Using co-production to increase activity in acute stroke units: the CREATE mixed-methods study

Stroke statistics: organisational issues and impact

Stroke, known in recent public communications as a ‘brain attack’, can have a devastating impact on people’s lives and equally on the lives of those who live with and care for those people. 17 The effects of stroke are wide-ranging depending on the location and extent of the brain damage, but they can include paralysis and cognitive and communication difficulties among many other problems, such as difficulty with vision, continence and fatigue. 18 Stroke continues to be the largest cause of disability in the UK, and 84% of people leaving hospital will require help with activities of daily living. 17

Stroke incidence is high; 100,000 people will experience a stroke in the UK each year. 17 Although the numbers of first stroke have fallen significantly since 1990, by 2035 the rate of people over the age of 45 years having a stroke is expected to rise by 59% and the absolute number of people living with stroke will rise by 123%. 17 Population studies show that stroke incidence is not equal across different populations; people of black African, black Caribbean or South Asian ethnicity are more likely to have a stroke at a younger age. 19 In London, black people are twice as likely as white people to have a stroke. In addition, people from more socially deprived groups are likely to experience more strokes earlier in life. 20

These UK trends are also reflected in global studies that show that the absolute numbers of people who have a stroke every year, stroke survivors and related deaths, as well as the overall global burden of stroke, are great and increasing. 21 Estimates vary depending on the population sample and data sources, but it is suggested that in the UK there are 950,000 people aged ≥ 45 years. 22 This costs the UK economy approximately £8.9B per year (5% of the NHS budget), of which £4B is spent on treatment, including organised inpatient stroke unit care. 23 This figure is set to treble by 2035. 22

Recent decades have seen significant developments in the organisation and management of stroke, particularly following the implementation of the National Stroke Strategy by the Department of Health and Social Care in 2007. 24 The role of organised stroke care is well established in significantly improving outcomes after acute stroke. 25 Most people experiencing a stroke in the UK will get early access to care provided by stroke specialist staff. Large-scale service reconfigurations such as the London and Greater Manchester models have fundamentally changed care pathways, and the average length of stay on an inpatient stroke unit is likely to be around 17 days. 26,27 Large-scale reorganisation has also seen the case mix on stroke units change; patients with mild disability are discharged earlier as a result of the expansion of early supported discharge services, while critically ill patients with more complex and severe disability are likely to require a considerably longer inpatient stay. 1

Settings and sampling

We set out to recruit four stroke units: two in London and two in Yorkshire. We included stroke units that met the classification of a specialised stroke service defined as ‘capable of meeting the specific health, social and vocational needs of people with stroke of all ages’ set out in section 3.2 of the 2012 National Clinical Guideline for Stroke (reproduced with permission). 62 A stroke unit is classified as either a ‘routinely admitting stroke unit’ with hyperacute stroke units and acute stroke units or a ‘non-routinely admitting stroke unit’. 63 All stroke units provide acute and rehabilitation care, but only hyperacute units admit patients within the first 72 hours post stroke and return discharge data in SSNAP. 62 We also aimed to recruit stroke units with evidence of previous participation in research so that we could ensure that the units had an interest in delivering the research planned.

Stroke units were purposively selected following discussions held with senior staff and local stroke research networks. As advised by NIHR, we aimed to include not those stroke units that were based in large teaching hospitals and already taking part in clinical trials, but those that showed a willingness and commitment to take part in a study such as CREATE (Collaborative Rehabilitation in Acute Stroke) that comprised multiple stages over at least 12 months.

The two stroke units selected for the first stage and full EBCD were based in London and Yorkshire; we refer to these stroke units as sites 1 and 2. In the second stage, stroke units taking part in accelerated EBCD were also based in London and Yorkshire (sites 3 and 4). Each of the four sites was a non-routinely admitting unit that received patients only after they had been cared for in a hyperacute unit either in the same hospital (as at sites 2 and 4) or at a nearby major stoke centre (as at sites 1 and 3). More detail about each site is provided in Table 1.

Each of the four sites was included in the most recent biennial SSNAP Acute Organisational Audit report,64 published in 2016, which includes site-level and national performance against 10 key indicators (see Appendix 1, Table 15). The sites also return data continuously for the SSNAP Clinical Audit, which measures performance against standards for 10 key domains reflecting processes of care provided to patients. The clinical audit includes an overall performance score for 3 months, made up of a combined total indicator score derived from the average of patient- and team-centred key performance indicators, case ascertainment and audit compliance. Performance is graded A–E, with A indicating first-class service, B indicating good or excellent in many aspects, C indicating reasonable overall – some areas need improvement, D indicating several areas need improvement and E indicating substantial improvement required.

The SSNAP Acute Organisational Audit reports and the prospective clinical audit data provide an indicator of each site’s performance at a point in time, but reporting on these data carries several caveats. First, CREATE sites 1–4 did not treat patients within 72 hours (known as the hyperacute stage) following stroke and audit data also included results from these units providing hyperacute care. Second, Acute Organisational Audit data provide a snapshot of staffing at 10.00 and 22.00 and a whole-time equivalent at each grade against national indicators and the national medians on any given day, but this is likely to vary from day to day and does not include data about the severity of disability of the patients cared for. Appendix 1, Table 15, shows an overview of SSNAP Acute Organisational Audit site-level data and performance of all participating stroke units (1–4) against the 10 key indicators.

Stage 1: setting up the project

Stage 1 involved a period of stakeholder mapping with the core staff team to identify key contacts and services and/or staff who could help champion the co-design process in the trust and expedite approvals required on each unit for changes in layout, décor, or activities. We formed study oversight groups with trust leads at sites 1 and 2, meeting in person and communicating by e-mail. Core staff, together with researchers, developed posters and flyers about the project and held briefings with larger staff groups at different times during the day, for example nursing handover, goal-setting meetings with therapists and weekly multidisciplinary team meetings.

Stage 2: gathering staff experiences through observational fieldwork and in-depth interviews

In stage 2, staff were interviewed to explore their experiences of working on the stroke unit, particularly with respect to routines, structures and interactions in the team and with patients (see Table 3 for numbers of staff interviews). We also explored how ‘activity’ was perceived and what staff felt were the barriers and limitations to activity that could be addressed through co-design. Researcher-led ethnographic observations, which contributed both to EBCD and to pre-implementation evaluation data collection, were carried out at each site. Observations were carried out over a 3-week period or less and included weekday and weekends between 07.30 and 12.30, between 08.30 and 13.30 or between 15.00 and 20.00. The purpose, described in more detail in the evaluation section of this chapter, was to develop an understanding of the social and organisational processes linked to activity and the regularities and irregularities of the organisation of work and of social interaction in order to enhance our understanding of how and why stroke patients may be active or inactive during the inpatient day.

Stage 3: gathering patient and carer experiences through observation and filmed narrative-based interviews

In stage 3, patients and families were observed as part of the researcher-led ethnographic observations and staff observations described above. Patients and families were recruited for filmed narrative interviews that were edited into one composite film specific to sites 1 and 2. Most patients were filmed in their homes, with one patient choosing to be interviewed in a university building. Interviews lasted between 1 and 2 hours, during which time their experiences of being a patient on the stroke unit were explored. Some patients chose to be interviewed with their family member or on occasion separately for practical reasons. Family members reflected on their experiences of visiting and supporting their relative during the admission. The interviews explored routines and structures that either helped or hindered activity; interviewers encouraged patients to reflect on their activity during a usual day and across the whole week, including weekends. The composite films were produced for sites 1 and 2 and comprised nine patients and three family members (site 1) and seven patients and four family members (site 2).

Stage 4: bringing staff, patients and carers together at separate and joint events

In stage 4, the interview and observational data from stages 2 and 3 were summarised to draw out key themes and help orientate discussions towards priorities for change. The composite films comprised a narrative of patients’ and family members’ reflections and experiences of being on the stroke unit and several touch points. The duration of the composite film was 20 minutes and 24 minutes for sites 1 and 2, respectively.

Stage 5: sustained co-design work in small groups formed around priorities

In stage 5, co-design groups were held over 4–5 months, with groups meeting up to five times. Co-design groups were researcher supported and co-led with clinicians, and they typically lasted 1.5–2 hours. The groups were held in accessible spaces, usually on site at the hospitals, and timed for after the midday meal to gain maximum attendance from nursing staff. Refreshments and transport for patients and families were provided if required. Each group explored the ideas within their own priority area and developed action plans; researchers made notes and shared these with all participants after each group to confirm actions. Tasks such as contacting estates or local voluntary groups were delegated and shared among the group.

Stage 6: a celebration and review event

An important part of the EBCD process is the opportunity for staff, patients, carers and researchers to come together and celebrate their involvement in developing, implementing and sustaining the co-designed changes. 51,61 Celebration events were held in both sites 1 and 2 and were attended by approximately 60 patients, staff and families. The events included presentations and informal sharing of experiences by those staff and patients involved in the co-design process, reflecting on the changes that had occurred and the lessons learnt and summarising the post-implementation data from observations and staff and patient interviews. The research team gave an overview of the progress of the CREATE project across sites and the plans for the next stage. A number of additional events enabled further dissemination of the project, including a mayoral visit and an official launch of the changes in the stroke unit at site 1, as well as more detailed feedback of the results to small groups of staff and other stakeholders.

‘Accelerated’ experience-based co-design sites 3 and 4

Our research questions asked how feasible it would be to adopt EBCD as a particular form of co-production for improving the quality and intensity of rehabilitation (activity) in acute stroke units. The full EBCD carried out at sites 1 and 2 took 9–10 months. Our methods for sites 3 and 4 were informed by Locock et al. ,60 who showed that it was possible to accelerate the process using a national video and audio archive. We were keen to use the stroke-specific trigger films developed at sites 1 and 2 and contextualised at two further sites using the methods outlined below. Consequently, at sites 3 and 4 we sought to reduce the length of the process by making two distinct changes from the methods of Locock et al. 60

First, we used the trigger films already generated by stroke participants at sites 1 and 2. Although staff and patients were still interviewed as part of our pre- and post-implementation data collection, we chose not to film the interviews or edit the narratives to produce a composite film; instead, we used the trigger films from sites 1 and 2.

Second, we progressed straight to a joint event after site set-up, interviews and observations without holding separate staff and patient events. This meant that staff, patients and carers saw the film together for the first time but the same methods were used in the joint event to explore ideas and priorities and to develop co-design groups. A total of 82 staff, patients and families attended the joint events at sites 3 and 4, and again co-design groups were formed around agreed priorities (Figure 2).

FIGURE 2.

Accelerated EBCD used at sites 3 and 4.

Celebratory events were held at sites 3 and 4 in a similar way as at sites 1 and 2. Staff, patient and family members gave presentations and shared experiences informally, and the research team gave a summary of the project and progress. Further dissemination and spread of project findings happened at sites 3 and 4, including a mayoral visit and official launch of the new common room at site 3, and an open day at site 4, as well as more detailed feedback of the results to small groups of staff and other stakeholders.

Recruitment

We aimed to recruit participants who reflected the population of stroke patients admitted and discharged from our sites, who would naturally include patients with different levels of stroke severity, gender, age and ethnicity. We also aimed to include participants who had communication and/or cognitive impairments in order to reflect the stroke population, and encouraged family members to provide support when patients were unable to complete the questionnaires or take part in interviews. Several strategies were used to estimate our target numbers for recruitment.

Based on stroke admission data across London and Yorkshire, we estimated that it would be possible to collect PROM/PREM data from an independent sample of 30 patients from each unit pre and post implementation of co-produced interventions.

Behavioural mapping data collection took place during non-consecutive days. We aimed to recruit a minimum of four and a maximum of eight patients who met the inclusion criteria and were able to provide consent on the day before the observation.

Pre implementation of EBCD cycles, we aimed to purposively recruit a sample of approximately 10–15 staff, stroke participants and family carer members (30–45 in total from each unit) to take part in interviews as part of the co-design process. Stroke participants and family members or friends (carers) were also recruited 3–6 months after discharge from the stroke unit to allow time for adaptation to life at home to begin, but sufficiently soon after their inpatient care episode to allow reasonably accurate recall.

Post implementation of EBCD cycles, we aimed to recruit a further sample of up to 10 members of staff/patients and carers in each of the four sites to participate in semistructured interviews to explore their experiences post implementation of co-designed interventions. We also aimed to carry out interviews with a sample of staff members, patients and families who took part in the co-design groups to explore their experiences of being part of the whole process. The size of the sample was also informed by reaching thematic saturation during data analysis.

Qualitative data analysis

We first describe the processes used to analyse the qualitative data generated from non-participant observations of staff, pre EBCD training and activity, during EBCD (i.e. of the separate patient and carer, staff meetings, joint meetings and co-design meetings) and from interviews with patients, carers and staff pre and post EBCD. The integration of these data in the EBCD evaluation and also in the linked process evaluation involved an iterative approach to analysis that focused initially on the data generated at each site and then progressed, using team half-day analysis meetings, to a comparison between sites, as described below.

Interview data video files (patients and carers at sites 1 and 2) and audio files (staff all sites and patients and carers at sites 3 and 4) were transcribed verbatim. The research fellows and research lead for each site completed an initial thematic analysis of the data at each site (London and Yorkshire) and prepared summary memos identifying the main themes and summarising the key issues related to the presence or absence of activity outside therapy, the opportunities to make changes and the attitudes towards possible changes. These summary memos were then compared and reviewed iteratively in a series of half-day face-to-face meetings (held approximately every 3 months in London or Leeds) by all four researchers, before the summaries were presented to and discussed with Study Steering Committee members.

