Preliminary Randomised Evaluation of Singing in Dementia (PRESIDE)

Background and study aims Dementia affects about 800,000 people in the UK, with the number expected to rise. As well as searching for medical treatments for dementia, there is a need for helpful social activities to support people with dementia to live well with the condition. People with dementia are at risk of social isolation and mental health problems, and family carers can feel unsupported and over-burdened by their role. Studies have suggested that group singing can improve mood, memory and relationships for people with dementia, and establish support networks which help carers. The shared activity of singing together may have benefits for the relationship between person with dementia and carer too. However, to date no large scale studies about community singing and dementia have been conducted. This study is a feasibility study, meaning that it aims to try out a study design to see if it would work on a larger scale. In particular, we want to see if we can recruit enough people to take part in the study, and whether they will remain in the study for long enough to collect all the data we need. Who can participate? Patients who have received a diagnosis of dementia, who are willing to join a singing group, and who have a carer who is willing to join the study with them. What does the study involve? Participants will be randomly assigned to either attend group singing straight away, or to wait for 10 weeks before attending group singing. We will collect data about their quality of life, mood, and cognitive function at several time points, so we can compare the differences between people who attend singing straight away and those who wait. The data we collect from this feasibility study will allow us to plan a larger trial of singing for people with dementia. What are the possible benefits and risks of participating? Participants in the study will have the opportunity to attend singing groups. Many attendees at these groups report finding them stimulating and enjoyable. However, not everyone will necessarily enjoy the singing groups even if others do (for example, if they do not like the songs chosen for a certain week). Participants will also be asked to complete questionnaires and tests three times during the study. Every effort will be made to ensure this experience is not too tiring or difficult, but some participants may find it burdensome.

