This protocol has been registered in Clinical Trials.org (NCT05525377) and follows the SPIRIT reporting guidelines (27) (Chan et al., 2013) adapted with supplemented and replaced sections borrowed from the CONSORT extension for pilot trials, as recommended by Lancaster and Thabane, 2019 (28). The patient and public involvement (PPI) component of this study is reported following GRIPP2-SF (29).
Study design
This is the protocol for a pilot randomised wait-list controlled feasibility trial followed by a qualitative evaluation. Participants randomly allocated to the intervention group will receive access to the educational programme for eight weeks. Those in the wait-list control group will receive treatment as usual (i.e., pointed to Rare Dementia Support website (https://www.raredementiasupport.org/)) and access to the programme at the end of data collection. The wait-list aimed to decrease attrition and also meet ethical demands (e.g., giving caregivers access to tailored educational resources that may potentially have a beneficial effect, thus denying treatment has ethical implications). Outcome measures will be collected at baseline, eight weeks and three months post-randomisation. Semi-structured interviews will be conducted after completion of the three-month follow up baseline measures as part of a process evaluation. The study participant Flow chart is shown in Fig. 1.
Participants and research setting
People identifying themselves as caregivers of someone with PCA, PPA or bvFTD will be recruited from the Rare Dementia Support Impact project (30) and the wider Rare Dementia Support service (RDS) based in the UK. See Brotherhood et al (30) for details on the ethical procedures and consent.
Inclusion criteria:
a. Adults (18+) who self-identify as an unpaid caregiver (e.g., partners, children, friends) of someone with a diagnosis of PPA, PCA or bvFTD.
b. The person with dementia is not living in a full-time care facility.
c. The care recipient must have a confirmed diagnosis of one of these types of dementia through self-report of the caregiver (to reflect the ‘real world’ application of the intervention).
Exclusion criteria:
a) Poor comprehension of written English
b) No access to the internet
Power calculation
As this is a feasibility study a formal power calculation is not required, however, the sample size does require justification. Originally, we aimed to recruit 135 participants comprising 45 caregivers from each of the three diagnostic groups that the intervention is designed for (PCA, PPA, bvFTD) with a 30:15 intervention control split in a 2:1 randomisation. This would have provided both a precise estimation of the overall retention rate but also within diagnostic group. Reviewing this target once recruitment was underway, we acknowledged that this would be a large feasibility study and with a rarer disease population would require recruiting a large proportion of the available population. This would not be the best use of the recruited population; therefore, we revised the target sample down to 30 participants (randomised 20:10 in favour of the intervention) this would allow an understanding of the trial processes, allow for adaptation and refinement of the intervention and understand acceptability and fidelity within the population. The confidence intervals on our estimates will be wider but this is an acceptable concession given the sparsity of the potential population to recruit. We will achieve a 95% confidence interval of ± 18.7% around our expected value of 76% of those completing the intervention and pre-post measures whilst viewing 70+% sessions. In terms of retention to the study we will achieve a 95% confidence interval of ± 15.5% around our expected value of 75%.
Progression criteria
Progression criteria for the trial will be considered in a holistic way with consideration given to possible refinements and adaptations that can be made to the intervention and the trial processes to mitigate any barriers encountered. Whilst we can set quantitative thresholds indicating success on elements of the feasibility trial these will be taken into consideration with the context of the conditions under which these were obtained; in addition to qualitative information collected. Non-progression will only be considered on the basis as to whether any of the conditions are unlikely to be resolved. Quantitative criteria will be fixed on recruitment and retention, whilst trial processes, suitability of proposed outcome measures, acceptability and fidelity will be assessed by a qualitative approach understanding the context of these. Successful recruitment will be denoted as achieving greater than 75% of proposed sample, less than 25% recruited of the proposed sample will be defined as unsuccessful and unlikely to be remedied. Successful retention will be denoted as greater than 70% of the recruited sample with unsuccessful retention denoted as less than 20% of the recruited sample.
Recruitment
Participants will be recruited from two sources:
1. Caregivers who have already consented to be involved in the wider RDS Impact study(30) and
2. Caregivers who contact us following various communications about the study (e.g. via the RDS service newsletter and social media, discussions between RDS members and RDS workers).
