Time to reflect is a rare and valued opportunity; a pilot of the NIDUS‐professional dementia training intervention for homecare workers during the Covid‐19 pandemic

Abstract Most people living with dementia want to continue living in their own home for as long as possible and many rely on support from homecare services to do so. There are concerns that homecare often fails to meet the needs of clients with dementia, but there is limited evidence regarding effective interventions to improve its delivery for this client group. We aimed to assess whether a co‐designed, 6‐session dementia training intervention for homecare workers (NIDUS‐professional) was acceptable and feasible. Facilitated training sessions were delivered over 3 months, followed by 3, monthly implementation meetings to embed changes in practice. Two trained and supervised facilitators without clinical qualifications delivered the intervention via group video‐calls during Oct 2020–March 2021 to a group of seven homecare workers from one agency in England. Participants provided qualitative feedback 3‐ and 6‐months post intervention. Qualitative interview data and facilitator notes were integrated in a thematic analysis. Adherence to the intervention and fidelity of delivery were high, indicating that it was acceptable and feasible to deliver in practice. Thirty of a possible 42 (71.4%) group sessions were attended. In our thematic analysis we report one over‐arching theme: ‘Having time and space to reflect is a rare opportunity’. Within this we identified four subthemes (Having time to reflect is a rare opportunity; Reflecting with peers enhances learning; Reflection and perspective taking can improve care; Recognising skills and building confidence) through which we explored how participants valued the intervention to discuss their work and learn new skills. Attendance was lower for the implementation sessions, perhaps reflecting participants’ lack of clarity about their purpose. We used our findings to consider how we can maintain positive impacts of the manualised sessions, so that these are translated into tangible, scalable benefits for people living with dementia and the homecare workforce. A randomised feasibility trial is underway.


| INTRODUC TI ON
An estimated 850,000 people live with dementia in the United Kingdom (UK), two-thirds in their own homes (Prince et al., 2014).
Most want to continue living as independently as possible and are often supported to do so by family carers and paid homecare workers ). An estimated 400,000 people with dementia and their families rely on homecare services for support (Carter, 2016).
For people living with dementia without a regular family carer, homecare services are often their only support. The Alzheimer's Society's (Carter, 2016) Fix Dementia Care campaign found that only 2% of people affected by dementia believed homecare workers had sufficient dementia training and nearly half (49%) of people affected by dementia disagreed that 'homecare workers understand the specific needs of people with dementia'.
Interventions have improved quality of life and quality of care for care home residents living with dementia (Ballard et al., 2020;Lawrence et al., 2016), but evidence regarding effective interventions to improve homecare for people living with dementia is limited (Cooper et al., 2017). Homecare workers often work alone in clients' homes, so their role and associated training needs are somewhat different to staff working in teams or building-based communal settings. Heavy workloads and tight schedules are a common barrier to staff training for this population (Su et al., 2021).
Two randomised controlled studies have evaluated dementia training interventions aimed specifically at homecare workers. A 12-week dementia care training programme in Taiwan improved dementia care knowledge, attitude, and competence of homecare workers (Su et al., 2021). The programme combined e-learning, and daily online support from a homecare supervisor, who also led monthly face-to-face peer support groups. Research suggests that peer support (face-to-face or virtual) could be especially important for homecare workers, who often work alone (Yeh et al., 2019). In a study in Japan, the Behaviour Analytics and Support Enhancement (BASE) programme trained homecare workers in a 2-day course to explore unmet needs and address 'challenging behaviours' of people living with dementia, reporting a significant reduction in 'challenging behaviour' 6 months after implementation (Nakanishi et al., 2018).
The NIDUS-professional (New Interventions for Independence in Dementia Study) training intervention is, to our knowledge the first training and support intervention for homecare workers to be co-designed by this staff group, their managers, health professionals, people living with dementia and their family carers  in the UK.
Our objective was to assess how acceptable and feasible a group video-call training and support intervention was to deliver in one homecare agency and whether the NIDUS-professional intervention is acceptable and feasible to deliver in practice, from homecare worker, homecare manager, and intervention facilitator perspectives; and how homecare workers perceived its value to them and their work.

| Design
We delivered a single pilot of the NIDUS-professional intervention with qualitative evaluation at 3 months (interviews with homecare workers, homecare manager and intervention facilitators) and 6 months (brief telephone follow-up interview with homecare workers and facilitator reflective logs) post-intervention. scalable benefits for people living with dementia and the homecare workforce. A randomised feasibility trial is underway. homecare, independence, interventions, person-centred care, training What is known about this topic?
• People living with dementia often rely on paid homecare workers to support their independence.
• The homecare workforce is undervalued and undersupported, and few homecare workers receive dementia-specific training.
• There have been few trials of interventions to support homecare workers and none to date in the UK that have been co-designed with homecare workers and people with lived dementia experience.

