Nurses' AI training acceptance and tool usage: a structural equation model of individual and hospital-level factors.
With the rapid advancement of artificial intelligence (AI), it has exerted a profound influence on the medical field. Currently, AI applications in nursing remain nascent in China. This study aimed to investigate nurses' attitudes and anxiety levels toward AI in general hospitals in western China, and to analyze the association of these psychological factors with their AI training acceptance and clinical AI tool usage behavior.
A multicenter cross-sectional design was employed. In February 2026, a questionnaire survey was conducted among 620 registered nurses in public general hospitals across western China (Sichuan, Guizhou, and Qinghai provinces). The questionnaire collected data on nurses' demographic information, whether they had received AI training and whether they had used AI tools in the workplace, as well as their responses to the General Attitudes Toward AI Scale (GAAIS) and the AI Anxiety Scale (AIAS). Participation was voluntary and anonymous. Descriptive statistics, reliability and validity testing, correlation analysis, and structural equation modeling were performed using R software (version 4.5.1).
Of 603 returned questionnaires (97.3% response rate), 591 were valid. Negative attitudes were significantly associated with lower training acceptance (β = -0.140, p = 0.015) and lower tool usage (β = -0.141, p = 0.003). Positive attitudes and AI anxiety showed no significant associations with either outcome. Older age and longer work experience were associated with higher rates of both outcomes, with work experience showing the strongest association with tool usage. No significant association was found between training acceptance and tool usage after covariate adjustment (β = 0.070, p = 0.356).
Hospital-level resources, older age, and longer work experience were the factors most strongly associated with nurses' AI training acceptance and AI tool usage, with work experience showing the strongest association with tool usage. Negative attitudes were associated with lower engagement in both outcomes, whereas positive attitudes and anxiety showed no significant independent associations. These findings suggest that institutional support, experience-based peer learning, and targeted reduction of negative perceptions may be key priorities for AI integration in nursing practice.
A multicenter cross-sectional design was employed. In February 2026, a questionnaire survey was conducted among 620 registered nurses in public general hospitals across western China (Sichuan, Guizhou, and Qinghai provinces). The questionnaire collected data on nurses' demographic information, whether they had received AI training and whether they had used AI tools in the workplace, as well as their responses to the General Attitudes Toward AI Scale (GAAIS) and the AI Anxiety Scale (AIAS). Participation was voluntary and anonymous. Descriptive statistics, reliability and validity testing, correlation analysis, and structural equation modeling were performed using R software (version 4.5.1).
Of 603 returned questionnaires (97.3% response rate), 591 were valid. Negative attitudes were significantly associated with lower training acceptance (β = -0.140, p = 0.015) and lower tool usage (β = -0.141, p = 0.003). Positive attitudes and AI anxiety showed no significant associations with either outcome. Older age and longer work experience were associated with higher rates of both outcomes, with work experience showing the strongest association with tool usage. No significant association was found between training acceptance and tool usage after covariate adjustment (β = 0.070, p = 0.356).
Hospital-level resources, older age, and longer work experience were the factors most strongly associated with nurses' AI training acceptance and AI tool usage, with work experience showing the strongest association with tool usage. Negative attitudes were associated with lower engagement in both outcomes, whereas positive attitudes and anxiety showed no significant independent associations. These findings suggest that institutional support, experience-based peer learning, and targeted reduction of negative perceptions may be key priorities for AI integration in nursing practice.
Authors
Guo Guo, Ding Ding, Tang Tang, Wei Wei, Chen Chen, Cao Cao, Zhang Zhang, Zheng Zheng
View on Pubmed