Hostname: page-component-54dcc4c588-54gsr Total loading time: 0 Render date: 2025-09-25T00:54:29.117Z Has data issue: false hasContentIssue false

Group cognitive stimulation therapy for people with intellectual disability and dementia: feasibility randomised controlled trial

Published online by Cambridge University Press:  01 August 2025

Afia Ali*
Affiliation:
Wolfson Institute of Population Health, Queen Mary University of London, London, UK
Cheryl Francis
Affiliation:
Department of Clinical, Educational & Health Psychology, University College London, London, UK
Sarah Hoare
Affiliation:
Department of Clinical, Educational & Health Psychology, University College London, London, UK
Joanna Carter
Affiliation:
Department of Clinical, Educational & Health Psychology, University College London, London, UK
Nia Goulden
Affiliation:
North Wales Organisation for Randomised Trials in Health and Social Care, Bangor University, Bangor, UK
Caroline S. Clarke
Affiliation:
Research Department of Primary Care and Population Health, University College London, London, UK
Georgina Charlesworth
Affiliation:
Department of Clinical, Educational & Health Psychology, University College London, London, UK Research and Development, North East London NHS Foundation Trust, London, UK
Zoe Hoare
Affiliation:
North Wales Organisation for Randomised Trials in Health and Social Care, Bangor University, Bangor, UK
Danny Acton
Affiliation:
Millenium Centre, Cheshire and Wirral Partnership NHS Foundation Trust, Chester, UK
Shafia Khanum
Affiliation:
Services for People with Learning Disability, East London NHS Foundation Trust, London, UK
Akinwande Onafalujo
Affiliation:
Services for People with Learning Disability, East London NHS Foundation Trust, London, UK
Adebayo Jejeloye
Affiliation:
Services for People with Learning Disability, East London NHS Foundation Trust, London, UK
Kate Brackley
Affiliation:
British Institute of Learning Disabilities, Birmingham Research Park, Edgbaston, UK
Elisa Aguirre
Affiliation:
Psychology Department, Universidad Europea de Madrid, Madrid, Spain
Aimee Spector
Affiliation:
Department of Clinical, Educational & Health Psychology, University College London, London, UK
*
Correspondence: Afia Ali. Email: afia.ali@qmul.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

Background

Group cognitive stimulation therapy (CST) has been shown to improve cognition and quality of life of people with dementia in multiple trials, but there has been scant research involving people with intellectual disability and dementia. This study aimed to assess the feasibility of conducting a randomised controlled trial of group CST for this population.

Aims

To assess the feasibility of participant recruitment and retention, the appropriateness of outcome measures, and the feasibility of group CST (adherence, fidelity, acceptability), as well as the feasibility of collecting data for an economic evaluation.

Method

Participants were recruited from six National Health Service trusts in England and randomised to group CST plus treatment as usual (TAU) or TAU only. Cognition, quality of life, depression, and use of health and social care services were measured at baseline and at 8–9 weeks. Qualitative interviews with participants, carers and facilitators were used to explore facilitators of and barriers to delivery of CST. Trial registration number: ISRCTN88614460.

Results

We obtained consent from 46 participants, and 34 (73.9%) were randomised: 18 to CST and 16 to TAU. All randomised participants completed follow-up. Completion rates of outcome measures (including health economic measures) were adequate; 75.7% of sessions were delivered, and 56% of participants attended ten or more. Fidelity of delivery was of moderate quality. CST was acceptable to all stakeholders; barriers included travel distance, carer availability and sessions needing further adaptations. The estimated cost per participant of delivering CST was £602.

Conclusions

There were multiple challenges including recruitment issues, a large dropout rate before randomisation and practical issues affecting attendance. These issues would need to be addressed before conducting a larger trial.

Information

Type
Paper
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2025. Published by Cambridge University Press on behalf of Royal College of Psychiatrists

People with intellectual disability are at increased risk of developing dementia, often with earlier onset, compared with the general population. Reference Strydom, Chan, King, Hassiotis and Livingston1,Reference MacDonald and Summers2 The consequences of dementia in people with intellectual disability include faster progression and elevated mortality rates. Reference Strydom, Chan, King, Hassiotis and Livingston1 The National Institute for Health and Care Excellence (NICE) guidelines for dementia (NG97) emphasise person-centred care and psychosocial interventions. 3 People with intellectual disability are often excluded from person-centred care initiatives, and this extends to the management of dementia. Reference Hahn, Fox and Janicki4Reference Watchman6 Two published reviews of psychosocial interventions for individuals with intellectual disability and dementia found a limited number of direct therapeutic interventions with the individual, with interventions being largely carried out with carers. Reference MacDonald and Summers2,Reference Muralidhar, Spector, Hui, Liu and Ali7

Cognitive stimulation therapy (CST) is a NICE-recommended manualised psychosocial intervention for individuals with mild to moderate dementia. 3 It is a 14-session group intervention incorporating 18 key principles including mental stimulation and person-centredness, with each session covering a different theme (e.g. physical games, childhood and food). Clinical trials worldwide have shown consistent improvements in a range of outcomes including cognition, language, working memory, depression, communication, neuropsychiatric symptoms and quality of life. Reference Desai, Leung, Fearn, John, Stott and Spector8 However, the presence of intellectual disability has been an exclusion criterion for the majority of published CST trials. Reference Spector, Thorgrimsen, Woods, Royan, Davies and Butterworth9

Our team previously evaluated individualised CST in 40 individuals with dementia and intellectual disability and concluded that it was feasible and acceptable, with improved quality of life at 21 weeks but no changes in adaptive functioning or cognition. Reference Ali, Brown, Tsang, Spector, Aguirre and Hoare10 However, significant barriers were identified, including some activities being unsuitable for individual needs or cognitive ability and carers struggling to deliver individualised CST within their busy work schedules. There have been no randomised controlled trials (RCTs) of group CST in people with intellectual disability and dementia. As group CST is recommended by NICE for people with dementia, 3 exploration of group CST for those with intellectual disability is warranted.