What this paper adds?
• Our intervention was acceptable and feasible to deliver in one agency, by facilitators without clinical training, using remote delivery.
• Homecare workers valued the intervention groups as a rare opportunity to have time and space to reflect on their work, reporting increased confidence and finding the interactive, group-based discussions particularly helpful.

London-Camden and King's Cross National Research Ethics
Committee approved the study (20/LO/0567); and we registered the protocol (ISRCTN15757555).

| Sampling and participants
We recruited homecare workers from one homecare agency in England. The NIDUS programme manager (LD) initially approached the homecare agency training manager, with whom the NIDUS team had a pre-existing relationship. The training manager identified and approached potential participants, provided them with a study information sheet, and asked for their permission for a researcher to make contact. We included homecare staff who were providing hands-on care for at least one client with dementia, who understood spoken English, and gave written informed consent to participate.

| The NIDUS-professional intervention
The NIDUS-professional intervention is a 6-session manualised training programme for homecare professionals working with people living with dementia. Initially designed as a face-to-face group training programme, NIDUS-professional was adapted for remote delivery via video-call in response to the Covid-19 pandemic. The group sessions for 6-8 homecare workers last 60-75 min and focus on exploring practical changes that they can try out to support their client's independence and developing self-care strategies to manage job stresses (see Table 1 for topics covered). This is followed by a 3-month implementation period, where the facilitators meet monthly with the group to support them to put their learning into practice, share challenges or successes, access peer support and group problem solving, and discuss the practical application of the training.
The NIDUS-professional intervention is person-centred. It aims to optimise wellbeing by treating people living with dementia as unique individuals, meeting their psychological needs and respecting their rights. The intervention focuses on peer learning and sharing of experiences, acknowledging and drawing on the homecare workers' different experiences, and levels of training. We co-designed the NIDUS-professional training and support intervention in workshops, with people living with dementia, family carers, homecare workers and managers and health practitioners, using existing interventions (Kales et al., 2015;Livingston et al., 2019;Low et al., 2015;Polacsek et al., 2020), our ethnographic studies (Leverton, 2020;Leverton et al., 2021) and lived experience of co-designers.
We learned from the Promoting Independence Through quality dementia Care in the Home (PITCH) intervention, a person-centred intervention designed for and with homecare workers in Australia (Dow et al., 2019). We also drew on materials from a person-centred intervention designed for care home workers for people living with dementia (Livingston et al., 2019). We presented the PITCH intervention at the first NIDUS coproduction meeting. The PITCH focus on developing empathy by considering how it might feel to have dementia and need homecare was endorsed by the NIDUS coproduction group as being particularly valuable and was therefore incorporated into the NIDUS-professional intervention.
The intervention co-design process  and its theoretical basis, and how this drew on concepts of person-centred care are described elsewhere .

| Intervention delivery and facilitator training and supervision
NIDUS-professional was designed to be delivered by graduates in psychology or relevant social science disciplines who do not have formal clinical training to increase scalability. PR and CC trained two members of the research team (KL and DK), without clinical qualifications, but with experience of working with people living with dementia, to facilitate NIDUS-professional (for facilitator demographics, see Table 2). Facilitator training and intervention delivery were completed remotely, using Zoom video conferencing. Training focused on clinical skills, NIDUS-professional content, and practical challenges, including adapting to the remote delivery method.
Sessions were role-played by facilitators as part of this training and clinicians (CC and PR) formally assessed these for adherence to the manual.
The intervention was delivered to homecare workers (none of whom was involved in co-producing the intervention) during Oct-Dec 2020. All sessions were recorded to assess fidelity. Participants were offered monthly sessions during Jan-Mar 2021, to support implementation of learning into practice. If participants were unable to join (intervention or implementation sessions), they were offered a 'catch-up' session with one facilitator, either individually or as a small group with other participants. Intervention groups were scheduled during a 'break' between shifts when most homecare workers were not visiting clients. We reimbursed the homecare agency for staff costs to ensure that participants were paid for any time spent attending the NIDUS-professional training and completing the qualitative interviews. Facilitators received weekly supervision with PR to consider and manage challenges around delivery of the intervention.
A clinically trained member of the research team (PR/CC) was available for support between supervisions.

