A Conversational Robot for Cognitively Impaired Older People Who Live Alone: An Exploratory Feasibility Study

摘要

BACKGROUND: Social isolation and loneliness are significant risk factors for poor mental health in older adults, particularly those living alone with cognitive impairment. Socially assistive robots (SARs) offer a promising approach to enhance social connection and emotional well-being in this population. This study aimed to examine the feasibility, acceptability and potential psychosocial impact of a 16-week, home-based use of a conversational SAR in cognitively impaired older adults living alone. METHODS: This was a single-arm, exploratory feasibility study involving home installation of a humanoid conversational robot (RoBoHoN) for 16 weeks. Usability was assessed with the System Usability Scale (SUS). Psychosocial outcomes, including loneliness (UCLA-LS3) and depressive symptoms (GDS-15), were measured pre- and post-intervention. Semi-structured interviews were conducted and analysed using hybrid content analysis to explore user experiences. RESULTS: A total of 11 participants were enrolled, including individuals with mild cognitive impairment, mild dementia and late-onset psychosis; nine completed the study. The mean age was 81.9 years, with an age range of 71-89 years. The mean SUS score was 81.7, indicating high usability even among cognitively impaired users. Small-to-moderate effect sizes were observed for reductions in loneliness and depressive symptoms (UCLA-LS3: r = 0.40; GDS-15: r = 0.64), although not statistically significant. Qualitative findings highlighted perceived companionship and increased opportunities to speak, while also noting limitations such as poor conversational depth, delayed responses and technical issues like battery depletion or sensitivity to background noise. CONCLUSIONS: The 16-week home-based use of a conversational robot was feasible and well accepted by older adults with cognitive impairment living alone. Although preliminary, findings suggest potential psychosocial benefits of SARs in this population. These results support the rationale for conducting a controlled trial targeting loneliness and depression in cognitively impaired, socially isolated older adults.