Monitoring quality of dementia care
Research type
Research Study
Full title
Monitoring quality of dementia care - preliminary evaluation of the MABLE interface for automated evaluation of caregiver interactions.
IRAS ID
233667
Contact name
Robert Stewart
Contact email
Sponsor organisation
King's College London
Clinicaltrials.gov Identifier
17/LO/1734, London - Camden & Kings Cross Research Ethics Committee
Duration of Study in the UK
1 years, 0 months, 3 days
Research summary
The aim of the project is to provide early information on a computer software product, MABLE, under development. The software seeks to identify strain in dementia care by recording conversations between people with dementia and their carers, by converting those conversations into text, and then analysing the text for features suggesting strain. Ultimately the aim of the product will be to provide assistance (potentially both direct advice and signposting towards sources of help); however, at this early stage, we are simply seeking to investigate whether people find the technology acceptable, whether they have any difficulties using it, and whether the automatic analysis of text is able to identify strain with sufficient accuracy. We will seek to recruit 10-15 consenting people with early stage and recently diagnosed dementia who are living alone in their own homes and who have at least one informal carer. The person with dementia will be provided with a specifically designed tablet (iPad) which allows them to make easy contact with their carer (and any other preferred contacts) and the carer will have an app installed on their smartphone to connect with the tablet and collect the information required for the study. The audio data will be automatically transcribed to text via the MABLE software and the calls will be automatically analysed to determine trends in mood, cognitive and behavioural patterns and indications of anxiety, depression, anger, agitation, confusion, etc. We hope to gather information routinely over about 12 months for each participant group. Questionnaire information will be collected at the beginning of the study and at 3-monthly intervals. In addition, the app and tablet will also ask both parties to provide general ratings about the most recent conversation. These and transcripts of sample conversations will be examined and results fed back in order to improve accuracy.
REC name
London - Camden & Kings Cross Research Ethics Committee
REC reference
17/LO/1734
Date of REC Opinion
8 Feb 2018
REC opinion
Further Information Favourable Opinion