肿瘤专科医务人员大语言模型使用行为影响因素研究OA

Objective To investigate key factors and barriers influencing medical staff's adoption of large language mod-el tools,propose strategies for enhancing effectiveness,and promote efficient implementation of the technology.Methods Based on the integrated Technology Acceptance and Usage Theory,a mixed-methods approach was employed to conduct a cross-sectional questionnaire survey among all medical staff at a specialized cancer hospital.Structural equation modeling was used to validate theoretical hypotheses.Interview data on the usage of large language model tools among 12 respondents were collected to supple-ment the validation of quantitative findings.Results Quantitative analysis indicated that perceived cost negatively impacted us-age intention;performance expectations and effort expectations positively influenced usage intention;performance expectations,effort expectations,and social influence positively affect usage behavior.Qualitative interviews revealed that demand for large lan-guage model tools exhibited dual-dimensional differentiation based on professional roles and job titles,with demand satisfaction showing a gradient pattern;usability and usage cost were primary barriers;organizational promotion and peer influence were key to large-scale adoption.Conclusion Tool functions should be tailored to different user groups,with application scenarios devel-oped by hierarchical levels to reduce usage costs.Usability should be enhanced through technical ecosystem integration,and or-ganizational support strengthened through outcome quantification and incentive mechanisms.Finally,pathways from efficiency im-provement to intelligent decision-making should be explored to achieve efficient application and continuous improvement of large language model tools.