In this study, we aimed to assess the feasibility and acceptability of group CST for people with intellectual disability and dementia compared with treatment as usual (TAU), with a view to informing the design of a future definitive RCT. The objectives were to assess: (a) feasibility of recruitment and retention of participants; (b) suitability of outcome measures; (c) feasibility of the CST intervention (acceptability, adherence and fidelity, and serious adverse events); and (d) feasibility of collecting data to inform a future health economic analysis. The study protocol has been published and provides further information about our methods and analysis. Reference Ali, Aguirre, Carter, Hoare, Brackley and Goulden11

Method

Design

This was a single-blind feasibility RCT with a qualitative evaluation. The study was randomised to enable us to assess whether participants were willing to take part in a randomised study and to test randomisation procedures before a full RCT. No formal power calculation was conducted. We set a target sample size of 50 to achieve adequate precision around our expected retention rate of 75% (95% CI: 62 to 86%). The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2013. All procedures involving human subjects/patients were approved by the East of England – Essex Research Ethics Committee (reference number: 21/EE/02/47).

Recruitment and procedures

Participants were recruited from community intellectual (‘learning’) disability teams from six National Health Service (NHS) trusts in England. To recruit to the study, professionals screened case-loads for possible participants and approached them via their carer. If interested in participating, potential participants were provided with an information sheet, and a researcher subsequently contacted the individual and their carer to discuss the study and to assess capacity to consent to taking part. If the individual agreed to participate, written consent was obtained. If they lacked capacity, a relative or friend (personal consultee) was approached, provided with a consultee information sheet and asked to sign a declaration form. If the individual did not have a personal consultee, a nominated consultee (e.g. a clinician not involved in the study) was approached instead. Carers completing proxy and/or informant measures were asked to provide consent to take part in the study. A demographic questionnaire was completed at baseline and included information about the participant’s age, gender, ethnicity, level of intellectual disability, severity and type of dementia, living arrangements, comorbid medical conditions and medication. Information about the carer’s age, gender, ethnicity and relationship or role was also collected. If a different carer was present at the follow-up appointment, their details were recorded at that appointment.

The inclusion criteria were: (a) premorbid mild or moderate intellectual disability (based on service records); (b) age 18 years or over; (c) clinical diagnosis of mild or moderate dementia (based on clinical notes); (d) ability to provide informed consent or (if the participant lacked capacity) availability of a personal consultee who has agreed for the individual to participate; (e) ability to communicate in English. The exclusion criteria were: significant visual or hearing impairment, physical illness or disability, or significant behavioural problems (e.g. aggressive behaviour) that could affect participation.

Randomisation and blinding

Participants were randomised at each site with a minimum of five participants. Randomisation was undertaken by the coordinating trials unit (North Wales Organisation for Randomised Trials in Health and Social Care) using a dynamic adaptive algorithm via a secure online interface. Randomisation was stratified by site or centre using a 1:1 allocation ratio. Participants and carers were informed of their group allocation by an unblinded researcher. Research assistants conducting quantitative follow-up assessments were blind to group allocation.

CST intervention arm

Participants in the intervention arm were expected to receive 14 face-to-face sessions delivered over 7 weeks. The target group size was between three and five participants, that is, the groups were smaller than those used for CST in people without intellectual disability. This was because people with intellectual disability and dementia are more likely to need support and reasonable adjustments to enable them to participate owing to communication and sensory impairments. Groups were facilitated by health professionals from the participating community intellectual disability team. These professionals had to have experience of working with people with intellectual disability and included psychologists, psychiatrists, occupational therapists and nurses. The groups took place within clinical services (and not at participants’ usual day care setting). Facilitators received standard CST training (1 day) followed by an additional session on adapting CST for this population. They used the CST treatment manual with a supplement outlining adaptations for individuals with intellectual disability, which was developed for the study. Examples of adaptations included simplification of activities, replacement of words with pictures where possible and activities that provided multisensory stimulation. The intervention group continued to have access to TAU (see below).

TAU arm

Participants allocated to the TAU arm received their usual care, including support from psychiatrists, nurses and psychologists; access to day care; and medication including antidementia (cognitive-enhancing) medication (e.g. acetylcholinesterase inhibitors).

Outcome measures

Outcome assessments were conducted at baseline (within 2 weeks before randomisation) and follow-up (8–9 weeks post-randomisation) via face-to-face assessments. The following measures were completed:

  1. (a) the Severe Impairment Battery (SIB) to assess cognitive functioning; Reference Saxton, McGonigle, Swihart and Boller12

  2. (b) the Dementia Questionnaire for people with Learning Disabilities (DLD), an informant measure used to assess cognitive and social functioning; Reference Evenhuis, Kengen, Eurlings and Prasher13

  3. (c) the Quality of life in Alzheimer’s Disease (QOL-AD) – proxy version, an informant measure of the participant’s quality of life; Reference Logsdon, Gibbons, McCurry and Teri14

  4. (d) the Glasgow Depression Scale for People with Learning Disability (GDS-LD) proxy version, an informant measure used to assess the presence of depressive symptoms; Reference Cuthill, Espie and Cooper15

  5. (e) EQ-5D quality of life questionnaires: proxy version, Reference Herdman, Gudex, Lloyd, Janssen, Kind and Parkin16 completed by carers on behalf of participants; and EQ-5D-5L self-report, completed by participants in the initial phases of the study and replaced partway with a modified version of EQ-5D-3L for adults with intellectual disability; Reference O’Dwyer, Meads, Bryant, Kind and Hulme17,Reference O’Dwyer, Bryant, Hulme, Kind and Meads18

  6. (f) a modified version of the Client Service Receipt Inventory (CSRI), tailored to this patient group and context in collaboration with patients, parents and carers, to collect information about health and social care resources. Reference Beecham, Knapp and Thornicroft19

Feasibility outcomes

The feasibility outcomes were assessed according to ‘Go/Review/Stop criteria’, which were used to determine the success of the trial and whether a future definitive RCT was feasible. The specific criteria are listed in Table 1.

