Delivering Cognitive Stimulation Therapy (CST) Virtually: Developing and Field-Testing a New Framework

Introduction

There are 55 million people living with dementia across the globe so there is a significant need to develop interventions that tackle the physical, mental, social and financial impacts of the condition.¹ With no treatments that prevent the progressive neurological deterioration resulting from dementia, the need for supportive interventions is pressing.

Cognitive Stimulation Therapy (CST) is the best-established psychological intervention for people living with dementia and has demonstrated benefits in relation to cognition, quality of life and overall cost-effectiveness.² CST was developed by Spector et al.³ as a brief, 14-session face-to-face intervention for people living with dementia, comprised of group activities which stimulate cognition including memory and language. In the United Kingdom (UK), CST is recommended by the National Institute for Health and Social Care Excellence (NICE) as a key psychosocial intervention for dementia.⁴ Culturally adapted versions of CST have been successfully implemented across the globe and CST is recommended internationally for people with early-stage dementia.^5–7

Access to in-person interventions like CST was immediately and unexpectedly restricted when the COVID-19 pandemic spread throughout the world in early 2020.⁸ Not only did this prevent access to services, treatments and interventions for people living with dementia who had been attending them previously, it also highlighted the gap in service provision for those who could not access services outside of the pandemic context.⁹ This includes people living with dementia residing in rural communities who are unable to readily access transport or those with reduced mobility. In response to the pandemic, many services have had to adapt rapidly in order to keep services accessible whilst maintaining the need for social distancing. This has seen services turn to developing treatments and interventions that can be delivered using digital technology.⁹

Digital Technology in Psychological Interventions

The use of digital technology or “e-Health” has become popular in clinical populations, such as adults with anxiety and depression, even before its mass adoption during the pandemic.¹⁰ E-Health is defined by the World Health Organization as ‘the cost-effective and secure use of information and communications technologies in support of health and health-related fields’.¹¹ Leading up to the pandemic, most of the digital psychological interventions on offer were based on Cognitive Behavioural Therapy (CBT), although other psychological therapies have been demonstrated to be beneficial when delivered virtually.^12–14

There is increasing interest in the use of e-health interventions for people living with dementia, and evidence of feasibility and their impact on psychological, social and cognitive domains.^15–17 Existing research does highlight the challenges that older adults face in accessing e-health interventions. Older adults can find access more difficult than working age adults due to barriers including limited computer literacy and lack of trust in digital interventions.^18,19 This is further compounded in people living with dementia who face declining cognitive ability and independent day-to-day functioning, and reduced help-seeking due to beliefs that cognitive symptoms are a normal part of aging.^20,21 Despite these barriers, it appears to be both feasible and beneficial to continue to deliver psychological interventions during the pandemic and beyond.^22,23

These findings pave the way for further expansion of e-health application to a wide variety of existing programmes for people living with dementia. CST is one such intervention that could benefit from adaptation for online use, especially in the context of the COVID-19 pandemic where in-person services may be restricted. The existing literature related to virtual delivery of CST is limited to a single case study series.²⁴ Furthermore, a workshop hosted by the [redacted for review] identified the demand from CST practitioners for a standard protocol for delivery of virtual CST, to enable service continuation during the pandemic. Therefore, the main aims of this study are:

1. To develop a protocol for virtual CST (vCST).
2. To evaluate the feasibility and acceptability of the vCST protocol through field-testing in a range of settings and countries.
3. To provide a framework for offering vCST virtually across global settings.

Materials and Methods

Overview

The study followed the Medical Research Council’s ‘developing and evaluating complex interventions’ framework, made up of four phases: Development, Feasibility, Evaluation, and Implementation.²⁵ Here we focus only on the Development and Feasibility phases. The project was split into two parts: (a) Development of the vCST intervention using the existing CST group manual and stakeholder consultation;²⁶ and (b) Feedback and validation from CST facilitators following vCST protocol field-testing in Brazil (Rio de Janeiro and São Paulo), China (Hong Kong), India (Chennai), Ireland and the UK.

