Large-scale implementation of stroke early supported discharge: the WISE realist mixed-methods study

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 16/01/17. The contractual start date was in September 2017. The final report began editorial review in February 2021 and was accepted for publication in June 2021. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Disclaimer

This report contains transcripts of interviews conducted in the course of the research and contains language that may offend some readers.

Copyright statement

Copyright © 2021 Fisher et al. This work was produced by Fisher et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaption in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2021 Fisher et al.

The NHS landscape

Stroke has a devastating and lasting impact on people’s lives and on the nation’s health and economy. Stroke is one of the top four causes of death and one of the largest causes of adult disability. 3 Recovery can continue for many years after an individual has a stroke; therefore, it is important that consideration is given to how to provide stroke rehabilitation services. The NHS Long Term Plan4 recommends increased investment in community health-care services. The provision of rehabilitation for stroke survivors following hospital discharge has also been highlighted as a national priority.

In England, the National Clinical Guidelines from the Royal College of Physicians5 and the National Institute for Health and Care Excellence6 recommend the provision of early supported discharge (ESD) as part of an evidence-based stroke care pathway. ESD is a multidisciplinary team (MDT) intervention that facilitates discharge from hospital and the delivery of stroke-specialist rehabilitation at home and at an intensity equivalent to that provided in an acute stroke unit. There is currently widespread implementation of stroke ESD services in the UK. However, despite research and policy drivers, national audit reports showed that the type of ESD services that stroke patients receive is variable and, in some regions, ESD is still not offered at all. 7 Alternative models of operation have been adopted, but it is not known how close they are to the evidence-based models that have demonstrated effectiveness in clinical trial settings. It remains unclear whether or not health and cost benefits of the ESD intervention are achieved when services are implemented in practice.

This report presents findings from a multimethod study investigating what, why and how ESD services are implemented and operate in real-world settings. The study responds to the second translation gap between clinical trials and clinical practice, with the view to facilitate further implementation of ESD nationally and internationally. We drew on realist evaluation (RE) principles to unpick the interplay between ESD services and the context within which they operate, and to determine whether or not and how they are effective and in what conditions. 8

Existing evidence

There is strong research evidence to show that the provision of stroke-specialist rehabilitation enhances recovery. 9,10 Service delivery models that offer home-based stroke rehabilitation have gained increasing interest, particularly as health-care services face the challenges of cost reduction and integrated care provision. 4,11 Cumulative evidence from clinical trials has shown that ESD can reduce the length of hospital stay and the risk of dependency of stroke survivors. 10

Our research to date, together with national clinical guidelines, has defined core evidence-based components of ESD,12,13 essential characteristics that theory suggests need to be implemented for the intervention to remain effective in practice. These include that ESD should be delivered by a stroke-specialist MDT, and consist of co-ordinated discharge from hospital and rehabilitation provided to eligible patients in their own home. 5,6,12–14 Findings so far have informed the successful implementation of a number of evidence-based ESD services in areas across the East Midlands. 12 What remains to be investigated is the impact of ESD in other geographical areas, given that these areas may differ in the ways that may influence how ESD operates.

In most other high-income countries, ESD has not been well developed in practice, resulting in some locally established services but a lack of large-scale implementation. 15 In many countries it is unclear what provision of rehabilitation there is for stroke survivors beyond the hospital setting, with the majority of national audits or registries focusing on acute stroke care. 16 There are also unanswered questions relating to the implementation of ESD in rural settings, with the original clinical trials mainly conducted in urban sites. 10,13 It has been reported that the role of ESD services in more dispersed rural communities has not really been adequately addressed. 17

In England, with outcome-based commissioning being a priority in the NHS, it is important to investigate whether or not ESD services are still effective in real-life clinical settings. Distinguishing between effective and ineffective implementation is crucial to addressing inequities in service provision and plan service improvements. 18 It is also vital with regard to informing the implementation of sustainable evidence-based service models.

Study aims

1. To investigate the effectiveness of ESD services when implemented at scale, in practice.

2. To understand how the context within which they operate influences the implementation and effectiveness of ESD schemes.

3. To identify transferable lessons to drive effective implementation of stroke ESD in clinical practice.

Research questions

1. What adopted models of ESD exist and how do these relate to evidence-based recommendations?

2. Can realised benefits of implementing ESD be quantified by defined measures of effectiveness: reduction in length of hospital stay, responsiveness of the service, amount of rehabilitation delivered and changes in patient dependency?

3. What site, model and patient-level characteristics influence the effectiveness of ESD services?

4. What are the cost consequences of adopted ESD models?

5. What contextual elements influence whether or not ESD is implemented in the first place and how do they shape the model of the service adopted?

6. What are the perceived outcomes of implementing ESD from the perspective of service users, clinicians, managers and commissioners, and how are these achieved in practice?

7. What are the conditions that contribute to the successful implementation and sustainability of ESD in practice?

Theoretical implications

Large-scale implementation of ESD needs to be considered in the light of the complexity inherent in the delivery of this type of intervention. ESD is a multidisciplinary and multicomponent health-care intervention that involves a critical mass of stakeholders working across different organisation settings along the stroke care pathway. It is acknowledged that successful community-based care requires not only a shift in budget investment from acute to community services but also dissolving traditional occupation and organisational boundaries. 4,19

Early supported discharge services do not operate in the controlled environments of experimental settings but in a complex and multilevel system, such as the NHS. They are, therefore, exposed to a range of contextual influences at different levels of the health-care system that act synergistically or antagonistically to evidence-based implementation. Decoding the observed variability requires distinguishing between receptive and non-receptive contexts, as well as understanding the interplay between these environments and the programme’s ‘active ingredients’. 14 This process of inquiry permits a better understanding of why the programme works in certain settings and not in others, what mechanisms underlie the programme’s success and what steps we need to make to achieve the desired outcomes. Findings can then inform future implementation, reconfiguration and improvement of ESD services to facilitate provision of evidence-based stroke care.

Implications for the NHS

This research was designed to benefit patients affected by stroke by investigating the effectiveness of services that they experience and to identify ways to improve the provision of care. The research focuses on provision of care at a particularly distressing time: when stroke survivors leave the hospital and face the consequences of stroke back at home. By investigating models of community service in practice and identifying the consequences of adopting different ESD models in variable contexts, we hope to provide clear guidance to commissioners with regard to the outcomes and impact of their decision-making, and, hence, to influence the commissioning of stroke services and address inequity in service provision.

Patient and public involvement

The Nottingham Stroke Partnership group is a group of stroke survivors and caregivers who meet bimonthly to provide a patient and public perspective to stroke research at the University of Nottingham (Nottingham, UK). This group is co-chaired by stroke survivors and Dr Rebecca J Fisher. This group have provided guidance and support throughout this study. Their concerns about inequitable access to stroke care in the community, particularly in rural areas, informed the study protocol. Trevor Gard and Frances Cameron have been members of our steering group, offering guidance on interpretation of findings, approaches to patient interviews and lay summaries of our results. They advised us on our approach to sending a feedback package to our patient participants, in which we summarised our findings and offered our thanks.

Overview of the study

Methodology

Study design

We adopted a mixed-methods design to draw information from multiple complementary sources. Our previous research and quantitative data analysis informed mainly the ‘Context’, ‘Mechanisms-resources’ and ‘Intended Outcomes’ and qualitative data used to elicit ‘Mechanisms-responses’ and ‘Unintended outcomes’. 12,13,23 The study was conducted in stages corresponding to two interlinking WPs undertaken sequentially (Figure 1).