| Interviews and measures
The researchers obtained written or verbal audio-recorded consent from eligible homecare workers and collected baseline demographic data (Table 3). Qualitative interviews were conducted with homecare worker participants after they completed their initial training; this included three individual interviews and one 'group' interview with two homecare workers (originally intended as a focus group).
An additional qualitative interview was also undertaken with the homecare manager. Interviews were conducted by researchers who had not delivered the intervention or collected outcomes with interviewees, using topic guides (Appendix S1), who asked attendees whether and how the intervention had impacted upon client, family carer and homecare workers' wellbeing and homecare workers' practice. The topic guide also helped the researchers gather practical and aesthetic feedback regarding the training materials, to aid development in preparation for a larger randomised feasibility trial.
DK conducted brief follow-up telephone interviews with homecare workers at the end of the implementation period (6 months), from which he recorded detailed notes including verbatim participant quotes. The NIDUS-professional facilitators (KL, DK) participated in a qualitative interview at 3 months and also recorded Session 1 -You and your role: looking after yourself as care workers and ways to do this

TA B L E 3 Baseline characteristics of participants
We took a reflective thematic approach to the analysis of qualitative data (Braun & Clarke, 2019). DK and KL systematically and independently coded transcripts of the interviews held after the 3-month initial training period, and reflective notes from 6-month follow up telephone interviews. Since DK and KL were intervention facilitators, CC also coded the facilitator interview transcript.
The researchers read the transcripts and notes to check for accuracy, anonymity and to familiarise themselves with the data, and then labelled meaningful fragments of text with initial codes. They inductively open and double-coded material to generate a coding framework. We refined and defined themes through discussion within the NIDUS team.
We integrated findings from the different data sources by exploring how codes from one dataset followed into the other, and vice versa (Moran-Ellis et al., 2006), developing one interwoven framework. Although we considered data sources equally valid, the majority of material reported comes from the 3-month interviews as this data form the bulk of the feedback regarding the intervention sessions. The 6-month telephone interview data were incorporated primarily to provide insights regarding the implementation period.
We adhered to COREQ guidelines to ensure methodological rigour of the qualitative analysis (Tong et al., 2007).
To analyse fidelity of NIDUS-professional delivery, two researchers independently applied checklists to each of the 6 group recordings. We calculated the proportion of expected intervention components (Table 1) delivered. We adopted established thresholds to rate fidelity (Noell et al., 2002): 81%-100% constituted high fidelity, 51%-80% moderate and <50% low fidelity. The researchers rated on a 5-point scale (1-not at all to 5-very much) whether the facilitator kept the group focused on the manual, and participants engaged, for each intervention component, and for each session, whether the facilitators kept to time.

| Recruitment and retention
The homecare agency identified 14 potential homecare worker participants for the researchers to approach. Of these, eight agreed to participate and completed baseline assessments (Figure 1; Flow diagram). Three participants declined to participate, all stating that they were 'too busy' to attend the training sessions. The remaining three participants did not respond to contact from the researcher and after several attempts, were assumed to have declined.
Of the eight participants who consented, one decided to leave the agency pre-intervention, so was no longer eligible. Seven participants therefore began the intervention in October 2020. Of these, five completed all six sessions (via the main group or catch-up sessions; see Table 4 for attendance details). Two participants attended a single group session and one individual catch up session. Of these two, one withdrew after the second session, stating that they did not have the time to fit the training into their day, between their double shift. The other participant did not wish to withdraw but was unable to attend any further sessions due to workload pressures. Of the five participants who completed the intervention, four attended at least one implementation session, as either a group session or an individual catch-up. Two participants attended one session, and two participants attended two sessions. The fifth participant did not attend any of the implementation sessions due to a change in job role within the agency, meaning she was no longer available to attend. Time pressures and lack of availability were the main reasons that participants reported being unable to attend all three of the implementation sessions. Some participants also stated that since they were so busy, they would only join sessions if they had something that was necessary to discuss with the group.
Sociodemographic characteristics are summarised in Table 3.
For most participants, this was their first dementia training course, beyond a mandatory e-learning module provided by the homecare agency.

