Structural equation modeling for influencing factors on behavioral intention to adopt medical AI among Chinese nurses: a nationwide cross-sectional study.

backgroundArtificial intelligence (AI) shows great potential to improve clinical nursing practices. However, concerns and challenges related to its implementation have led to resistance among nurses, hindering the widespread use of AI tools in healthcare. This study aimed to apply the Unified Theory of Acceptance and Use of Technology (UTAUT) model to medical AI in Chinese nursing and to examine the relationships specified within the established theoretical framework.

methodsThis study employed a nationwide cross-sectional survey to assess Chinese nurses' demographic characteristics, knowledge, and attitudes toward medical AI. Guided by the UTAUT framework, key constructs like performance expectancy, effort expectancy, social influence, facilitating conditions, and behavioral intention were included, along with perceived risk and trust as additional variables. Data were collected from 8,514 valid questionnaires. Structural equation modeling (SEM) was employed for analysis. Confirmatory factor analysis (CFA) was conducted to construct validity of the extended UTAUT model, followed by evaluation of model fit and estimation of path coefficients. The Bootstrap method was utilized to examine the mediating effects.

results84.0% (7151/8514) of nurses had heard of medical AI, 79.5% (6771/8514) of nurses held optimistic views regarding its prospects, only 17.8% (1512/8514) of nurses had actually utilized it in clinical nursing work. The results of the SEM analysis revealed that performance expectancy (β = 0.074, P < 0.001), effort expectancy (β = 0.158, P < 0.001), social influence (β = 0.039, P < 0.001) and trust (β = 0.670, P < 0.001) had significant positive impacts on behavioral intention of nurses to adopt medical AI, while perceived risk had a significant negative impact (β=-0.013, P = 0.037). In contrast, facilitating conditions did not significantly affect behavioral intention (β=-0.010, P = 0.334). Additionally, the influence of performance expectancy (β = 0.148, P < 0.001), effort expectancy (β = 0.206, P < 0.001), social influence (β = 0.130, P < 0.001) and facilitating conditions (β = 0.074, P < 0.001) on behavioral intention was mediated by trust, highlighting the core role of trust in promoting intention to use AI. Notably, trust did not significantly mediate the relationship between perceived risk and behavioral intention (β=-0.007, P = 0.229).

conclusionsChinese nurses generally express high acceptance of medical AI, but its practical application rate remains low. To facilitate the widespread adoption of AI in nursing settings, future efforts should prioritize improving the accuracy, usability, and accessibility of medical AI. Moreover, initiatives should focus on creating a supportive environment, addressing perceived risks, and fostering greater trust in AI technologies.