For observational data, field notes were prepared by each researcher conducting an observation and shared among the researchers for that site. On completion of the series of observations (pre and post implementation and during EBCD activities), summary memos were developed to identify recurring themes and to compare and contrast findings from pre- and post-EBCD activities within and then between sites in London and Yorkshire. Again, these were compared and reviewed iteratively by all four researchers in a series of half-day meetings (as described above) before being shared with Study Steering Committee members. The memos included references to contextual factors considered relevant to service delivery, and to patient experiences of the EBCD process in each site. These processes were used for sites 1 and 2 and then repeated for sites 3 and 4.

Following these half-day meetings and the discussions resulting from presentations of the ongoing data analysis, the core and cross-cutting themes reported in Chapter 5 were developed and agreed by the research leads for London and Yorkshire and shared with Study Steering Committee members.

Integration of data in the experience-based co-design evaluation and process evaluation

The data used in the process evaluation were not generated separately from those used in the main evaluation of the feasibility of using full and accelerated EBCD in the four sites; rather, the same data were critically examined using NPT’s four core constructs and associated components. A data collection plan linked to NPT’s four constructs was developed prior to data collection. The purpose of the plan was to engage with the NPT constructs as data were analysed at each time point, and to identify evidence of (in the summary memos described above) examples such as staff progression from coherence to cognitive participation. This might comprise staff making sense of the EBCD approach, then thinking about what introduction of and support for increased patient activity outside planned therapy would mean for them individually and for the routine service provision currently in place. Once the EBCD activities had ceased at sites 3 and 4, the summary memos and the researcher reflections were reviewed by the research lead for the process evaluation and a draft single integrated account was constructed. This was reviewed by the research team as a whole, and the final agreed account is presented in Chapter 5. Our approach comprised both an ongoing integrative analysis of these data focused on staff and patient engagement with the EBCD process and on designing and implementing changes to promote or directly support increased activity, and a post hoc review of the full integrative data set.

Confirmability of analysis was further enhanced through a process of independent, joint and team half-day analysis and review cycles, after which the emerging analysis was discussed with the Study Steering Committee members. Credibility and transferability of the analytical approach are evident in how we have used detailed data extracts and interview quotations to support plausible explanations of the observational and interview data in terms of participants’ engagement with EBCD and the facilitators of and barriers to its introduction and use in the four sites. We incorporated researcher reflection and reflexivity in the data collection process and used these insights in the team analysis of the data.

Quantitative data analysis

Patient-reported outcome measure and patient-reported experience measure data

These data were entered into a SPSS file and reported as descriptive statistics (or frequency counts) for each item. These data provide insight into patients’ perceived functioning post stroke (PROM) and their experiences on stroke units (PREM). Some of the PREM items sought responses directly related to opportunities and resources for activity.

Figure 3 depicts our integrative approach to analysis across the whole data set (qualitative and quantitative) process evaluation.

FIGURE 3.

Data analysis for both the evaluation and the intervention (EBCD).

Ideas that drove the changes

At sites 1 and 2, filmed patient interviews carried out pre implementation were synthesised to produce a composite film shown at all joint events. Staff interviews and ethnographic data were also analysed and contributed to the data corpus; these were presented at separate staff and patient events at sites 1 and 2. These data underpinned and drove the discussions, which led to the priority areas for co-design groups. The ideas generated were largely similar across all four sites and focused on space, activity and communication. With minimal local variation, ideas were revised or shaped at the joint meetings and then interpreted and adapted in co-design groups.

Space

I often thought, on the stroke unit, it would have been nice to have a little group where they all could be interactive with each other instead of sitting staring at the walls . . . make the day spread out a bit better . . . I was always sitting watching things and thinking they could have a little group in the corner where they could be doing little things.

Carer, site 2, pre

The above quotation illustrates the inter-relatedness of space (environment) and activity, which became increasingly clear during the EBCD process. The four sites differed in physical layout, colour scheme, room size and use, and available social and outdoor space. However, across the sites, staff members’, patients’ and family members’ suggestions to improve the environment were very similar. Site 1 had no shared space (day room) in which patients and visitors could meet, and visitors were cramped by the bedside. Site 2 had a day room but it was used mainly for staff meetings and equipment storage; this was replicated at site 3, which had a garden but it was accessible only through the day room, which patients did not use. At site 4, the day room had become a storage area for wheelchairs and specialist stroke chairs; another seating area was routinely used by staff for taking breaks and storing cleaning equipment. Our observations confirmed the restrictions posed by limited space and the impact on patient use; there was little or no space or opportunity to socialise in groups with peers or visitors. This issue was compounded by areas that could have been used by patients and visitors instead being used as storage for clinical items. Space and furnishings appeared to affect patient and carer morale; this issue was depicted in data from all four stroke units, although less so at site 3, but generally environments were felt to be uninspiring:

And it looks very much like a hospital environment, I think the colours are very hospital-y, I think the decor is a bit shabby and when you’re spending, when you spend, you know, you’re spending months in this hospital sometimes it just feels like there’s nothing that looks particularly homely.

Patient, site 1, pre

Restrictions on activity caused by the environment were also noted across all sites, as illustrated by how the way in which a ward is set up can limit communication between patients:

When she [the carer’s wife] first went in . . . a really elderly lady was in the next bed and she couldn’t communicate at all. If they could have rotated the beds round so that the ones who could talk to each other [were next to each other], rather than have to talk over them to another patient, that would have been better for them.

Carer, site 2, pre

Patients and staff felt that space could be repurposed to allow more opportunities to move about, play games and socialise. They suggested getting rid of clutter in therapy gyms, day rooms and corridors. The restrictions on activity caused by equipment storage was evidenced in observations (Figure 4):

The corridor is cluttered with table, trash bins etc. so we have to take turns [to pass] . . . The corridor is cluttered with a hoist, 3 wheelchairs, various trolleys and trash bins.

Field notes, site 1, pre

FIGURE 4.

Equipment storage in corridors at site 1. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

In addition, the space around each bed was limited, and frequently tables were used for clinical equipment such as wipes or syringes, leaving little room for personal items (Figure 5). Frequently staff complained about the ward not being fit for purpose and restricting independent activity:

On the movable table next to his bed, I see a little jar with water and a cup half-filled with orange juice, tissues and a blue folder like the ones used for patients’ notes. On the bed side table there are two framed pictures – one is at the front and partially covers the other. The rest of the table is occupied by medical equipment.

Field notes, site 1, pre

I don’t think this ward is suitable for stroke rehab, we need a bigger ward definitely. We’re always falling over stuff. I think because, you know, a simple question of, ‘Can I go to the toilet?’ we’ve got to move everything . . . So the poor little patient’s waiting an extra 5, 10 minutes and they might have peed themselves so that’s not beneficial to them because that’s just promoting incontinence and, you know, they’re on a rehab ward.

Staff, site 4, pre

FIGURE 5.

Typical bed space and décor at site 1, described by patients and staff as cluttered and uninspiring. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

The benefits of the bedside having a personal feel developed as a theme through our observations and interview data. We observed that when personal items were around the bedside, staff tended to interact in person-centred ways to a greater degree, asking the patient questions about a family member or a photograph:

I noted that the patient in bed 19 had a different bedside than other patients in the bay, as the family had brought in things that were important to her into the ward. She was keen to show me everything they had brought in, and felt it was immensely therapeutic to be able to look at pictures of family instead of the blank wall – she said she had spent time gazing at them and thinking about each of them.

Field notes, site 1, pre

Additionally, some carers had done their best to enhance and personalise the spaces around their relatives who were unable to get out of bed:

I took photographs and pictures and put them all on the wall to try and brighten it up, because it was just horrible, it was a horrible ward.

Carer, site 2, pre

The sense of being trapped in the ward and the lack of light seemed to add to the feeling of dreariness and loss of connection with the outside world:

[. . .] is sitting in bed. The lights are off and the blinds down. The room is quite dark. She is sitting in silence. The TV is still positioned in front of her but it’s off.

Field notes, site 1, pre

The necessity of having somewhere else to go apart from the bedside became a strong priority for change:

A table maybe in the middle of the ward to bring people together.

Carer, site 3, pre

It would have been good to mix with other people, so you can get to know them.

Patient, site 3, pre

One carer at site 4 mentioned that she struggled when using a wheelchair to take her husband to the hospital café, which was a considerable distance away, and that she would have preferred to have been able to go for a drink and change of scenery to a nearby day room. She also felt that being in the day room, rather than off the ward, would have made it less likely that her husband would miss therapy sessions. When space was available, it was often used only by staff, and the impact on activity was evidenced at sites 2, 3 and 4 (site 1 had no day room) (Figures 6 and 7):

I think the day room has, has been a bone of contention in terms of the nurses are upset they can’t use it as a restroom. The therapists and everybody else view, their view it’s not a restroom, it’s a day room, how do we try and accommodate everybody? It would be nice to have a bit more office space for everybody else, it would be nice to, you know, kit out the gym a little bit better so you could use it more. The garden I think, you know, we don’t use our garden space enough, yeah.

Doctor, site 3, pre

FIGURE 6.

Day rooms at sites 2 and 3, which tended to be used only as meeting rooms for staff. Reproduced with permission of Fiona Jones, University of London, and David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, and David Clarke, University of Leeds, personal communication, 2020.

FIGURE 7.

Day room at site 4, which was used for storing specialist chairs and wheelchairs. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

Communication

The final issue was communication. There was an overall impression of missed opportunities to enable activity reflected in our ethnographic observations and the limitations of communication with staff was raised by patients and families. Each unit constituted a specific type of space where communications were structured around the tasks at hand, and outside this communications were related to information-giving (e.g. feedback on progress and prognosis), with very few ‘small-talk’ interactions. Unstructured phatic communication was not the norm. We saw very little evidence of personal and social interaction between staff and patients, or between patients; when this did happen, patients commented on how much it inspired or encouraged them to do more. Communication between staff and patients was again shaped by structures and routines, and the lack of interaction outside those times was noticeable. Patients also felt that communication could enable their activity and wanted more specific ideas from staff, and feedback on what they could do independently or with a carer. Patients and families had a shared perception that the staff group were very busy and had limited time. Our observational data largely supported these comments and we rarely saw anything other than task-orientated interactions. As noted in field notes, there was also a lack of ‘chatter’ between patients and members of staff, which created a rather quiet, uninteresting atmosphere:

Hardly anyone else interacted with the patient. I heard no hand over of information about what the patient could do for themselves, or any aspect of their recovery, or activity.

HCA [health-care assistant] walks in and push her trolley next to him. She records some routine observations of pulse and blood pressure. They don’t talk. He closes his eyes while she performs the procedure.

The group moved on without speaking to this patient. There was no other exchange of information other than about his PEG [percutaneous endoscopic gastrostomy] tube.

A nurse came in to attend to the patient in bed 21, the curtains were pulled but I heard no chatter.

Field notes, site 1, pre

It was not always what was said but what was not being said. One patient stressed the importance of ‘good morning welcomes’ (patient, pre, site 1) and others mentioned the importance of staff introducing themselves to patients:

I just found it horrible when no one even says a good morning, but they are discussing my condition . . . Even down to saying things like ‘I am waiting for a letter from the council’. It feels like your private life is not private at all . . . Because I have had a stroke you lose everything, dignity, the ability to talk and walk, it feels horrible.

Patient, site 3, pre

There were positive examples of communication. Those members of staff who spoke to the patients ‘as a person’ stood out during our ethnographic observations, and some of them were explicitly mentioned during the interviews:

She [health-care assistant] just made the unbearable bearable, she really did. We all hated the weekends when she wasn’t there, you’re just bored anyway and there was mostly agency staff at weekends and, you know, it was just, we just so missed having her around. She just, you know, she used to say to us like ‘we’ll laugh together and we’ll cry together’, you know, and she was just amazing.

Patient, site 3, post

Patients and families gave very few examples of knowing how to be more active; they could not remember being given advice and we rarely witnessed communication that would encourage activity:

It was only the second day, as soon as you got to the rehab ward you wanted your laptop, didn’t you, to check e-mails? [As] the relative, you think, ‘Oh, should I be doing that, will it harm him any more?’, so I went and asked the sister and she was very enthusiastic, ‘Oh yes, anything that will stimulate him, bring it in’, so she was really encouraging . . . You feel you need permission to do these things.

Carer, site 4, pre

Overall, our observations and experiences of communication varied, depending in part on how busy staff were or on the staff who were on the unit that day. However, patients felt that there was inconsistency in receiving information and in staff members’ understanding of their needs:

I mean the first day I got there [. . .], it’s just like, ‘this is your bed, get your clothes off and try and get into bed’. You’re not helped to even get in the bed and then I remember, I had to call, I went to call a nurse, one of them said ‘Why didn’t you call me by the bell?’ I said, ‘because all I was told was, we’re going to get weighed and straight into bed’, no one said a word to me, so I didn’t know anything, so you’re left in limbo. And you feel like a nuisance and don’t ask. And you shouldn’t have to be asking, you know. I remember that night crying and one of the nurses had to come and comfort me because I said I want to go home, I’d rather go home and not be able to do nothing, because that’s how I felt.