PURPOSE The purpose of PRESIDE is to test the feasibility of conducting a randomised trial of community singing in dementia that investigates the effects of singing on quality of life and explores the impact singing groups have on the relationship between the person with dementia (PwD) and carer. PRIMARY OBJECTIVE The primary objective of the PRESIDE study is to test the protocol for a two-arm, parallel group superiority randomised trial of community singing in dementia with a waiting-list control design. The relevant indicators are recruitment, retention, and the acceptability of the trial design. SECONDARY OBJECTIVES The secondary objective of the study is to explore the impact of group singing on PwD and their carers. This would improve understanding of the mechanisms of action in group singing for PwD and potentially indicate suitable outcome measures for full-scale trial. DATA COLLECTION A mixed-method design will be applied to collect data using questionnaires, semi-structured interviews and observations. PRIMARY END POINT The attainment of the feasibility criteria at 6 months is the primary end point for PRESIDE. Proposed criteria for proceeding to a full trial: 70% recruitment; 70% retention of participants over 3 months; and acceptability of waiting-list controls shown by 70% singing group take-up post waiting-list period. SECONDARY ENDPOINT The secondary endpoint is the measurement of social independence, loneliness, quality of life, activities of daily living, mood, musical engagement and cognition of PwD; the measurement of quality of life, loneliness, mood, caring relationship and experience of carers. Each dyad's use of health social care services and resources will also be documented. Follow-up data will be collected at 3 months after baseline to compare the difference between the intervention and control groups, and 6 months to investigate the sustainability of any treatment effect. Further, to better understand the experiences of dyads who attend the singing group, qualitative data will be collected using interviews and participant observation. STOPPING RULES AND DISCONTINUATION The intervention will be stopped if a Serious Adverse Event (SAE) occurs which is judged to be directly related to the trial and likely to affect other participants. This decision will be taken by the DMEC. Individual participants may withdraw from the trial at any time and their reasons for withdrawal will be recorded if they are willing to give them. The Trial Steering Committee would decide whether to continue if recruitment figures were poor. RANDOMIZATION AND BLINDING Participants will consist of a PwD and their carer, with stratification on MMSE score and minimisation on age, sex and MusEQ score. Baseline data will be collected by members of the Clinical Research Network (CRN) for all recruited pairs prior to randomisation. Randomised pairs will be assigned to either the intervention group or the waiting-list control group. Follow-up data collection at 3 and 6 months will be undertaken by members of the CRN, who will be blind to group allocation. However, there is the possibility that since the study participants will know their own allocation, they may reveal this during the follow-up data collection. Any unblinding of outcome assessors which occurs in this way will be recorded, as it has implications for the feasibility of the study design. Randomisation will be performed by a web application created by the trial programmer at the University of Nottingham Stroke Trials Unit, using web and database servers provided by University of Nottingham Information Services (IS). The research team will be divided in the data management group, who will not be blinded to group allocation (intervention or waiting list control) for logistical reasons (data collection, anonymising data and running of the study), and the analysis group, who will be blinded throughout the study and have no contact with study participants. Allocation will only be revealed to blinded members of the research team in the case of emergencies (i.e. SAEs). In this case the Chief Investigators and Data Monitoring and Ethics Committee will be informed. PARTICIPANT AND STUDY DURATION The study will last two years. Participant involved is anticipated to be between 9 and 12 months. The end of the trial will be the final assessment of the last participant with dementia and their carer. After recruitment and baseline data collection, participants will either attend a singing group for ten weeks or wait for 12-13 weeks before attending a singing group for ten weeks. They will receive follow up at 3 months and 6 months after baseline. RECRUITMENT Recruitment will be multi-channelled: PRESIDE posters and a cover letter will be sent for distribution to people who have recently been diagnosed with dementia via NHS Memory Assessment Services and PwD who are currently on the waiting list for Singing for the Brain sessions via the Alzheimer's Society. Local community organisations, charities and local press (Radio Nottingham, Nottingham Post and similar) will also be utilised, and social media advertising will be part of the recruitment campaign. Recruitment will be undertaken using the online platform "Join Dementia Research". Regional recruitment may also take place through having PICs within primary care, with the support of the NIHR CRN for GP surgeries. Eligible participants will be screened from a database by a member of the direct care team at that practice. Participants that meet the inclusion and exclusion criteria will be invited to join the study through letter invitation from their GP surgery. Participants who contact the research team post invitation, will then be provided with study specific information by the research team. Participants may contact the research team directly to express interest in taking part in the study, for example, in response to advertisement. EXPRESSION OF INTEREST AND PARTICIPANT INFORMATION Potential participants who are eligible and interested in taking part will receive a PRESIDE Participant Information Sheet (PIS). They will have the opportunity to read the information and to speak to members of the research team. If they wish to proceed with involvement in the study, they will receive a visit from a PRESIDE research member. This will take place at a convenient location, such as their home or other venue. The researcher will fully explore the implications of participation in the study and confirm their eligibility. They will be provided with a consent form which they will sign if they choose to continue with participation. Following consent, the researcher will enrol the participant dyad and complete baseline outcome measures at the same visit. The research or their nominee (such as the participant's usual care team) will inform the participant, of all aspects pertaining to participation in the study. If needed, the usual hospital interpreter and translator services will be available to assist with discussion of the trial, the PIS, and consent forms. However, consent forms and information sheets will not be available in other languages since understanding English is an inclusion-criteria for this study. It will be explained to the potential participant that entry into the trial is entirely voluntary, that their treatment and care will not be affected by their decision and that they can withdraw at any time. Participants will be made aware that should they withdraw, anonymised data already collected cannot be erased and may still be used in the final analysis. REMOVAL OF PARTICIPANTS AND PARTICIPANT WITHDRAWAL Temporary discontinuation might happen if either participant (PwD or carer) from the dyad is unavailable, indisposed or does not want to join the session on a specific day. Participants may withdraw from the study at any time, without providing any reason. Carer participant withdrawal would result in the participant with dementia being re-paired with a different person. Withdrawal from a PwD. would result in withdrawal of the dyad. However, the carer participant would still be able to attend the allocated singing sessions if they wished. Participants may also be withdrawn from the study at the discretion of the Investigator. Participants who have lost capacity to consent but have remained part of the study (see section on Informed Consent below) will be immediately and permanently withdrawn from the study if they show observable signs of dissent or discomfort or it is reported from their carer. Participants will be made aware that withdrawal will not affect their future care. Abrupt termination of the study is unlikely to affect the participant safety as the study focus on an entertainment social intervention. Failure to recruit sufficient participants could lead to termination of a specific regional group. In this instance, participants would be offered another group to join. INFORMED CONSENT The Consent Form will be signed and dated by the participant before they enter the trial. PIS and, when appropriate, a dementia friendly booklet will be provided for PwD and carers. Fully informed consent will be taken at the initial study visit by trained and experienced members (CRN or PRESIDE) that hold a valid Good Clinical Practice Certificate (NIHR). Absolute care will be taken to ensure that participants fully understand the study and their involvement. PwD who do not have capacity to provide informed consent at the start of the study will be excluded from the study. Due to the progressive nature of dementia, Dewing's 'process consent model' is applied so there is periodical assessment of capacity and consent is continuous and ongoing. If a participant with dementia loses capacity during the trial period, where appropriate, the carer participant in the dyad would be asked to act as consultee. If the carer participant is a paid carer, then this would not be appropriate and a consultee declaration from the most appropriate person for this individual would be sought. The consultee will be provided with information about their role within the parameters of the study and will be asked to make a judgement about whether they would have wanted to continue to take part in the study. If they judge that the PwD would have wanted to continue to take part, the consultee will sign a declaration form to indicate this. In the unlikely event that a PwD was judged to have regained capacity during the trial period, the research team would discuss with them to ascertain whether they wished to continue in the study. We shall treat the consent of carers equally rigorously. Should a carer lose capacity to consent during the trial period, a consultee will be sought for them. In this situation, the continuation of a dyad in the study will be at the CI's discretion. METHODS Data analysis will be conducted by a statistician using Stata 16 software. All measures will be summarised with mean (S.D.) for normally distributed variables, median (IQR) for skewed variables and frequency (%) for response of categorical variable, by intervention arm across follow-up time. ANCOVA will be performed to quantify treatment effect size and its precision by means of regression modelling with baseline measures as covariate, and to derive the change from baseline score for each arm. Due to the non-independence of patient-carer dyad measures and data being measured repeatedly, multivariate multilevel modelling will be performed to account for the association between same kind of outcome measure of patient-carer data and repeated data. Data will calculate recruitment rate, retention rate and attrition rate, exploring patterns of missing data and drop out from the waiting list control group. All results will be used to inform future definitive trial design. All data will be analysed on UoN computers and backed up to the UoN servers.