The consent procedure is as follows: 1) participants should first consent to participate in the wider RDS Impact study(30) of which this educational programme is a sub study, 2) participants will be sent a link to a Qualtrics (Qualtrics, Provo, UT) page at which point they will access screening questions as to their suitability for inclusion, 3) if they are screened as eligible, they will be directed to an online opt-in consent form that they should complete before given access to the survey collecting demographics and baseline measures, 4) If they are screened as ineligible, they will be directed to a page thanking them for their interest in the study and giving them email contacts for the Rare Dementia Support service in case they want to reach out for support. Both consent forms can be found in Appendix I. Using Qualtrics display logic, those who don’t consent will automatically receive a message thanking them for considering involvement in the study and asking them why they chose not to participate (being clear that they don’t have to answer this if they don’t want to).
We will recruit from each diagnostic group as screened. Up to fifteen semi-structured interviews will explore {what] in a subgroup of participants, Individuals will be purposively sampled (by diagnostic subtype and allocation group) and approached via email asking if they want to participate. Interviews will be conducted post follow up via Zoom.
Randomisation
Randomisation will be provided via secure online platform hosted by NWORTH CTU, Bangor University. The randomisation procedure will be carried out by the project manager of the study (EB). Once consent and baseline measures have been completed, the participant can be entered into the randomisation system. Randomisation will be stratified by diagnosis (bvFTD, PPA, PCA). A dynamic adaptive randomisation algorithm will be used to maintain the allocation ratio of 2:1 in favour of the intervention and the balance within stratification variables. We intend to use an unequal allocation ratio during this feasibility study due to there being more uncertainties about the interventional arm than the control arm. Development of the intervention and adjustment to the different diagnosis groups will benefit from providing more data from the feasibility study on the intervention arm specifically.
Intervention
The intervention group will receive the Better Living with Non-memory led Dementia programme, an 8-week duration 6-module educational programme covering the following topics: 1) welcome to the programme and what to expect from it, 2) understanding the disease, 3) how to provide better support for the person with dementia, 4) how to look after the caregiver’s own mental health, 5) where to find additional sources of support and 6) an introduction to the value of support groups. At the end of modules 2, 3, and 4 participants will be asked to complete a real-life task to put in practice skills learned in the specific module (e.g., approaching a friend and explaining the disease in lay terms). After that, every participant will be encouraged to reflect on that experience and share it with a programme facilitator via email. The facilitator will be a member of the team, with no clinical training, who will interact with the participant via email sticking to principles of active engagement (a script has been developed on this purpose based on principles of active listening adapted to written communication (see Appendix II). The facilitator will not act as an advisor as to the use of therapeutic techniques. All course modules will be printable using the pdf download button on the course’s page.
Intervention development
This educational programme was developed following the MRC guidance for development of complex interventions (25,26,31) comprising the following stages:
Stage 1 Identifying existing evidence and interventions.
The initial idea about the development of this educational programme was prompted by the experience of the research team (consisting of academics, clinicians and support workers specialised in non-memory led dementias) interacting with people with PCA, PPA and bvFTD and their families over the years. Rare Dementia Support, a specialised national support service for these types of dementia, is run by our group at UCL. In this first stage of the development the authors conducted a revision of the most up to date literature on the broad topic of provision of online support for caregivers of people with dementia, extracting information about relevant variables such as: intervention content, theories used for logic models, information about skills taught to caregivers in previous interventions, resources used to support learning, outcome measures, barriers and facilitators to caregiver engagement, details about how other blended interventions have been articulated, whether they work and if they are sustainable and have achieved successful implementation. A second step involved the revision of existing tools (both for online and face to face delivery). Since most online interventions fail to reach the implementation phase(32) it is not surprising that we found far more papers describing the efficacy of the tools than tools themselves which were available for the research team to test. The quality of the studies conducted also varied in quality. For the purpose to inform our intervention, we scrutinized a selection of studies that met the following criteria 1) had a comparator group and 2) had a N above 20 in each arm (either RCT or pilot studies) (21,33–47). In a third step, we drew on the experience of colleagues leading the development or administering online interventions. We therefore held a series of meeting during the evidence gathering phase. We met with research colleagues involved in the development of Communication Bridge (ref), Partners in Balance (ref) and Rhapsody (ref), three tools recently developed or adapted for people with young onset and non-memory led dementias. Meetings with NHS clinicians working in memory services and IAPT services and experience in the use of digital tools were also held. Multiple websites offering online information, training, therapy and support were reviewed to inform web design, participant flow and user experience (e.g., https://www.dementiacarecentral.com, https://www.beatingtheblues.co.uk, https://www.partnerinbalans.nl/statics/de/).
