Assessing people with dementia participating in cognitive stimulation activities—A qualitative pilot video analysis exploring the importance of facilitating the participation

Abstract Background This pilot video analysis was part of a feasibility control study, which aimed to gain information about the size and variability of the changes in outcome measures to plan a substantive effect study. It compared a cognitive stimulation programme named Lifelong Learning with other existing dementia services. Objective The pilot video analysis explored how facilitation is performed, when assessing people with dementia with standardized measures, to ensure their participation in research. Design A test battery of five measures (Mini‐Mental State Examination (MMSE), Quality of Life in Alzheimer's Disease Scale (QoL‐AD), General Self‐Efficacy Scale, Rosenberg Self‐Esteem Scale and Hawthorn Friendship Scale) was used. Each assessment was video‐recorded. The findings from a microanalysis of 10 videos are presented in this article. Setting The study involved 55 active participants with mild‐to‐moderate dementia in six municipalities in Northern Denmark. Results The identified themes related to supportive facilitation: Positive facilitator strategies; Creating a safe and comfortable environment; and to dilemmas in facilitation: Balancing multiple dilemmas and Balancing the MMSE test. Discussion Results are discussed in relation to using standardized measures. Conclusion The quality of facilitation when using standardized measures is of great importance as it may influence the participant, the assessment and the answers given. The facilitation role needs to be thoroughly planned and executed with ethical consideration to improve the participation of vulnerable groups in research and ensure a person‐centred approach. Patient or public contribution The identified measures were chosen based upon previous qualitative results and user‐involvement workshops with people with dementia.


| INTRODUC TI ON
Dementia research has developed over time with various areas channelling the focus of research to explore causes of dementia, treatment, understanding and relational aspects of dementia care. 1 'Treatment-oriented' research explores different types of treatment, including how they minimize any decline or support maintenance of a person's condition. 1 Another aspect that is rarely discussed is how participants with dementia experience an assessment. Dementia is a progressive cognitive condition. When considering the design of research with people with dementia, the impact on that person's memory, decision making, understanding, concentration, mood and problem-solving may be pertinent to ensure the assessment does not become a burden, 5,14 causing fatigue or distress, as they may have difficulties in paying attention for a longer period. 15 The design may also be a barrier to participation if not all measures are completed. 6,16,17 Furthermore, when using measures with people with dementia, ethical factors, such as ways of supporting vulnerability and dignity, are relevant to consider. 18 This influences the identification of reliable and appropriate measures. 5 A recent review of cognitive stimulation interventions identified research uses 5-6 measures on average, but this can be as many as 10-15 in a single study. 19 No comment is made in these studies as to what the appropriate number of measures is to use, and little is stated about the relevance of the measures chosen or how the person with dementia experiences them.
Furthermore, little information is provided on how assessments are conducted in accordance with the instructions, 20 how the tests are administrated and whether the total time of the battery is taken in consideration to minimize the burden for participants. 20,21 It is therefore difficult to judge the quality of the assessments, which may potentially influence the quality of the results. This paper trials a new approach, using video recordings, to explore the facilitation of measures in the assessment process, providing new insight into this little-studied aspect of conducting research together with people with dementia to ensure their participation in research.

| ME THODS
This paper presents findings from a pilot video analysis, which was part of a feasibility study, to gain information about the assessment of a Lifelong Learning Service for people with dementia. The study compared an intervention group receiving Lifelong Learning with a control group, participating in treatment as usual (services at day-care centres, etc). The study was conducted in six Danish municipalities. The Lifelong Learning concept is an on-going cognitive stimulation programme, aiming to support cognition, decision making, activities of daily living and social engagement. [22][23][24][25][26] The study followed the Standards for Reporting Qualitative Research (SRQR checklist for qualitative studies).

| Recruitment
Recruitment was supported through the staff at each of the services, who were guided on the study´s inclusion criteria that participants should: have a dementia diagnosis; participate in a service; and be able to consent. In total, 88 participants were recruited for the feasibility study. The dropout/exclusion rate was 37.5% due to progression of dementia, hospital admission, relatives' illness, non-dementia diagnosis and death. Participants were excluded if they attended less than 10 sessions. This resulted in 55 participants (n = 30 intervention group and n = 25 control group) with a median age of

| Measures
The measures were identified by a user-involvement workshop with people attending the lifelong service and interviews with service staff. This guided the choice of the measures used, which were the: Mini-Mental State Examination Test (MMSE-2) 27,28 ; Quality of Life in Alzheimer's Disease Scale (QOL-AD) 29,30 ; General Self-Efficacy Scale 31 ; Rosenberg Self-Esteem Scale 32 ; and Hawthorn Friendship Scale. 33 The assessments were facilitated by first, second and third author, all with backgrounds in nursing.

| Video analysis
Each participant was assessed pre-and post-intervention, over 5-6 months. Each assessment was video-recorded. An adapted 25 version of Ridder's 34 video analysis was used for the data analysis.
The aim of this pilot analysis was to answer the research question: How is facilitation performed, when assessing people with dementia with standardized measures, to ensure their participation in research?
During the analysis phase, it was important to consider what the participants responded verbally and non-verbally. The rationale for using video recordings was to ensure these nuances of social interaction were captured. The value of video is the ability to watch and re-watch interactions in a way that observation alone does not enable. 35 After reviewing all videos (n = 55 pre-assessment and n = 55 post-assessment), a stratified sample (videos divided into subgroups based upon characteristics) of ten videos (n = 10) was used to include the following: pre-assessment videos; five control and five intervention participants; example from each site; level of dementia (high and low MMSE score); and diversity of gender. The  Table 1) to avoid recall or familiarity with the measures.
All videos were watched multiple times to identify codes for a video graph. An analysis framework was developed and tested using one video. This framework was adapted with additional coding options included for the remaining videos to identify moments with facilitation. 34,36,37 A graph was created for all videos to log each interaction and questions asked, and identify moments for a deeper analysis (see Figure 1).
The video graph enabled the selection of clips for a microanalysis. Notes had been made on the graph of moments, which showed typical/atypical situations of facilitation. 37 Therefore, 13 clips were chosen. The length of the clips reflects that interactions were often short, with participants responding quickly (see Table 2).
Ridder's 34 microanalysis process was as follows: watching each clip to get an impression; identifying 'meaningful events' and writing what is seen and heard; writing a 'subjective assessment'; writing a reflection; and writing an evaluation. A section was added to allow quotes to be included. 25 Table 3 shows an example of this microanalysis. The final analysis stage was to draw the themes from across each microanalysis and the video graph.

