CognoSpeak: a Cognitive Health Assessment Tool (CcHAT)
Study Code / Acronym
CcHATAward Number
NIHR202911Programme
Invention for InnovationStatus / Stage
ActiveDates
1 February 2022 -31 January 2025
Duration (calculated)
02 years 11 monthsFunder(s)
NIHRFunding Amount
£1,357,247.00Funder/Grant study page
NIHRContracted Centre
University of SheffieldContracted Centre Webpage
Principal Investigator
Dr Daniel BlackburnPI Contact
heidi.christensen@sheffield.ac.ukPI ORCID
0000-0001-8886-1283,0000-0003-3028-5062Principal Investigator
Dr Heidi ChristensenWHO Catergories
Methodologies and approaches for risk reduction researchRisk reduction intervention
Disease Type
Cognitive ImpairmentCPEC Review Info
| Reference ID | 14 |
|---|---|
| Researcher | Reside Team |
| Published | 12/06/2023 |
Data
| Study Code / Acronym | CcHAT |
|---|---|
| Award Number | NIHR202911 |
| Status / Stage | Active |
| Start Date | 20220201 |
| End Date | 20250131 |
| Duration (calculated) | 02 years 11 months |
| Funder/Grant study page | NIHR |
| Contracted Centre | University of Sheffield |
| Contracted Centre Webpage | |
| Funding Amount | £1,357,247.00 |
Abstract
How to develop a quicker and more accurate way to assess people with memory concerns? Background: Early diagnosis of dementia and consistent monitoring of people with mild cognitive impairment (MCI) can improve treatment and reduce anxiety for patients and families. However, the current NHS memory assessment pathway (MAP) is expensive, slow and hugely variable. This project will deliver a low-cost, accessible and repeatable tool that will provide early and accurate detection and sensitive monitoring of cognitive impairment (CI). The potential to deliver assessments remotely (e.g. in patients homes) is attractive and could make MAP more accessible to underrepresented groups (e.g., ethnic minorities). CognoSpeakTM is an online tool for assessing cognition using automated speech analytics. It engages the user in a conversation, placing a high demand on multiple cognitive domains. Extensive PPIE work has guided the development throughout. Users find the tool trustworthy and easy to use. Clinicians value the projected improvements to MAP. Accuracy exceeds that of current pen-and-paper tests. Aims & Objectives: This research aims to advance the integration of CognoSpeakTM into MAP in order to improve detection and monitoring of cognitive impairment (CI) for all. The project will deliver a market-ready CognoSpeakTM tool with UKCA Class I approval that has been co-designed with MAP-clinicians and a range of end-user groups. The main objectives are: Demonstration of feasibility of stratification in primary care using CognoSpeakTM. Co-design of CognoSpeakTM with end-users to ensure no cultural or language barriers. Demonstration of feasibility of using CognoSpeakTM with underrepresented groups. Development of AI-algorithms agnostic to language background. Tripling of data resource to allow validation in people with memory problems (primary care), ethnic minority, and in PwMCI (secondary care). Production of a health economic (HE) model to aid commercialisation and NHS implementation pathways. Delivery of a trial platform to demonstrate HE effectiveness of CognoSpeakTM. Methods Phased research plan with two cycles of software development, AI refinement and PPIE feedback. Determination of tool accuracy with sensitivity and specificity using detailed neuropsychology, structural brain imaging and clinical assessment. Regular co-production workshops with patients and families, ethnic minority groups and clinicians will improve the tool ensuring usability for all and an accessible and acceptable reporting system. Deliverable timeline 36 months project: M0-M6: Initial PPIE and CognoSpeakTM co-design. M6-M18: Phase-I data collection and evaluation; AI development; PPIE co-design of next CognoSpeakTM version. M18-M30: Phase-II using refined CognoSpeakTM. M30-M36: Final dissemination (mixed-media co-production; academic papers; end-of-project showcase); regulatory approval; AI finalising. Post-project: M36: Market-ready CognoSpeakTM. Follow-on study to demonstrate HE-benefit of CognoSpeakTM at beacon sites. Impact and Dissemination Key measures of success: Completion of evaluation to time and target. CognoSpeakTM roll-out in three beacon sites. Refined commercialisation plan informed by PPIE, HE and AI. Improved understanding of use of CognoSpeakTM in primary care. Better understanding of under-provision of diagnostic healthcare to ethnic minority groups. Future impact: Reformed, easier-to-access and faster assessment yielding benefits for patients and carers More efficient MAP (improved waiting times, accuracy and PwMCI follow-up). NHS cost savings (projected at £86.3million over 5 years).
Aims
This research aims to advance the integration of CognoSpeakTM into MAP in order to improve detection and monitoring of cognitive impairment (CI) for all. The project will deliver a market-ready CognoSpeakTM tool with UKCA Class I approval that has been co-designed with MAP-clinicians and a range of end-user groups. The main objectives are: Demonstration of feasibility of stratification in primary care using CognoSpeakTM. Co-design of CognoSpeakTM with end-users to ensure no cultural or language barriers. Demonstration of feasibility of using CognoSpeakTM with underrepresented groups. Development of AI-algorithms agnostic to language background. Tripling of data resource to allow validation in people with memory problems (primary care), ethnic minority, and in PwMCI (secondary care). Production of a health economic (HE) model to aid commercialisation and NHS implementation pathways. Delivery of a trial platform to demonstrate HE effectiveness of CognoSpeakTM. Methods Phased research plan with two cycles of software development, AI refinement and PPIE feedback. Determination of tool accuracy with sensitivity and specificity using detailed neuropsychology, structural brain imaging and clinical assessment. Regular co-production workshops with patients and families, ethnic minority groups and clinicians will improve the tool ensuring usability for all and an accessible and acceptable reporting system.