This study involved two main aspects: (1) the ‘intervention’, consisting of ‘full and accelerated’ EBCD, to generate and implement a number of co-designed changes to increase supervised and independent activities within four stroke units, and (2) the evaluation, which was carried out pre and post implementation of the co-designed activities in each unit. For the purpose of the report, we first document the methods used in the intervention development and then present the evaluation components in Chapter 3.
Parts of this chapter are based on Clarke et al.16 © Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/.
The intervention: full and accelerated experience-based co-design
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.
| Characteristic | Site 1 | Site 2 | Site 3 | Site 4 |
|---|---|---|---|---|
| Number of stroke beds | 24 | 24 | 26 | 26 beds (14 and 12) |
| Hospital | District general hospital with 629 beds | District general hospital with 500 beds | City-centre teaching hospital with 700 beds | District general hospital with 600 beds |
| Layout | The stroke unit is on the third floor of the hospital. It has an L-shaped layout with five bays, each containing four beds, and four single rooms. The end of one corridor connects directly to another medical ward and the nursing desk is at the end of the other corridor. Visitor and staff catering services are on the ground floor of the hospital | The stroke unit is on the third floor of the hospital. It has an L-shaped layout, a long main corridor with five bays, each containing four beds, and three side rooms off this corridor. Day room and one-bed pre-discharge flat are in the smaller L-section. Visitor and staff catering services are on the ground floor of the hospital | The stroke unit is on the ground floor of the hospital. It has a central desk and one wide corridor with two male and two female bays, each with six beds, and two single rooms at the entrance to the ward. Visitor and staff catering services are on the ground and first floors of the hospital | The stroke unit is on the second floor of the hospital. It has a circular layout around four ‘pods’ that make up the stroke service. Three pods, 7a, 7b and 7d, are rehabilitation wards and 7c is the hyperacute ward. There is access by lift to outside spaces including a small therapy garden and walkways around the main hospital site. Visitor and staff catering services are on the ground floor of the hospital |
| Shared space | No day room; no outside access | Day room used by staff as a storage and meeting area at start of study. No outside access | Day room used also by staff; access to outside garden | Day room on 7d in use as a chair store (32 chairs) at the start of the study Day room on 7b accessible but used frequently for staff meetings |
| Visiting hours | 14.00–20.00 | 11.00–20.00 | 14.00–17.00/18.00–20.00 | 11.00–20.00 |
| Meetings | Nursing handover between the day and night staff each morning. A brief morning multidisciplinary meeting, known as the ‘whiteboard meeting’, and weekly MDT meetings to review discharge plans. MDT meetings on Tuesdays | Nursing handover between the day and night staff each morning and at 12 o’clock for the late shift. Therapists and nurses meet each Monday morning. This ‘board round’ is at 09.00; MDT meeting every Wednesday | Nursing handovers every day at 08.00. and 20.00; MDT meetings on Friday afternoon | Nursing handover between the day and night staff each morning; therapist handovers follow this Monday to Friday |
| 7-day therapy service | No | An OT and a PT work on Saturday covering the acute and rehabilitation units. On Saturdays and Sundays one stroke rehabilitation assistant is on the unit | No | No |
| Performance in last Acute Organisation Report 201664 when units were graded against 10 key indicators (see Appendix 1, Table 15) | Achieved 7 of the 10 key indicators | Achieved 4 of the 10 key indicators | Achieved 8 of the 10 key indicators | Achieved 5 of the 10 key indicators |
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.
Project governance and management
The project needed Health Research Authority approval, including an independent ethics review. In each of the four sites, a senior clinician was identified who acted as principal investigator; they negotiated site access, supported local approvals and took day-to-day responsibility for the study, including identifying potential participants.
In addition to the senior clinician, each stroke unit nominated a group of core clinical staff, which included senior nurses, therapists, dietitians and psychologists. They played a key role in EBCD by helping facilitate introductions and communications with local stakeholders such as head of estates, volunteer co-ordinators, matrons, general operation managers, and communication leads for the trust. After receiving training from The Point of Care Foundation about the six stages of EBCD, the core groups assisted the research team with communications about the stages of co-design and explained to all staff how this might advance in their own stroke units. The Point of Care Foundation training was delivered by an independent facilitator who had experience in EBCD and it was carried out before EBCD commenced with teams from across the two stroke units in stage 1 and then later with the further two stroke units in stage 2. Training consisted of a full day for sites 1 and 2 but was reduced to a half-day for sites 3 and 4 following feedback from clinical teams. Further detail of training is in Report Supplementary Material 1.