Patient, site 3, post

We found alignment between staff and patients/families on many issues relating to communication.

The sense that there was ‘no time for chat’ came through in many staff interviews, which resonated with patients’ and families’ interviews and our observations. Talking positively to patients and wanting to get to know them as a person was not necessarily considered part of the ‘work’. Staff reported that they felt limited by time and clinical priorities:

I just feel like there’s just limited time, like I’m sure everyone would love to sit down and have a conversation but they’ve got so much to do and I think that probably makes people feel more, more stressed and probably more afraid to get into like a conversation because then you have to be rude to get out of it, so maybe if you feel like you never get into it then you don’t have to be rude in that way.

Staff, site 1, pre

In addition to the lack of social communication, we observed, without exception and across all sites, that notices by patients’ beds described information for the purpose of conveying clinical information between staff and very little about personal needs, experiences or activity goals or plans. The sense of ‘the person’ was absent in this information (Figure 8).

FIGURE 8.

A typical method used at site 1 of conveying abbreviated clinical information written on whiteboards above beds. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Staff cited many challenges with finding time in their existing schedules to prompt, set up and participate in activities with stroke survivors. These comments were mentioned by all staff groups but more commonly by nurses, who did not appear to see ‘increasing activity’ as a key nursing function. However, the challenges of doing anything other than routine tasks contrasted with the following comment from one staff member:

I do make a massive effort with everybody, I try to give them a bit of my time even if it’s me sitting down with them for 5 minutes, but I mean there’s no facilities to do much else with them than talk. Or if you do, you’ll probably told off for sitting down and talking to them. We [health-care support worker and a student nurse] were getting [patient name], up one morning, he’d just come onto the ward, anyway we were asking him about his life and, you know, joking with him because he had a really good sense of humour, he was only 42, 43 and he really enjoyed that, and we were getting him up, because he was like ‘I’ve been so bored since coming in here’ and he said ‘Thank you for you two coming in and cheering me up and making a difference’. And then one of the nurses popped their head in and was like ‘Less of this and more work’, and I was like ‘Have you actually listened to what he was actually saying to me? We were actually discussing his life with him’.

Staff, site 2, pre

In conclusion, although these units were delivering on national targets for therapy provision, the extent of patients’ inactivity outside scheduled treatment was concerning. This was compounded by spaces that limited activity opportunities and by staff routines that were structured to ensure that caseload was managed but not to promote activity. Communications between staff and patients were task oriented and not patient centred, so there were few instances of focusing on what patients would want to – and could – do on their own. There was a small number of instances of activity promotion (e.g. the art group at site 2), but in the main there was a clear mismatch between staff members’ understanding of the need for and benefits of increased activity and the current service provision in these stroke units. Participation in the EBCD process at all sites led to the identification of and agreement on priorities for improvements that it was considered would lead to increased activity. The impacts of some of these changes are summarised in Chapter 5.

Space

I mean the clock thing was something that really struck me. You know, you kind of assume that everyone has a watch or a phone or something so everyone knows the time but I guess, you know, people don’t and people of a different generation maybe don’t have a phone or whatever or you know.

Staff, site 1, post

Changes in the stroke unit ‘space’ included rethinking and repurposing existing spaces (all sites); repainting corridors and bays and creating and introducing artwork in the stroke unit (sites 1 and 3); reclaiming, repainting and equipping day rooms as spaces for activity and patient, family and volunteer use (sites 2, 3 and 4); personalising bed spaces (site 1); enabling greater access to kitchens and bathrooms; and introducing a tool to get to know patients as people: the ‘a little something about me’ board (site 2).

Different priorities did emerge. At site 1, for example, co-design groups prioritised creating the feel of ‘bringing the outside in’, and all of the bays were redecorated with a theme such as lavender or the seaside. Paintings or prints were put up; the walls were painted; digital clocks were put in each room visible to every patient; and hooks for visitors’ coats, photo-hangers and extra shelves to declutter patients’ bedside tables were added (Figure 9). Our interview and observation data support the positive impact that these changes had:

It is a picture of a beach and a cave, similar colours to the photograph from Cornwall that is on F’s wall. He [family member] also shows the photo to F and tells her this is very similar. She smiles.

Field notes, site 1, post

He also uses his photo-hanger. There are get well cards and photos.

Field notes, site 1, post

FIGURE 9.

One of the new colour schemes in a four-bedded bay at site 1. Clinical items were put on new shelves behind each bed, local art work was put on the wall, and a photo-hanger was put up beside each bed. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

At site 2, the unit initially agreed that changes to the colour schemes and appearance of the unit should be made and worked with an artist to do this, but the changes were put on hold when estates raised concerns about costs, suitability and access. Corridors in the site remained dark and unwelcoming, in stark contrast to the bright, accessible and very well-used refurbished day room on this unit. Site 3 also worked with a local artist to redesign the colour scheme in the newly refurbished day room, and new communication about use of the space was developed. Site 4 already had light and bright corridors with some artwork in place, and the bedside curtains had local scenes and landmarks imprinted on the materials, so these areas were less of a priority for co-design groups.

All of the wards had emphasised creating a common space for patients and carers. Site 4 had a day room but it was used for wheelchair and seat storage. However, as the project began, the chairs were removed and the space become usable for individual and group activity. At site 4, a seating area (for four to six people) facing floor-to-ceiling windows looking out across the main car park and beyond to the hills surrounding the town was also repurposed. This area had previously been used mainly by staff taking their meal/rest breaks and for a display of Stroke Association materials. During the EBCD process, a magazine and newspaper stand was added, jigsaws were donated and several house plants were placed on an accessible window ledge. This space was used regularly by patients and carers both ad hoc or opportunistically and as part of planned therapy-linked activity, for example getting patients who liked gardening to play a part in watering and caring for the plants and encouraging patients with minor cognitive impairment to work on jigsaws with other patients (Figure 10):

We had a gentleman who was really disengaged, he wouldn’t really engage in therapy, but I gave him the job of watering the plants [in the window area] every day and he started doing that and apparently he did better in therapy after the engagement sessions. One of the physios said ‘I wasn’t up for this woolly hippy stuff [CREATE] that you lot are up to but, I could see how it worked, it worked really well’.

Staff, site 4, post

FIGURE 10.

New repurposed seating area at site 4. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

Another carer mentioned the value of having the window seating area as an alternative to the day room(s):

I think we did one or two [jigsaws] in there but people don’t all go in the dayroom you see, people tend to like the corners [the window seating area] there and people tend to join, you’ve got a nice view over the hills, you can look over the car park and you used to love sitting by the window and looking out and seeing what were going on.

Carer, site 4, post

The patients and carers at site 2 appreciated the newly refurbished day room. One patient, who had been on the unit before the room was available but had since seen it as part of a co-design group, felt that it would give patients a choice about what they could do and give them a change of scenery:

They [patients] can go there [to the day room] and there’s people there that will have had the same as they have had, and you can either talk to them or not talk to them . . . they can go to the dayroom and read a newspaper . . . or speak to somebody . . . but it’s just getting out of that bed space and moving. [Previously, when he was on the unit] we had a choice – sit in our bed space or walk to the end of the ward, or walk up and down and that was it.

Patient, site 2, post

The room was used for both scheduled and unscheduled activities. Staff reportedly encouraged patients to go in and participate in newly introduced groups. These groups included a regular Sunday lunch club run by volunteers and a breakfast club and a baking club run by the OTs. Other groups were also held there from time to time, such as music groups and bingo (Figure 11).

FIGURE 11.

New social space (day room) used for scheduled and unscheduled activities at site 2. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

At site 1, observations showed how patients used the new common area. This site previously had had no dedicated space for therapy-related independent practice or social interaction. Prompted by the co-design groups, the end of one corridor, which had been used to store equipment, was transformed into a new open seating area (Figures 12 and 13):

I moved through the ward to the new social corner where I met patient F. who was doing a puzzle. Someone had brought it in for him, it was not his own he told me. He told me he thought the ward had become more inclusive over the time he was there. And he said he liked the social space.

Field notes, site 1, post

FIGURE 12.

End of a corridor at site 1, previously used for storing chairs and hoists. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

FIGURE 13.

New space at site 1, now an area for patients and families to meet and socialise. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Redesigning the space also involved taking stock of what was available to support increasing patient activity, and opening up areas previously inaccessible to patients and visitors:

Yeah because there’s patients’ relatives, instead of coming and ask you now, they go there, they make a cup of tea, they can make a cup of tea for their relative so there’s no point in going to the kitchen because they’re not even allowed, or sometimes we’re busy, you say oh I’m busy, but now they just walk straight and do, yeah, and do that which is good.

Staff, site 1, post

At all four sites, redecoration involved decluttering and rethinking how the space could be used. At site 3, funds were raised to redecorate the shared day room and add a new kitchen area for making drinks and snacks. A local artist painted the walls to make the space look less like a staff meeting room.

At site 4, the co-design groups explored a range of options for making iPads (Apple Inc., Cupertino, CA, USA) more accessible; towards the end of the project, purpose-built flexible stands on wheels were purchased, which meant that iPads were secure but could be used both in therapy sessions and independently (Figure 14):

I think it’s great. Those rooms, that [ward D day] room is particularly good because, you know, when it was full of chairs we couldn’t even go in there. Now it’s, now it’s a free resource we’re lucky to have it, you know, and anything is better, yeah, I think it’s great. I think those pagers are great, the iPads, you know, they’re great on the stands, wonderful.

Patient, site 4, post

FIGURE 14.

New mobile iPad stands used at the bedside at site 4. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

The changes to how the space was used also had an impact on communication. Having familiar home items in the space to make it more homely and placing a photo-hanger by patients’ bedsides contributed to conversations and to learning about the patient as a person:

I said that it is impressive as a fireman he must have saved many lives, and as an electrician bring light to people’s lives. Patient in bed 3 says ‘and electric shocks’ and he laughs. We talk about their photos in the photo hanger. Patient in bed 4 has a holiday photo from Germany, a river, in his photo hanger. It is black and white. He said he went with his wife and it was a long time ago. Patient in bed 4 is very good with names, I realised this before. He knew the names of members of staff when man in bed 4 didn’t. He asked him again about someone’s name and he knew it. I tell him he is very good with names, the man in bed 3 agrees, says yes, isn’t he. I spent about 20 minutes with them. The patient in bed 3 asks me to pass him his blanket, it is red and cosy. He says it is very nice, he is a bit cold. His granddaughter brought it for him. His grandson used to play rugby and they went to watch him on the weekends. He was very good he says. They say thank you and say it is very nice of me to put the programme on. They don’t want a newspaper to look at whilst waiting for the rugby to start.

Field notes, site 1, post

At site 2, a therapist commented on how the ‘a little something about me board’, which contained personal information about previous hobbies, activities or work that patients wanted to share with staff, had shaped the therapy activity she had decided to pursue with a patient:

I was planning some upper limb work with this chap but he was a little hard to motivate. I noticed that the board said was a keen domino player, so that’s what I suggested to him, we did the session focused around playing dominoes, something that he liked to do and we still did the upper limb work, so that’s something that’s happened because of the project for me.

Staff, site 2, post

The nature of the changes made to space meant that these were most visible and generated the majority of positive comments from patients and staff. We were also able observe the impact that the changes to spaces and the different ways space was being used had on patient activity. Staff members’ comments were orientated around the enabling impact of the changes, as well as the change in aesthetics, which gave a more homely, welcoming feel:

So I think the ward is different, I think definitely things look a lot better and a lot more inviting in the environment it’s engaging in, and I think that, that makes a difference in how we communicate on the ward and how we interact with that patient, because there’s stimulus to talk about, there’s pictures now on the wall to talk about and engage and help patients find their way in colours that help, signpost where they’re going and it’s something to talk about, a neutral that we can talk about that everyone has an opinion on and it doesn’t often require a lot of complex language as a speech therapist it’s nice to have something that people could gesture a thumbs up or a thumbs down, they can nod or shake their heads to say that they like a painting or a photograph. So in that respect I think that’s really helped how the ward, the ward interacts and changes.

Staff, site 1, post

Activities

Changes initiated by activity co-design groups included regular visits from therapy dogs (sites 1 and 2); new activity boxes for every bay (sites 1 and 3); increased structured group and individual therapeutic activities (all sites); and increased volunteer and external group supported activities (music, gardening, massage, pet therapy, art and crafts, social Sunday lunches) (all sites) (Figure 15):

We have huge gaps in the day where your patient’s doing nothing, they’re bored, they become institutionalised, so with these extras, like your volunteers coming in, you’ve got various groups, you’ve got your cooking group, your breakfast club, your lunch club, it just makes for a, well it’s a more positive experience isn’t it, well I feel it is.

Staff, site 2, post

FIGURE 15.