FIGURE 1.

Flow chart of the study, including WPs 1 and 2.

Work package 1 began with a literature review that aimed to identify key contextual determinants to the implementation of ESD services and elicit potential mechanisms. Quantitative analyses of historical prospective Sentinel Stroke National Audit Programme (SSNAP) data were then performed. The nature of the statistical modelling approaches used meant that context was explored in terms of defined and measurable variables, some of which were treated as ‘confounding’ (or ‘controlled for’) factors. We recognise that this is somewhat at odds with the philosophy of RE. However, the decision was made to use this quantitative analysis primarily to test the causal relationship between the core components of the intervention and the outcomes and, hence, findings were written up in line with recognised statistical convention.

These analyses did enable an investigation of the predictor variables of interest (as mechanisms-resources) in terms of our measure of adoption of evidence-based core components. By investigating the relationships between context, mechanisms-resources and outcomes as measurable ‘factors’, we were able to test our preliminary programme theories from a quantitative perspective. Findings informed the development of CMO hypotheses, which were tested through WP2.

Work package 2 generated insights from the perspectives of actors and stakeholders who make ESD happen in the real world, enabling mechanism-responses to be explored. An exploratory, multiple-case study design allowed ESD sites featured in WP1 to be investigated qualitatively, drawing information from individual interviews, focus groups and documentary evidence. A cost–consequences analysis (CCA) was performed to investigate the costs associated with ESD implementation, drawing information from interviews and documentary evidence. Finally, data from each WP were synthesised in relation to programme theories (with underpinning CMOs) that address the study’s questions.

Work package 1: how effective is early supported discharge when implemented at scale in practice?

Evaluating the effectiveness of early supported discharge service provision (see Chapter 4)

To investigate our study hypotheses, information about each ESD service, collected by SSNAP in the 2015 post-acute organisational audit, was collated by the research team. ESD team models were then analysed by comparing ESD service information with ESD consensus statements and national clinical guideline recommendations. 5,6,13 An ESD consensus score was then applied to each ESD team. Multilevel modelling enabled us to appreciate the variation in outcomes as a mixture of patient variability nested within ESD service provision variability. This approach is displayed graphically in Figure 2.

FIGURE 2.

Schematic representation of the WISE (What is the Impact of Stroke Early supported discharge?) study multilevel modelling approach. The triangle represents the fact that many patients are nested within fewer ESD/hospital units in a multilevel structure.

Additional ‘contextual’ variables at the patient level in the model, obtained from the 2016 SSNAP clinical data, included age at admission, sex, pre-stroke independence, comorbidities, National Institutes of Health Stroke Scale (NIHSS) score on admission, type of stroke and modified Rankin Scale score at discharge from hospital. 29,30 These reflect previously validated stroke case-mix models. 31,32 A key factor of interest at the ESD team level (in addition to the ESD consensus score) was the level of rurality. 33 Effectiveness of ESD service provision was measured with the following outcome variables: responsiveness (time from hospital discharge to first contact), rehabilitation intensity (total number of treatment days/total days with ESD) and stroke survivor outcome (modified Rankin Scale score after ESD delivered).

Work package 2: how do contextual factors influence the implementation and effectiveness of early supported discharge in practice?

Description of overarching programme theories

Rapid evidence synthesis

Conclusion

The RES was a first step in assessing our initial programme theories and informing the development of CMO hypotheses to be tested through data collection and analysis. In line with our third programme theory, findings further supported the need to consider the services’ wider networks and communication processes. They also identified certain patient-level factors that deserve further exploration, including patients’ beliefs and expectations of their own recovery and how ESD could contribute to their recovery process. One of the included studies suggested that certain features of physical geography (i.e. hilly terrain) were a common patient-reported barrier to outdoor physical activity. 53 We did not identify any other studies investigating the impact of rurality on ESD, highlighting a paucity of research in relation to the influence of geographical factors.

We are aware that this synthesis was pragmatically tailored to resources and time constraints and, therefore, very focused. As a result, it did not consider findings from a wider body of evidence (e.g. different population groups, grey literature) that could have further enriched insights from this review. Although our intention was to include quantitative studies, very few studies contained any description of contextual factors relevant to the purpose of this study. The decision to include studies based on their relevance to our research question, rather than their scientific quality, was a pragmatic one, considering the small sample size of relevant studies. Although this decision could have introduced bias, we need to note that in the context of the wider project this synthesis was a starting point to further data collection and hypotheses testing. Although we attempted to configure findings across studies into CMO hypotheses, this study was not a realist synthesis and, therefore, we did not engage in the process of distinguishing between contextual features and underlying mechanisms. None of the included studies had attempted to make these links, reiterating the need to better understand the interaction between contextual influences and the evidence-based active ingredients of the intervention.

Introduction

The SSNAP is the national stroke register of England, Wales and Northern Ireland in which all acute admitting hospitals and post-acute stroke teams are mandated to participate. 61 SSNAP has played a key role in monitoring the performance and improving the provision of acute stroke care. The collection of SSNAP data from community stroke services now offers a unique opportunity to investigate the large-scale impact of ESD.

Clinical guidelines recommend that ESD services should provide responsive and intensive rehabilitation (with treatment at home beginning within 24 hours of hospital discharge), with the aim to promote stroke survivor recovery. 5,6,62–65 By investigating if and how these aspects of an effective ESD service can be realised in practice, this study aims to inform the provision of evidence-based care for stroke survivors.

Our previous research has hypothesised that the active ingredients of ESD can be defined with evidence-based core components,13 and that these core components are essential characteristics that need to be implemented for the ESD intervention to be effective in practice. 12 The aim of this study was to determine if such core components had been adopted by ESD teams in real-world settings in England and whether or not these related to realised benefits of ESD.

Methods

Study design

We present results from this observational cohort study (Figure 3), which was conducted as part of the larger mixed-methods study. 37 We determined a priori a sample size of 4750 patients for a study power of 80% to detect standardised effect sizes of 0.25 for each outcome.

FIGURE 3.

The ESD study design flow chart. WISE, What is the Impact of Stroke Early supported discharge?

Results

Figure 4 shows the variation of ESD consensus scores across the 31 ESD teams and four English regions. The total ESD consensus scores across the 31 teams varied between 5 and 15 {mean 10.6 [standard deviation (SD) 2.4]}, with no team achieving 100% adherence, reflecting that a range of ESD models had been adopted. In terms of the English regions, adherence to the core components of ESD service delivery was greatest in the East of England.

FIGURE 4.

Distribution of ESD consensus score by (a) study team; and (b) region.

All 31 ESD teams included in this study reported that they provided a stroke-specific service (see Table 5). Only three teams reported having at least the recommended level of input from doctors and only four teams said that they had access to social workers. For the range of ESD models, there was a mixture of urban and rural settings [mean level of rurality 35.6 (SD 21.8)], as well as varying performance of associated referring hospitals [mean SSNAP hospital rating score 72.2 (SD 12.1)].

Data from 6260 patients with a completed NIHSS score were included in the primary analysis, and their characteristics are shown in Table 6. The majority of patients (91.9%) had a mild or moderate stroke (NIHSS score of < 15). The most common age group was 70–79 years (30.8%) and 4151 (66.3%) patients were functionally independent prior to their stroke (modified Rankin Scale score of 0).