In each stroke unit, the core staff and principal investigators helped researchers by identifying patients and family members who might want to engage with co-design and inviting them to do so. Following the guidance given in the EBCD toolkit,65 patients were recruited if they had been an inpatient on the unit in the previous 3–6 months and ranged in terms of ethnic group, gender and stroke severity. We set out to recruit patients with and patients without family members and a similar number of staff.
Steps of the process (full and accelerated experience-based co-design)
We followed all six stages of EBCD at sites 1 and 2, completing the full co-design process. This enabled staff, patients and families to reflect on their experiences of the acute stroke unit, work together to identify improvement priorities and devise and implement changes, and then jointly reflect on their achievements. At sites 3 and 4 we began an ‘accelerated’ process at the joint staff and patient event, using the composite films from sites 1 and 2 to trigger discussion about priorities for co-design. EBCD in CREATE was also embedded in our mixed-methods case evaluation and we undertook pre- and post-implementation data collection that is not part of a standard EBCD approach; this comprised extended ethnographic observations, behavioural mapping and patient-reported outcome measure (PROM)/patient-reported experience measure (PREM) questionnaires.
The six stages of the EBCD cycle are described as typically taking 9–12 months to complete.51 The full EBCD cycle, contextualised to CREATE and used at sites 1 and 2, is shown in Figure 1 and described in more detail below.
FIGURE 1 Full EBCD cycle used at sites 1 and 2. (Figure not reproduced here; see the original article.)
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.
Patient events
Through a staged process of facilitation and discussion, following guidance and methods suggested by trainers from The Point of Care Foundation, patients and families viewed the composite film and explored their ideas for change. These methods included an icebreaker exercise and working in small groups to brainstorm ideas and emotions after viewing the film. Emotional mapping was used to rank and prioritise ideas for change from most to least important and to refine the final list to be shared with staff at the joint event.
Staff events
The staff events were structured in a similar way to the patient events but without the use of a composite trigger film. Discussion and ideas were generated following the research team’s presentation of observational data and staff interviews. Through a staged process of facilitation and emotional mapping exercises, they explored ideas about areas for change, and generated a list of ideas and priorities as a group to present to patients and families at the joint event.
Joint events
Patients and staff then came together for a joint event. The numbers of attendees at each event are shown in Table 2. Other stakeholders from the trust, including volunteer co-ordinators and senior nursing and therapy managers, also attended these events. Attendees watched the composite film, and the staff and patient/family groups then separately presented their list of priorities. Facilitated by the researchers alongside the core team members/champions, staff, patients and families worked in small groups through a staged process of sharing and discussing what they had heard, what resonated and what they perceived as the most important priorities for change. During several stages of discussion, each small group chose their joint priorities; through stages of voting and discussion, the wider group agreed a final list. Participants then indicated which priority areas they preferred to work on and signed up for one (or more if they wished) of the co-design groups at each site.
| Site | Co-design participants (n) | Number of co-design group meetings | ||
|---|---|---|---|---|
| Patients | Carers | Staff | ||
| Site 1 | 5 | 6 | 13 | 10 |
| Site 2 | 4 | 4 | 12 | 9 |
| Site 3 | 11 | 7 | 14 | 15 |
| Site 4 | 3 | 4 | 15 | 9 |
| Total | 23 | 21 | 54 | 43 |
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.
Break point
Our original proposal aimed to have a break point between phases 2 and 3 (full and accelerated EBCD) to review the results and evaluate changes in behavioural mapping data and experiences of implementing the co-produced interventions from qualitative findings. If a positive change was found in supervised and independent therapeutic patient activity following the implementation of co-produced interventions, evidenced in either behavioural mapping data or qualitative data from post-implementation interviews and feedback events, then we would proceed to test the interventions in two further stroke units in phase 3. Subsequently, ethics approvals (see Report Supplementary Material 3) stated that we would have to submit substantial amendments detailing the range of changes in phase 2 (sites 1 and 2) that might be expected, and this approval would need to be in place before recruiting and commencing at sites 3 and 4. We were able to demonstrate to the Study Steering Group that the qualitative data indicated positive change but changes in behavioural mapping data were inconsistent. Following guidance and discussion with our Study Steering Group and NIHR manager, we had agreement to proceed to phase 3. On reflection, the use of the break point should have been defined not as ‘potentially stopping’, but rather as using the findings from phase 2 (sites 1 and 2) to plan and inform the accelerated EBCD at sites 3 and 4.
‘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. (Figure not reproduced here; see the original article.)
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.