Work

Human or LLM as Standardized Patients? A Comparative Study in Medical Education

Year: 2026

Type: conference-paper

Abstract: Standardized patients (SPs) are indispensable for clinical skills training but remain expensive and difficult to scale.Although large language model (LLM)-based virtual standardized patients (VSPs) have been proposed as an alternative, their behavior remains unstable and lacks rigorous comparison with human standardized patients.We propose EasyMED, a multi-agent VSP framework that separates case-grounded information disclosure from response generation to support stable, inquiry-conditioned patient behavior.We also introduce SPBench, a human-grounded benchmark with eight expertdefined criteria for interaction-level evaluation.Experiments show that EasyMED more closely matches human SP behavior than existing VSPs, particularly in case consistency and controlled disclosure.A four-week controlled study further demonstrates learning outcomes comparable to human SP training, with stronger early gains for novice learners and improved flexibility, psychological safety, and cost efficiency.

Source: Proceedings of the conference - Association for Computational Linguistics. Meeting

DOI: 10.18653/v1/2026.acl-long.1243

Authors: Bingquan Zhang, Xiaoxiao Liu, Yuchi Wang, Zhou Lei, Qianqian Xie, Benyou Wang

Institutions: Chinese University of Hong Kong, Shenzhen, Wuhan University

Language: English

Topic: Artificial Intelligence in Healthcare and Education

Subfield: Health Informatics

Field: Medicine

Domain: Health Sciences

SDGs: Good health and well-being, Quality education

Open Access status: hybrid

Funders: Tencent, National Natural Science Foundation of China, Chinese University of Hong Kong, Guangdong Science and Technology Department