In terms of the outcomes, 69% of the sampled patients were seen after ≥ 1 day, with 31% of patients seen within 1 day for the ‘days to ESD’ variable. The median rehabilitation intensity value of the sampled patients was 0.38 treatment days for every day with the ESD team, with the 25th percentile being 0.19 and the 75th percentile being 0.59. For the stroke survivor outcome measure, 9% of sampled patients were classified as moderate to severe at ESD discharge (modified Rankin Scale score of 4–5), with the percentages of patients with a modified Rankin Scale score of 0, 1, 2 and 3 being 9%, 31%, 31% and 20%, respectively.

Results of the multilevel modelling are presented below. The degree of clustering was greater for the process measures ‘days to ESD’ and ‘rehabilitation intensity’ than that for the patient outcome measure of modified Rankin Scale score (adjusted intraclass correlation coefficients: 0.56, 0.26 and 0.08, respectively). Figure 5 shows the amount of variation among the 31 ESD teams in relation to the outcomes of this study.

FIGURE 5.

Distribution of study outcome variables across study teams. (a) Proportion of EDS teams that make patient contact within 24 hours; (b) mean rehabilitation intensity; and (c) mean modified Rankin Scale score.

Results for the association between the total ESD consensus score and the ‘days to ESD’ variable are shown in Table 7, which are unadjusted and adjusted for all patient characteristics, the level of rurality and weighted average SSNAP hospital score. Odds ratios are also presented in Table 7, with percentage odds reported here. From the adjusted results, a 1-unit increase in the ESD score was associated with an odds ratio of 0.71 [95% confidence interval (CI) 0.51 to 0.99] or, in other words, with a reduced odds (by 29%) of the ESD team seeing the patient after ≥ 1 day following hospital discharge. Hence, an increase in ESD consensus score was associated with a more responsive ESD service. Exploring the effect of components, this association appeared to be driven by having more core team members meeting or exceeding the recommended WTE level per 100 stroke patients [a 1-unit increase was significantly associated with a 47% reduction in the odds of the ESD team seeing the patient after ≥ 1 day (95% CI 14% to 67%)]. There was some evidence at borderline significance of an effect of access to other team members (reduced odds of 70%, 95% CI –8% to 92%). Further investigation at an individual item level showed that having access to a social worker was associated with more responsive ESD service, with a 97% reduced odds of the ESD team seeing the patient after ≥ 1 day (95% CI 61% to 99%).

Table 8 presents the linear multilevel model results for the rehabilitation intensity outcome measure, which are unadjusted and adjusted for all patient characteristics, the weighted level of rurality, the average SSNAP hospital score and the total ESD consensus score. Focusing on the adjusted results and coefficients (presented as percentages here), the ESD consensus score was significantly associated with treatment intensity, such that a 1-unit increase in ESD consensus score increased treatment intensity (total number of treatment days/total days with ESD) by 2% (95% CI 0.3% to 4%). With respect to this significant association, holding weekly MDT meetings with the core team attending (see Table 5) and a member of the ESD team attending the acute meetings were all positively associated with increased rehabilitation intensity: specifically an average of 8% (95% CI 0.9% to 16%) improvement in rehabilitation intensity.

Table 9 presents the ordinal logistic multilevel model results for the patient outcome measure, which are unadjusted and adjusted for all patient characteristics, weighted average SSNAP hospital score and level of rurality. There was no significant association between the ESD consensus score and the stroke survivor outcome measured by the modified Rankin Scale score at ESD discharge. Site-level control variables, namely percentage rurality and hospital SSNAP rating score, had no statistically significant relationship with any of the outcomes.

Discussion

This study was designed to inform the large-scale implementation of ESD by evaluating its effectiveness in real-world conditions, at scale, using recommended methodology. 71 This addresses recent recommendations for investment in stroke rehabilitation made in NHS England’s Long Term Plan4 and the lack of large-scale development of ESD worldwide. 15 The study found that a variety of ESD service models have been adopted in regions of interest, as reflected by the variability in the ESD consensus score. The ESD consensus score was significantly associated with a more responsive ESD service (reduced odds of patient being seen after ≥ 1 day) and increased rehabilitation intensity when controlling for patient characteristics and other confounding variables, but no effect on stroke survivor outcome, as measured by the modified Rankin Scale, was demonstrated. We conclude that adopting defined core components of ESD was associated with providing a more responsive and intensive ESD service, suggesting that adherence to evidence-based criteria is likely to result in more effective services in practice. This builds on methods used to investigate the organisation of stroke unit care, bringing a much needed focus on community-based stroke care. 70

There are limitations inherent to observational data that we aimed to address with the study design. Although the study used a large sample of stroke patient data, it must be acknowledged that data from a relatively small sample of ESD services were used in this study; further research would be required to confirm wider transferability, particularly beyond England. A key feature of this study was the development of the ESD consensus score. Although we acknowledge that more in-depth investigation of ESD model features is required to make definitive conclusions, this approach offered a useful way to quantify the adoption of core components for quantitative analytical purposes. 37 It provided a simple means by which to evaluate services based on international consensus and clinical guidelines relating to ESD. 5,6,13 We attempted to control for a number of confounders, however we cannot rule out the possible influence of unobserved variables. Outcomes of interest were reliant on a relatively small SSNAP data set, which was entered by community stroke service staff. Findings are reliant on accurate reporting and the possibility of bias cannot be excluded. It should also be noted that previous studies have suggested that ESD reduces length of hospital stay; investigation of this, using hospital SSNAP data, is reported in Chapter 5. 37

Clinical guidelines emphasise the importance of seamless transfers of care and previous studies have reported the negative impact of delayed or un-coordinated transfers on patients. 23,38 In addition to teams with higher total ESD consensus scores being more likely to see patients sooner, findings highlighted the importance of the ratio of staff to patients. Hence, teams that met (or exceeded) consensus-recommended WTE levels of staff per 100 stroke patients were more likely to be responsive, emphasising the need for ESD services to be appropriately resourced. 13 Previous studies have also highlighted the transfer problems relating to lack of joint working between health care and social care. 23,38,72 This study adds to this debate by highlighting the importance of access to a social worker as part of the ESD team.

Early supported discharge has been recommended as a high-intensity rehabilitation intervention, with guidelines and systematic reviews referring to daily visits or four or five visits per week. 5,6,10 In this study, the intensity of rehabilitation delivery was measured by calculating the percentage of treatment days in relation to the patients’ total time with the ESD service. In addition to the total ESD score, the MDT working component was associated with increased intensity of rehabilitation delivery. This resonates with previous studies emphasising the importance of MDT working in the delivery of stroke care and, in particular, MDT meetings. 73–75

Routine collection of patient outcomes in SSNAP is currently limited to the use of the modified Rankin Scale. Findings could be interpreted such that the model of ESD adopted did not influence patient outcomes, as measured by the modified Rankin Scale; however, caution is required. Robust modified Rankin Scale data were available only at discharge from the ESD service (instead of at a later follow-up stage) and, therefore, it is possible that there was not sufficient time to investigate ESD effects. There was also a lack of variability of this outcome measure in the study, possibly reflecting a focus of ESD services on treatment of mild to moderate stroke survivors. There have also been concerns from teams about the reliability of the use of this score across the stroke care pathway. 76 We suggest that routine collection of additional validated patient outcome measures (e.g. measuring activities of daily living, general health/mood and quality of life) at longer follow-up periods in national stroke audits or registries is required. 10,12,13

Finally, at the site level, the lack of effect of rurality was surprising. It is encouraging that we found examples of evidence-based ESD models in rural regions; however, reported challenges with health-care provision in these settings cannot be overlooked. 77,78 Further investigation of the impact of geographical location on implementation of ESD is required.