Ethics

This study was conducted in accordance with the Declaration of Helsinki. Ethical approval was granted in all countries (UK and Ireland, UCL Research Ethics Committee, project 17127/001; India, Schizophrenia Research Foundation [SCARF], SRF-DC/18/OCT-2020; Brazil, Federal University of São Paulo [UNIFESP] and CONEPE, process no. 4895729 and Federal University of Rio de Janeiro Institute of Psychiatry , ID 57019616.5.1001.5263; Hong Kong, University of Hong Kong, reference no. EA2004006).

Part 1: Development of vCST Protocol

The vCST protocol was developed by adapting the existing group CST manual, informed by stakeholder consultations with four groups: people living with dementia; caregivers of people living with dementia; service managers; and CST group facilitators who had previously run vCST groups.²⁶ People with dementia and caregivers were service users of organisations in the UK, and service managers were recruited from these same services. Facilitators were recruited by email from third-sector organisations in the UK and Hong Kong.

Questions for the consultations were developed using the Consolidated Framework for Implementation Research (CFIR).²⁷ Interview questions related to the five main domains of the framework: intervention characteristics, outer setting, inner setting, characteristics of individuals and process.

All focus groups took place over the online video conferencing platform “Zoom” and were recorded for post-group analysis. Two researchers then took field notes from these recordings, which were clustered together into key ideas. The data gathered from the focus groups was used to develop guidelines for delivering vCST in collaboration with researchers at The University of Hong Kong working on a trial of vCST. Both teams worked closely throughout the project to ensure the vCST intervention and protocols were aligned, whilst also accounting for specific cultural adaptations.

Part 2: Field-Testing

Design

The vCST protocol was field tested in five countries and feedback was gathered from facilitators. vCST was delivered through a randomised controlled trial in the UK and Ireland. In Brazil, Hong Kong, and India, vCST was delivered as a pre-post study. The empirical results from these studies will be reported separately. Here we focus on the content of the vCST protocol and its feasibility through field-testing.

Procedure

The template version was co-developed in the UK, Ireland and Hong Kong as described above, then translated and culturally adapted by teams in India and Brazil, based on previously culturally adapted versions of in-person CST.^5,28 In India, additional stakeholder engagement was carried out to ensure the intervention was acceptable and culturally appropriate. Interviews with psychologists, psychiatrists and healthcare professionals took place prior to vCST groups beginning. Interviews with people living with dementia and caregivers took place prior to and midway through the delivery of the intervention, where delivery could be adjusted according to feedback.

Facilitators all had experience of working directly with people living with dementia and had experience of either delivering or observing in-person CST. Facilitators were asked for feedback following completion of the vCST sessions, which comprised:

1. Which sessions/activities went well?
2. Which sessions/activities did not go as well?
3. Did any activities need adapting further?
4. Did you develop any new ideas?
5. Any other comments.

Results

Part 1: Development of vCST Protocol

Stakeholder Consultation

Twenty participants across four focus groups took part in the consultation. Each focus group lasted approximately 90 minutes and was facilitated by two researchers. Group one was attended by three people living with dementia from the same third-sector organisation in the south-west of England. An additional individual session was conducted by the researcher with a fourth person with dementia from a different third-sector organisation in the north of England. All four had attended in-person group CST sessions before the first UK national COVID-19 lockdown, before switching to virtual CST groups midway through the intervention.

Group two was attended by four caregivers of people living with dementia. Three were recruited from a third-sector organisation in the south-east of England and one from a third sector organisation in the south-west of England. Only one caregiver had cared for a person living with dementia who had previously attended vCST before.

Group three was attended by eight vCST facilitators. One was from Hong Kong team and all others were from England. All were working in third-sector dementia organisations and had facilitated vCST sessions online previously.

Group four was attended by four service managers recruited from third-sector dementia organisations across England - one of whom worked for a service that had already implemented vCST sessions. The key ideas relating to each of the CFIR domains are outlined in Table 1.

