Can a speech-based tool improve early detection of memory problems in the NHS?
| ISRCTN | ISRCTN82732915 |
|---|---|
| DOI | https://doi.org/10.1186/ISRCTN82732915 |
| Integrated Research Application System (IRAS) | 354094 |
| Central Portfolio Management System (CPMS) | 72001 |
| National Institute for Health and Care Research (NIHR) | 510914 |
| UK Science and Technology Facilities Council (STFC)-UK Research and Innovation (UKRI) grant number | NHSPOC01 |
| Sponsor | Punto Health Ltd |
| Funders | National Institute for Health and Care Research, UK Research and Innovation, Invention for Innovation Programme, Science and Technology Facilities Council |
- Submission date
- 10/08/2026
- Registration date
- 28/08/2026
- Last edited
- 28/08/2026
- Recruitment status
- Not yet recruiting
- Overall study status
- Ongoing
- Condition category
- Nervous System Diseases
Plain English summary of protocol
Background and study aims
Nearly one million people in the UK have dementia, but only about two-thirds have been formally diagnosed. Early detection of memory problems is difficult in routine practice, and waiting times for specialist assessment are often very long — often exceeding 13 weeks when the recommended benchmark is 6 weeks. This puts pressure on NHS memory services, and delays care.
PuntoTest is a simple speech-based tool using artificial intelligence to screen for memory problems. It can be used by patients themselves or with support from a clinician, offering a flexible and accessible approach to early detection. The tool has already been tested in other countries with promising results. This study tests whether it works well in NHS memory clinics and GP practices, and whether it's practical to use in the UK.
Who can participate?
Patients aged 50 years or over who have concerns about their memory, mild memory problems, mild dementia, or no memory concerns (as a healthy comparison volunteer) and have had a recent memory assessment (ACE-III within the last 3 months), or can do this as part of the study
What does the study involve?
The study involves a single visit lasting about 30 minutes. Participants will answer a few basic questions about themselves at the start, then speak into a tablet to complete a simple memory screening tool, and also take a standard memory test (ACE-III) so we can compare results. No additional visits needed.
What are the possible benefits and risks of participating?
Benefits: Participation will help us understand whether PuntoTest could support earlier and more accurate memory assessments in the NHS. This could benefit many people with memory concerns in the future and help improve the way care is delivered. Taking part may not directly benefit participants personally. However, the ACE-III assessment, which is completed during the same visit, is already part of the standard care.
Risks: Minimal. There are no known medical risks associated with taking part. The assessment involves speaking into a tablet and completing a standard memory test. Both are low-risk activities. Some people may find a memory assessment slightly tiring or stressful, or may feel worried about the results. If participants find it difficult at any point, they can speak to the research team. Participants are free to stop at any time.
Where is the study run from?
The study is sponsored by PuntoHealth Ltd and conducted across four NHS sites in England.
When is the study starting and how long is it expected to run for?
The study is expected to start in September 2026, depending on site setup. Enrolment of participants is planned over approximately 12 months (from September 2026 to July 2027). Data analysis is anticipated to be completed by December 2027, although this date is indicative and may change depending on recruitment progress and data collection completion.
Who is funding the study?
1. The National Institute for Health and Care Research (NIHR) Invention for Innovation (i4i) FAST Programme.
2. The UK Science and Technology Facilities Council (STFC)-UK Research and Innovation (UKRI).
Who is the main contact?
Maria Rollano Corroto, Project Manager, PuntoHealth Ltd, maria@puntohealth.com
Contact information
Scientific, Public
Project Manager, Punto Health Ltd
Health Foundry, 1 Royal Street
London
SE1 7LL
United Kingdom
 ORCID ID |
0009-0005-8185-4474 |
|---|---|
| maria@puntohealth.com |
Public, Scientific
CEO & Co-founder, Punto Health Ltd
Health Foundry, 1 Royal Street
London
SE1 7LL
United Kingdom
| ORCID ID |
0009-0004-0250-287X |
|---|---|
| anna@puntohealth.com |
Principal investigator
Clinical Reader in Ageing and Dementia Studies, Consultant Old Age Psychiatrist
King’s College London and South London & Maudsley NHS Foundation Trust, 16 De Crespigny Park
London
SE5 8AB
United Kingdom
| ORCID ID |
0000-0002-7712-930X |
|---|---|
| latha.velayudhan@kcl.ac.uk |
Study information
| Primary study design | Interventional |
|---|---|
| Allocation | Non-randomized controlled trial |
| Masking | Open (masking not used) |
| Control | Active |
| Assignment | Single |
| Purpose | Screening |
| Scientific title | PuntoTest: speech-based AI cognitive screening in NHS primary and secondary care |
| Study acronym | PuntoTest |
| Study objectives | Primary Objective To assess the feasibility of implementing PuntoTest within NHS primary and secondary care settings. This includes: 1. The ability to recruit the target number of participants across sites 2. Participant completion rates of the PuntoTest assessment 3. The feasibility of recruitment pathways from both primary and secondary care Secondary Objectives 1. To compare performance of PuntoTest with current gold standard cognitive assessment, ACE-III (Addenbrooke's Cognitive Examination-III) and further develop the model. 2. To assess variation in performance across demographic groups, including age, sex, and education 3. To explore participant experience and acceptability of the PuntoTest assessment through qualitative interviews 4. To generate preliminary health economics data to inform future NHS commissioning decisions and grant applications |
| Ethics approval(s) |
Approved 04/08/2026, London - Stanmore Research Ethics Committee (2 Redman Place, Stratford, London, E20 1JQ, United Kingdom; -; stanmore.rec@hra.nhs.uk), ref: 26/LO/0440 |