Conclusion

Original clinical trials of ESD were conducted across the world and implementation of ESD is recommended in many countries’ stroke guidelines. 5,6,62–65 This study supports the use of an international ESD consensus document as a means to guide the implementation of effective, evidence-based ESD in practice. 13 We suggest that the extension of national stroke registries with the inclusion of community stroke data could offer important opportunities to evaluate stroke service delivery beyond the hospital setting. 16 This could go some way towards addressing the current gaps in the provision of stroke rehabilitation that exist globally, moving towards the goal of ensuring that stroke survivors receive the evidence-based care that they deserve.

Introduction

One of the attractive features of the ESD intervention, as reported in clinical trials, has been the reduction in length of hospital stay, with the most recent Cochrane systematic review suggesting a reduction of 6 days. 10 This reduction has contributed to the reported cost-effectiveness of combined stroke unit care and ESD. 11 Small-scale evaluation of ESD operating in localised areas has suggested smaller, but significant, reductions in length of acute hospital stay, but it remains unknown what impact ESD is having over a larger scale. 79 The average length of hospital stay for all stroke patients who survived fell from 40 days in 2001 to 20 days in 2013, and has remained around this length since then. 80 This trend over time has meant that more stroke unit beds are available for new patients, but it has also increased the demand for patient care in the community post hospital discharge.

The study reported here, similar to that reported in Chapter 4, explored the effectiveness of ESD service models. We investigated if trial-based benefits of ESD are realised in practice, with the focus here on the impact of ESD on the length of hospital stay.

Methods

Study design

We present the results from an observational cohort study (Figure 6), which was conducted as part of an overall mixed-methods study. 37 We determined a priori a sample size of 21,760 patients for a study power of 80% to detect standardised effect sizes of 0.25 for each outcome.

FIGURE 6.

Length of hospital stay study design flow chart. WISE What is the Impact of Stroke Early supported discharge?

In this chapter, we have used two different study designs: (1) a cross-sectional analysis (using SSNAP data from 2015–16) comparing the length of hospital stay between patients who did and patients who did not have ESD on their care pathway, with adjustment for hospital and individual confounding factors; (2) repeated cross-sectional analysis using two sets of SSNAP data (2013–14 and 2015–16) to establish whether or not length of hospital stay changed over time and whether or not any change was attributable to ESD.

Results

Analysis of cross-sectional data from 2015–16

Table 10 presents descriptive statistics of the patient- and admitting hospital-level variables and the median number of days spent as an inpatient disaggregated by patients who received ESD (ESD, n = 8648) or did not receive ESD (non-ESD, n = 22,908) on the care pathway. Compared with patients who did not receive ESD, those who did were younger and more likely to be independent (lower pre-morbid modified Rankin Scale score) before stroke, and were more independent at inpatient discharge (lower post-morbid modified Rankin Scale score). ESD patients were also more likely to have been admitted to a hospital with a higher SSNAP score rating and higher rate of delayed transfers than non-ESD patients.

TABLE 10 - Characteristics of the 2015–16 patient cohort

IQR, interquartile range.

Patient characteristic	ESD (N = 8648)	Non-ESD (N = 22,908)
Age (years), n (%)
< 60	1422 (16.4)	3674 (16.0)
60–69	1692 (19.6)	3932 (17.2)
70–79	2646 (30.6)	6080 (26.5)
80–89	2373 (27.4)	6873 (30.0)
> 89	515 (6.0)	2349 (10.3)
Sex: male, n (%)	4869 (56.3)	11,785 (51.4)
Already inpatient at time of stroke: yes, n (%)	264 (3.1)	995 (4.3)
Congestive heart failure prior to admission: yes, n (%)	348 (4.0)	994 (4.3)
Hypertension prior to admission: yes, n (%)	4743 (54.9)	12,299 (53.7)
Atrial fibrillation prior to admission: yes, n (%)	1370 (15.8)	4016 (17.5)
Diabetes prior to admission: yes, n (%)	1770 (20.5)	4834 (21.1)
Stroke/transient ischaemic attack prior to admission: yes, n (%)	2075 (24.0)	6100 (26.6)
Modified Rankin Scale score before stroke, n (%)
0	5688 (65.8)	12,796 (55.9)
> 0	2960 (34.2)	10,112 (44.1)
NIHSS score on arrival, n (%)
0	1041 (12.0)	3267 (14.3)
1–5	4640 (53.7)	11,252 (49.1)
6–14	2253 (26.1)	5596 (24.4)
15–24	636 (7.4)	2322 (10.1)
> 24	78 (0.9)	471 (2.1)
Type of stroke, n (%)
Infarction	7804 (90.8)	20,772 (91.2)
Primary intracerebral haemorrhage	789 (9.2)	2002 (8.8)
Modified Rankin Scale score at inpatient discharge, n (%)
0	1046 (12.1)	4016 (17.5)
1	2259 (26.1)	5024 (21.9)
2	1928 (22.3)	3420 (14.9)
3	2189 (25.3)	3838 (16.8)
4	1118 (12.9)	4154 (18.1)
5	108 (1.3)	2456 (10.7)
Total length of hospital stay (days), median (IQR)	6.9 (2.9–18.8)	6.0 (2.2–24.8)

Figure 7 shows the amount of variation among the 55 admitting hospitals in relation to their average inpatient length of stay and the variation in the proportion of stroke patients who received ESD on their care pathway. In this study, almost one-quarter of stroke survivors attended admitting hospitals with no ESD on their care pathway and more than three-quarters of the stroke survivors attended admitting hospitals with less than a 40/60 split of ESD/non-ESD being available on the care pathway.

FIGURE 7.

Distribution of (a) mean inpatient length of stay; and (b) the proportion of patients who received ESD across admitting hospitals.

Before adjustment, ESD patients had a longer length of hospital stay than non-ESD patients: median 6.9 days [interquartile range (IQR) 2.9–18.8 days] compared with median 6.0 days (IQR 2.2–24.8 days), respectively. Following multilevel model analysis, after controlling for patient- and hospital-level characteristics, the mean length of stay for patients who received ESD along their care pathway compared with those who did not remained significantly longer by 15.8% (95% CI 12.3% to 19.4%) (Table 11). Similarly, from the sensitivity analysis using propensity score matching, patients who received ESD on their care pathway, on average, stayed in hospital longer than patients who did not receive ESD (when controlling for all other patient characteristics and hospital SSNAP and ASCOF scores) (Table 12). The within admitting hospital propensity score-matching process (n = 10,449) suggested that ESD patients stayed in hospital longer than non-ESD patients by an average of 8.7% (95% CI 4.8% to 12.6%). The between-admitting hospital propensity score matching process (n = 30,791) found that ESD patients stayed in hospital longer by an average of 7.3% (95% CI 3.0% to 11.8%).