Stage 2 Identifying theory.
This educational programme is informed by theories of self-efficacy (48), behaviour change (49), coping theory (50) and social learning theory (48). Self-efficacy is the person´s belief that one can perform competently and capably in given situation (48). Caregivers with low self-efficacy beliefs tend to focus on their personal inefficiencies, the difficulty of the task and the negative consequences of failure. Self-efficacy is an important component of behaviour change. The COM-B model of behaviour change (49) argues that behaviour comes from an interaction of three factors: capability to perform the behaviour, opportunity, and motivation to carry out the behaviour. Interventions for caregivers need to alter one or more of these three elements to achieve change. To select the most suitable behaviour change techniques (BCTs) for our intervention we used the Theory and Techniques Tool (51) (https://theoryandtechniquetool.humanbehaviourchange.org/tool). This Tool also provides potential links between the BCTs and mechanisms of action, which informed the format and content of the manuals of our programme. Coping refers to the strategies that individual use to manage stress (50) and these strategies may in turn affect the overall well-being of caregivers (52). In the context of caregivers of people with dementia, coping strategies and social support can act as potential mediators between caregiver stressors and health outcomes (53). Lastly, social learning theory (48) posits that people learn from each other from observing, modelling, and imitating their behaviour. This latter theoretical framework influenced the format of the resources used in this educational programme (e.g, abundant vignettes and case studies, links to recordings of support group meetings and interviews with caregivers).
Stage 3. Modelling process and outcomes.
Information from stages 1 and 2 was pulled together to develop a logic model (shown in Fig. 2). It was decided that the educational programme would consist of 6 modules or manuals following the structure of previous interventions for caregivers (ref). The broad content of the educational programme manuals was decided based on the intervention targets and the co-production work with people with lived experience described in the section below. The specific way to bring that content to life was shaped by the behaviour change techniques and mechanisms of action that were deemed to drive effective change and are listed in Fig. 2. According to our logic model, the outcomes of the educational programme are expected to include changes in caregiver self-efficacy, relationship with the care recipient, psychological status, quality of life and health and social health. The group of people with lived experience who co-produced the programme manuals also contributed to inform the clinical outcome measures used in the pilot feasibility trial (see section below).
Development of the manuals: coproduction with experts by experience
The 6 modules/manuals of Better Living with Non-memory Led Dementia have been coproduced with a group of experts by experience (EbE) (people with dementia and/or their relatives) through an iterative process with investigator ASG. Members of the organisation Rare Dementia Support were invited to join the co-production group. Distribution of a Call for Experts by Experience fliers was used for this purpose (these flyers were co-produced with a couple of EbE and can be found in Supplementary material). Twenty-one EbE joined the co-production group between end of 2019 and 2021. Due to the Covid-19 pandemic, most meetings took place over videoconference and, exceptionally, phone or face to face, (17 videoconferences, 1 phone call and 1 face to face meeting). There was also extensive correspondence by email between the EbE group and ASG. The subgroup of EBE working on the manuals for bvFTD was entirely composed by relatives of people with bvFTD and meetings took place as a group. This was due to the difficulty to engage people with bvFTD in the process given the behavioural features and lack of insight characterising the condition. In the groups working on the PCA and PPA, there was representation of people living with the condition. To support the participation of the person with dementia, these meetings took place one-to-one between ASG and the person living with dementia and their relatives. Briefly, the procedure for the meetings involved 1) sending supporting material before the meeting and information about how to prepare (created following easy-to-read and inclusive guidelines), 2) asking about preferred forms of communication prior to the meeting, 3) sending questions in advance to be asked during the meeting, so the person with dementia could have time to think and write down their responses if needed, 4) sending minutes after the meetings in an accessible format for the person with dementia and their relatives to review together. This last step proved an important opportunity to correct inaccuracies and misunderstandings.