| Health condition(s) or problem(s) studied | Early detection of memory problems |
| Intervention | Study Design This is a multi-site feasibility study designed to assess the implementation of PuntoTest within NHS primary and secondary care settings. This includes embedded health economic assessments, as well as exploratory evaluation of performance against the current gold standard. What Happens to Participants Eligible participants are identified by clinical staff at each site as part of their routine care pathway. The Patient Information Sheet (PIS) is provided in advance by the treating clinician or GP, allowing participants at least 24 hours to consider participation and ask questions. If they are interested, a study visit is arranged, where a trained member of the research team will explain the study and obtain written informed consent prior to any study procedures. At the clinic or GP visit: 1. The research nurse answers any questions and takes written informed consent (15-30 minutes). 2. The participant completes the 10-minute PuntoTest assessment on an iPad in the clinic room, which includes a brief collection of demographic information (such as age, sex, and education level) followed by the speech-based assessment. 3. The standard ACE-III assessment will then be administered by the clinician as part of routine care. No additional appointments are required. At the GP site (Greenwich), Dr James Rogers will be trained in ACE-III administration for this study. 4. A subset of participants will be invited to take part in a qualitative interview, either in the form of an individual interview or a focus group. These will be conducted by a qualitative PPIE expert from UCLP. Depending on the number of participants who opt in to participate. These will be delivered by either video call or in person, to explore their experience of using PuntoTest. They will be compensated £25 per hour for this additional optional qualitative component. Number of Participants and Sites - NHS participants: approximately 150 (spread across Oxleas Memory Service, Enfield Memory Service, Lewisham Memory Service, and Greenwich GP Practice) Patient and Public Involvement A lived experience advisory panel were involved in the review of the protocol and patient consent materials, as part of a series of PPIE workshops led by UCLPartners. Select members from this panel will also be invited to sit on the Steering Committee. Co-design workshops will be held during study delivery to inform ongoing participant engagement and materials. |
| Intervention type | Device |
| Phase | Not Applicable |
| Drug / device / biological / vaccine name(s) | PuntoTest |
| Primary outcome measure(s) |
|
| Key secondary outcome measure(s) |
|
| Completion date | 17/12/2027 |
Eligibility
| Participant type(s) | |
|---|---|
| Age group | Mixed |
| Lower age limit | 50 Years |
| Upper age limit | 120 Years |
| Sex | All |
| Target sample size at registration | 150 |
| Key inclusion criteria | 1. Adults aged > = 50 years 2. Capacity to provide informed consent 3. Individuals in one of the following groups, as determined through routine clinical assessment and application of the following criteria: 3.1. Subjective Cognitive Decline (SCD; also referred to as Subjective Memory Complaints, SMC). The following criteria will be used to define SCD (Jessen et al., 2014): 3.1.1. Self-experienced persistent decline in cognitive capacity compared with a previously normal status 3.1.2. Normal performance on standardised cognitive assessments 3.1.3. No significant impairment in activities of daily living 3.1.4. Absence of mild cognitive impairment or dementia 3.2. Mild Cognitive Impairment (MCI). An individual meeting the following diagnostic criteria will be identified as having MCI (Albert et al., 2011): 3.2.1. Concern regarding a change in cognition, whether reported by the self, informant, or clinician. 3.2.2. Objective evidence of impairment in one or more cognitive domains, typically including memory. 3.2.3. Preservation of independence in functional abilities. 3.2.4. Absence of dementia. 4. Mild dementia. An individual meeting the following diagnostic criteria will be identified as having mild dementia (DSM-5, 2013): 4.1. Evidence of cognitive decline from a previous level of performance, reported by the individual, an informant, or a clinician 4.2. Objective impairment in one or more cognitive domains (e.g. memory, attention, language, executive function) 4.3. Interference with independence in everyday activities (e.g. requiring assistance with complex tasks) 4.4. Not better explained by delirium or another mental disorder 5. Healthy controls. An individual meeting the following diagnostic criteria will be identified as a healthy control: 5.1. No self-reported cognitive concerns 5.2. Normal performance on a standardised cognitive assessment, defined as a score of > = 27 on the Mini-Mental State Examination (MMSE) (Molloy & Standish, 1997\) 5.3. No clinical diagnosis of mild cognitive impairment or dementia 5.4. Independent in activities of daily living 6. Availability of a recent ACE-III assessment (within the previous 3 months), or ability to complete the assessment as part of the study procedures 7. Sufficient proficiency in English to complete the study procedures and digital assessment |
| Key exclusion criteria | 1. Moderate or severe dementia, or any condition associated with lack of capacity to provide informed consent 2. Inability to complete the ACE-III assessment required for study participation 3. Severe sensory, motor, or cognitive impairments that would prevent completion of the study procedures, including use of the digital assessment tools, even with support 4. Acute or unstable physical or mental health conditions that, in the opinion of the clinical team, would make participation inappropriate |
| Date of first enrolment | 28/09/2026 |
| Date of final enrolment | 31/07/2027 |
Locations
Countries of recruitment
- United Kingdom
- England
Study participating centres
London
SE18 3RG
England
London
EN2 8JL
England
London
SE13 7DW
England
London
SE28 8BE
England
Results and Publications
| Individual participant data (IPD) Intention to share | No |
|---|
Editorial Notes
10/08/2026: Study's existence confirmed by the National Institute for Health and Care Research (NIHR) (UK).