TABLE 11 - Association between having received ESD on the care pathway and inpatient length of stay: 2015–16

Adjusted for age, sex, pre-stroke independence, comorbidities, NIHSS score on admission, type of stroke and modified Rankin Scale score at discharge from hospital (patient level); ASCOF rate and hospital SSNAP rating score (admitting hospital level).

Note

Model coefficients are on the natural log-scale; significant results were back-transformed, that is, exponentiated to obtain percentage change in length of hospital stay.

Inpatient length of stay models	ESD model
	Unadjusted	Adjusteda
Patients (n)	30,975	30,791
Admitting hospitals (n)	55	55
Patients per admitting hospital (n)
Minimum	27	26
Mean	563.2	559.8
Maximum	1331	1310
Intraclass correlation coefficient	0.05	0.08
ESD on the care pathway, coefficient (95% CI); p-value
Received ESD	0.08 (0.04 to 0.12); < 0.001	0.15 (0.12 to 0.18); < 0.001

TABLE 12 - Matched analysis between patients who received ESD on the care pathway and similar patients who did not: 2015–16a

This is a sensitivity analysis using propensity score matching to complement the multilevel model results presented in Table 11.

Between means patients in all hospitals were matched even if the hospitals did not offer ESD.

Adjusted for age, sex, pre-stroke independence, comorbidities, NIHSS score on admission, type of stroke and modified Rankin Scale score at discharge from hospital (patient level).

Within means that only hospitals with patients who did and patients who did not receive ESD were matched.

Note

Model coefficients are on the natural log-scale; significant results were back-transformed, that is exponentiated, to obtain percentage difference in length of hospital stay.

Inpatient matching models	Betweenb,c	Withinc,d
Patients (n)	30,791 (ESD 8593; non-ESD 22,198)	10,449 (ESD 4572; non-ESD 5877)
Admitting hospitals (n)	55	14
ESD on the care pathway, coefficient (95% CI); p-value
Received ESD	0.07 (0.03 to 0.11); < 0.001	0.08 (0.05 to 0.12); < 0.001

Analysis of repeated cross-sectional data from 2013 to 2014 and 2015 to 2016

Table 13 presents descriptive statistics of the patient- and admitting hospital-level variables used in the repeated cross-sectional analysis, as well as the median number of days spent as an inpatient disaggregated by patients who received ESD (n = 11,376) or did not receive ESD (n = 39,417) on the care pathway across the two time periods. This descriptive, repeated, cross-sectional analysis showed that the number of patients who received ESD significantly increased over the two time periods. Moreover, across the two time periods, ESD patients were younger, more likely to be male and more likely to have pre-morbid independence, as evidenced by a lower modified Rankin Scale score. Before adjustment, ESD patients were recorded as having a longer length of stay in hospital than non-ESD patients [median 7.7 days (IQR 3.2–23.8 days) vs. median 6.3 days (IQR 2.5–22.1 days) in 2013–14, respectively; median 6.8 days (IQR 2.9–18.5 days) vs. median 5.7 days (IQR 2.1–24.6 days) in 2015–16, respectively].

TABLE 13 - Patient characteristics from the 2013–14 and 2015–16 cohorts

Patient characteristic	ESD (N = 11,376)	Non-ESD (N = 39,417)
Age (years), n (%)
< 60	1847 (16.2)	6212 (15.8)
60–69	2239 (19.7)	6647 (16.9)
70–79	3467 (30.5)	10,543 (26.8)
80–89	3126 (27.5)	12,087 (30.7)
> 89	697 (6.1)	3928 (10)
Sex: male, n (%)	6352 (55.8)	19,959 (50.6)
Already inpatient at time of stroke: yes, n (%)	361 (3.2)	1766 (4.5)
Congestive heart failure prior to admission: yes, n (%)	481 (4.2)	1558 (4)
Hypertension prior to admission: yes, n (%)	6261 (55)	21,313 (54.1)
Atrial fibrillation prior to admission: yes, n (%)	1840 (16.2)	7299 (18.5)
Diabetes prior to admission: yes, n (%)	2345 (20.6)	8074 (20.5)
Stroke/transient ischaemic attack prior to admission: yes, n (%)	2740 (24.1)	10,482 (26.6)
Modified Rankin Scale score before stroke, n (%)
0	7389 (65)	22,847 (58)
> 0	3987 (35.1)	16,570 (42)
NIHSS score on arrival, n (%)
0	1430 (12.6)	6884 (17.5)
1–5	6020 (52.9)	18,802 (47.7)
6–14	2955 (26)	9162 (23.2)
15–24	861 (7.6)	3833 (9.7)
> 24	110 (1)	736 (1.9)
Type of stroke, n (%)
Infarction	10,315 (91.2)	35,989 (92.1)
Primary intracerebral haemorrhage	998 (8.8)	3079 (7.9)
Modified Rankin Scale score at inpatient discharge, n (%)
0	1440 (12.7)	8474 (21.5)
1	2928 (25.7)	8649 (21.9)
2	2562 (22.5)	5456 (13.8)
3	2817 (24.8)	6090 (15.5)
4	1471 (12.9)	6733 (17.1)
5	158 (1.4)	4015 (10.2)
Time period, n (%)
2013–14	3065 (26.9)	21,841 (55.4)
2015–16	8311 (73.1)	17,576 (44.6)
Total length of hospital stay (days), median (IQR)
2013–14	7.7 (3.2–23.8)	6.3 (2.5–22.1)
2015–16	6.8 (2.9–18.5)	5.7 (2.1–24.6)

Figure 8 shows the amount of variation across the two time periods in relation to their average inpatient length of stay, as well as the variation in the proportion of stroke patients who received ESD on their care pathway. Although the mean inpatient length of stay fell only slightly between the two time periods, the prevalence of ESD availability more than doubled across the 41 admitting hospitals in this study.

FIGURE 8.

Distribution of (a) the mean inpatient length of stay; and (b) the proportion of patients who received ESD across time periods.

Table 14 presents the results from the linear multilevel model. Again, we controlled for patient characteristics, SSNAP admitting hospital score and average ASCOF rates for each admitting hospital. There was no significant interaction between receiving ESD and time period (p = 0.79); therefore, the effect of ESD was similar in both time periods, despite more patients accessing ESD over time. Therefore, the main effect of ESD across both time periods was estimated to be an increase in the length of hospital stay by an average of 18.8% (95% CI 13.9% to 24.0%) compared with patients who did not receive ESD on their care pathway. Furthermore, the main effect of the 2015–16 time period compared with the 2013–14 time period was found to significantly reduce the length of hospital stay by an average of 14.0% (95% CI 12.2% to 15.9%).

TABLE 14 - Association between having received ESD on the care pathway and inpatient length of stay: 2013–16

Adjusted for age, sex, pre-stroke independence, comorbidities, NIHSS score on admission, type of stroke and modified Rankin Scale score at discharge from hospital (patient level); and ASCOF rate and hospital SSNAP rating score (admitting hospital level).

Note

Model coefficients are on the natural log-scale; significant results were back-transformed, that is exponentiated to obtain percentage change in length of hospital stay.

