/Published in: JMIR Medical Informatics
Date Submitted: Nov 22, 2025
Date Accepted: Jun 18, 2026
Chinese Pathologists' Adoption of Large Language Models: A Cross-Sectional Survey on Influencing Factors
ABSTRACT
Background:
China's aging population and rising cancer awareness have increased demand for pathological diagnoses, exacerbating the pathologist shortage.
Objective:
This study evaluates Chinese pathologists' perceptions of and willingness to adopt LLMs and identifies key adoption factors.
Methods:
A cross-sectional survey of pathologists at the 2024 National Pathology Academic Conference in Xi’an examined factors influencing LLM adoption using a structured questionnaire and statistical analysis.
Results:
A total of 339 pathologists participated, 69.6% female and 30.4% male, mostly aged 30-40 (48.4%). Most held a bachelor's degree (63.1%), and 46.6% were attending physicians. Pathologists were from various regions, with 69.6% from third-tier cities or above. Average weekly working hours were 43.01, and 34.2% read over 5,000 slides annually. Regarding LLMs, 73.2% were familiar, mainly through social media, and 25.8% had used them. ChatGPT was the most recognized model. Users rated LLMs lower in clinical applications, particularly in patient data analysis (3.61 vs. 3.90, p<0.05) and understanding disease mechanisms (3.66 vs. 3.95, p=0.042). Factors influencing LLM usage included working in fourth-tier cities (ad-OR=1.99, p=0.030), reading over 5,000 slides (ad-OR=2.12, p=0.047), research involvement (ad-OR=2.94, p<0.001), teaching (ad-OR=2.37, p=0.001), and prior LLM experience (ad-OR=2.45, p=0.002). Ethical concerns included health inequality, plagiarism, and data privacy. Suggested improvements included integrating LLMs with diagnostic tools, enhancing datasets, and updating knowledge bases regularly.
Conclusions:
Chinese pathologists are highly interested in integrating LLMs into clinical practice to enhance diagnostic accuracy, streamline workflows, and improve service quality. Future efforts should focus on refining LLM capabilities, addressing ethical concerns, and ensuring smooth integration.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.