Case-only participant groups comprising of those with dementia and carers. The sessions will be delivered by an experienced leader of singing in dementia, supported by volunteers, and an instrumentalist (guitar or piano). The intervention will be delivered on one day per week to 3 groups of up to 27 couples (not necessarily spouses) in each of which pairs one person has dementia. Singing will last 60 minutes, with 30 minutes for socialising before or after that. The intervention groups will start singing immediately; the control groups will wait 12-13 weeks before starting. The study will last about two years, with participant involvement expected to last between 12-15 months. The intervention will be delivered in three geographical areas, with a singing intervention group set up in each of these areas. The intervention groups may continue into the second round.

In pragmatic terms, masking or trial 'blinding' is not possible nor desirable for this study; instead we will aim to keep assessors and trial statistician in ignorance of participants' allocations, and to monitor success by asking assessors to judge those allocations & their confidence therein.

Intervention involves singing groups with person with dementia and carer. Please see earlier descriptions.

Recruitment rate is the number of participants recruited throughout the study compared to the set number of participants required to reach statistical power. To assess the feasibility of gaining recruitment into community singing groups with the potential of having a waiting-list control group is investigated. Recruitment of all participants into the study against target figures will be reported as a recruitment rate (as a percentage or ratio) and form an independent unit of measure to provide an indication whether or not recruitment targets are likely to be achieved should a randomised control trial be pursued.

Retention rate is calculated by measuring the number of participants who were recruited into the study against the number of participants that completed the duration of the study (either by percentage or ratio). Retention rate of all participants as a unit of measure throughout the two-year period will be analysed. Waiting list control group participant retention data will be compared against intervention group participant retention data which will provide an understanding as to whether participants would be willing to wait to participate in the group singing activities. This provides a unit of measure to calculate the feasibility of the waiting list control study design. Participant retention rates (using the unit of measure outlined above) will be analysed throughout the two-year period to further investigate if and when participants may be at risk of study withdrawal (reported as either a percentage or a ratio).

Scales measuring the impact of Dementia on CARers (SIDECAR) is a standardised psychometric questionnaire that demonstrate robust measurement properties, meeting COSMIN quality standards for study design and psychometrics. SIDECAR provides a theoretically based needs-led Quality of Life profile specifically for dementia carers and uses a scale of 0-100. SIDECAR is free for use in public health, social care, and voluntary sector services, and not-for-profit organizations. In this study SDIECAR is used to measure the impact of dementia on carers. Data is collected from carers at baseline, 3 months and 6 months.

The Geriatric Depression Scale is a psychometric standardised questionnaire designed specifically for the aged, as a screening instrument for depression with a scale from 0-30 (higher score shows a greater risk of depression, with scores between 20-30 showing severe depression). In this study, The Geriatric Depression Scale is administered to all participants and data is collected at baseline, 3 months and 6 months.

Wellbeing is measured using the psychometric standardised questionnaire Short Warwick-Edinburgh Mental Wellbeing Scale. T-Scores range from 7 to 35 and higher scores indicate higher positive mental wellbeing. In this study, Short Warwick-Edinburgh Mental Wellbeing Scale (SWEMWS) is administered to all participants and data is collected at baseline, 3 months and 6 months.

The Three-item Loneliness Scale is a psychometric standardised questionnaire designed to measure indirect loneliness. The total score ranges from 20 to 80. Higher scores indicate higher loneliness. The most commonly used categorization is the following: 20-34 denotes a low degree of loneliness, 35-49 a moderate degree of loneliness, 50-64 a moderately high degree of loneliness, and 65-80 a high degree of loneliness. In this study, the Three-item Loneliness Scale s administered to all participants and data is collected at baseline, 3 months and 6 months.