Inpatient length of stay models	Unadjusted	Adjusteda
Patients (n)	49,667	49,266
Admitting hospitals (n)	41	41
Patients per admitting hospital (n)
Minimum	115	106
Mean	1211.4	1201.6
Maximum	2512	2446
Intraclass correlation coefficient	0.04	0.08
ESD on the care pathway, coefficient (95% CI); p-value
Received ESD	0.14 (0.09 to 0.20); < 0.001	0.17 (0.13 to 0.22); < 0.001
2015–16 time period	–0.06 (–0.09 to –0.04); < 0.001	–0.15 (–0.17 to –0.13); < 0.001
Received ESD × (2015–16 time period)	–0.05 (–0.11 to 0.02); 0.162	0.01 (–0.04 to 0.06); 0.790

Using the parameter estimates from the adjusted multilevel model results in Table 14, we calculated the predicted average length of hospital stay for patients who did and patients who did not receive ESD in both time periods. We exponentiated the predictions to convert the unit back to days and present the results in Figure 9. This visual representation highlights our main findings from this repeated cross-sectional analysis, which is a reduction in average length of hospital stay for both ESD and non-ESD patients and that the ESD patients stayed in hospital longer than non-ESD patients by about 1 day on average across the two time periods.

FIGURE 9.

Predicted mean length of hospital stay by time period and whether or not patients received ESD.

The degree of clustering was low in both adjusted multilevel models (adjusted intraclass correlation coefficients 0.08 and 0.08, respectively) implying that the total variance that we observed in the length of hospital stay data was more greatly accounted for by the patients than by the admitting hospitals.

Discussion

This study has investigated the effect of ESD on length of hospital stay in real-world settings, with the aim to inform evidence-based improvements in stroke care. Unadjusted findings indicated that stroke patients who received ESD spent approximately 1 day longer in hospital than all other patients; however, those who received ESD differed from those who did not in their characteristics. With adjustment using two different designs and alternative analyses for the differing case mix, the length of hospital stay remained longer for those who received ESD on their care pathway. This contrasts with findings from original ESD randomised controlled trials, in which a reduction in length of hospital stay was reported. 10

This is an important finding, particularly when considering the implementation of ESD in practice. The promise of reducing the length of hospital stay made ESD attractive with regard to cost savings and is likely to have contributed to the widespread implementation in England. 7,11 The economic implications of ESD in the real world have previously been investigated through simulation modelling (informed by SSNAP data similar to data used here). 85 The authors reported an estimated mean length of hospital stay for ESD patients that was 8 days less than that for non-ESD patients and, thereby, derived cost savings associated with modelling the increased use of ESD.

However, an important difference with that economic analysis and the study reported here relates to the fact that we used actual and total inpatient length of stay data combining different inpatient settings where applicable on a per-patient basis. Analysis showed that ESD patients were younger and more likely to have a lower modified Rankin Scale score at inpatient discharge. The simulation modelling findings drew on sample statistics derived from patient distributions for each separate inpatient setting, stratified by age category and modified Rankin Scale score, and their results were driven by assuming that more severe stroke survivors received ESD. What our study adds is that when comparing ESD with non-ESD patient groups matched for important patient characteristics, receiving ESD does not relate to a reduction in length of hospital stay.

Another important consideration is that the average length of hospital stay for stroke survivors overall has reduced dramatically over the last 20 years, from 40 days to 20 days, the latter decade of which our study captures between the two cohorts in this chapter (i.e. 2013–14 and 2015–16). 80 This makes it less likely for an intervention like ESD to have an effect (than the years 1997–2004 when the original UK clinical trials were conducted). 10 That ESD in this study was associated with an average increase in length of stay of 1 day could be interpreted as the need for additional time to facilitate the transfer of care to the ESD team, rather than stroke survivors simply leaving hospital without support (if ESD was not available). Previous studies have also highlighted transfer problems relating to lack of joint working between health care and social care. 23,38,72 Chapter 4 added to this debate by highlighting the importance of access to a social worker as part of the ESD team. 1

What was not possible to investigate in this study was whether or not access to ESD (despite an additional day in hospital) was associated with improved patient outcomes over the longer term. Recent observational studies in Sweden suggest patient and care-giver benefits related to the provision of ESD in regular clinical practice, in line with our previous study findings from England. 86,87 We suggest that the routine collection of additional validated patient outcome measures (e.g. measuring activities of daily living, general health/mood and quality of life) at longer follow-up periods in national stroke audits or registries is required. 10,12,13

Like other observational studies, this study had its limitations. Although the numbers of patients were sufficient for the analysis undertaken, by focusing on specific regions of England the transferability of findings could be questioned, particularly outside England. Other limitations include the potential sources of residual confounding and the reliance on SSNAP data being accurately reported by hospital teams. Length of hospital stay data are inherently variable and, although we tried to include a proxy measure for other sources of delays in discharge, analysis would have benefited from social care-related patient-level variables. 85,86 Patients also experienced a series of hospital stays (e.g. transfer between acute and rehabilitation wards), relying on accurate transfer of SSNAP records between hospitals and resulting in us focusing on total length of hospital stay and ESD provision on a care pathway. 88 What this means is that the impact of ESD on a particular hospital may not be apparent, which may be of more interest to providers of hospital services.

Conclusion

Original clinical trials of ESD were conducted across the world and implementation of ESD is recommended in many countries’ stroke guidelines. 5,6,62–65 This study investigated the impact of ESD in real-world settings and focused on the length of hospital stay. ESD was not associated with a reduction in length of hospital stay as previously reported in clinical trials, although the increased average length of stay was just 1 day. This highlights the importance of investigating whether or not trial-based outcomes of interventions are realised in the real world. What remains to be investigated and reported are the benefits of accessing ESD over the longer term. In addition, by focusing the analysis on patients who would have been eligible but did not receive ESD, we have highlighted an important gap in service provision.

Introduction

Findings from WP1 (see Chapter 4) suggested that a range of ESD models had been adopted. In WP2, our aim was to gain a deeper understanding of the types of model in operation. Using a purposive sampling approach, case study sites from WP1 were selected based on the level to which evidence-based ESD was implemented (contrasting ESD models) and the influence of rurality on ESD effectiveness (urban vs. rural sites). To create service descriptions of each ESD model in operation, we collected documentary evidence from the six purposively selected sites. We also collected data about staffing and adopted models of care from each team leader using a questionnaire (also used in Chapter 7). These data were complemented and validated by qualitative data collected in ESD team interviews (details in Chapter 8).

This chapter provides a brief overview of the six services that participated in our WP2 fieldwork research.

More detailed information about each individual service is provided in Appendix 3. Service descriptions are based on data collected during WP2, information gathered at meetings with services during site visits, documentary material and interviews with staff and commissioners within the data collection process.

Site summary

The sites were labelled A–F, with site A being the most urban through to site F being the most rural. Sites A and B were clearly distinguished as hybrid services providing both an ESD service and a community rehabilitation service. The more rural services were not providing a clearly defined hybrid service but they did offer additional time and service to their patients based on need beyond the level of a strict ESD service.

The individual service descriptions in Appendix 3 highlight variation in the way that services had been developed and configured to meet the needs of the local stroke pathways and geography. Some had been established for a much longer period of time than others, and had gradually adapted over time to meet the changing needs of their population and wider health services.

Models of operation were tailored to each service, with some teams taking referrals widely, including from GPs and by self-referral, whereas other teams adhered rigidly to acute trust referrals only. Some teams in-reached to the acute trusts and others did not. Eligibility criteria also varied depending on the way that services were configured and again were related to the wider stroke pathway, capacity and resources (Table 15) in the team, and the availability of other services to take patients with more severe disability. Patient length of stay varied from the recommended average of 6 weeks to 20 weeks.

