Voice Assistant to Provide Cognitive Stimulation Therapy at Scale

PROJECT SUMMARY
Cognitive stimulation therapy (CST) is one of the most effective interventions for maintaining cognitive
functioning and wellbeing of persons living with dementia (PLwDs). CST is currently recommended by the
United Kingdom government for the management and support of PLwDs. Alzheimer's Disease International
(ADI) also recommends CST as an effective intervention. However, providing CST to individuals in need
remains a serious challenge. CST is traditionally provided by trained health professionals, which is costly and
not scalable. Recent work has focused on adapting CST to be provided by care partners at home. However,
current approaches add considerable burden on care partners, which results in low adherence and engagement.
Such adherence and engagement issues undermine the efficacy of CST and consequently, lead to non-optimal
outcomes. As such, irrespective of the established evidence in favor of CST, there remains considerable
challenges in disseminating it at scale, specifically when it comes to individuals living in remote locations.
We argue that voice assistants (VAs) provide an ideal platform to deliver CST at scale. VAs are widely available
given their integration with most recent consumer devices. Recent VAs can support visual displays and voice
interactions — best of both worlds when it comes to designing simple, intuitive, and effective interfaces. More
importantly, significant numbers of older adults actively use VAs. The high adoption and acceptance of VAs
among older adults provides a unique opportunity to support PLwDs and their care partners.
This project aims to advance this vision of leveraging VAs for dementia care support. We will develop a VA to
deliver customized and tailored cognitive stimulation therapy for PLwDs. We will minimize care partner
burden and provide effective support to deliver CST sessions at home. We will follow the Center for eHealth
Research (CeHRes) roadmap to develop the VA, which will result in an iterative and community-driven
implementation process. We will also deploy the VA with 15 dyads (PLwDs and care partners) for 12 weeks.
This will allow us to collect granular interaction, usability, and acceptance data from real-world settings. We
will also conduct focus group interviews to collect in-depth qualitative data.
This project will establish a preliminary evidence base regarding the feasibility and acceptability of using VAs
for dementia interventions. Given the wide adoption of VAs, the project can positively impact a large
population of PLwDs and care partners including those living in remote and rural areas.

Grant Number: 1R21AG085439-01
NIH Institute/Center: NIH
Principal Investigator: Saeed Abdullah