New activities (art groups and therapy dogs) available at site 1. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

The co-design process often led to discussions about how to connect with community and voluntary groups to a greater extent. At site 1, for example, a family member of one of the patients worked as a local artist. Together with her colleagues, she applied for funding and organised art workshops on the ward at the weekends. At site 2, the patient experience office worked with the volunteer co-ordinator to train new volunteers to provide mealtime and social lunch support following discussion with the co-design group members. In addition, the local university’s music department was contacted, which led to two or three singers coming to the unit every 2 weeks to sing songs both they and the patients had chosen. At sites 3 and 4, co-design groups, aware of some of the activities that had been developed at sites 1 and 2, utilised this learning much earlier by connecting with community organisations, and at site 3 a programme of concerts by local choral and music groups was organised. Observations showed the impact of these activities:

One of the ladies told me she had to cry as she found it really moving and said in a good way. She also walked past the group again later and said that it was really nice and she smiled.

Field notes, site 3, post

Additionally at site 3, a retired PT and ex-patient had begun to hold fortnightly ‘exercise groups’. The feedback she had received was positive and our observations showed that the groups had almost got ‘too popular’. Many patients wanted to attend but there was limited space available, even in the repurposed day room. At site 4, a new breakfast group was set up and by the end of the co-design group activity it was running 7 days per week. This group was led by therapists and supported by nurses, who continued the activity at weekends when therapists were not present. The co-design group initiated other planned (arts and crafts groups) and ad hoc events (open afternoon, evening meal with other stroke survivors, a small group watching a favoured television programme) and some patients and families noticed the impact of these:

I do see evidence, psychological evidence [of being willing to engage in more activity] . . . and I also think patients are doing more now.

Patient and family members, site 4, post

Despite an increase in the opportunity to access activities, our observations showed that there were still long periods of time when patients were inactive. Opportunities to access activities were lost for patients who needed help to get out of bed:

I think certain members of staff are really great at getting patients out of bed and really proactive with that and will look at what patients are signed up to do, whether it be therapy sessions or whether it be something like the reading group which has gone on timetables. And then there are other members of staff who don’t look at the timetables at all.

Staff, site 3, post

Co-design groups wanted to focus attention on those patients who were unable to get out of bed and attend groups by designing and providing new activity boxes that could be brought to patients’ bedsides, but we rarely saw anyone using these:

Patient in bed 8 is alone. She is lying in bed. She is not very mobile, but there is nothing else e.g. radio, TV for her. She always smiles when she sees me walking past. The patient, gentleman, 92, is sitting alone. He is saying ‘no no no’ as he quite often does. The doctors are standing around the nurses’ station. 2 of them are looking on their phones, talking about something else, work unrelated.

Field notes, site 1, post

However, observations and interview data also supported the notion that a small number of patients did not necessarily want to engage even when increased activity opportunities were provided. Their pre-stroke behaviour, activity and leisure preferences were factors in their willingness to participate in what was offered:

There was various things going on that the volunteers organised. There was afternoon teas and stuff like that but you’d no desire to go had you?

Carer, site 4, post

Members of staff reported that they were more aware about the activities on the wards post implementation. This was also the case for members of staff who did not take part in the co-design groups:

There, everything, everything is there, everything is there, I’ll go there and give it to them if they are sitting down and there’s nothing to do, give them something to do, even if it is just to practice writing their name or practice signing, I’ll give them a piece of paper and pen so that they can practice and write. [. . .] Yeah, sometimes you think, you can see they want to write, do you want to write, you see them, sort of, practising, oh I don’t know how to write my name again, OK just practice, just keep like that. [. . .] Well before we, yeah now we are doing it but before, as I said, before we were not, no, we are not doing all these sort of things so, but now I don’t know because of the CREATE, the awareness, we’ve got the awareness now so we are trying to encourage them to do all these things.

Staff, site 1, post

In addition to being more aware of the need to promote activity, members of staff reported some examples of how their approach to activity had changed:

Well, we just, we’ve got to get the people out [of bed], washed and everything like that, and then we bring them, or they bring them in here [day room] and help with, anybody that’s available, really, help with breakfast . . . because we still do it on a weekend, even when therapy staff aren’t here, the nursing team still do it.

Staff, site 4, post

Behavioural mapping

As outlined in Chapter 3, behavioural mapping was used to provide an indication of activity levels in each unit on 3 separate days at agreed time points before and after EBCD cycle implementation. The approach used in CREATE recorded episodes of social, cognitive or physical activity. These data were from separate groups of patients in each site; we did not seek to compare ‘before and after’ scores for individual patients but rather used the behavioural mapping data as a broad indicator of activity levels in each site. All patients recruited for behavioural mapping were observed at 10-minute intervals between 08.00 and 17.00 or between 13.00 and 20.00 on 3 separate days, including weekends. This allowed up to 60 observations of each patient per day. During each 10-minute interval, the data for each patient was based on an observation made by the researcher over a period no longer than 5 seconds (see Appendix 7 for the full behavioural mapping protocol).

Of note is that the percentages stated in tables relate to ‘inactivity’, so a reduction is a positive finding.

Within-site comparisons

Cross-site comparisons

Pre implementation, site 1 had the highest level of physical inactivity and site 3 had the lowest level of social inactivity. However, overall levels of inactivity across the three domains were broadly similar across the sites. The largest change in levels of inactivity occurred at site 1, where physical inactivity levels fell by 29%. At site 2, levels of cognitive inactivity fell by 13%, mostly accounted for by an increase in participants reading independently or with others. Social inactivity fell by 10% at site 4, accounted for by an increase in talking with other patients, relatives and staff. When comparing overall activity, site 1 had the lowest level pre implementation, at 62.1%, and levels at sites 2 and 4 were 72.7% and 80.6%, respectively, with site 3 having the highest overall level of activity, at 82.4%. Post implementation, all sites increased overall levels of activity, with sites 1 and 2 showing increases of 15%, site 4 showing an increase of 11% and site 3 showing an increase of 3% from its relatively high baseline level.

Summary

Although our behavioural mapping data are largely consistent with previously published work in terms of levels of inactivity, they do suggest that the overall time spent inactive on these stroke units was not as much as that seen in previous studies. 7,86 The behavioural mapping data, although from a small number of participants at each site, highlighted the differences in the provision of individually focused activity opportunities (which remained limited) and the changes we saw in the space/environment of the units, which led to an increase in structured group and informal social activities not easily captured by behavioural mapping. In addition, the behavioural mapping process highlighted the long periods of time that patients spent largely alone, despite the changes to space and activity opportunities, and was one of the main ways in which we observed the limited and task-oriented staff communication processes that largely did not change in the sites between the pre- and post-EBCD periods.

Overall, we question the relevance and validity of using behavioural mapping as a reporting tool in CREATE in relation to anomalies in reporting; the small number of patients who were able to consent the day before behavioural mapping took place; not including scheduled therapy as part of our reporting; and finally the impact of contextual issues such as staff shortages and the severity of disability of the inpatient caseload. These methodological considerations are discussed in more detail in Chapter 6.

Patient-reported experience measures and patient-reported outcome measures

Participants were sent a PROM/PREM pack containing a survey about their experiences as well as the impact of their stroke. Demographic details of all respondents can be found in Appendix 9.

We provide a summary of pre-implementation and post-implementation PREM and PROM data below and the full data set can be found in Appendices 10–12. First, we report on PREM items that scored particularly high or low, and on broad indicators of the quality of the patients’ experiences in these units. As these data are from different cohorts of stroke survivors at each time point and in each site, they provide an indication of patients’ overall experiences of these stroke services during their inpatient stay.

For reference, Box 1 shows the items referred to in the following commentary and Table 14 shows the PROM/PREM response rates.

Pre-implementation patient-reported experience measure data: summary of findings

In terms of the specific focus on the CREATE study, responses to statement 5 were most relevant. Pre implementation, between 30% and 55% of respondents disagreed that there were enough things for them to do in their free time, with only 4% of respondents indicating agreement with the statement at site 1, 27% agreeing at sites 2 and 4 and 44% agreeing at site 3. By contrast, across all sites, ‘facilities’ were considered to be good, with between 74% and 82% agreeing with statement 2.

Across all units, some dissatisfaction was evident in areas of information provision, with < 50% agreement with statement 1 (range 36–48%). Similarly, there were low levels of agreement with statement 10 (range 27–48%), suggesting that limited attention was given to identifying personal goals for recovery. At the same time, these respondents had a higher frequency of agreement with statements 11–13, which focused on feeling as though they and the staff were partners in the process of care (statement 11: range 50–75%), that they had been kept informed during the care process (statement 12: range 41–61%) and that family or carer involvement in treatment discussions could occur if wanted (statement 13: range 67–82%).

There was strong agreement with the statements that there was a friendly atmosphere on the units (statement 6: range 82–91%) and that staff really cared about them (statement 7: range 68–87%), and there were similar levels of agreement that the staff worked well as teams (statement 8: range 67–85%). The respondents largely agreed with statement 9, that they felt that they could talk to staff if they had a problem (range 60–80%). Although the responses to statements 7–9 were positive, up to one-third of the participants disagreed with these, so experiences clearly varied within and across units.

The responses to statement 4 were more mixed, indicating some concerns related to feeling able to discuss personal matters in private (range 51–67%). This reflects to some extent the limited number of private rooms/spaces on the stroke units and echoes concerns raised by former patients in the filmed interviews at sites 1 and 2. At site 3, only 4% of respondents agreed that they received enough emotional support (statement 15), with 33% disagreeing. However, 59% of the respondents at this site indicated that this issue was not applicable to them; allowing for this, the responses still indicated a low level of agreement that enough emotional support was received. At site 1 only 35% agreed with this statement, whereas at sites 2 and 4 more than two-thirds of participants indicated that they had received enough emotional support (statement 15: both 67%).

In terms of respondents feeling happy with the amount of therapy or treatment received in relation to the five areas listed in statement 14, the main area of concern across sites was linked to problems related to continence. Here, agreement levels ranged from 26% at site 3 (although 41% indicated that this was not applicable to them) to 42% at site 2, 42% at site 4 and 57% at site 1. Site 3 had lower levels of agreement for swallowing, speech and mobility, at 33%, 48% and 52%, respectively. These differ quite markedly from those at sites 2 and 4, where levels of agreement relating to swallowing, speech and mobility ranged between 85% and 95%. At site 1, the levels of agreement ranged from 61% for therapy or treatment for swallowing to 80% for both speech and mobility. In respect of treatment or therapy to prepare people for independent living, levels of agreement ranged from 44% at site 3 to 56% at site 1 and 82% at sites 2 and 4, where, with the exception of continence, higher levels of satisfaction were also evident with other areas of therapy or treatment.

Post-implementation patient-reported experience measure data: summary of findings

The post-implementation response rates at sites 1 and 2 were higher than those at sites 3 and 4; the reasons for this were unclear, but the timing of the survey could be one factor, at sites 3 and 4 it was distributed close to the Christmas period. The response rates are detailed in Table 14; in all sites and at each time point when these were distributed, the response rate was < 40%.

We report here on the differences in scoring of PREM items between pre and post intervention. We cannot attribute these differences in responses directly to changes occurring in the stroke units during the EBCD process, but we report on the responses as broad indicators of the quality of the patients’ experiences in these units. The full data set can be found in Appendices 8 and 9.

Among the post-intervention responses, agreement with statement 5 increased at three sites; this was most marked at sites 1 and 2, where the proportion agreeing increased from 4% to 40% at site 1 and from 27% to 50% at site 2. At site 4, the increase was more modest, to 36% from 27%, and, despite these increases, the scores still indicate relatively low levels of opportunity to be independently active. However, the level of agreement with statement 5 at site 3 actually fell quite substantially, from 44% to 9%. These responses are not easy to explain and differ quite substantially from the qualitative data, which record that a range of increased activity opportunities were made available and occurred at this site.

As with the pre-implementation responses to statement 2, there were still quite high levels of agreement that the facilities on the sites were good, but, again, it is interesting to note that although the proportion agreeing with this statement increased at sites 2 and 3 by 4% and 8%, respectively, the level of agreement at sites 1 and 4 fell by 10% and 12%, respectively. In terms of information provision, an area of concern in the pre-implementation period, responses to statement 1 also varied post intervention, with sites 2 and 3 showing a reduction in agreement with this statement of 8% and 10%, respectively, and sites 1 and 4 showing no change.

Co-design groups at all sites had identified communication with patients and families (rather than information provision specifically) as an area for improvement. The post-implementation responses show that providing information early in the inpatient stay remains an area of concern for patients. At sites 3 and 4 the changes introduced (new unit information leaflets) came towards the end of the co-design period and may have occurred too late to be captured in these questionnaires, but at sites 1 and 2 communication-focused changes had been implemented for longer periods of time. At the same time, another statement related to information provision (12: ‘the staff kept me informed every step of the way’) showed improved levels of agreement at each site (range 56–64%, up from 41–61%).

Responses to statement 10 indicated higher levels of agreement at three out of the four sites (i.e. sites 1, 2 and 4) (range 40–64%, up from 27–48%), most notably showing a substantial increase at site 4, from 27% to 64%. At site 3, the level of agreement fell by 8%, from 48% to 40%. So, while acknowledging the lower overall response rates post implementation, some improvement was evident, but responses to these items in the PREM indicate that progression to a more patient-centred approach remains an issue for these sites.