There were some commonalities between the teams, including stated responsiveness times driven by national SSNAP targets. Systems, processes and ways of working were configured to ensure that patients were seen within 24 hours and had a therapy assessment within 72 hours of admission to the team, although mechanisms for achieving these targets varied widely. All teams held weekly MDT meetings and routinely collected data for audit and evaluation purposes over and above the nationally mandated SSNAP data set.

Introduction

Cost implications are likely to be an important consideration with regard to successful adoption and implementation of interventions, such as ESD. There is a drive for cost-effective, evidence-based practice that seeks to ensure that limited health-care resources are used to their maximum benefit. 89,90

Some of the more technical forms of economic evaluation (including cost-effectiveness, cost–utility and cost–benefit analyses) have caused concern in relation to the lack of impact on the practice of service development. 91–93 The use of QALYs (used in cost–utility analyses), a utility-based index measure of health benefit, presents particular interpretation problems in a policy-making context. Complex interventions, such as ESD, have multiple health- and non-health-related outcomes, and it is neither feasible nor meaningful to combine the costs and outcomes into a single measure, such as an incremental cost-effectiveness ratio, as is undertaken in cost-effectiveness or cost–utility analysis.

The CCA is advocated, which presents information about costs and outcomes (clinical and other) in the form of a balance sheet. 94 Outcomes or consequences are shown in their natural units (some of which may be monetary), and costs and outcomes are not combined into a single measure. This economic evaluation approach is very versatile and practical. 95 CCA allows the decision-maker to determine whether or not, overall, the treatment or intervention or service is worth carrying out or commissioning. 96 The CCA approach helps to refine economic methods, identifying relevant costs and outcomes, and helps to generate hypotheses for definitive cost-effectiveness studies. It provides a broad and rich source of economic information, which is increasingly needed by NHS decision-makers. 97

Given our overall realist approach and the recognised importance of context, the wider costs and benefits associated with the intervention need to be acknowledged. 34 Moreover, the resources associated with the delivery of a complex intervention, such as ESD, are likely to equate to different costs in different places. 35 Hence, we undertook a CCA in which the costs associated with different models of ESD service provision were considered in the light of the geographical context in which they were operating and the outcomes or benefits that were achieved.

Materials and methods

Results

Table 16 shows the 17-item ESD consensus score for each of the six ESD services and represents the level of adherence to the core components of ESD service delivery. The information presented indicates that no two services are configured in the same way even if they achieved the same consensus score, thereby underlining the variability in models of service delivery. In addition, the two services that provide both ESD and community rehabilitation did not meet or exceed the recommended WTE level of core staff per 100 patients.

Table 17 provides an overview of the six services involved in this study. The services varied in terms of the percentage of rurality of their location, which is in line with our purposive sampling strategy. There were also differences in the patient caseload and number of staff. We also noted some variability in their ESD consensus scores, which indicated different levels of adherence to an evidence-based service model.

The cost–consequences balance sheet is shown in Table 18. The services with the highest overall costs were located in the most rural areas. The largest element of any ESD service cost was pay. As the level of rurality increased, the travel cost on a per-patient basis also increased.

The pairwise linear correlation matrix is presented in Table 19. This analysis confirmed that higher costs per patient (overall, staff and travel costs considered separately) were significantly and positively associated with ESD services in areas of higher rurality, as well as higher ESD consensus scores. The aggregate costs (not on a per-patient basis) were not associated with ESD consensus scores. By contrast, the aggregate overall and staff service-level costs were significantly and positively associated with the number of staff in each service, that is the size of team but not on a per-patient basis. This contrast showcased how per-patient costs relate to the adherence to evidence-based core components of ESD service delivery, whereas aggregate costs relate to the size of team.

Discussion

This chapter has evaluated different stroke ESD service models in different geographical settings according to their operational costs and adherence to core components of service delivery using CCA. We present an analysis of the costs, consequences and their associations using data collected directly from six purposively sampled ESD services. This study has provided evidence of variability in terms of the adopted model, which was measured as the level of adherence to ESD core components using an ESD consensus score. No two services proved to be the same in terms of the ESD consensus score, level of rurality and size of team in terms of the number of staff and patients, as well as the overall, staff and travel costs.

This study has shown that higher ESD service costs per patient were significantly associated with the level of rurality in which the ESD service operated. It is known that health-care service provision in rural areas is more costly than in urban areas for a similar level of service provision. 99,100 Higher costs have been attributed to difficulties in staff recruitment and retention, higher travel costs and non-productive staff time when travelling, as well as difficulties in realising economies of scale while adequately serving sparsely populated areas. 100

In terms of consequences, higher service costs on a per-patient basis were associated with greater adherence to the evidence-based core components, as measured by our ESD consensus score. One important aspect of the ESD consensus score was to determine if a service meets the recommended WTE level of core staff per 100 patients. 1 Therefore, services that were resourced appropriately and had the adequate number of staff per patients were awarded a higher ESD consensus score. In this study, we found that it was the lower WTE amount of core staff per 100 patients within the most urban service, as displayed in Table 16, that contributed to this service having the lowest ESD consensus score. Furthermore, our correlation analysis in Table 19 showed how these staff to patient ratios were detected by the significant positive association between service costs per patient and the ESD consensus score. Therefore, resourcing a service that adheres to the core components of ESD service delivery requires costs to be appreciated on a per-patient basis rather than just in aggregate.

This study does suffer from some limitations. A major limitation is the small sample size. There were 142 ESD services in operation in the UK in 2015, but data from only six services (4%) were used in this study. Therefore, our findings are suggestive and the associations reported would benefit from being tested in a larger sample, including further exploration of different ESD and community stroke rehabilitation models. 98 Another limitation pertains to our use of CCA. This CCA provides a snapshot in time, which suffers from being unable to capture all of the consequences, especially indirect effects arising from ESD service provision, as well as failing to capture consequences that occur at other points in time. This limitation hinders decision-makers, who may still encounter difficulties in drawing unambiguous conclusions for service delivery. 94

The evidence relating to the cost implications of ESD is mixed, with previous studies reporting reductions, increases and no differences compared with treatment as usual. 10,79 Teng et al. 101 postulate that any reductions in hospital length of stay associated with ESD translate into negligible cost savings, as any beds made available are then used by other patients. From this perspective, the authors suggest that ESD can be viewed as an additional cost. In this chapter, we have looked at the cost of providing ESD across six services and presented evidence to show how this cost differs across the level of rurality, size of team and adherence to ESD core components of service delivery.

Conclusion

We conclude that additional resource and the associated costs were required by rural services to meet evidence-based standards (and appropriate staff to patient ratios). Our previous findings would suggest that additional cost implications may be associated with additional patient benefit via the ESD service’s ability to deliver a more responsive and intensive service. 1 We recommend further research to evaluate observed cost differences, across different geographical areas, from a patient outcome perspective.

Introduction

This chapter presents a qualitative interview study with staff members involved in the delivery of stroke ESD services. A realist approach was used to guide both data collection and data analysis, which was informed by three a priori-defined programme theories (see Chapter 3). The results section presents key CMO configurations identified for each of the programme theories. Key messages are discussed in the light of previous literature, including methodological strengths and limitations.