Table 1 Key Questions and Ideas from Focus Groups

The vCST protocol developed through stakeholder consultation follows the 14-session plan outlined in the group CST manual in terms of session structure and themes, with activities adapted for online delivery.²⁶ Sessions last 45–60 minutes as stated in the manual. The final order of sessions matched the group CST manual as outlined in Table 2.

Table 2 vCST Session Themes

All vCST sessions begin with introductions, welcoming members to the group, orienting participants to time, date and place and, doing a warm-up activity. Everyone then sings the chosen group song and discusses a recent newspaper article. Next, group members choose a main activity based on the session theme. Each session ends with a summary, feedback from participants, reminding participants of the next session theme, including activities and materials to bring, and then saying goodbye.

Session Format

Each group should run with four to five participants. This was deemed as the optimum group size given the number of people who would be visible on the screen at one time, and to enable the participation of all group members. Sessions should run for 45–60 minutes dependent on the group’s engagement levels during the session. Participants should be able to take breaks during session if they feel unable to stay engaged but should be encouraged to stay for the whole session if possible.

Session Activities

In-person CST uses physical objects, pictures, and music. For vCST, activities should be taken from the group CST manual and be transferred into a digital format. Text, images, videos, and sounds can be presented to all participants through software such as Microsoft PowerPoint, by using the share screen and/or share sound functions. For example, in the “Using Money” session interactive PowerPoint presentations of a price matching task can be used, and for the “Current Affairs” session, online news clips can be shown to the group by sharing the screen and sound.

Physical Resources

The initial protocol was based on the assumption that researchers were unable to send physical resources for sessions to participants. Any additional resources required, such as pens, paper, or other household objects, should be agreed with participants at the end of the previous session. This would give participants time to gather these objects if needed. Each session contained at least one activity that did not need additional resources, in case participants were not able to obtain them.

Platform and Technology

The video conferencing platform Zoom was chosen to deliver the sessions, as most people had experience with and preferred this platform. To support participants to access the sessions, a “How to use Zoom” guide was created, which should be sent to participants before the group. Participants should also be offered a one-to-one session on Zoom prior to attending the group to give them experience with using the platform, alongside additional telephone support as required. Participants should be advised to use a laptop or tablet and not a mobile phone to access the sessions, as the smaller screen of a mobile phone does not allow multiple participants to be viewed. Sessions should be set in “Gallery view” so that all participants are on screen simultaneously.

Role of Facilitators

Each group should run with two facilitators, one to lead on delivering the content and one to provide practical and other types of support as required. Reminder emails with the Zoom link should be sent to all participants the day before each session. Participants should be asked to sign into the session 10–15 minutes before the start time, to allow sessions to begin punctually and to allow time for the second facilitator to contact any participants who have not attended on time.

Support from Caregiver

If a participant is unable to access Zoom independently, the person living with dementia should identify a named caregiver with both of their consent who would support them to access the sessions and be contacted for any technical support. Any caregivers involved in giving support should be advised not to attend the sessions, but to be nearby (ie, in the next room) for the duration of the session in case they need to give technical support to the participant.

Part 2: Field-Testing

Facilitators

Field-testing took place across 10 groups in Brazil, Hong Kong, India, Ireland, and the UK. In Brazil a Postdoctoral Researcher, a Clinical Psychology student and a Gerontologist facilitated the sessions in a university setting. Three facilitators in Hong Kong and two in Ireland were Occupational Therapists in a university and care service setting respectively. In India, a Research Assistant and a Psychiatrist facilitated the sessions in a non-governmental organisation. In the UK, four facilitators were Trainee Clinical Psychologists, and one was a psychology PhD student, all based in a university. Facilitators were trained in CST, and were provided with guidelines for vCST delivery which had been developed from the focus group findings.²⁹ All 15 facilitators were contacted, and feedback was received from 14.

vCST Participants

vCST participant demographics are not outlined in this paper, as this was a field-testing study. However all participants met the inclusion criteria for group CST, which comprises:

1. Meeting the ICD-10 criteria for dementia.
2. Having mild to moderate dementia (confirmed by the person with dementia and their caregiver or rated on the Clinical Dementia Rating Scale).³⁰
3. Having sufficient hearing and vision to follow conversation and comment on visual material.
4. Having the ability to participate in a group for 1 hour.
5. Additionally for vCST, participants needed access to technology, and the ability to use video conferencing software, or a caregiver who could support.