Responses to statement 4 show higher levels of agreement in three out of the four sites (i.e. sites 2–4), with percentage increases of between 4% and 12% (range 55–73%). Site 1 had a 7% reduction to 60% agreement from 67%, so finding somewhere private to discuss personal matters was a concern for some patients. Statement 9 addresses broadly the same area of patients being able to talk to staff about any problems; agreement with this statement fell in three sites out of four from the relatively high pre-implementation levels, which ranged from 62% to 80% to 52% to 76% post implementation, a reduction of between 4% and 10%. Responses to statement 15, which explores receipt of emotional support, showed an increase in agreement in three out of the four sites (i.e. sites 1–3) (range 44–67%), with a substantial improvement at site 3 (up from 4% to 44% and with only 8% indicating that this was not applicable post implementation at this site, down from 59% pre implementation). At the same time as highlighting these areas where services still had room for improvement in communication and support, respondents indicated high levels of agreement with statement 6 (range 80–100% agreement), that there was a friendly atmosphere on the unit, statement 7, that patients felt that staff really cared about them (range 76–93%), and statement 8, that the staff worked well as a team (range 80–97%). Statement 17, relating to patients being satisfied with their progress during the inpatient stay, had higher levels of agreement at all four sites (range 76–86%). Statement 16, addressing patients feeling supported and prepared for discharge, showed small increases in agreement at sites 1 and 2 (to 56% from 54% and to 86% from 81%), whereas site 3 showed a small reduction, from 59% to 56%. The reduction in levels of agreement with statement 16 at site 4 is more marked, with this down 17% from 81% to 61%.

Responses to the five areas listed in statement 14 (respondents feeling happy with the amount of therapy or treatment received) present a mixed picture, as in the pre-implementation period. These post-implementation responses are from a smaller number of patients at all but site 1 and will reflect to some degree the individual post-stroke disabilities and impairments experienced by these respondents. At site 1 there were lower levels of satisfaction with therapy for swallowing problems (20% vs. 67%), but 64% of respondents indicated that this issue was not applicable to them and a further 22% indicated that they were unsure. A similar substantial reduction is seen in responses to the speech and communication statement at site 1 (36% vs. 80%) but again 48% indicated that this issue was not applicable. In addition, quite a large reduction in satisfaction is seen for continence (57% to 28%), but 52% indicated that this was not applicable, and the level of disagreement with this statement among patients at this site fell from 36% pre implementation to 12% post implementation. Improving mobility also showed a reduction but to a lesser extent (68% vs. 80%).

The overall PREM data analysis presented above shows a mixed picture across all sites, with inconsistencies across items and sites. Response rate fell significantly at site 3, which might have been related to the timing of data collection, either side of the Christmas period. Other than the changes in relation to item 5, which we would have expected following the implementation of co-designed activities, it is difficult to draw any firm conclusions from this relatively small sample. We make further commentary on our chosen PREM in our methodological consideration in Chapter 6.

Patient-reported outcome measures

The PROM questionnaire comprised the Subjective Index of Physical and Social Outcome, the EQ-5D and the Oxford Handicap Scale. We did not intend the PROM to be used to derive scores for each of the three subcomponents, but rather to review and describe participants’ perceived levels of disability and adjustment to life after stroke. As previously indicated, these self-completion questionnaires were distributed to two different cohorts of stroke patients, and returned by post. Those who returned questionnaires may not be representative of the stroke patients who received care in the participating stroke units during the time periods identified. Nonetheless, these data provide some insight into the post-stroke characteristics of patients who returned the questionnaires. Below, we describe the broad issues emerging from these data.

Questions 1–10 were from the Subjective Index of Physical and Social Outcome and were about physical functioning/mobility. All respondents had similar difficulties with dressing in the pre- and post-EBCD intervention cohorts; most reported some difficulty with dressing themselves, and on average 17% of respondents could not dress themselves fully post stroke. Across sites, up to 25% of respondents reported either a lot of difficulty with moving around their homes or not being able to move around their homes independently at all. Among the post-EBCD intervention cohorts, respondents at sites 1, 2 and 4 felt marginally more satisfied overall with their ability to perform daily activities around the home than did those at site 3.

With the exception of site 1, where rates were higher in the pre-intervention cohort, those participants responding in the post-EBCD periods reported that they had a lot of difficulty or could not shop for or carry a few items for themselves, indicating relatively high levels of dependency in this area. For question 5, those indicating they were completely dependent or needed assistance most of the time increased in the post-EBCD cohorts at sites 2, 3 and 4, but fell at site 1 from 58% to 32%.

Questions 6–10 related to social/emotional functioning. Except at site 2, where the percentage fell by 9%, response rates indicated that more people in the post-EBCD cohorts were never bored. In all but site 4, there was a small increase in the numbers of respondents reporting a good deal or quite a lot of communication with friends or associates. At site 4 there was a reduction of 30% in the response rate for these two options in the post-EBCD cohort. Again, in relation to levels of satisfaction with activities and interests, respondents were completely or mostly satisfied, whereas respondents at site 4 reported lower levels of satisfaction with responses to these two options in the post-EBCD cohort (down by 30%). There were no differences in response rates to the item relating to visiting friends in either the pre- or post-EBCD cohort.

Our data showed that, across sites, between one-third and half of respondents felt self-conscious to some degree about their appearance, with up to 11% reporting that they tried to avoid going out in public, which is a significant proportion. At the same time, between 45% and 60% of the remaining respondents indicated that they were perfectly happy with their appearance when out in public. There were no major differences in responses between sites in the pre- or post-EBCD cohort.

Questions 11–15 are from the EQ-5D. In this PROM we did not ask participants to rate their perceived level of severity in each of the five domains and so standard scoring is not applicable. We describe below the overall levels of agreement with the statements.

In all sites, > 46% of respondents reported some problems with walking, and between 5% and 18% reported being confined to bed. A similar pattern was evident for washing and dressing, with between 40% and 73% of respondents in the pre-intervention cohort at sites 1–3 reporting having some problems with washing and dressing or being unable to wash or dress themselves. In addition, a high proportion of respondents across all sites reported some problems performing usual activities or being unable to perform usual activities (range 55–84%), with some variation between sites. More than one-third of respondents experienced moderate pain or discomfort, with between 5% and 22% reporting extreme pain or discomfort. Responses also showed that 31% and 46% reported being moderately anxious or depressed. Rates differed slightly between the pre- and post-EBCD cohorts across sites, with a reduction in respondents selecting this response at site 1, no change at site 2 and increases of 5% and 13% at sites 3 and 4, respectively. Rates of those not reporting being anxious or depressed varied by site from 36% to 62%, but there was no consistent pattern between pre- and post-EBCD response rates.

Questions 16–21 are from the Oxford Handicap Scale. The first three items identified little or no impact of disability and impairment on everyday life and the remaining three items indicated moderate to severe impact on everyday life.

Overall, the responses indicate that between 55% and 89% of the respondents had no symptoms, symptoms that did not interfere with everyday life or symptoms that had caused some changes in the respondents’ lives but the respondent was still able to look after themselves. This level of response was largely similar in the pre- and post-EBCD cohorts at sites 1–3 but increased from 77% to 89% at site 4. However, responses to the remaining three items in this section of the questionnaire (questions 19–21) indicated that between 9% (site 3 pre- and site 4 post-EBCD cohorts) and 45% (site 2 pre-EBCD cohort) had moderate to severe symptoms that significantly changed their lives and resulted in their needing help to care for themselves. At the most dependent end of this spectrum, between 3% (site 2) and 27% (site 3) of participants in the post-EBCD cohorts at these sites reported needing constant attention day and night.

Commentary on patient-reported outcome measures

Acknowledging the limitations of the sample size and response rates across sites, these self-report data are largely consistent with the levels of mood disturbance reported in systematic reviews,21,75 and disability and impairment reported by the Stroke Association in the UK and in stroke data internationally. 17 There does not appear to be any great difference in levels of disability and impairment between sites and between the pre- and post-EBCD cohorts. There is, however, some indication that respondents in the post-EBCD cohort at site 3 reported higher levels of post-stroke impairment and disability than at other sites; however, the response rate is < 20% for this cohort. Similarly, any signals that respondents in the pre- and post-EBCD cohorts at site 4 reported less impairment and disability and lower levels of mood disturbance have to be viewed in the light of the very low response rate of only 12%. Overall, these data indicate that stroke patients treated in the participating stroke units before and during implementation of the EBCD intervention were not atypical and do not appear to be any more or less likely to have been receptive to activity-based interventions than other groups of inpatient stroke survivors in other UK stroke units.

Our target recruitment number was based on the anticipated numbers of patients likely to be cared for and discharged from the stroke unit in the 3 months prior to the commencement of the EBCD process and then cared for and discharged in the 3-month period during which EBCD activity took place in each unit. We recognised that this number was no more than a crude estimation influenced by external factors such as caseload factors and admission rate. As it quickly became apparent that response rates were low in almost all sites, we extended the period of time for distributing pre and post questionnaires to 6 months. The differences in the numbers of patients sent questionnaires reflects the number discharged from the units during each time point and known to be alive at that time. This partly reflects length of stay, which was often closely linked to stroke severity.

We had no direct contact with these potential participants, but at site 2 (Yorkshire) the consultant physicians encouraged their patients to complete PROM/PREM questionnaires at their 6-week follow-up appointment, which may have increased the pre-EBCD response rate at that site. We did not send repeat questionnaires at this site as the response rates were good; however, the consultants changed at that site so no similar process was in place for post-EBCD questionnaires. We requested that principal investigators use this method of completion at follow-up clinics at site 1 but they did not do so because of concern about participant burden. At site 3 there were no 6-week follow-up appointments and so, if no response had been received to the questionnaire, second copies were sent out 4 weeks later (after checking whether non-respondents were still alive). We judge that the difficulties with recruitment to and completion of the measures, as outlined above, influenced the validity and the contribution of the PROM/PREM data to our overall findings.

The ‘experience’ of taking part (staff, patients and families)

The themes below demonstrate our understanding of some of the more nuanced reasons why participants wanted to take part and found the whole experience positive. The themes demonstrate why they wanted to be involved and the benefits to individuals, as well as the perceptions of the advantages of working together in a group and seeing through the changes.

Making sense of CREATE and the experience-based co-design process (coherence and cognitive participation over time)

Four members of each stroke unit team were invited to attend CREATE and EBCD training in London. Sites nominated staff; our only stipulations were that those attending should include both therapists and nurses, and that these staff should be aware they would be taking a lead role in the EBCD process at their site. During the full-day training for staff from sites 1 and 2, participants met with researchers and heard about the aims and design of the CREATE study. They also heard directly from one of the originators of the EBCD approach (GR) as well as from an experienced clinician who had led an early EBCD project. We involved a designer (AM) in this training. The day addressed the theory of (a.m.) and practical applications of EBCD (p.m.). Feedback suggested that participants found the day to be lengthy and that the more theoretical overview of the methods was less engaging and less useful than the practical applied material. Based on this, we reduced the training time for staff from sites 3 and 4 to a half-day; we retained the input from the experienced former clinician and focused on practical applications of EBCD by reporting how sites 1 and 2 had worked with the EBCD process and the changes they were putting in place.

Accepting the comments regarding the preference for the more applied approach to training, most participants evaluated the training sessions positively. Staff from sites 3 and 4 heard first-hand accounts of how those at sites 1 and 2 were working to increase activity and discussed examples of changes already made by these teams. The groups also generated a series of practical implementation questions that were discussed during the training session (Figure 16).

FIGURE 16.

The EBCD training event for staff from sites 3 and 4. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

It was helpful meeting other people involved in the study on their wards. What they did and how it involved patients.

EBCD training feedback, sites 3 and 4

From researchers’ reflections on participating in the training sessions and from informal contacts with staff attending both sets of training, enthusiasm levels varied even at this early stage of the study. In both training sessions, staff from one site were much more engaged than staff from the other. In session 1, staff from site 2 (Yorkshire) were more engaged, whereas in session 2 staff from site 3 were more engaged (London). We explored whether this variation could be explained by whether staff had chosen to attend the training or had been selected; however, this was evident for nursing staff in one site only (who had been selected).

Attending training afforded these core staff groups the opportunity to start making sense of the EBCD approach and to think both individually and collectively about how to share this knowledge and understanding with their stroke unit colleagues, about what changes could occur in these settings and what such changes would entail practically for their day-to-day work and that of their colleagues. The training also challenged these staff to think about how working with patients and carers as part of the participatory EBCD approach would differ from their experience of using mainly written feedback on patients’ and carers’ experiences of the stroke units (e.g. through the Friends and Family Test and through surveys).

Making sense of the study and experience-based co-design in the wider stroke team (coherence and cognitive participation over time)

Stroke units are staffed by large teams of therapists and nurses, two or three physicians and support staff. Ensuring that all staff working in the study sites had an understanding of the study and of the EBCD approach remained challenging; one factor influencing this was the time between researchers introducing the study at sites and the data collection and EBCD activity commencing. At sites 1 and 2, major changes to the Health Research Authority and Research Ethics Committee approvals process, which occurred soon after approval for the study was sought, introduced a delay of almost 4 months between training the core groups and data generation and EBCD activity getting under way. Information about the study and the EBCD approach was disseminated through study posters and newsletters and by researchers explaining them at multidisciplinary team meetings and staff handovers at each site. However, the lengthy delay meant that early interest in the study shared by EBCD-trained core groups members quickly dissipated and the spread of understanding of what the study and the EBCD approach would mean for these units was minimal. We had anticipated that, after the training, the time before any study activity took place would have allowed core staff to make their own observations of patient activity/inactivity in their units and to identify and consider opportunities for increasing activity. However, as changes to working practices were not yet required and researchers’ access to sites was limited to planning meetings, only core team members at site 1 did this. The delay meant that in both sites the core staff and wider teams focused on routine work demands; the CREATE study and the EBCD process were not really ‘visible’ and, prior to data generation through observations and behavioural mapping, did not make demands on any staff group at this time.

