Effects of a Humanoid Companion Robot on Dementia Symptoms and Caregiver Distress for Residents in Long-Term Care

Abstract

Objectives: To test the effects of deploying a humanoid companion robot (Kabochan) in comparison with usual care for long-term care facilities’ residents with dementia.

Design: A 2-arm, randomized controlled trial with ABAB withdrawal design, lasting 32 weeks. After an 8-week baseline period, Kabochan was introduced in a nonfacilitated, individual approach with experimental-group participants (n = 52) for 8 weeks, then removed for 8 weeks, and then reintroduced for another 8 weeks. The control group (n = 51) received the usual standardized care.

Setting and Participants: Seven long-term care facilities in Hong Kong. 103 residents (76% women, 87.2 ± 7.4 years) with a clinical diagnosis of dementia.

Measures: Outcome assessments occurred at 5 time points: baseline (week 1) and the end of each phase (weeks 8, 16, 24, and 32). Primary outcomes were assessed with the Neuropsychiatric Inventory Questionnaire (symptom severity and caregiver distress subscales) and the Geriatric Depression Scale; secondary outcomes were measured by the Hong Kong Montreal Cognitive Assessment 5-minute Protocol, the Modified Barthel Index for Activities of Daily Living, and the Quality of Life–Alzheimer's disease scale.

Results: A multivariate analysis of variance indicated a statistically significant group × time interaction for neuropsychiatric-related caregiver distress at week 16 (F = 6.72, P = .011), with a moderate effect size (η p 2 = 0.06). When Kabochan was removed in the withdrawal phase (weeks 17–24), the neuropsychiatric symptoms became more severe at week 24 for the intervention group (F = 4.68, P = .003), although the effect size was small to moderate (η p 2 = 0.04). No statistical between-group differences were found in other health outcomes.

Conclusions and Implications: The Kabochan was potentially effective at reducing short-term neuropsychiatric symptoms and relevant caregiver distress for residents with dementia. An individualized care plan with continuous monitoring is required to integrate the humanoid robot into routine dementia care.