| E THI C S
The Ethics Committee of Northern Denmark was informed about the study. It was judged that no further application was needed in relation to LBK nr 1083 of 15/09/2017 definition of a Health Science Research Project and the Committee law § 14, stk. 1, jf. § 2, nr 1-3. The Helsinki Declaration was the ground upon the study was conducted. 38 All information material and consents were developed based upon the author's experiences of producing accessible documents for people with dementia. 23,25 This allowed the participants to be informed of the study prior to agreement to participate to sign the consent themselves in collaboration with their relatives or service staff. Furthermore, an on-going consent inspired by Dewing 39 was used at the post-measurement to ensure continued participation.
Due to the requirements of confidentiality and anonymity, the video recordings are not allowed to be shown. All names used within the article are pseudonyms.  The facilitator helped Lone to break down the required answer into a step-by-step process that was manageable and used the response paper as a visual aid. The highlighted words indicate where the facilitator placed emphasis on words to support the distinction between agreeing and disagreeing.

TA B L E 1 Details related to the video chosen
Another facilitating strategy was to explain that there were no right or wrong answers, with the aim to minimize any pressure in the

Supporting text Transcription of clip
Researcher (R) asks Participant (P) to repeat three words. R is looking at P as she speaks, and P looks at R. R and P then both look at the paper. P says yes and responds correctly straight away.
It seems a relaxed start to the MMSE.
P looks as though she is concentrating and repeats the words quickly, this could be a way of not forgetting them!
There is a small space between R and P, which may make the paper more obvious in the situation and for P to see what R is writing.
situation. However, for the MMSE, participants sensed there was a right or wrong answer, and facilitators found that reassurances enabled participants to recognize that it was okay if they could not remember, thereby reducing perceived stress during this assessment.

| Creating a safe and comfortable environment
Overall, the assessment seemed quite relaxed with a positive atmos-

| Balancing multiple dilemmas
Multiple dilemmas were identified in the facilitation. Some facilitators were seen to provide their own interpretation to some of the  However, this could put a strain on the facilitator, who wanted to support those who were showing signs they were struggling.

| Balancing the MMSE test
The facilitators experienced needing to carefully assess how best to react during the measurement. At times, the facilitators repeated a question, when there had been an extended pause, or the participants asked the question to be repeated. They knew this meant a lost point but felt that refusing to repeat the question could add discomfort for some, especially those already exhibiting signs of anxiety. This could be criticized for disregarding the MMSE test.

| D ISCUSS I ON
This study provides insight into the process of administering vali- Delphi study 42 identified that a 3-point scale was more accessible.
They found it was preferable to use a scale that asked participants the importance of a situation, rather than relying on extreme responses. Morbey et al 42 also found the design of the questionnaire, for example the use of colour, layout, font size and use of symbols, influenced on how a person with dementia responded. These factors may be relevant to incorporate in standardized measures, many of which are not yet taking this into consideration. As this study found, participants were drawn to the paperwork and making this more dementia-friendly could be a way to support their engagement and participation.
The MMSE used in this study was found to be challenging for both the participants, in providing the right answers within a limited time, and the facilitators, allowing limited support and feedback for the participants. In some situations, the participants showed signs of agitation. This is also identified by Hellström et al, 43  some of these to clarify meaning. These adaptive approaches were echoed in this study as facilitation was matched to individual needs and responses. This flexibility in facilitation is an important way for people with dementia to be part of such research projects and to feel supported in responding as accurately as possible. This highlights that awareness of how dementia symptoms may affect involvement, being relaxed and being flexible in the moment is important aspects of conducting research with people with dementia. 55 Training may therefore be needed as researchers may be faced with distress, anxiety, grief and declining health when involving people with dementia in research. [56][57][58] This ensures researchers perform effectively in research. 42,59

| LI M ITATI O N S
A limitation is that only ten videos were analysed in this pilot analysis. Thus, it is not possible to generalize these findings. However, the study highlights areas for future research, such as role of facilitation and its potential impact on outcomes. Another limitation is the difference in how facilitation was provided due to the facilitators' experience and knowledge in dementia research. More training in the measures and how to facilitate the process might have improved the quality of the study.

| CON CLUS ION
This study identifies the importance of careful facilitation when involving people with dementia, illustrating, the challenges and dilemmas that might occur during an assessment and how person-centred facilitation supports participation. This paper concludes that the quality of facilitation may influence the participant, the assessment and the answers given. It is therefore important that the facilitation role is thoroughly planned and executed with ethical consideration as its foundation.

ACK N OWLED G EM ENTS
We would like to thank all the participants in the study for their big interest and effort in completing each measure. We also want to thank all the professionals in the municipalities in Northern Denmark for their big effort in informing, recruiting and providing us with demographic data of each participant. Also a thank you to Professor Jackie Campbell for statistical advice and work.

CO N FLI C T O F I NTE R E S T S
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Anonymized data are available on request due to privacy/ethical restrictions.