The EuroQol-5D-5L (EQ-5D-5L) is the most widely used multi attribute utility (MAU) instrument for measuring health-related quality of life in cost-effectiveness analysis. The EQ-5D-5L index scores generated using this algorithm range from -0.224 to 1, with 0, 1, and negative values corresponding to death, full health, and health states worse than death, respectively. In this study, this standardised questionnaire is administered to all participants and data is collected at baseline, 3 months and 6 months.

DEMQOL/DEMQOL-Proxy is a widely used multi attribute utility (MAU) instrument for measuring disease specific quality of life analysis. DEMQOL and DEMQOL-Proxy can provide robust measurement of disease specific quality of life in people with a diagnosis of dementia when scores are derived from analysis (using the Rasch model). The DEMQOL assessment uses 2 4-point ordinal scales (1 to 4) with example possible scoring choices being (1) a lot to (4) not at all. The self-report/interview format total scores range from 28 to 112 and the proxy version scores have a range of 31 to 124. The measure can be completed in less than 10 minutes with higher scores indicating better Quality of Life. In this study, this standardised questionnaire is administered to all participants and data is collected at baseline, 3 months and 6 months.

The Client Service Receipt Inventory (CSRI) is a research instrument designed to collect information on service utilisation, income, accommodation and other cost-related variables. Its primary purpose is to allow resource use patterns to be described and support costs to be estimated using an appropriate unit cost. Therefore, there is no singular score but a comprehensive record of the support and services received by participants. This includes any item that has resource implications, such as accommodation and any on-site care, employment status and income, formal health and social care services used, and unpaid care. From this individual level data, an aggregate picture can be created for the study population. In this study, CSRI is administered to all participants to assess their use of health and social care. Data is collected at baseline, 3 months and 6 months.

The Mini Mental State Examination is a psychometric standardised questionnaires designed to assess cognitive impairment. Patients score between 0 and 30 points, and clinical cut-offs of 23/24 have typically been used to show significant cognitive impairment. In this study, all Persons with Dementia will complete the MMSE at prior to randomisation (as recruitment of participants will take place in pairs consisting of a person with dementia and their carer, with stratification on MMSE score). Administration and data collection of MMSE scores from people with dementia will also take place at 3 month and 6 months post study start.

Independence and social engagement are important outcomes for people with dementia. The psychometric assessment of the Engagement and Independence in Dementia Questionnaire (EID-Q), provides a measure of social independence and engagement. Each response has a corresponding score (0-4). Responses should be summed to provide an overall score for the EID-Q. Subscales for the EID-Q are: 1. Activities of Daily Living (items 1-6), 2. Decision Making (items 7-10), 3. Activity Engagement (11-13), 4. Support (14-20) and 5. Reciprocity (21-26). These subscales can be summed to provide a total subscale score. Items 6, 10 and 22 are reverse scored. Therefore, scores range between 0-104 with higher scores showing a greater level of independence and engagement. In this study, the questionnaire is administered to people with dementia at baseline, 3 months and 6 months.

measured using Music Engagement Questionnaire Music Engagement Questionnaire (MusEQ), is a standardised 35-item scale questionnaire, designed to measure engagement with music in daily life. Scores range from 11-55 on the basis of a 1-5 Likert scale questionnaire. Higher scores indicate higher levels of engagement in music. In this study, this will be administered to people with dementia, at baseline only.

The Music in Dementia Assessment Scale, an observational tool designed to rate response to and engagement with music, will be completed optionally by carers before and after a sample of sessions in order to monitor "in the moment" reactions and the short-term impact of singing on people with dementia. Previous qualitative research has shown that participants often mention these short-term impacts of music participation, but outcome measurement usually focuses on longer-term change. MiDAS asks the carer to rate the person with dementia's levels of interest, response, initiation, involvement and enjoyment on visual analogue scales. Once a carer is familiar with the measure, it is rapid to complete. We will ask carers to complete MiDAS scales on a voluntary basis immediately before and several hours after their first, fifth and ninth sessions.

Inclusion Criteria: Participant with dementia: To be diagnosed with dementia which is at a mild or moderate level Aged ≥18 years To have a carer who spends at least 2 h per week with them and who is willing to attend the group Willing to join a singing group and attend weekly Able to give informed consent Able to speak and understand English. Carer: Able to speak and understand English. Willing in principle to attend the group regularly Able to give informed consent Exclusion Criteria: Participants with dementia: Lacking capacity to give informed consent Has engaged regularly in a singing group (other than religious services) in the past 6 weeks Significant hearing impairment Simultaneous participation in any other interventional study. History of severe mental illness, or alcohol/drug addiction Carers: Lacking capacity to give informed consent Significant hearing impairment Simultaneous participation in any other interventional study. History of severe mental illness, or alcohol/drug addiction

Individual participant data will be anonymised so no one will be able to identify which participant data came from. For example, demographic descriptive data.