Table 1 Stop/Go criteria for the feasibility outcomes

CST, cognitive stimulation therapy.

Recruitment and retention were assessed using the following criteria:

  1. (a) adequate recruitment, defined as the number of participants recruited to the trial;

  2. (b) eligibility rate, defined as the percentage of participants who met the eligibility criteria from those referred;

  3. (c) consent/recruitment rate, defined as the percentage of participants recruited to the trial of those eligible;

  4. (d) retention, defined as the number of participants completing the trial (follow-up assessment) from those recruited.

The suitability of study outcome measures was assessed on the basis of:

  1. (a) completion rates of outcome measures, defined as the proportion of participants completing outcome measures at baseline and follow-up; and

  2. (b) whether outcome measures were sensitive to change in the population.

The feasibility of the CST intervention was assessed as follows.

  1. (a) Adherence of CST: overall attendance among CST group participants was determined on the basis of group attendance registers completed by facilitators and percentages of sessions delivered by facilitators according to session logs.

  2. (b) Fidelity of intervention delivery: fidelity was measured using a previously developed fidelity checklist, adapted for the study and based on CST principles and core intervention components. Reference Hui, Tischler, Wong and Gibbor L, Spector20 A four-point Likert scale was used to rate the extent to which each CST principle had been successfully incorporated into the session. The five intervention components (orientation, current affairs activity, main activity, use of themed therapy resources and obtaining feedback) were rated on a binary scale (‘1’ if a component had been completed and ‘0’ if it had not). The fidelity checklist was completed by group facilitators after each session. An independent observer listened to 50% of the available audio recordings of sessions and rated the sessions using the same checklist. A total fidelity score and a percentage fidelity score (obtained fidelity score/maximum checklist score (minus items that were non-applicable) × 100%) was calculated for each session for both facilitator and observer ratings. A percentage score of 80–100% adherence to the fidelity checklist was considered to indicate high fidelity, whereas 51–79% indicated moderate fidelity, and 50% or less indicated low fidelity. Reference Borrelli21

  3. (c) Acceptability of CST: semi-structured interviews were conducted with group participants, carers and facilitators to assess acceptability and satisfaction with the intervention and to understand some of the barriers and enablers affecting feasibility outcomes such as retention and adherence. All group participants and carers and group facilitators were eligible for participation in qualitative interviews. Group participants were interviewed if available on the day of one of the final two sessions (to maximise likelihood of recall of the group experience). A Talking Mats approach was employed for participants with intellectual disability, as this is appropriate for use with individuals with a basic level of communication. Reference Murphy, Gray and Cox22,Reference Stans, Dalemans, De Witte and Beurskens23 Symbols and/or images were used for each question and placed on a visual scale to represent the following responses: ‘Like’, ‘Unsure’ or ‘Don’t Like’. Participants were asked about their views relating to the activities, frequency and duration of sessions, and being with others; carers were asked about positive and negative consequences of attending groups for the participant; and facilitators were asked about their experiences of preparing and delivering the sessions. Carers and facilitators were also asked to provide suggestions for improvements. Interviews were audio-recorded. Carer and facilitator interviews (up to 1 h) took place remotely and were recorded via Microsoft Teams. All the interviews were transcribed verbatim.

  4. (d) Adverse events: numbers of serious adverse events during the study and whether they were related to the intervention were recorded.

Health-related quality of life was assessed using the EQ-5D-5L proxy Reference Herdman, Gudex, Lloyd, Janssen, Kind and Parkin16 at baseline and at follow-up, completed by the carer on behalf of the participant. The same questionnaire was administered to participants for them to self-complete during the early part of the study. Feedback suggested that this measure was hard for participants to complete; therefore, it was changed to the modified EQ-5D-3L for adults with learning/intellectual disabilities, which became available for piloting during the study. Reference O’Dwyer, Meads, Bryant, Kind and Hulme17,Reference O’Dwyer, Bryant, Hulme, Kind and Meads18 This measure has been specially adapted for use with individuals with intellectual disability (simplified text and easy-to-read images) with permission of the EuroQoL group. All participants were asked to complete the same measure at both time points; that is, they were asked either to complete the EQ-5D-5L at both time points, or to complete the modified EQ-5D-3L at both time points. To examine the feasibility of collecting healthcare and social care resource use, a modified CSRI Reference Beecham, Knapp and Thornicroft19 questionnaire was used. This captured use of services related to the intervention, TAU in both arms, and other treatment pathway costs, including primary, community and hospital care, medications, and use of social services (funded by government or privately). The cost of the group CST intervention was estimated from information provided by staff at sites, which included staff time for training, facilitators running the session and resources purchased for the groups. As this was a feasibility study, we did not calculate overall costs or quality-adjusted life-years, and no formal comparisons were made between the groups; the purpose was to inform the design of a future definitive RCT and not to perform a cost-effectiveness analysis.

Statistical analysis

Statistical analysis was conducted using Stata version 16 for Windows. Participant flow data were analysed, and values for eligibility, recruitment, attrition and withdrawal rates are presented, including reasons for ineligibility and withdrawals. Descriptive statistics were used to describe the data (e.g. mean and standard deviation for continuous data and counts and percentages for categorical data).

Outcome data were analysed using analysis of covariance. Adjusted means (adjusted for baseline values) were obtained, and 95% confidence intervals were calculated. Intraclass correlation coefficients were calculated to take into account the effects of clustering, which would need to be adjusted for in a sample size calculation.

