Anticipatory Care Tool
By identifying small but significant changes in physical and mental wellbeing, it is possible to reduce hospitalisations for people living with dementia.
Background
ACT, which stands for Anticipatory Care Tool, is the product of a research project at University for the Creative Arts funded by Innovate UK in their Zinc Catalyst Healthy Ageing strand.
The project began in 2023 under the leadership of Dr Harry Whalley and Mark Brill. Previous work included a 2018 initiative called Memory Tracks — connecting music and memory for people living with dementia. Development of cognitive stimulation therapy software led to exploration of anticipatory care: identifying physical and mental changes to reduce hospitalisations.
While the NHS Long Term Plan includes anticipatory care, its implementation remains informal among family and non-clinical carers. ACT aims to create an accessible tool for anyone in regular contact with a person living with dementia.
The Problem
Hospital admissions disrupt care continuity, block beds, and significantly reduce quality of life. Anticipatory care offers a proactive alternative.
What Is ACT?
An observational app designed for anyone in regular contact with a person living with dementia.
Works on mobile, tablet, or PC — wherever care happens.
Identifies changes in wellbeing without providing diagnoses.
Requires no specialist expertise. Designed for family, carers, and regular contacts.
Facilitates appropriate health service referrals when changes are detected.
Development
At its core, ACT comprises 16 observational questions across four sections covering physical health, wellbeing, behavioural, and cognitive domains.
These questions were derived from analysis of eight professional assessment tools used in clinical practice:
Research involved co-design workshops and interviews with Memory Matters (Plymouth) and Lifecare (Edinburgh).
How ACT Works
Choose an existing person or add a new one with their name and reference details.
Include any optional notes or context relevant to the observation.
Complete 16 questions across four sections using a 1–5 scale (1 = lowest concern, 5 = highest).
Skip inapplicable questions, navigate flexibly through sections, and submit your observation.
The algorithm applies weighted values per question and section, incorporating a red flag system that identifies deteriorations using a moving average. The algorithm is adjustable based on feedback and ongoing research.
Future enhancement: Machine Learning time series prediction to support anticipatory approaches — identifying patterns before they become crises.
Next Steps
Contact