In the early stages of the development of the manuals, the work with the EbE group focused on the structure of the manuals and main messages to come across. For instance, one of our EbE with PPA wrote “I would like people to understand that this is the way I speak, I have a condition, people need to be more patient”. The husband of a person with PCA said about the manuals “they should be written in a positive way and convey hope”. The EbE group provided extensive advice about what symptoms to address in the programme and how, top priorities in management of symptoms, what strategies were more important to consider, they shared real situations to be include as vignettes, made suggestions about options for respite and support, identified common barriers to implement symptom management strategies and shared tips of how they had overcome them, how they managed difficult emotions (e.g., shame, guilt), deal with legal issues, relationships with others, how they introduced companions and advice on what vocabulary to use in the manuals. The group preferred that all the material and resources from the course was also provided in a printable format (not only on the web) so people could share it with relatives and friends. The group also advised on the development of the website to host the education programme. Wireframes of the website were shared with the group either in videoconference meeting or by email and their feedback was incorporated. In the late stages, drafts of the manuals were circulated by email for EbE to review. It was predominantly EbE who were relatives, and not people with dementia, that contributed to this last round, and they sent feedback as notes in emails or direct corrections of the drafts of the manuals (either typed or handwritten). The group would also share resources they had found useful to understand their relative’s condition or to manage symptoms or their own well-being. Some members of the EbE group also contributed to the final proofreading.
The work of co-production took place between October 2019 and March 2022. The final draft of the manuals also received input from educational group facilitators and went through several rounds of proofreading with academic and clinical colleagues.
Content of the manuals
Six manuals per phenotype (PCA, PPA, bvFTD) resulted from the co-production phase, making 18 in total. The first 3 manuals for each phenotype are phenotype specific. The fourth one is common across phenotypes since it is focussed on caregiver wellbeing. The fifth manual in each set contains a list of resources relevant across phenotypes (e.g., charities and organisations that can provide support) and the sixth one is an introduction to phenotype specific support groups with links to three support group webinars each. Together, the 18 manuals contain 49 purpose-build illustrations, 53, vignettes and 44 links to external resources (e.g., interviews, videorecording of support meetings).
Comparison
The control group (wait-list) will receive treatment as usual, consisting of explicit signposting to the publicly available Rare Dementia Support website and will continue with any kind of support the participants may already being receiving (e.g. psychological support, online information, support groups, etc). They will be given access to the intervention at the end of the study.
Data collection, management, and analysis
Demographics
After giving consent, participants will fill in an online survey with demographic information before start completing the baseline measures.
Feasibility measures:
The following feasibility measurers will be collected:
a) Recruitment process
-
Number of people agreeing to be sent information about the study.
-
Proportion of participants who agree to participate.
-
Proportion of eligible participants who agree to participate.
-
Potential inequalities in recruitment feasibility will be assessed by comparing basic data on ethnicity age, gender, and diagnosis of care recipient between Rare Dementia Support members, those who are screened as eligible for, and those who consent to the study.
b) Retention
-
The analysis will consider the points from the CONSORT checklist for randomised pilot and feasibility trials (54) to ensure that all topics are being covered. Values for eligibility rates, recruitment rates, attrition rates and withdrawal rates will be reported using the participant flow data collected within the study. This will be evaluated overall and per group.
-
Furthermore, details on reasons for ineligibility and non-recruitment will be reported within a table along with their related patient frequencies and percentages. Information on withdrawals and non-respondents will be presented including reasons where applicable and time points during the trial.
c) Feasibility of measurement tools
-
Time taken to fill in questionnaires (with consideration for outlier times to be those where participants have left the questionnaire to complete at another time).
-
Missing data from questionnaires (questionnaires will be implemented using the Qualtrics function that reminds participants when they have not filled in an item, but will not force a choice)
-
Follow-up response rates (8 weeks and 3 months post randomisation).
Acceptability measures:
We used the definition of acceptability recently proposed by (55) to decide on the following acceptability measures:
Prospective acceptability:
-
Burden (reasons for not taking part or discontinuation)
-
Percentage who completed baseline measures agreeing to be randomised and reasons for not taking part in randomisation.
-
Answers to qualitative questions in baseline questionnaire about how they feel about the intervention (e.g., how do you feel about taking part in this course?), burden (e.g. how much effort do you think will be involved in taking part in this study?) and ethicality (e.g., do you have any ethical concerns about taking part in the course? If yes, what would these be?) (see Appendix III).
Concurrent acceptability:
Retrospective acceptability
It will be measured in two ways:
1. To capture views of everyone participating in the study, follow up measures will include a brief acceptability questionnaire designed for to assess opportunity costs and perceived effectiveness. The health economic questions used for the control and intervention group will differ: participants in the intervention arm will be asked questions on resource use and questions on willingness to pay while those in the wait-list control arm will be asked about resource use only. Participants in the wait-list control arm will be asked about acceptability of filling in the measures at each time point). Participants in the intervention group will be asked about acceptability of the intervention (see Appendix III).
