Attitudes and perceptions of the application of large language models among health professionals: A mixed-methods...

To explore the attitudes and perceptions of health professionals towards applying large language models (LLMs).

A mixed-methods systematic review.

Ten databases (PubMed, Web of Science, The Cochrane Library, Embase, CINAHL Complete, PsycINFO, the China National Knowledge Infrastructure, Wanfang Database, China Science and Technology Journal Database, and SinoMed) were searched from their inception till February 2025. The Joanna Briggs Institute's convergent integrated approach was adopted to extract data and synthesize it according to the theoretical framework.

A total of 30 articles were included. According to the Technology framework, the analysis revealed four themes: (1) Performance expectancy: Health professionals' perceptions of the effectiveness of LLMs application; (2) Effort expectancy: Health Professionals' perceptions about the difficulty and ease of LLMs application; (3) Social influence: Social environment's influence on health professionals' attitudes toward LLMs application; and (4) Facilitating conditions: Objective resource and environmental factors supporting or hindering health professionals' LLMs application.

This systematic review presents the first comprehensive synthesis of health professionals' attitudes and perceptions toward the application of LLMs. The majority of health professionals expressed generally favorable attitudes toward LLMs, albeit tempered by concerns about trust, integration, and ethical safeguards. Therefore, relevant departments need to formulate effective strategies and guidelines to address the issues and standardize the application of LLMs.