Personalising Domiciliary Dementia Care Through Co-Design: Developing a Preference-Based Matching and Scheduling App
Parallel 5
18 November, 2026 2:10 – 3:20 pmMini Presentations: Group 2
Rasa Mikelyte
Research Fellow
University of Kent
Summary
Delivering high-quality domiciliary dementia care is challenging due to competing demands in the care market and the diverse needs of people supported at home. Evidence indicates that aligning care workers’ skills with the preferences of people living with dementia and their families, together with scheduling that promotes continuity, can improve the quality of support. This project aimed to personalise domiciliary dementia care through the co-design and development of a research-informed prototype app that uses machine-learning approaches to align worker and client preferences and generate schedules that balance clients’ need for consistency with workers’ need for sufficient hours.
The project comprised (1) synthesising research to establish what constitutes high-quality domiciliary dementia care and for whom; (2) conducting 10 focus groups and 7 interviews (N=48) with people with dementia, family supporters, care workers and care managers; and (3) running 6 co-design workshop rounds with these stakeholder groups (N=20), which iteratively informed algorithm development and interface design. Health-economics oversight was integrated throughout.
The project shows that co-designed, preference-based matching and scheduling tools are feasible and valued. Treating prototype development as a research activity enabled extensive, iterative involvement of people with lived and practice experience, ensuring the tool reflected real-world needs. However, involving multiple stakeholder groups also revealed tensions and competing priorities that could not always be resolved without risking inequity. These insights highlight the need for ongoing dialogue, transparency and ethical consideration when developing person-centred technologies.