Interrater agreement between the facilitator- and observer-reported fidelity scores was calculated using percentage agreement and the weighted kappa statistic (κ). We used thresholds of κ > 0.4 for moderate and κ > 0.6 for good interrater reliability. Reference Walton, Spector, Williamson, Tombor and Michie24

Qualitative analysis

A thematic analysis was used to analyse data from facilitator and carer interviews. This analysis method combined inductive and deductive coding approaches Reference Fereday and Muir-Cochrane25,Reference Crabtree, Miller, Crabtree and Miller26 to identify relevant themes related to acceptability, as well as barriers to and facilitators of the intervention. Participant interviews were analysed using basic content analysis. Reference Weber27

Changes from the published protocol

There were a few changes to the analysis following publication of the study protocol. Reference Ali, Aguirre, Carter, Hoare, Brackley and Goulden11 In our results, we now include the percentage of sessions delivered by facilitators. The assessment of fidelity was amended to include coverage of CST principles as well as intervention components; therefore, the criteria for assessment were updated to reflect this change. Owing to time constraints, we were unable to interview participants and carers in the control arm.

Results

Participant characteristics

In the whole sample, the mean age of participants was 61.9 years (s.d. 8.5), 61.8% were male, and most were of White British ethnicity (88.2%). Nineteen (55.9%) had a moderate learning disability, 55.9% had Alzheimer’s dementia, 50% had mild dementia and half were taking antidementia medication. Most lived in 24-h supported housing (76.5%). The carers who completed baseline measures were mostly paid carers (85.2%) with a mean age of 45.3 years (s.d. 12.9), and were mostly female (70.4%) and of White English British ethnicity (55.6%). Details of the sample’s demographics according to group allocation are presented in Table 2. The CST arm also had more male individuals (66.7 v. 56.3%) and more people of Black and minority ethnic origin (16.7 v. 6.2%).

Table 2 Participant demographics at baseline in the intervention (CST) and TAU arms

CST, cognitive stimulation therapy; TAU, treatment as usual; INT, intervention.

Feasibility outcomes

Recruitment and retention of participants

Recruitment spanned from 1 May 2022 to 31 October 2023 (17 months); this was 5 months longer than initially planned owing to delays in opening sites and the need for multiple visits and telephone calls to enable each participant and/or their consultee to make a decision regarding participation in the study. Figure 1 shows the flow of participants through the trial. Of the 82 participants who were approached, 81 were screened and 60 were eligible. Of the 21 who were ineligible, six could not communicate in English, five had significant physical illness or disability, and 14 did not consent. Of the 60 eligible participants, 46 consented and 41 completed the baseline assessment.

Fig. 1 Recruitment and retention for the trial. TAU, treatment as usual.

Of the 41 participants who completed the baseline assessment, only 34 were randomised. Five participants withdrew before baseline completion, and seven withdrew before randomisation (12 in total). At one site, three individuals withdrew following significant delays between recruitment and the CST group, as the latter was postponed owing to a heat wave. Other reasons for withdrawal included admission to hospital (n = 1), death (n = 1), deterioration in dementia (n = 1), being unable to complete baseline assessment (n = 1), an insufficient number of participants for randomisation at one site (n = 2) and concerns about attending groups (n = 2). Of the 34 participants who were randomised, 18 were allocated to the intervention arm and 16 to TAU. There were no withdrawals post-randomisation, and all the randomised participants completed their 8–9-week follow-up.

Regarding outcome assessments, Table 3 shows the completion rates for the SIB total score, DLD, QOL-AD proxy and GDS-LD proxy outcome measures; these rates were good (less than 5% missing data) except in the case of the QOL-AD proxy, for which almost one-third of participants did not complete one item (the question about marriage, which was not applicable). Table 4 shows the scores at baseline and follow-up for the four listed outcome measures, and Table 5 shows the adjusted means. The DLD (adjusted mean −3.52; 95% CI −19.19 to 10.68) and the QOL-AD proxy (adjusted mean 1.41; 95% CI −2.44 to 5.27) showed positive changes in favour of the intervention group, whereas the results of the SIB and GDS-Proxy were in favour of the control arm. The intraclass correlation values for the GDS-LD proxy and QoL-AD proxy were negligible (<0.001), although those for the DLD (0.098) and SIB (0.29) were larger.

Table 3 Missing data

TAU, treatment as usual; INT, intervention; SIB, Severe Impairment Battery; DLD, Dementia Questionnaire for People with Learning Disabilities; QOL-AD, Quality of life in Alzheimer’s Disease; GDS-LD, Glasgow Depression Scale for People with Learning Disability.

Table 4 Raw scores for SIB, DLD, QOL-AD proxy and GDS-LD proxy outcome measures

TAU, treatment as usual; INT, intervention; SIB, Severe Impairment Battery; IQR, interquartile range; DLD, Dementia Questionnaire for People with Learning Disabilities; QOL-AD, Quality of life in Alzheimer’s Disease; GDS-LD, Glasgow Depression Scale for People with Learning Disability.

Table 5 Adjusted means for outcome measures from analysis of covariance

TAU, treatment as usual; INT, intervention; SIB, Severe Impairment Battery; DLD, Dementia Questionnaire for People with Learning Disabilities; QOL-AD, Quality of life in Alzheimer’s Disease; GDS-LD, Glasgow Depression Scale for People with Learning Disability.

CST group delivery and adherence

All the sites chose to deliver two sessions in 1 day rather than over 2 days a week. Four CST groups were delivered in total. Of these, two groups had 12 sessions delivered, one had 13 sessions and two had 14 sessions. Group sizes ranged from two to five participants. Eighteen participants were allocated to the CST arm, but two of these did not receive the intervention because there were not enough participants at one site for the group to be run. Of the possible total of 214 available sessions, 137 sessions were attended by the 16 participants, representing an attendance rate of 64.0%. Two participants who were offered CST groups did not attend any sessions, and nine (56%) attended ten or more sessions.

