Latent Profile Analysis of Patients' Acceptance of Companion Robots in Unaccompanied Wards: A Cross-sectional Study

Abstract Aims To understand the current status of inpatients’ acceptance of companion robots in no-accompanying wards through latent profile analysis, identify potential subgroups and their population characteristics, and analyze the influencing factors of different profile categories. Design A cross-sectional study. Methods The study utilized 425 inpatients from the no-accompanying ward of a tertiary hospital in Sichuan Province, China as research subjects. The study employed a number of research tools, including demographic characteristics, the Almere Technology Acceptance Questionnaire (ATAQ), the Medical Outcomes Study Social Support Survey (MOS-SSS), the PROMIS Emotional Support Short Form, the Negative Attitude towards Robots Scale (NARS), and the General Self-Efficacy Scale (GSE). A latent profile model of inpatients’ companion robot acceptance was analyzed using Mplus 8.3. The influencing factors of each profile model were analyzed by multinomial logistic regression analysis. Results 425 inpatients were finally included. Three latent profiles were identified: the wait-and-see acceptance group (n = 83, 19.53%), the cautious acceptance group (n = 139, 32.71%), and the active acceptance group (n = 203, 47.76%), with Entropy = 0.868, indicating good classification accuracy. Age, education level, and household income showed statistically significant differences among the three groups (P < 0.001). Medical social support, emotional support, and general self-efficacy scores showed a gradient distribution (active > cautious > wait-and-see), while negative attitudes toward robots showed the opposite trend (P < 0.001). Multinomial logistic regression revealed that advanced age, lower education level, lower household income, insufficient emotional support, low general self-efficacy, and negative perceptions of robots were independent risk factors for classification into the low-acceptance profile. Conclusion The findings revealed significant heterogeneity in inpatients’ acceptance of companion robots in no-accompanying wards. Elderly patients with lower education levels, farmers, or retirees constitute high-risk groups for low acceptance. Adequate social support and higher self-efficacy can enhance patients’ willingness to accept robots. Clinical practice should implement stratified health education and optimize robot functions to better meet the needs of older adults. Clinical trial number Not applicable.