Feedback from Field-Testing

Which Sessions/Activities Went Well?

There was variety of feedback relating to the sessions and activities that worked best, due to group preferences or levels of engagement which could vary from session to session. However, there were some common themes across groups. Facilitators reported that sessions involving physical objects or “show and tell” had worked well, as the items sparked conversation and supported participant engagement. Some reiterated that participants needed to be reminded at the end of the previous session to bring physical objects for the following week, and suggested strategies such as encouraging participants to mark this in their diaries or sending a reminder email before the session. Structured activities or those with visual stimuli were reported to have run successfully. Examples of such activities include in “Faces and Scenes”, participants compared photos of famous places in the past and present and discussed differences and similarities. In “Word Games” participants played “hangman” or completed the gaps of common sayings or idioms. In “Identifying Sounds” participants matched images of items and their associated sound clips presented through PowerPoint.

Which Sessions/Activities Did Not Go As Well?

Facilitators reported that, for some groups, discussion-based activities that were less structured were not as successful, especially in groups where participants had specific language or communication difficulties. For example, activities in the “Current Affairs” session, and the warm-up activity that involved discussing a recent news article. In some cases, this was attributed to group dynamics, where more reserved group members did not participate as much in discussions. Furthermore, some facilitators reported that it was challenging to find articles of interest, as the news was focused on the pandemic, which many participants and facilitators did not want to discuss every week.

Some activities had more complicated instructions, for example, the bingo activity in “Number Games” and the paper folding activity in “Being Creative”, which also required fine motor dexterity and was more challenging for some participants. Some facilitators also fed back that it was more difficult to prompt or guide participants on activities where the task is completed outside the camera’s field of view – for example searching for a number in “bingo” or colouring in – as the camera generally focuses on the face of the participant rather than on the table in front of the participant. Facilitators felt that these challenging activities worked better if a caregiver was available to support.

The group song that starts each session also had mixed feedback. Some facilitators fed back that groups disengaged with this activity, and one group wanted to play different songs each week instead of listening to the same song.

Did Any Sessions Need Adapting Further?

In India and Hong Kong, facilitators posted or couriered resource packs to participants, instead of relying on the “shared screen” function to show visual stimuli. These packs included images, worksheets and resources needed for the activities such as colour pencils and maps, and in India there was an accompanying instruction leaflet for caregivers sent over email. The printed resources could also include larger text and pictures for those who struggle to see the PowerPoint.

Some facilitators reported that overuse of the “shared screen” function reduced the quality of interaction, as participants’ faces become small icons when function is enabled, which can be hard to see and make sense of. However, in other groups, the shared screen function worked effectively, and the visual stimuli enhanced engagement. Facilitators advised that “shared screen” is used for a limited period and that all resources for each session should be on one PowerPoint to avoid having to move from one screen to another.

Did Any Specific Activities Need Adapting?

Poor quality of internet connections was an issue for some groups, which could result in a lag between the images and sound when sharing a video. This was confusing for some participants, so some facilitators reverted to using a text or images as stimuli. Some facilitators adapted the “Current Affairs” section to include newspaper headlines, articles and images instead of videos. One facilitator reported that they had adapted the team quiz in Session 14 to an individual quiz, as they found it was harder for participants to confer in teams over Zoom.

Did You Develop Any New Ideas?