Once the study was under way at sites, maintaining awareness of EBCD activity remained a challenge as, after the initial enthusiasm generated by the joint meetings dissipated, often some weeks would pass before the actions planned by co-design groups came to fruition and the changes became visible. Early ‘wins’ such as reclaiming the day room at site 4 and opening the new ‘Social Corner’ at site 1 ensured that participants and wider staff and inpatient groups became aware that change was happening. Researchers circulated newsletters on several occasions in each site to raise awareness among staff who were not directly involved in co-design group activity (see Report Supplementary Material 4).

Infection control restrictions on what materials could be displayed in public and staff circulation areas in some sites sometimes negated this strategy, and it was not uncommon in post-intervention interviews to hear some staff say that they had never read a newsletter and that they did not know what was happening in the study. Therapists at sites 2 and 4 reported routinely sharing information about the EBCD-driven changes during their daily handovers or weekly team meetings, so lack of awareness in these smaller therapy teams was not reported:

Because there’s been a couple of therapy staff involved from physiotherapy and occupational therapy mainly, we meet in a morning for a board meeting and we’ve handed over what the progressions have been from the CREATE study, so if they see anything on the wards this is what it is. So say for example we’re telling them about the updating of the garden and if we need to take anybody to the garden.

Staff, site 4, post

Nursing work patterns on stroke units are different from those of therapists, with larger numbers of nurses and health-care support workers in each team. Communication across nursing teams working early, late and at weekends, and on extended day and night duty, is more difficult to achieve, but it was notable that CREATE information was not included routinely in nursing handovers at any of the sites.

However, a factor that increased staff awareness at all sites was researchers commencing data generation; once this occurred, staff interest gradually increased. Some awareness of the study’s purpose also became evident among the patients being cared for on the units at that time through informal conversations and patient/carer questions during the ethnographic observations and through patients’ willingness to participate in behavioural mapping.

Raising awareness and engaging staff, patients and carers through researcher presence on units (coherence and cognitive participation)

At sites 3 and 4 there was no delay between training core staff groups and regular researcher presence on the units as data generation got under way. In all sites this regular researcher presence during the pre-EBCD observation and interview periods was a factor in enabling a wider group of staff to start to make sense of the study. Staff were curious about behavioural mapping and general observational processes, and researchers used this opportunity to provide brief verbal explanations of the study and EBCD to staff and patients on every occasion on which they were generating data. For staff, researchers supplemented verbal explanations with written study summaries, posters and newsletters indicating that the study was under way and explaining how it would proceed. In some units the presence of the researcher prompted staff trained in EBCD to discuss the study during routine team meetings, and this, in turn, reportedly encouraged team members to generate ideas for changes to increase activity. When researchers were conducting observations and behavioural mapping, staff commonly expressed their views about the low levels of current patient activity outside therapy and who could do what differently. Informal comments were very similar to those expressed in the semistructured interviews conducted after the observations:

The presence of researchers on the ward during ethnographic observations seems to have led to staff discussing the CREATE project and the subject of patient activity on the ward. Informally staff tell me they are aware that patients can become very bored and experience low mood if they do not engage in any activity for long periods. This, they report is especially true for patients who are on the ward for a long time; most staff say they know the worst times for inactivity were weekends and evenings. They comment [registered nurses, health-care support workers and therapists] that the study is necessary and that they hope that the situation can be improved. Some staff expressed an interest in being involved in the planned EBCD activities. There were different thoughts about whether this was something HCSWs [health-care support workers] would have more time to do or whether this should be something all staff contributed to. It’s possible that this increased awareness of inactivity led to the observed small number of efforts to increase activity on the ward during the weeks of formal observation.

Field notes, site 2

Although the initial researcher presence was helpful in raising awareness of the issue of inactivity and prompting wider staff groups to think about what would need to happen to change this, informal and pre-EBCD interview comments suggested that some staff were sceptical because they had experiences of good ideas for change being identified but not progressing. It was also difficult for most staff at this stage, namely prior to any EBCD activity, to envisage whether or not, or how, increasing patient activity may have an impact on their existing roles and working practices. Despite some enthusiasm for increasing activity levels, even at this early stage staff routinely expressed the view that their existing workloads and staffing levels, which were often below those recommended in the National Clinical Guideline for Stroke, would not allow them to spend time completing activities with patients. Nurses and health-care support workers most commonly expressed this view. In NPT terms, at this stage, for most staff, the ‘intervention’ was still undefined and not widely understood; commitment to change and comprehension of the possible benefits of changes were not widely evident. In turn, at all sites, there was no expectation or direction from senior therapists or ward managers at this stage that individual staff members or groups should review or change their working practices.

Patients and carers at each site also became aware of the study as a result of the observations and behavioural mapping and, as with staff, they sought information about the study and expressed their views on their current levels of activity outside therapy. This early contact and initial relationship with researchers appears to have influenced several inpatients later (after they had been discharged home) in consenting to participate in pre-EBCD interviews and, subsequently, in EBCD activities. In contrast to some EBCD studies identified in our rapid evidence synthesis, in which the recruitment and retention of service users was problematic, researcher presence proved beneficial in terms of both recruitment to these parts of our study and sustaining patient/carer involvement in EBCD activity throughout the study. 16

The influence of participation in experience-based co-design elements (coherence, cognitive participation and collective action)

At sites 1 and 2, the structured and facilitated approach of EBCD was a major factor in staff making sense of CREATE and the EBCD approach and progressing to a more engaged position, readier to commit to thinking about how change could happen in their site, to revised or new working practices and to thinking through who would need to be involved (cognitive participation). At these sites the initial (separate) staff meetings and then joint meetings enabled quite large groups of staff to experience participation in some of the EBCD cycle stages (see Table 2). They viewed edited trigger films, heard first-hand accounts of patients’ and carers’ experiences, and then jointly identified ‘touch points’, discussed experiences that patients said mattered most to them and considered some solutions to these issues. These were high-energy meetings at which there were strong levels of enthusiasm for change; they were a powerful catalyst for larger groups of staff beginning to share the view that changes to stroke unit environments, access to resources and routine working practices to increase patient activity were not only desirable but also possible (Figures 17 and 18):

It felt quite exciting – it will be interesting to see how it develops, keen to be involved and contribute.

Feedback after joint event, staff, site 3

FIGURE 17.

Staff event at site 2: thinking differently about activities for different patient groups and about how activity could occur in existing spaces. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

FIGURE 18.

Joint event at site 2. Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020.

Reproduced with permission of David Clarke, University of Leeds, personal communication, 2020

Other features of EBCD that facilitated the development and implementation of changes were the defined and time-limited nature of the EBCD process and its use as a previously tested patient-centred service improvement model. This seemed to legitimise the staff time and resource allocation committed to participation in the study and the associated intervention. Staff noted that the EBCD approach contrasted with previous attempts to introduce change in working practices, which were referred to as ‘poorly defined’ in terms of timescales, roles and responsibilities and as operating without additional resources. The patient-centred and participatory approach inherent in EBCD also appeared to have added a sense of responsibility for the core staff groups to deliver on agreed actions, and not to ‘let down’ the patients and carers with whom they worked in the co-design groups.

For patients and carers at sites 1 and 2, separate meetings provided the opportunity to explore their shared experience of stroke and stroke unit care. The trigger films demonstrated commonality of as well as differences in experiences, and watching the films began the process of identification by these former patients and their carers as a group of not only the shared experience of stroke, but also a shared belief in the importance of improving the experiences of patients and carers in their local stroke units. There were some differences in how the trigger films were received; at site 2 staff showed visible emotion at the experiences highlighted by patients, whereas at site 1 there was a mixed reaction:

During the film itself, there was intense attention among the participants and signs of recognition of what patients and family members were saying – for example I noticed some staff members nodding, laughter at some of the observations made by participants. I had expected the film to raise visible emotions but I didn’t see any [tears, sighs, shaking of heads]. There seemed to be recognition and some laughter. On the table of mostly staff where I was sitting there was quiet attention but no visible signs of emotion. I don’t know if on other tables there were more visibly emotional responses.

Researcher reflections, joint event, site 1

Former patients and carers at sites 3 and 4 did not have the trigger film participation in common but they recognised the issues described in the films by site 1 and 2 participants as similar to their own. It seemed that the shared experience of stroke and a desire to improve the experiences of others in these units were sufficient to form a common bond.

Researcher reflections on the joint meetings and co-design group meetings highlighted the importance of former patients articulating their stroke experience and the impact that their impairments had on their daily lives. This seemed to inform an ongoing consideration of why activity outside therapy mattered to patients and how, practically, it could be increased. This common sense of purpose and shared experience is often referred to in discussion of the concept of a ‘community of practice’. 77 A community of practice has been defined as: ‘a process of social learning that occurs when people who have a common interest in a subject or area collaborate over an extended period of time, sharing ideas and strategies, determine solutions, and build innovations’. 77 Although not ‘practitioners’ in the health-care sense, patients and carers participating in the EBCD activities in all of our sites shared a common experience not only with their peers, but also, to a greater or lesser extent in the different units, with the staff with whom they developed an ongoing relationship, shared ideas about activity post stroke and problem-solved in the co-design group meetings.

The structured approach of EBCD and the facilitated events and activities of the EBCD cycle provided former patients and carers with a way to express their experiences and involved them in a group where they worked as partners in bringing about improvements to the stroke units that would increase patient activity:

[. . .] I think, yeah, we felt able to say what we wanted to say and what we wanted to say has turned out to have a valuable effect, so, yeah, very happy to say that we didn’t feel intimidated in any way.

Carer, site 4, post

The same was true of participation by those who were ‘external’ to the day-to-day work of the stroke units. These opportunities contributed to participants’ progression in thinking from a shared understanding of the importance of change (coherence) to working together to define ways to implement, and ideas generated into existing working practices (cognitive participation). This progression includes thinking through how the changes designed to increase activity could be actioned, who would take responsibility for action and, in the longer term, who in everyday practice would deliver or support the actions. These mechanisms represent collective action in NPT terms, and, although the work of implementation on the stroke units commonly fell to staff, the findings section of this report identifies actions completed by carers and former patients to support change. These included designing and painting wall murals (sites 1 and 3) and redesigning patient- and carer-facing documentation (sites 2 and 4), actions designed to increase activity by providing stimulating spaces and ensuring that inpatients and their families were aware of spaces for and opportunities to undertake joint and independent activity.

The above may suggest that EBCD elements, including co-design group activity, progressed without incident. In terms of core group members and participation in co-design groups, this was not the case at most sites. At site 1, for various reasons, three of the four team members who received the EBCD training were not able to participate in the co-design work. However, other team members expressed an interest and volunteered to take on these roles, so staff commitment to the process evolved more naturally from the outset. At site 2, a core team member leading one of the co-design groups experienced problems with taking time away from their ward-based role and then had an extended period of sickness. This had an impact on and slowed the progress of this co-design group, but, as at site 1, other staff members, in this case supported by the ward manager and site researcher, maintained the activity and the agreed actions were progressed. Actions included securing funding for one of the health-care support workers to provide regular hairdressing and personal grooming sessions outside their normal ward role; this realised both person-centred and social activity goals.

At site 3, a champion emerged quickly who led the process at site level who encouraged staff to attend training and sign up for events and attended most if not all co-design groups. Site 3 had the most co-design groups and the most attendees, which we believe was largely because of the enthusiasm and persistence of the local champion (see Table 2). At site 4, two EBCD-trained core team members, both nurses, attended only the joint meeting and did not participate directly in any of the work of the co-design groups they had agreed to. Despite having support from their ward manager, both maintained that they could not take time out of their routine work to participate in the co-design meetings; other members of the co-design groups agreed to conduct the review of priorities and action-planning. Progress was slow in the three co-design groups and few actions were completed at this site until a combined meeting of the groups led to rapid realisation of several agreed activity support actions. The additional contribution of the site researchers as facilitators in supporting and sustaining co-design work is explored later. In our view, the development of communities of practice to a greater or lesser extent in the co-design groups at each site provided a support mechanism that sustained the work and progress of these groups when they were faced with the challenges outlined above.