Assessment of intervention fidelity

Fidelity ratings and audio recordings were available for three of four groups. Facilitators from two groups completed the maximum of 14 ratings, and the other completed 13. Only 27 audio recordings were available (12, ten and five, respectively, from each group). The average total fidelity score rated by facilitators was 26.3 (of a possible total of 38), with a mean percentage fidelity score of 74.3%. Fifteen audio recordings were rated. The mean observer rated fidelity score was 25.4, and the percentage fidelity score was 75.9%. The agreement between facilitator and observer ratings was good at 80.6%, with an average κ of 0.79.

Acceptability of CST: qualitative findings

Twenty semi-structured interviews were conducted with nine facilitators, six carers and five participants (three males and two females, all White British, four with mild intellectual disability and one with moderate intellectual ability). The findings from the analysis of facilitator and carer interviews are presented in Table 6, summarised according to the following themes: group attendance, positive and negative experiences and impact of CST, barriers and enablers to participation and CST content. In summary, group attendance was affected by carer availability and access to transportation, as many participants had to travel far; positive benefits included social interaction and improvements in memory, communication and sleep, but participants became fatigued due to two sessions being held in 1 day. The main enablers were facilitator skill and ability to further adapt sessions and carer involvement; the main barriers were distance and travel time for participants and the amount of time needed to prepare for sessions. Findings from the content analysis of participant interviews are summarised in Table 7. Participants were largely positive about the sessions, although one participant indicated that the sessions were too long. Facilitators and carers were asked to make suggestions for improvements to the delivery and content of CST, and these are summarised in Table 8.

Table 6 Themes from the qualitative interviews with facilitators and carers

CST, cognitive stimulation therapy; CST-IDD, CST for people with intellectual disability and dementia.

Table 7 Basic content analysis of participant interviews

P, participant.

Table 8 Summary of recommendations for future groups from facilitator and carer interviews

Adverse events

There were six serious adverse events during the trial (two deaths and four hospital admissions), which were all unrelated to trial participation.

Health economics data completeness results

Analysis of health economic data was conducted using StataNow release 18.5 for Windows. All 18 participants in the TAU arm and 15 of 16 participants in the CST arm completed the health-related quality of life questionnaires that were administered to them. The EQ-5D-5L proxy questionnaire completed by carers had no missing items, whereas there were a few missing items for the self-reported measures: EQ-5D-5L (up to three missing items per arm) and the new modified EQ-5D-3L for adults with intellectual disability (up to five missing items per arm). As the sample sizes were small, no conclusions could be drawn regarding the different levels of completion of the different measures. The item responses, including the visual analogue scale score, and the calculated utility scores can be found in Supplementary Tables 13 available at https://doi.org/10.1192/bjo.2025.10764. In the modified CSRI, most participants reported some general practitioner visits, and smaller numbers reported seeing a psychiatrist, a learning disabilities nurse or professionals from the other categories that were listed in the questionnaire (Supplementary Tables 4 and 5). Information on reported use of prescription and over-the-counter medications is given in Supplementary Tables 6 and 7.

Cost of CST-IDD group sessions

The total cost of running the sessions, including staff time for training, staff time for delivering CST, manual purchase and consumables costs, was calculated as the sum of these items (see Tables 9 and 10 and Supplementary Table 8). The total was calculated to be £9636 across all groups and participants, corresponding to £2409 per group (four groups) and £602 per participant (16 participants).

Table 9 Costs of delivering cognitive stimulation therapy

Table 10 Intervention cost, calculated as the sum of staff time spent on training and on session preparation and delivery, and the expenditure on resources for the sessions (from Supplementary Table 8)

PSSRU, Personal Social Services Research Unit.

Progression criteria

Table 11 summarises the feasibility outcomes and the extent to which the progression criteria were met. The only feasibility outcomes to meet the ‘Go’ criteria were recruitment (adequate recruitment, and recruitment rate), percentage of sessions delivered by facilitators and serious adverse events. All the other outcomes, including retention, group attendance and acceptability, were in the ‘Review’ category.

Table 11 Summary of the feasibility outcomes and whether they met progression criteria

CST, cognitive stimulation therapy; QOL-AD, Quality of life in Alzheimer’s Disease.

Discussion

Summary of findings

This was the first feasibility RCT of group CST in people with intellectual disability and dementia. We achieved the ‘Go’ criteria for recruitment (total number of participants recruited and consent/recruitment rate). However, a longer period of recruitment than planned was needed (17 months instead of 12 months), and 12 participants dropped out following consent, with only 34 randomised. Reasons included one group having to be postponed owing to an extreme heat wave. There were often several months between participants providing informed consent and the baseline assessment being conducted, as groups could not go ahead until the minimum number of participants had been recruited. The large attrition rate between consent and randomisation affected the retention rate, which was just below our target of 75% (‘Review’ category).

There was an indication that the results from the DLD and QOL-AD proxy favoured the intervention group, whereas those from other measures, the SIB and GDS proxy, favoured the control group; however, none of these differences was significant. These results need to be interpreted with caution given the small sample size. However, all measures appeared to be appropriate for this population.

Although most of the sessions were delivered (‘Go’), group attendance was poor, with only 56% of participants attending ten or more sessions (‘Review’). Factors affecting group attendance, as highlighted by the qualitative interviews, included lack of carer and transport availability, location and travelling distance, and groups clashing with other commitments. However, one group reported very good attendance. The strategies used by facilitators to promote attendance included calling carers and participants to remind them about the session and helping to resolve issues with transport. Other factors that contributed to feasibility outcomes such as group attendance included facilitator skill, appropriate adaptations and social interactions, which fostered positive engagement in sessions.

The fidelity of delivery of CST was of moderate quality (‘Review’). The interviews with participants, carers and facilitators highlighted positive experiences such as enjoyment of activities and sharing experiences with other group members, as well as benefits of CST with respect to participants’ memory, communication and sleep. Overall, the groups were perceived positively by participants, albeit with some barriers including fatigue due to two sessions on 1 day and the need to adapt and modify CST sessions further.