In Ireland, the facilitators fed back that participants had enjoyed the exercise videos from “Physical Games” so much, they incorporated physical activity into the warm-up of each session. Facilitators in Hong Kong made use of online resources and websites to provide multisensory stimulation, including a virtual tour of an art gallery for “Being Creative”, or the use of street view in Google Maps to go for a “virtual walk” around a city or neighbourhood for “Orientation”. Other activities were carried out that were similar in aims and delivery to the suggested activities, such as creating an “odd one out” game in the “Categorising Objects” session or listing as many words as possible from a category beginning with the same letter in the “Word Association” session. In India, groups took part in chair yoga during the “Physical Games” session and, based on the suggestion from a caregiver, participants were encouraged to wear traditional clothes in the last session to celebrate the cultural and ethnic diversity of the group, which promoted conversation amongst the group. In Brazil, the “Being Creative” session coincided with Mothering Sunday, so the activities were themed around motherhood. These new activities demonstrate the flexibility that CST facilitators can employ when planning sessions.

Any Other Comments

The majority of facilitators reiterated the importance of a trial session using videoconferencing technology for participants and caregivers, anda second facilitator to support participants join sessions and address any issues with technology. Caregivers were required to support with technology and, in some cases, with more complicated activities. In India, where participants only had access to tablet computers, it was recommended that groups comprised of three participants due to the smaller screen size and limited space for more faces.

During orientation at the start of each session, a facilitator in India reminded participants that the session was taking place virtually and gave a brief overview of the functions of Zoom. This helped to address disorientation related to using videoconferencing platforms and challenges posed by lower levels of digital literacy in older adults with dementia. The facilitators in Ireland noted that it was more challenging to use orientation sensitively using Zoom, as they had to rely on sharing images or using verbal cues, as opposed to facilitating in-person groups where orientation can be done subtly using props in the room, for example, using flowers such as daffodils to orientate to spring.

A pair of facilitators in the UK gave participants the option of staying on the Zoom call after meetings. They reported that the participants in one group requested this as a way of socialising, but it may not be suitable for all groups. In Ireland, the facilitators sent a follow up email with a summary of the session and included any materials and worksheets used. This also provided participants an opportunity to ask questions and share feedback between sessions.

After all feedback from facilitators was collated, the final version was reviewed and approved by all co-authors and a consensus was reached to develop the final vCST session protocol, which is outlined in Table 3. The general guidelines for vCST delivery which had initially been developed from the focus group findings were updated based on further insights from field-testing, and are freely available online (https://www.ucl.ac.uk/international-cognitive-stimulation-therapy/cst-research/virtual-cst).²⁹

Table 3 Final vCST Session Protocol

Discussion

This study aimed to investigate the feasibility of a new vCST protocol for people living with dementia through stakeholder consultation and feedback from 14 facilitators following field-testing in the Brazil, China (Hong Kong), India, Ireland, and the UK. The qualitative findings indicate that the adapted protocol was feasible, acceptable to facilitators and had the flexibility to be adapted across cultures. The study therefore supported the proposal of a new delivery framework.

The proposed protocol and framework are vital, given that services have not been able to deliver face-to-face interventions during the COVID-19 pandemic. A recent survey found that during the height of the COVID-19 pandemic in the UK, many National Health Service (NHS) memory clinics and third-sector dementia organisations did not have guidance on how to adapt CST for virtual delivery. This resulted in individual adaptation of CST by different facilitators, which was time-consuming and could result in inconsistencies across services (Fisher et al., In preparation).

Despite CST being a popular, widely available, and beneficial psychosocial intervention for people living with dementia, it still remains inaccessible to those who may be unable to travel due to health, socioeconomic or geographical reasons, as well as to those shielding during the COVID-19 pandemic. Access to virtual services is key to filling this gap. However, low levels of digital literacy and access to technology may be a barrier for older adults with dementia to access vCST. Access to virtual care is limited in low- and middle-income countries, and in the UK, there are 5.3 million “internet non-users”, with those over the age of 75 years old making up over half of these non-users.^31,32 Despite the shift to digital delivery of services, there is there is little evidence that those who were digitally excluded before the pandemic have become internet users since the pandemic.³³ There is a clear need for top-down investment and engagement with older people to prevent digital exclusion. To support access to vCST on an individual level support from a caregiver and facilitator is necessary, including resolving technical issues and conducting a trial run using the video conferencing software. These approaches were also highlighted in a recent review of barriers and facilitators to telemedicine in dementia care.²³

In vCST, unlike in face-to-face CST, participants of the same group can reside in different regions of a country. This not only makes CST accessible to a wider range of participants but may be a point of conversation and more stimulating cognitively for participants. However the facilitator should ensure that no participants are excluded and that different cultures are celebrated.