Leadership, managerial authority and support (cognitive participation and collective action)

Previous EBCD studies have identified the central importance of senior staff support and leadership in service improvements generated through participation in EBCD cycles. Closely allied to this is the presence or absence of managerial authority, resource allocation and support for proposed changes. Based on our review of facilitators for co-production and EBCD-like projects, we aimed to set up project oversight groups with membership including directorate-level managers, senior nurses and those with cross-organisational roles. 16 At sites 1 and 2, we set up these groups, but the commitments of these often very senior managers meant that interaction was largely through e-mail updates or one-to-one meetings with researchers. We could not engage equivalent staff and establish an oversight group at site 3 or 4. The same processes of offering to meet with these staff individually and repeated invitations to attend EBCD activity were adopted, but these were largely not responded to or the promised attendance did not actually occur. One impact of the accelerated EBCD process was that researcher time at sites 3 and 4 was also reduced. Researchers enlisted the support of unit-based EBCD champions in seeking to involve more senior managers, but this did not lead to the involvement in oversight of the project that we sought. Although oversight group members at sites 1 and 2 were often not active participants, they were important in expressing their support for project activity, in helping to navigate often complex trust organisational structures and in specific situations for providing financial resources to enable co-design group ideas. For example, at site 3, staff raised £8000 from participating in a sponsored 10-km run. The chief executive at the site matched the sum raised and these funds were used to redecorate the day room and install a kitchenette. In our view, the oversight groups at sites 1 and 2 provided direct and indirect support for unit-based staff. Having defined senior managerial contacts in these organisations ensured that unit-based staff could activate these potential lines of support. This was not impossible to achieve, but it was made more difficult at sites 3 and 4, as only the chief executive and the therapy lead, respectively, engaged with the EBCD groups and the researchers.

In all sites, a range of staff working in the trust in roles including therapy service managers, matrons, patient experience managers, patient safety officers, volunteer co-ordination and estates management joined patients and carers in different co-design groups. Typically, these individuals were not involved in every co-design meeting held but their involvement was often significant (see below) and had an impact in terms of both cognitive participation and collective action. NHS staff are acutely aware of the resource constraints affecting their services, and often have experienced frustration at barriers they have encountered when seeking to improve or change services. These include infection control and patient safety requirements, lengthy delays or inaction when estates work is requested and lengthy and bureaucratic processes to work through when looking to include volunteers in unit-based activity. Collectively, we found that staff who were otherwise dynamic and enthusiastic advocates for increasing patient activity often shared the view that such barriers were fixed and would, inevitably, limit what could be achieved. The perceptions that ‘the trust’ would not ‘allow’, for example, murals to be painted on walls, shelving to be added in patient bays, hot drinks facilities to be added for patients and carers to use independently or volunteers supporting patients with social eating were pervasive and initially were shared by some of the clinical managers who attended joint meetings at site 1. These perceptions proved to be largely inaccurate.

The contributions of participants from the wider hospital, patient services and volunteer services were significant and these participants typically explained how changes could in fact be realised, and, importantly, provided examples of such changes already operating in the same trust. These participants were used to working with departments and services that were often criticised in the routine discourse of ‘infection control/estates will not allow it’ and understood what was required to introduce, generate support for and bring about changes involving these departments and services. Sometimes these participants undertook to make things happen themselves; for example, at site 2 the patient experience officer worked with the volunteer co-ordinator to identify new volunteers, develop a mealtime support training programme and follow this through to the provision of a regular volunteer-run social eating group on Sundays. At site 3, the volunteer co-ordinator attended the joint event and many of the co-design groups to provide a rapid link to possible volunteers for activities such as reading support and exercise groups. At site 4, the therapy services manager supported the idea of a breakfast club to increase social contact between patients and provide an opportunity for meaningful functional rehabilitation activity. This support was evidenced by changing the working times and early-morning responsibilities of two therapy assistants who were assigned to establish and run the breakfast group (collective action). Therapists and nurses supported the group, quickly establishing it as a 7-day-a-week activity. In this case the work of implementing change was consistent with staff members’ beliefs about the unit as a rehabilitation space, and the work required of the therapy assistants was defined as rehabilitation support integral to their job purpose and consistent with the kind of one-to-one kitchen activity interventions they already completed with patients prior to discharge (cognitive participation and collective action).

However, support from wider trust services was not always evident. At sites 1 and 4, volunteer co-ordinators either focused on the obstacles to volunteer involvement or ignored repeated requests to discuss routine volunteer involvement on the stroke unit. Co-design groups responded differently to such challenges; some decided not to pursue volunteer activity, whereas others, such as at site 4, made direct contact with existing volunteers to request support for arts and crafts sessions and contacted the volunteer programme of a local college, which already had permission to work in the trust, to set up their involvement in regular sessions of activity support on the unit.

At each site, unit-based leaders emerged; often, but not always, these were core group staff who had been through the EBCD training. They were not always the most senior or most experienced members of staff in the units. For example, at site 2, two members of the co-design groups who were particularly active in increasing patients’ social, cognitive and physical activity were a health-care support worker and a stroke rehabilitation assistant. A similar situation was apparent at site 4, where a therapy assistant worked across co-design groups and formed an alliance with the ward manager (who was not EBCD trained). Together, these two staff members drove the initial changes, which included reclaiming and equipping the day room, repurposing a large window seating area and encouraging therapists and nurses to use these. At site 1, one member of staff who had not been to the initial training emerged as the EBCD lead for the ward after the first joint meeting. A general manager and a hospital matron at site 1 also helped expedite changes such as new curtains, clocks and redecoration; their involvement was prompted entirely by a meeting at which unit staff presented the aspirations for the ward and asked them for help with the barriers being experienced. The meeting was organised by the lead facilitator (a dietitian) and the researcher for site 1. Direct actions related to the requests made by the EBCD champion suggested that these staff (a matron and a general manager) also developed some commitment to the project’s anticipated benefits, evidenced by their allocation of resources and continued involvement in seeing these planned changes through to completion. Although these staff were not necessarily part of the community of practice developed through the co-design groups, they made a defined contribution to the collective action needed to improve patients’ experiences at this site.

In terms of leadership style, perhaps the predominant approach evident across sites was an action-centred approach,78 with most staff (but, again, not all) leading co-design groups focusing on and reviewing progress against agreed tasks in each meeting but at the same time actively including, as far as they were able to, patients and carers as partners in the planning and decision-making of the groups. There were no instances in which patients or carers were invited to lead the co-design groups in any of the sites, although, as indicated above, their contributions were actively sought, and our findings confirm that these were influential in the changes that occurred across the four sites. Medical staff largely did not participate in the co-design meetings, but they continued to show interest in the EBCD approach and supported the work in practical ways. At site 4, the stroke consultant sourced and purchased secure, mobile iPad stands and accessed unused hospital bleepers. At site 1, the medical consultant helped facilitate and set up the staff and joint events and secured funds for equipment from his research budget. In recognition of this, the co-design groups chose to name the new social corner after him when he left the unit.

Experience-based co-design activity was welcomed but added to the workload of core group staff (cognitive participation, collective action and reflexive monitoring)

In common with findings from our rapid evidence synthesis of co-production and EBCD studies,55 core staff group members in the stroke units in CREATE were not expressly allocated any time out of their usual working day to participate in and action co-design activity. Instead, staff were encouraged to work flexibly, and their line managers largely encouraged and supported their staff to participate in the EBCD work. This was somewhat easier for therapists at each site than it was for nursing and health-care support workers, for whom taking time out to attend meetings in the course of a shift was often problematic. Ward managers at sites 2 and 4 tried increasing the number of nurses or health-care support workers on shift on days when EBCD meetings occurred but, on some occasions, the additional staff were moved to other wards perceived to be understaffed on those days, so this strategy was not repeated. These ward managers felt that this represented a lack of understanding in the wider organisation of the evidence about the staffing levels required to provide safe and effective care for stroke patients and was an understandable organisational response to recurring staff shortages in different areas of the trusts. However, the enthusiasm for the project at sites 2 and 3 was such that some staff came to the joint meeting on their days off. This kind of commitment was also evident later in the study at these sites, with staff coming in to work on their days off or before a shift commenced to participate in the co-design meetings. Core team members differed in the amount of time they were able to allocate to EBCD activity and progressing the actions agreed in co-design meetings, but most identified that they needed to complete some activities in their own time. These staff typically explained their willingness to undertake this extra work by referring to patient benefits, which were becoming more obvious as the co-design activity progressed in each site (collective action). Where the work of implementing increased patient activity was perceived to be enhancing the patient experience in ways consistent with staff members’ rehabilitation ethos, and where implementation was not requiring substantial change in staff members’ roles, core group members indicated that the extra effort was worthwhile (reflexive monitoring). Other, less involved, team members’ feedback on changes, and on how these were having a positive impact on patients being more independent, completing personal care tasks or engaging in therapy, increased the likelihood that the changes would be sustained:

No, I think it’s given me a massive workload, I think it’s doubled it . . . to be fair. But I was committed, I mean I took it on, I was committed, but I’ve enjoyed that, I’m glad for the changes. [. . .] it’s what I wanted for the ward, from the minute I got there I knew we needed changes, so I was happy to help bring the changes.

Staff, site 4, post

At each site there was some delay before the ideas and actions agreed in co-design meetings came to fruition. In most sites, changes began to happen, sometimes rapidly after the second co-design meetings and were then visible to the wider staff groups. Changes to the environment and use of space in each of the units were positively received. These spaces then began to be used for structured group activities. As these changes became visible to wider staff groups and involved more patients more regularly, these staff in turn appeared more receptive to further changes involving external partners or providers, for example complementary therapy from a local health network (site 1), singers from a local university (site 2) and young musicians from a community arts group (site 3). When changes became visible and change gained momentum across sites, this elicited both positive and negative responses. For example, at site 2 there were complaints that the ‘something about me board just appeared’, with no real consultation with the wider staff group about why it was needed or who would complete the information required. However, the same staff also said that they liked what the board was designed to do. Staff experiences and comments such as these highlight that it is difficult to secure cognitive participation and commitment to collective action from those stroke team members who have had limited engagement with the implementation of EBCD-driven co-designed interventions.

Overall, however, shifts in staff members’ perceptions and behaviours suggested that collective action and reflexive monitoring were becoming more widespread in most units as activity-focused space changes or opportunities became more evident. However, at sites 1 and 4, continuing the work to improve patients’ activity opportunities and engagement relied largely on core group members or those who had taken on these roles as other staff left or did not participate as expected in EBCD activity.

Did the reduced elements/accelerated approach impact on the implementation process (coherence, cognitive participation and collective action)?

In an earlier study of accelerated EBCD, Locock et al. 60 focused on the impact of using an edited film of patients’ experiences compiled from an existing national archive. In our study, at sites 3 and 4, we used existing films from our first two sites. However, we also removed the separate patient and staff meetings. As a result, all participants in these sites had less opportunity to explore experiences and to prioritise areas of change in working practices or resource availability. The other important impact of removing the separate staff meeting was that the opportunity for a larger group of staff members to discuss the project and gain insight into the EBCD process was lost. This appeared to have had an impact on the numbers of attendees at the joint meeting at site 4, which was much smaller than that at the meeting at site 2. The result was that the number of staff, particularly nursing staff, who knew of and sought to participate in study-related activities remained small. However, this was not the case at site 3, where the accelerated EBCD also occurred; here, staff participation was equal to that at site 1. In fact, members of staff asked for a separate staff meeting prior to the joint event so that they could brainstorm ideas for change beforehand, and thus somewhat initiated a replication of the staff event themselves, although this was much more brief. Their ideas were captured and formed a guide for their priorities at the joint event (Figure 19).

FIGURE 19.

Ideas captured at pre-meeting: staff at site 3. Reproduced with permission of Fiona Jones, University of London, personal communication, 2020.

The absence of the edited film and separate patient meeting did not appear to have an impact on the commitment and engagement of former patients, carers and volunteers who participated in accelerated EBCD. Most of these participants had been recruited after being interviewed about their experiences in the stroke unit, so a basic relationship had been established with the researcher at that site. These former patients and carers participated in the majority of the EBCD activities and thus repeatedly engaged with dialogue about, action-planning for and, later, evaluation of changes to the environment and, to some extent, changes to the working practices of stroke unit staff.

Co-production approaches such as EBCD enable participants to focus on and make the changes they care most about in terms of their experience of a service;61 in the EBCD work in the CREATE study, former patients and carers spoke repeatedly of wanting to improve the experience of people who, like them, would need to spend an often extended period in a stroke unit. In this sense, the coherence and cognitive participation of stroke survivors, carers and volunteers was enabled by their continued participation in the work of co-design groups. Because former patients, carers and volunteers in each site were active contributors to designing changes, they played a key role in ensuring that EBCD champions and other staff involved in the work of the co-design groups maintained the momentum and that they saw through the work planned. Former patients and carers became an integral part of co-design groups in each site as they believed in and were committed to making changes happen. As indicated above, these participants were part of the communities of practice that developed following the joint events and through the regular co-design meetings. Opportunities for former patient and carer involvement in reflexive monitoring occurred to some extent as co-design meetings progressed. However, as these participants largely stopped being directly involved after the last co-design meeting, the celebration event was the main opportunity they had to engage with staff they worked with on the EBCD and to review what had been achieved and what was required to sustain these achievements.

Reviewing the qualitative data generated in our study using the lens of NPT ensured that we adopted a systematic approach to exploring the factors and organisational processes that acted as either barriers to or facilitators of successfully implementing, embedding and sustaining EBCD-driven co-produced quality improvement in the form of increases in social, cognitive and physical activity in these stroke units. NPT proved valuable for reviewing our interpretation of the findings from this large data set. In relation to barriers to and facilitators of using full or accelerated EBCD, and implementing complex interventions in health-care settings more generally, using NPT helped identify the challenge of ensuring not only that staff have the time and opportunity to comprehend the nature of the intended change in practice, but also that the mechanisms of change, in this case the facilitated, structured participatory co-production methods inherent in EBCD, are communicated to and involve as many staff in the change process as is feasible. The process evaluation highlighted that, despite its participatory approach, EBCD is subject to the same implementation challenges as other complex health-care interventions in terms of variable organisational buy-in, differences in the commitment of professional leaders, and differences in the engagement of staff whose primary concern is ensuring that their responsibility for care and treatment is discharged and for whom existing routine working practices are familiar and get the required work done.