The healthcare and social care resource use questionnaires were well completed. The proxy version of the EQ-5D yielded the highest completion rates of the three questionnaires, although the sample sizes for both self-completed versions were small and therefore no meaningful conclusions could be drawn regarding completion rates. The average cost of CST was £602 per participant, which was somewhat higher than the cost for people with dementia without intellectual disability. Reference Knapp, Thorgrimsen, Patel, Spector, Hallam and Woods29 The higher cost was largely due to the smaller group sizes.

Results in context

The results of this study suggest that group CST was associated with small improvements in cognition and quality of life (although the study was not powered to detect a significant difference); this was consistent with CST studies in people with dementia without intellectual disability, which have found improvements in cognition and quality of life. Reference Desai, Leung, Fearn, John, Stott and Spector8 Our previous feasibility RCT of individual CST found improvements in quality of life but not cognition. Reference Ali, Brown, Tsang, Spector, Aguirre and Hoare10 The views of participants regarding being in the CST group were largely positive, and their carers reported positive benefits outside the groups, similar to the findings of studies of CST in people with dementia without intellectual disability. Reference Spector, Gardner and Orrell30,Reference Gibbor, Yates, Volkmer and Spector31 Other themes that emerged, such as the importance of carer engagement (e.g. for transport) and facilitators having appropriate skills and the time and resources needed for preparation, were shared across a number of studies. Reference Gibbor, Yates, Volkmer and Spector31 Fidelity was particularly poor in our study of individual CST, which was delivered by family and paid carers; fidelity in the present study was of moderate quality, indicating that professionals are more likely to deliver the intervention as intended, which might in turn lead to improved outcomes.

Strengths and limitations

One of the strengths of the study was the inclusion of the voices of people with dementia and intellectual disability about their experiences of CST groups, something that is often lost within research processes. Reference Muralidhar, Spector, Hui, Liu and Ali7 This was a multicentre study across six NHS sites in England and included sites from both urban and more rural areas. The inclusion of such diverse sites provided information about potential barriers relating to group attendance, which was one of the key issues highlighted in the study.

Despite the study taking place in ethnically diverse regions of England, the vast majority of participants with intellectual disability were from White ethnic backgrounds. This may reflect broader issues such as delayed diagnosis of dementia in people from minority ethnic groups, Reference Dodd, Pracownik, Popel, Collings, Emmens and Cheston32 but there needs to be a concerted effort to include these groups in dementia research. Another issue encountered in this study was the heterogeneity among group participants, with groups comprising individuals with both mild and moderate intellectual disability and dementia. It is preferable to have CST groups of individuals with similar abilities, as this promotes better engagement and inclusion. It was not possible to run separate groups based on abilities owing to issues with recruitment and not having a sufficient number of participants at each site. It would have been preferable to have delivered CST groups at participants’ usual day services such as day centres, as is the case for CST studies in people with dementia without intellectual disability, and this may have reduced burden and transportation costs.

To estimate the intervention cost, we did not include costs for room bookings or refreshments, as these were not recorded. We also did not include costs to participants, families and/or carers such as transport costs, and we did not include time spent by paid carers who facilitated participants’ attendance, although with hindsight we would recommend that this latter cost in particular should be included in a future full trial. Therefore, the overall cost of CST per participant is likely to be higher. We also did not include follow-up assessments beyond the end of the intervention period, nor did we collect data on the cause of intellectual disability (e.g. Down syndrome), although many of the participants did have Down syndrome, owing to its association with dementia.

Recommendations and implications

A future study would need to consider strategies to mitigate the issues highlighted in this study. Recruitment was affected by the lack of eligible participants. At one site, this was because of a long waiting list for dementia assessments. In addition, dementia diagnoses for people with intellectual disability usually require longitudinal assessments, Reference Dodd, Pracownik, Popel, Collings, Emmens and Cheston32 which can result in diagnosis at later stages of the condition, when the individual may no longer be eligible for CST. In our study, almost all the participants had either Alzheimer’s dementia, vascular dementia or a mixed dementia. This was probably because of the strong association between Down syndrome and Alzheimer’s dementia, but it may also reflect challenges in diagnosing other types of dementia in this population. It is crucial that the diagnostic pathway is improved, and it may be necessary to include additional resources to support dementia assessments in any future research.

There were some difficulties around attendance, and any future research should consider strategies to make groups more accessible for participants and their carers. This might include careful consideration of the location (e.g. a day centre) and allocation of a budget to support participants who may not be able to afford transport or to pay for carers’ time. A future study also could explore the feasibility of attending an online CST group. Feasibility studies of people with dementia without intellectual disability, across a number of countries, have found online delivery to be acceptable, although data on effectiveness are not yet available. Reference Perkins, Fisher, Felstead, Rooney, Wong and Dai33 Remote delivery of CST groups will have specific challenges for people with intellectual disability and dementia, but it may improve access and attendance.

The CST supplement was found to be helpful in providing suggestions for modifying activities. However, facilitators commented on the need for further adaptations based on the ability of participants and for challenges with engagement to be addressed; some of the CST content and resources in the supplement were not suitable. In light of these comments, a further revision of the supplement is warranted. The supplement and training should also include guidance on modifications to the session format, such as shorter sessions incorporating more breaks, varied group sizes depending on abilities, and how to work with and manage carers who are present during the groups. A longer training session and supervision of facilitators (e.g. every 2 weeks) might be helpful in addressing and overcoming some of the challenges in delivering the intervention.

The DLD and the QOL-AD proxy showed some sensitivity to change in favour of the intervention, in line with findings of other CST studies, indicating that they were appropriate measures to assess treatment effect within this population. There is no evidence to distinguish between the appropriateness of the EQ-5D-5L proxy or modified EQ-5D-3L; therefore, we would recommend using both in a potential future main trial.

The sample size calculation for a definitive trial will need to include approximately 25% attrition owing to the high observed rate of withdrawal between consent and randomisation, as well as accommodating expected attrition in a larger sample to follow-up. The sample size calculation would also need to account for clustering in the intervention group by considering an appropriate design effect.