| Intervention fidelity
Both raters who reviewed audio recordings of the intervention sessions, reported that all 33 intervention session components recorded were delivered; one component, the introduction to session three was not recorded and so could not be evaluated. Thus, overall fidelity (33/34:97%) was in the range specified a priori to be high.

| Thematic analysis
We identified one overarching theme that responded to our research questions: 'Having time and space to reflect is a rare and valued opportunity'. The context of this work, a remote group via video-call in a pandemic, is evident in narratives. Most participants felt that video conferencing worked well under the circumstances but brought challenges. The remote delivery made it easier to fit the training into their working day, but participants described difficulties in reading social cues and one participant described the impact of this on group dynamics: We're all waiting and then we all talk. (P7, group interview).
Within our overarching theme, we describe four linked subthemes below.

| Subtheme 1: Having time to reflect is a rare opportunity
In the first, Having time to reflect is a rare opportunity, we consider how homecare workers perceived having time to step back from the relentless pace of homecare work as rare and valued. The Unable to aƩend sessions due to workload and requiring a break between shiŌs (n=2) Began the intervenƟon (n=7) Withdrawn before intervenƟon beganhomecare worker leŌ the agency (n=1) Consented into study and completed baseline (n=8) Declined to take part (n=3) Male, less than 6 months Female, 10 years or more Female, less than 6 months The DICE model (see Table 1  , we built links between the two interventions; homecare workers were encouraged to bring goals for care that their clients were working on within the NIDUS-family intervention.
Unfortunately, due to the challenges of recruiting in the pandemic, we were not able to test NIDUS-family alongside NIDUS-professional as originally planned. We are now doing this in a randomised feasibility trial that is under way.
There is a robust evidence base across different settings (Ballard et al., 2020), and increasingly in homecare settings (Nakanishi et al., 2020;Su et al., 2021), that where staff can be facilitated to build skills in person-centred care through peer support, case discussion and discussion of care principles, this can improve outcomes for clients (e.g. quality of life, quality of care, reduced agitation). It can also improve retention of care workers, through increased sense of worth, confidence and job satisfaction.
A particular challenge is to develop implementable and sustainable training models, in a setting where staff turnover is high, funding-limited, time-pressured (Leverton, 2020;Leverton et al., 2021) and workloads heavy (Su et al., 2021). Participants reflected on how these barriers could be overcome with greater organisational support, for example, through the allocation of protected time for training by employers.
To be scalable, interventions need to be cost-effective . The shift online necessitated by the pandemic may have led us to one workable solution to this -a remote group intervention.

| LI M ITATI O N S
This pilot study was conducted within one care agency, with a small sample of homecare workers, and therefore our findings may not be generalisable to other agencies or contexts. Participants were initially approached by the agency training manager who may have approached staff with particular qualities not necessarily representative of the wider workforce, e.g. who were more likely to engage with the training or who were systematically different in some way-whether more highly skilled or in particular need of trainingthough from facilitator perspectives the group appeared diverse in experience and previous training. The intervention took place midpandemic in the UK and this unusual context may have increased commitment to the group at a time when few other resources were available, or decreased commitment due to multiple, competing demands, personally or professionally.
While four of five participants who attended the main intervention sessions did come to at least one implementation group, attendance was not as high as for the main groups. We will consider how to maintain attendance during the implementation phase in our future trial.
Greater clarity around the purpose of this phase (to sustain and utilise peer support at a time when new ways of working, developed through the intervention, are being implemented), clear support from management, and protected time for attendance would be helpful.
We plan to conduct a larger randomised feasibility trial, that should address most of the above limitations.

| CON CLUS ION
In this small pilot study, a manualised person-centred dementia training intervention, specifically co-designed for homecare workers in the UK, was feasible to deliver and acceptable in the short term, though we identified a challenge in maintaining commitment to the intervention after the formal structured sessions ended. Homecare workers valued the time and space to reflect offered by our intervention. As well as demonstrating the many benefits that such training offers for homecare workers and the nature of the care they provide, our findings highlight the challenging working conditions and the dearth of support for people within this profession, which is critical to not only the wellbeing of people living with dementia and their families, but to society.
Our study has implications for policy and practice, in highlighting the value of regular group training and supervision for homecare workers.