2. A more in-depth qualitative interview focussed on the above headers (as well as mechanisms of change) will be given to a smaller number (n ~ 15) of purposively sampled individuals.
Proposed clinical measures.
The range of clinical outcomes encompassing self-efficacy, relationship with care recipient, psychological status, quality of life and health and social health, will be assessed through the following measures at baseline line, 8 weeks, and 3 months post randomisation. An additional section about health economics will be added to explore the feasibility of evaluating the costs of the educational programme in a larger trial. We will be using the following measures:
- WHO 5 Wellbeing Index(56)
- GAD-7 (57)
- PHQ-9 (58)
- De Jong Gierveld Loneliness Scale (59)
- Lubben Social Network Scale (60)
- Pearlin Mastery Scale (61)
- Caregiver self-efficacy Scale (62)
- Dementia Management Strategies Scale (63)
- The Quality of carer-partner relationship scale (64)
- Health economics questions (developed by the health economic collaborators, RTE and BA) (See Appendix III).
The recommended SPIRIT schedule for the enrolment or participants, administration of the intervention, and assessment time points is shown in Table 1.
Table 1
SPIRIT participant timeline with time schedule of enrolment, intervention implementation, and assessment time points.
|
STUDY PERIOD
|
|
Enrolment
|
Allocation
|
Post-allocation
|
Close-out
|
TIMEPOINT**
|
-t1
|
0
|
8 weeks
|
3 months
|
tx
|
ENROLMENT:
|
|
|
|
|
|
Eligibility screen
|
X
|
|
|
|
|
Informed consent
|
X
|
|
|
|
|
Allocation
|
|
X
|
|
|
|
INTERVENTIONS:
|
|
|
|
|
|
Better Living with Non-memory led Dementia
|
|
![](data:image/png;base64,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) |
|
|
Waitlist control
|
|
|
|
|
|
ASSESSMENTS:
|
|
|
|
|
|
Demographics
|
|
X
|
|
|
|
Feasibility measures
|
|
|
|
|
|
Recruitment
|
X
|
|
|
|
|
Retention
|
|
X
|
X
|
X
|
X
|
Measurement tools
|
|
X
|
X
|
X
|
|
Acceptability measures
|
|
|
|
|
|
Prospective
|
|
X
|
|
|
|
Concurrent
|
|
|
X
|
X
|
|
Retrospective
|
|
|
X
|
X
|
X
|
Clinical outcome measures
11. WHO 5 Wellbeing Index
12. GAD-7
13. PHQ-9
14. De Jong Gierveld Loneliness Scale
15. Lubben Social Network Scale (
16. Pearlin Mastery Scale (
17. Caregiver self-efficacy Scale
18. Dementia Management Strategies Scale
19. The Quality of carer-partner relationship scale
|
|
X
|
X
|
X
|
|
Health economics
|
|
|
X
|
X
|
|
*Recommended content can be displayed using various schematic formats. See SPIRIT 2013 Explanation and Elaboration for examples from protocols. |
**List specific timepoints in this row. |
Statistical analysis
Quantitative data
All analysis will be guided by the principle of intention to treat. There will be a preliminary analysis of intervention outcomes. Point estimates and 95% confidence intervals will be calculated using adjusted means from the analysis and used to estimate standard deviations and effect sizes for continuous data. Means and standard deviations of response rates for count data and proportions in each category will be provided. These will be used to confirm the sample size calculation for a definitive study. Dependent on the recruitment within each diagnosis group we may consider presenting preliminary results within diagnosis groups. Exploratory analysis will be performed to determine the most appropriate model of analysis for a definitive RCT, including consideration of the further possible covariates and factors to be included in an analysis model. All quantitative analysis will be completed using Stata 17, SPSS v25 and R version 3.
Qualitative data
Semi-structured online interviews for collecting data on participant experience and programme usability will be recorded via Zoom and transcribed verbatim. Transcripts will be analysed using thematic analysis (65). Members of the research team with experience in thematic analysis will be involved in the qualitative analysis of data and will assess data saturation throughout the Zoom interview process.
Progression criteria
Progression criteria to a definitive trial for participant recruitment will be decided ad hoc (66,67) and be as follows: less than 25% of the target samples size (30) recruited, will be a criteria for not to progress to a definitive trial, 75–100% of the target sample recruited, will support progression to a definite trial, 25–74% of the target sample recruited will warrant progression to a definitive trial only if necessary adjustments to barriers of recruitment are feasible. Procedures and processes will be reviewed and adapted to support proceeding to a full trial.