However, moving beyond NPT’s focus on implementation and the work this brings, it has also been important to reflect on EBCD as a process and the ways in which many stroke unit staff engaged with this process. We found that, rather than being conceptualised by stroke unit staff as an additional burden to those already having an impact on their practice, such as the continuous clinical audit process of SSNAP and the continuing challenges associated with high workloads and staff shortages, the facilitated and structured approach inherent in EBCD provided a legitimate and, to some extent, protected space in which to jointly plan and co-design changes in unit spaces, unit processes and some staff practices. These changes were focused on issues that former stroke patients identified as important and that staff recognised and endorsed as areas that would improve patients’ levels of activity and overall experiences in stroke units. In the context of the CREATE study, although a significant number of changes became evident in all sites post EBCD, differences in enacting what was often referred to as the rehabilitation ethos in the participating stroke units also influenced the involvement of different staff members in promoting and facilitating increases in social, cognitive and physical activity. These, together with the cessation of the time-limited, facilitated and structured EBCD process and with the movement of staff, are factors that will influence the sustainability of the changes now evident in the participating units. Changes involving structured therapy-led activity, such as breakfast clubs and other group activities, are more likely to become embedded in staff members’ routine work and depend less on individual staff members. Activities that depend on volunteer and external group engagement are more at risk of cessation when key individuals move on. We have highlighted the challenge of sustaining changes that occur in co-production projects previously. 55

Interventions to promote increased activity on stroke units

CREATE is similar to other studies that have set out to increase activity levels on stroke units, and we have shown mixed results. However, most previous research has focused on dose-driven interventions, including circuit class therapy and 7-day-a-week therapy. 37 These have increased the amounts of therapy provided but have not increased meaningful patient activity outside therapy sessions. A more general approach to the environmental enrichment of stroke units with attention given to cognitive and social activity as well as physical activity is now gaining interest. 2,87 The main studies,2,87 which have been conducted in Australia and have utilised controlled pre and post designs and are similar to CREATE, have evaluated the impact of a more stimulating environment on inpatient activity. The sample size was greater (n = 30) and the measures were focused largely on quantitative outcomes. Behaviour mapping results showed an increase in activity levels across all domains and sustained changes at 6 months post intervention. However, in these studies, the stimulating activities and changes made were driven by the perspectives of professionals, with no evidence of patient and carer involvement. The authors did highlight the benefit of using change management methods to address implementation, but, unlike in CREATE, they did not use a specified improvement approach.

Co-design approaches in acute health care

In accordance with the findings of our rapid evidence synthesis, we also found that co-design in acute health-care settings can be challenging and time-consuming to set up. Similar to other studies, engaging with co-design and implementing changes were viewed positively by patients, carers and staff. However, unlike many projects that have used EBCD, we incorporated more ‘designerly thinking’ into our project through the activities and support of our co-applicant Alastair Macdonald. We believe that this enabled greater creativity to emerge through the co-design groups and from the site champions.

We found no appreciable difference between the full EBCD cycle used at sites 1 and 2 and the accelerated form used at sites 3 and 4. This mirrors the finding of Locock et al. ,60 although, unlike in that study, we developed our trigger films locally in stroke units instead of using a national database.

Strengths

We believe that the strengths of CREATE lie in the design and the multiple data sources used in our evaluation and intervention (EBCD), and include a considerable range of data from interviews with stroke patients, families and staff (n = 155) and ethnographic field work (365 hours). This has enabled us to reach a deep level of understanding of the experience for patients, families and staff taking part in co-design, and the staff groups who were less directly involved in co-design, as well as the priorities and impact of the improvements on patient activity. We believe that our staged and iterative approach to qualitative analyses has facilitated an understanding of both the similarities and the discrete differences between sites, which have informed our recommendations about the use of EBCD as an improvement approach in other stroke units.

Our process evaluation also drew on multiple data sources, including researcher reflections, which were kept by all members of the research team (FJ, DC, KG and SH) throughout the project. This collaborative approach to sharing experiences, insights and analysis has provided a new level of insight into ‘the work’ of co-design in acute health care and what it takes to succeed and to sustain change.

We believe the cyclical approach used through EBCD enabled a close relationship to be gained with the research teams based in each site. Although this could be viewed as a limitation, the strategies used, such as encouraging teams to seek ways to overcome barriers, involve local community groups to support their work and celebrate their achievements, all helped staff, patients and families to gain a sense of momentum and change.

Limitations

The added value of using co-design to initiate change

Traditional approaches to improving activity on stroke units have focused largely on setting national targets for therapy intensity. These have failed to have an impact on the stroke unit environment and the range and quality of activity opportunities outside formal therapy provision. We now question the narrow focus of this ‘top-down’ approach to therapy and its lack of consideration of the broader rehabilitation concern of increasing activity. Outside the narrow focus of clinical audit, action is required at both national (guideline recommendations) and local levels to increase therapeutic activity outside therapy. This needs to be based on a refocusing on rehabilitation post stroke, needs to be addressed at stroke unit team level (i.e. taking a multidisciplinary approach), and will benefit from the use of participatory service improvement methods such as EBCD.

The success of participatory approaches and EBCD used for the first time in acute stroke services lies in the knowledge that patients, families and staff decided and agreed on what mattered most to them as well as what could be achieved. We believe that the strength of EBCD in both its full and its accelerated form is the facilitated, structured, participatory and time-limited process. The nature of the ‘work’ in CREATE was fundamentally different from usual staff or externally driven quality improvement initiatives in stroke, and it prioritised the participation by stroke survivors and their families in more creative, tactile and relational interactions and outputs to improve opportunities for independent and supervised activity.

It was also evident that CREATE provided distinctly different experiences for staff who engaged in EBCD and that the changes agreed were driven not by external policy pressures or by local organisational demands, and not by external audit, but rather they were shaped very clearly by the experiences of patients and carers whom staff had recently cared for or treated, and were locally determined by small groups who took responsibility for bringing change about. This motivated groups to work together in co-design meetings to define ways to implement the new working practices, what Wenger refers to as ‘communities of practice’ in each site who were central to bringing about change. 88 The involvement of patients and carers increased the accountability of the staff who participated and made it more likely that the planned changes would proceed. Co-design also facilitated carers’ and volunteers’ continued involvement in activities and directly contributed to the sustainability of the changes made to the day-to-day working of these stroke units.

Co-design activities in CREATE also initiated new and ongoing engagement with local people and/or organisations for whom the hospital is a key part of their community. The positive aspect of greater community involvement was notable in one of the London sites threatened with closure at various times. As a result of increased engagement with art communities, CREATE was viewed locally as a ‘good news’ story in terms of providing examples of how co-produced changes improved inpatient services in the ‘local’ hospital. Figure 20 shows a mayoral visit that celebrated the opening of the new post-EBCD ward at site 1. We believe that acknowledgement and celebration of the efforts made by staff, patients and carers – the final part of the cycle of EBCD – is an important factor in raising awareness of change and sustaining achievements.

FIGURE 20.

Mayoral visit to celebrate changes to ward at site 1. 89 Reproduced from the Sutton & Croydon Guardian89 with permission from Epsom and St Helier University Hospitals NHS Trust.

Reproduced from the Sutton & Croydon Guardian89 with permission from Epsom and St Helier University Hospitals NHS Trust.

The ongoing challenge of (in)activity in stroke units

Culture change in any organisation is challenging, and in our project this was no different. The process evaluation highlighted the importance of organisational buy-in but also the challenges of achieving this at unit level and beyond. It was also clear that EBCD provided both a structure and a space for changes aimed at directly improving the experiences of stroke patients and their family carers. As a result of making these changes, the morale and commitment of staff also seemed to improve. Nonetheless, although improvements in activity opportunities were observed, the process evaluation also highlighted the challenge of involving and keeping engaged large groups of staff in the participatory change process. Similarly, differences in staff members’ perceptions of their role in rehabilitation and enacting this in day-to-day working practices highlighted that interdisciplinary rehabilitation and increased activity promotion across a whole stroke team would require more sustained intervention targeted at how teams work in this context, which was beyond the scope of CREATE. Although the tangible improvements to the space and activity opportunities was visible across all stroke units and validated by our qualitative findings, when behaviour mapping was used as an objective measurement tool no consistent change was found in the proportion of time on weekdays and weekend days that stroke patients in the participating units spent on physical, social and cognitive activities. The effectiveness and long-term impact of lengthy periods of co-design work (ranging from 4 to 9 months in this project) on the independent and supervised activity of stroke patients is still uncertain. Currently, work routines in participating units and day-to-day interactions between staff and patients are largely task driven, focusing on care needs and delivering scheduled therapy. Interactions that facilitate greater social exchange, cognitive activity or physical exercise are relatively uncommon, and we believe that to increase such activity opportunity requires a fundamentally different approach to how therapy and nursing contribution to rehabilitation is viewed and delivered so that stroke unit environments and work practices become more conducive to activity and so that activity is seen as a priority and part of the rehabilitation work of all staff, not only therapists.

At all CREATE sites we found concern that ‘something must be done’ and a willingness for staff, patients and families to work together to make improvements. We believe that, for this approach to be used across other stroke units, local facilitation needs to be provided by a member of staff who has protected time to be allocated to this work. Change required buy-in and commitment from multiple stakeholders including senior management to validate the shift from achieving national performance targets to a cultural recognition of the therapeutic value of a stroke unit. Change was facilitated when stroke unit staff worked with volunteers and people from the community. This will take time, but trigger films such as those developed in this project can help gain traction to make changes, some of which can be initiated quickly and relatively cheaply (e.g. photo-hangers by bedsides, and the ‘a little something about me’ board).

Research support

Tino Kulnik conducted pre-implementation interviews for site 1 and supported several co-design events and the input of behavioural mapping data.

Alessia Costa supported ethnographic fieldwork at site 1.

Louisa-Jane Burton and Sadia Ahmed supported the behavioural mapping and PROM/PREM analysis and interpretation.

Carole Pound supported co-design events at site 1.

Parminder Dhiman carried out data collection at site 1 (PROM/PREM and behavioural mapping).

Study Steering Committee

Dr Rebecca Palmer (chairperson)

Dr Gillian Cluckey

Amanda North

Pamela Beaumont

Jeanette Gledhill

Dr David Meads.

Site and support

Principal investigators at all sites: Dr Paul O’Mahoney and Dr Janet Putteril (site 1), Dr Liz Iveson (site 2), Dr Camille Julien and Dr Karen Kee (site 3), and Dr Anand Nair (site 4).

Participants and co-design groups at all collaborating sites – members of staff, patients and family members.

Community support

Hana Horack and Team Imagineers have applied for funding and run art workshops for the patients at weekends at our first site. Cherilyn Yeates from Croydon Arts and Music gave time to paint a mural at site 3.

Croydon Arts and Music give concerts at the weekends at site 3.

The Reader charity and the Stroke Association, in particular Jack Blaney, supported the co-design process at site 3. The UK Stroke Forum acknowledged CREATE through the Patient, Carer and Public Involvement Prize.

Performance-level indicators

• Brain scanning.

• Stroke unit.

• Clot-busting treatment.

• Specialist assessments.

• Occupational therapy.

• Physiotherapy.

• Speech and language therapy.

• Team working.

• Standards met by discharge.

• Discharge process.

Aims

To estimate what proportion a weekday and weekend day, stroke survivors recovering on ward [. . .] spend in physical, social and cognitive activities.

The day before

Check with the staff which patients are suitable for observation (n = 10 are required).

The criteria for inclusion are:

1. Inpatient with a confirmed diagnosis of stroke.

2. Not being managed palliatively.

3. Able to understand English or have a family member who can act as an interpreter.

4. Able to give full informed consent on the day before the behavioural mapping.

5. Patient is happy to participate.

Check that the patients will consent to be observed.

You may want to explain to them about the research and tell them that you will be watching them every 10 minutes throughout the day to gather important information about their activity so that we may ultimately improve the care of people with stroke.

Tell staff which patients have consented to participate.

Confirm what bed the patients will be in the following day.

You will need to check this again in the morning as patient locations change. Also check that the patients are not going home that day. If there is a plan for the patient to be discharged in the morning then do not bother including them. If the patient may be discharged later that day, it is worth including them just in case.

Spreadsheets

Prepare the behavioural mapping spreadsheet so that no copying is required on the day of observations, including adding the date and time, in advance.

Procedure

The following procedure reflects what was performed as a part of the Enriched Environment Post Stroke Trial. We have adapted this protocol and the following procedures to meet the requirements of the CREATE study. Any modifications to the protocol must be agreed with Fiona Jones and David Clarke.

On the day

Explanation of terms used on spreadsheet

Distinguishing between categories

Reading and writing: is to be categorised as a cognitive activity.

Group activities: if observed in group activities, there is the potential for the patient to be engaged in more than one type of activity. For example, talking while performing upper limb tasks in an upper limb group. If though, on observation the patient is in a position or situation conducive to interaction (as per the definition of ‘people present’), and no physical activity is observed, only social activity will be recorded.