Overall, this RCT of group CST in people with intellectual disability and dementia provides valuable information about the feasibility and acceptability of a future definitive trial. On the basis of our findings, it may only be possible to conduct a future RCT if there are improvements to the dementia diagnostic pathway within clinical services, coupled with considerable revision of trial processes and procedures such as the recruitment strategy, reductions in delays between consent and baseline assessments, and strategies to improve group attendance.

Supplementary material

The supplementary material is available online at https://doi.org/10.1192/bjo.2025.10764

Data availability

The data that support the findings of this study are available from the corresponding author, A.A., upon reasonable request.

Acknowledgements

We thank John O’Dwyer and colleagues for providing us with advance access to their new modified EQ-5D-3L questionnaire for adults with learning/intellectual disabilities. We also thank all the clinicians who facilitated the CST groups, and the patients and carers and National Health Service sites that took part in the study, as well as members of our trial steering committee.

Author contributions

A.S., A.A., E.A., G.C., C.S.C. and Z.H. conceptualised the study and were responsible for its design. S.H., J.C., C.F., S.K., A.O. and A.J. were responsible for data collection. N.G. and C.S.C. carried out the statistical and health economic analyses, respectively, and all authors were responsible for interpretation of the data. A.A. wrote the first draft of the manuscript. All authors revised the paper critically for important intellectual content and gave final approval of the version to be published.

Funding

This study was funded by the National Institute for Health and Care Research (NIHR; grant reference: NIHR 201934). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

Declaration of interest

None.