Previous research has found that memory, comprehension of syntax, and orientation are the cognitive domains the most impacted by face-to-face CST.³⁴ However vCST may not have a comparable impact on these domains, due to the different method of delivery. Participants attend virtual sessions from home, remaining close to their caregivers and separate from the other participants and facilitators. This may impact the quality of the interaction with others and could affect the way that participants perceive themselves and fellow participants.

Strengths and Limitations

There are several strengths identified in the study, including the use of stakeholder consultation in the development of the final vCST guidelines and protocol. This is especially important given that both the people living with dementia and the group facilitators had both experienced vCST sessions prior to attending the focus groups. Stakeholder consultation is essential in creating new interventions as it can help to identify problems, solutions and priorities in the development and implementation process that researchers may not be aware of.³⁵ Secondly, the protocol was field-tested in a variety of settings across five countries. CST is an internationally implemented intervention, and it is promising that virtual CST was acceptable and feasible across the five countries. Finally, the vCST protocol is adapted from the original CST in-person protocol developed by Spector et al.,²⁶ which has a strong evidence base for benefitting people living with dementia.^2,36 As the vCST protocol adheres so closely to the original it makes comparisons across modality easily viable.

Despite these clear strengths, some limitations were also identified. The field-testing highlights that the protocol is feasible and acceptable, however the field-testing took place with only a selected group and many people will not have access or be able to engage in vCST groups due to lack of access to technology, income, or limited digital literacy. Sampling bias may have prevented reaching people with additional barriers for access to vCST, and future studies should try to recruit more inclusive samples. We did not use thematic analysis or other formal methods to analyse qualitative data, because there was a need to collect and analyse data rapidly at a time of limited staff capacity. Additionally, stakeholder consultation for the development of vCST was restricted to Hong Kong, India, and the UK. Due to limited time and staff capacity, this was compounded by having only one focus group per stakeholder group. Further focus groups could take place, adding more sites and using more detailed surveys, to elicit a wider range of views. Finally, the current study represents pilot work, with empirical testing of the intervention needed, including outcome measures and qualitative interviews.

Future Research

Reporting of results from ongoing RCTs is required to establish the efficacy of vCST and to establish feasibility and acceptability, and the impact on different dementia populations. Trials of face-to-face CST groups have demonstrated benefits on mood and quality of life, and it is necessary to examine if these benefits remain when the intervention is delivered virtually. Future research could also measure the facilitators’ fidelity to the vCST protocol. Fidelity to an intervention increases the reliability and validity of the data as all participants are more likely to receive the same intervention. Data on participant engagement in vCST sessions could also be explored, which would help to inform if the vCST protocol needs to be adjusted further, for example if people living with dementia are not able to engage in 45 to 60-minute session. Participant self-report measures may not be an accurate indicator of engagement, so observational data could be collected through the development of a coding system for researchers to use when watching video recordings of sessions, or by using eye-tracking technology to assess eye movements which can correlate with participant attention.³⁷

Conclusion

Overall, a 14-session vCST protocol developed in this study was feasible and acceptable as a psychosocial, e-health intervention for people living with dementia. We therefore recommend that vCST is offered as an intervention across dementia services to increase access to a CST programme for those who are otherwise unable to access CST in-person, for reasons including health, mobility, and transport problems. This is especially important when services are not able to offer in-person CST due to social distancing needs during current and future pandemics. vCST may not replace in-person CST sessions but can be a useful alternative in situations where in-person CST is not possible. Larger trials on vCST are required to further investigate benefits on mood and quality of life.