References

Strydom, A, Chan, T, King, M, Hassiotis, A, Livingston, G. Incidence of dementia in older adults with intellectual disabilities. Res Dev Disabil 2013; 34: 1881–5.10.1016/j.ridd.2013.02.021CrossRefGoogle ScholarPubMed
MacDonald, S, Summers, SJ. Psychosocial interventions for people with intellectual disabilities and dementia: a systematic review. J Appl Res Intellect Disabil 2020; 33: 839–55.10.1111/jar.12722CrossRefGoogle Scholar
National Institute of Health and Care Excellence (NICE). Dementia: Assessment, Management and Support for People Living with Dementia and their Carers (NICE Guideline NG97). NICE, 2018 (https://www.nice.org.uk/guidance/ng97 [cited 27 Oct 10 2024]).Google Scholar
Hahn, JE, Fox, S, Janicki, MP. Aging among older adults with intellectual and developmental disabilities: setting national goals to address transitions in health, retirement, and late-life. Inclusion 2015; 3: 250–9.Google Scholar
Tuffrey-Wijne, I, McLaughlin, D, Curfs, L, Dusart, A, Hoenger, C, McEnhill, L, et al. Defining consensus norms for palliative care of people with intellectual disabilities in Europe, using Delphi methods: a white paper from the European Association of Palliative Care. Palliat Med 2016; 30: 446–55.10.1177/0269216315600993CrossRefGoogle Scholar
Watchman, K. Reducing marginalisation in people with a learning disability and dementia. Aust J Dement Care 2015; 2: 31–4.Google Scholar
Muralidhar, M, Spector, A, Hui, EK, Liu, L, Ali, A. A systematic review of psychosocial interventions for people with intellectual disabilities and dementia. Aging Ment Health 2024; 28: 385–95.10.1080/13607863.2023.2265322CrossRefGoogle ScholarPubMed
Desai, R, Leung, WG, Fearn, C, John, A, Stott, J, Spector, A. Effectiveness of Cognitive Stimulation Therapy (CST) for mild to moderate dementia: a systematic literature review and meta-analysis of randomised control trials using the original CST protocol. Ageing Res Rev 2024; 97: 102312.10.1016/j.arr.2024.102312CrossRefGoogle ScholarPubMed
Spector, A, Thorgrimsen, L, Woods, B, Royan, L, Davies, S, Butterworth, M, et al. Efficacy of an evidence-based cognitive stimulation therapy programme for people with dementia: randomised controlled trial. Br J Psychiatry 2003; 183: 248–54.10.1192/bjp.183.3.248CrossRefGoogle ScholarPubMed
Ali, A, Brown, E, Tsang, W, Spector, A, Aguirre, E, Hoare, S, et al. Individual cognitive stimulation therapy (iCST) for people with intellectual disability and dementia: a feasibility randomised controlled trial. Aging Ment Health 2022; 26: 698708.10.1080/13607863.2020.1869180CrossRefGoogle ScholarPubMed
Ali, A, Aguirre, E, Carter, J, Hoare, S, Brackley, K, Goulden, N, et al. Group cognitive stimulation therapy versus usual care for people with intellectual disabilities and dementia (CST-IDD) in the UK: protocol for a mixed-methods feasibility randomised controlled trial. BMJ Open 2023; 13: e072391.10.1136/bmjopen-2023-072391CrossRefGoogle ScholarPubMed
Saxton, J, McGonigle, KL, Swihart, AA, Boller, F. The Severe Impairment Battery. Pearson, 1993.Google Scholar
Evenhuis, HM, Kengen, MM, Eurlings, HA. The Dementia Questionnaire for People with Intellectual Disabilities. In Neuropsychological Assessments of Dementia in Down Syndrome and Intellectual Disabilities (ed Prasher, VP): 39–51. Springer, 2009.10.1007/978-1-84800-249-4_3CrossRefGoogle Scholar
Logsdon, RG, Gibbons, LE, McCurry, SM, Teri, L. Quality of life in Alzheimer’s disease: patient and caregiver reports. J Ment Health Aging 1999; 5: 2132.Google Scholar
Cuthill, FM, Espie, CA, Cooper, SA. Development and psychometric properties of the Glasgow Depression Scale for people with a learning disability. Individual and carer supplement versions. Br J Psychiatry 2003; 182: 347–53.10.1192/bjp.182.4.347CrossRefGoogle ScholarPubMed
Herdman, M, Gudex, C, Lloyd, A, Janssen, M, Kind, P, Parkin, D, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res 2011; 20: 1727–36.10.1007/s11136-011-9903-xCrossRefGoogle ScholarPubMed
O’Dwyer, J, Meads, D, Bryant, L, Kind, P, Hulme, C. The development and testing of a modified EQ-5D for intellectually disabled adults. Value Health 2023; 26: S315.10.1016/j.jval.2023.03.1798CrossRefGoogle Scholar
O’Dwyer, JL, Bryant, LD, Hulme, C, Kind, P, Meads, DM. Adapting the EQ-5D-3L for adults with mild to moderate learning disabilities. Health Qual Life Outcomes 2024; 22: 37.10.1186/s12955-024-02254-xCrossRefGoogle ScholarPubMed
Beecham, J, Knapp, M. Costing psychiatric interventions. In Measuring Mental Health Needs (ed Thornicroft, G): 200–24. Gaskell, 2001.Google Scholar
Hui, EK, Tischler, V, Wong, GH, Gibbor L, Spector, AE. The impact of Virtual Individual Cognitive Stimulation Therapy (V-iCST) on cognition, quality of life, mood, and communication in dementia: a feasibility randomized controlled trial. Alzheimer’s Dement 2018; 18: e060636.10.1002/alz.060636CrossRefGoogle Scholar
Borrelli, B. The assessment, monitoring, and enhancement of treatment fidelity in public health clinical trials. J Public Health Dent 2011; 71: S52–63.10.1111/j.1752-7325.2011.00233.xCrossRefGoogle ScholarPubMed
Murphy, J, Gray, CM, Cox, S. Communication and Dementia: How Talking Mats Can Help People with Dementia to Express Themselves. Joseph Rowntree Foundation; University of Stirling, 2007.Google Scholar
Stans, SEA, Dalemans, RJP, De Witte, LP, Beurskens, AJHM. Using talking mats to support conversations with communication vulnerable people: a scoping review. Technol Disabil 2019; 30: 153–76.10.3233/TAD-180219CrossRefGoogle Scholar
Walton, H, Spector, A, Williamson, M, Tombor, I, Michie, S. Developing quality fidelity and engagement measures for complex health interventions. Br J Health Psychol 2020; 25: 3960.10.1111/bjhp.12394CrossRefGoogle ScholarPubMed
Fereday, J, Muir-Cochrane, E. Demonstrating rigor using thematic analysis: a hybrid approach of inductive and deductive coding and theme development. Int J Qual Methods 2006; 5: 8092.10.1177/160940690600500107CrossRefGoogle Scholar
Crabtree, B, Miller, W. A template approach to text analysis: developing and using codebooks. In Doing Qualitative Research (eds Crabtree, B, Miller, W): 163–77. Sage, 1999.Google Scholar
Weber, R. Basic Content Analysis 2nd ed. Sage, 1990.10.4135/9781412983488CrossRefGoogle Scholar
Jones, K, Weatherly, H, Birch, S, Castelli, A, Chalkley, M, Dargan, A, et al. Unit Costs of Health and Social Care 2023 Manual. Personal Social Services Research Unit, University of Kent, and Centre of Health Economics, University of York, 2023.Google Scholar
Knapp, M, Thorgrimsen, L, Patel, A, Spector, A, Hallam, A, Woods, B, et al. Cognitive stimulation therapy for people with dementia: cost-effectiveness analysis. Br J Psychiatry 2006; 188: 574–80.10.1192/bjp.bp.105.010561CrossRefGoogle ScholarPubMed
Spector, A, Gardner, C, Orrell, M. The impact of Cognitive Stimulation Therapy groups on people with dementia: views from participants, their carers and group facilitators. Aging Ment Health 2011; 15: 945–9.10.1080/13607863.2011.586622CrossRefGoogle ScholarPubMed
Gibbor, L, Yates, L, Volkmer, A, Spector, A. Cognitive stimulation therapy (CST) for dementia: a systematic review of qualitative research. Aging Ment Health 2021; 25: 980–90.10.1080/13607863.2020.1746741CrossRefGoogle ScholarPubMed
Dodd, E, Pracownik, R, Popel, S, Collings, S, Emmens, T, Cheston, R. Dementia services for people from Black, Asian and Minority Ethnic and White-British communities: does a primary care based model contribute to equality in service provision? Health Soc Care Commun 2022; 30: 622–30.10.1111/hsc.13167CrossRefGoogle ScholarPubMed
Perkins, L, Fisher, E, Felstead, C, Rooney, C, Wong, GHY, Dai, R, et al. Delivering Cognitive Stimulation Therapy (CST) virtually: developing and field-testing a new framework. Clin Interv Aging 2022; 17: 97116.10.2147/CIA.S348906CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Stop/Go criteria for the feasibility outcomes

Figure 1

Table 2 Participant demographics at baseline in the intervention (CST) and TAU arms

Figure 2

Fig. 1 Recruitment and retention for the trial. TAU, treatment as usual.

Figure 3

Table 3 Missing data

Figure 4

Table 4 Raw scores for SIB, DLD, QOL-AD proxy and GDS-LD proxy outcome measures

Figure 5

Table 5 Adjusted means for outcome measures from analysis of covariance

Figure 6

Table 6 Themes from the qualitative interviews with facilitators and carers

Figure 7

Table 7 Basic content analysis of participant interviews

Figure 8

Table 8 Summary of recommendations for future groups from facilitator and carer interviews

Figure 9

Table 9 Costs of delivering cognitive stimulation therapy

Figure 10

Table 10 Intervention cost, calculated as the sum of staff time spent on training and on session preparation and delivery, and the expenditure on resources for the sessions (from Supplementary Table 8)

Figure 11

Table 11 Summary of the feasibility outcomes and whether they met progression criteria

Submit a response

eLetters

No eLetters have been published for this article.