Aims and objectives

The aim of this qualitative study was to obtain a better understanding of the interaction between contextual influences, core intervention components and the reasoning and actions of staff members involved with ESD delivery. Through the process of interview collection and analysis we assessed and refined the three programme theories that we developed at an earlier stage:

• Programme theory 1 – ‘The adoption of evidence-based core components is important for the intervention to be effective in practice’.

• Programme theory 2 – ‘Core evidence-based components of ESD will operate differently in urban versus rural settings’.

• Programme theory 3 – ‘The quality of communication processes between key stakeholders and services in the pathway will influence the implementation of an evidence-based service’.

Data collection

The site selection process is described in Chapter 2. Information was collected through both one-to-one and group interviews with up to 20 NHS staff participants who were associated with each ESD service. Purposive and snowball sampling was employed to recruit a cross-section of key informants for each of the participating services.

Participants had to meet the following eligibility criteria:

• adults (aged ≥ 18 years)

• NHS clinical staff participants involved in the clinical delivery of the ESD service (i.e. consultants, OTs, RAs, physiotherapists, social workers, nursing staff and psychologists)

• NHS non-clinical staff participants involved in managing the delivery of the ESD service (i.e. office managers and administrators)

• stroke service commissioners (i.e. members of a CCG involved in the commissioning of ESD services).

Recruitment to the study started in September 2018 and data collection continued until August 2019. The initial approach to ESD staff members was through their line manager, who gave permission for their staff to be approached to participate in the study. The service leads were approached by the research team (RF, NC and AB) and were invited to participate in the study using an e-mail introduction in which we included the participant information sheet. Researchers also offered to visit the sites and present the study to clinical teams to give them the opportunity to have their questions answered. All participants provided written informed consent and were given a unique ID number to ensure anonymity and confidentiality.

Data analysis

Data were analysed following an iterative, retroductive process moving between deductive and inductive phases. 21 We started our analysis using a top-down approach, the starting point being three key programme theories and related CMO configurations. Staff participants’ narratives and documentary data from each site were examined to identify connections between contexts, mechanisms and outcomes, which were coded into CMO strings. 102 These were then further developed and refined through new insights from participants’ narratives during the early stage of data collection. The three programme theories formed the overarching framework that guided the types of hypotheses that we investigated through the interview data and, therefore, what we coded in participants’ narratives. However, we also looked for examples that both confirmed or disconfirmed the theory or pre-existing CMOs or led to further refinements, allowing for an inductive process to take place. The coding process of the interviews took place in the following stages:

1. Participant-level data.

   At this stage, the units of analysis were individual interviews and focus groups. The same coding and analysis processes were followed for both of these types of interview. Analysis started by reading through each transcript to map out where different topics and theories were being introduced and identify potential connections between CMOs. Service descriptions synthesised from interviews and documentary evidence served to provide background information on the operation of each service. Coding was guided by the overarching framework of the key programme theories. Only text relevant to one of the three areas described by each programme theory was considered for coding. Text extracts were coded as candidate CMO strings under the relevant core programme theory/theories, where the connections between these elements could be discerned from participants’ narratives. CMO configurations were coded as either triads (where all three elements are present) or dyads (e.g. C–M, M–O combinations). 102 At this stage of preliminary coding, it was not always clear whether certain processes operated as contexts or mechanisms. We were conscious that these initial configurations may need to be revisited/refined at a later stage of analysis as the connections between these features became better understood. Where it was not possible to identify dyads or triads but the inclusion of a text extract was considered important, the extract was coded as referring to distinct CMO processes. Information coded this way was retained only if it provided insights towards the development of CMO patterns when combined with information from other interviews within the same site.

2. Site-level data.

   Once the analysis of interviews from each site was completed and insights from both one-to-one and group interviews were brought together to identify key CMOs, cross-site comparisons were conducted to identify confirming and disconfirming cases and explore how pertinent mechanisms interact with site-specific contextual conditions to generate variation in outcomes. The identified CMO configurations were related back to the original programme theories and further refinements were made.

The analysis was conducted by two members of the research team (NC and TC). They initially coded data independently, producing tables of CMO configurations related to each overarching programme theory. They then discussed and compared their analysis and insights with the rest of the research group until agreement was reached that no further revisions and refinements were required.

Results

As shown in Table 20, we spoke to 117 staff members in total, through either one-to-one or focus group interviews. In the following section, we present our CMO configurations and provide supporting interview quotes. Within each programme theory, subsections describe key CMOs organised thematically under relevant headings. To facilitate the reader, descriptions corresponding to context, mechanism and outcome elements are indicated in the text in brackets. The narrative is complemented by figures showing how the interaction of identified contexts and mechanisms leads to intended (in coral boxes) or unintended outcomes (dark purple boxes). The participants are identified in the following quotations by the letter denoting the site that they came from, an ‘S’ for staff and a number showing the order of interview at that site. Focus group quotations are prefixed with an ‘FG’ and include an ‘R’ plus a number to indicate the initial responder order where required.

Programme theory 1: ‘the adoption of evidence-based core components is important for the intervention to be effective in practice’

This programme theory investigated how participants responded to the adoption of evidence-based intervention characteristics that they identified as key to the operation of the service. We remind the readers that, according to research13 and clinical guidelines,5 core components include the provision of a responsive stroke-specialist rehabilitation service to eligible stroke survivors (mild to moderate disability) and at an intensity similar to what they would receive at hospital. The intervention is delivered by a co-ordinated MDT (e.g. physician, physiotherapist, OT and SLT) that has regular weekly MDT meetings and regular meetings with stroke unit hospital staff. Each of the components is described in a separate section through relevant hypotheses (CMOs) about how prominent contextual factors interact with intervention resources and staff’s reasoning to lead to intended and unintended outcomes.

Patient eligibility criteria

These CMOs concerned how teams responded to clinical guideline recommendations for admitting patients with mild to moderate disability. Services had differing approaches in relation to adherence to selection criteria. CMOs described in Adhering to eligibility criteria reflect staff’s responses to the adoption of eligibility criteria regarding disability levels. In Flexibility in applying eligibility criteria, CMOs concern the flexibility demonstrated by some services in the application of selection criteria in response to pressures posed by certain contextual characteristics, such as the availability of community stroke rehabilitation in the local stroke care pathway.

Adhering to eligibility criteria

Most respondents agreed that adhering to selection criteria regarding disability levels (M) helped the services to manage capacity and promote responsiveness (O) (Figure 10). Services struggled to offer the therapeutic intensity that is required to address the needs of more severe patients (C), and attempting to cater for their needs could dilute the intensity of the service for the rest of the patients or reduce the flow of new patients into the service:

We need to be targeted to be able to take new patients – to have capacity to take on new patients. Once upon a time, they would have seen patients for up to a year within this team. They would just be on the books.

BS20

Yes we are very strict with eligibility.

Even though you want to change it and you want to make exceptions, the moment you start doing it the more you’ll do it, you’ve just got to be cruel sometimes and say no to people.

Yeah. And I think not forgetting the fact that it’s supposed to be a virtual ward and it should be an intense service. So being able to provide that intense service by not offering too many slots depending on how many staff you have.

FGCS1

FIGURE 10.

Adhering to eligibility criteria CMOs.

One of the team leads (ES1) explained that focusing on the mild to moderate patients meant that hospital discharge became a less complex process (O); reasons included this patient group not requiring care packages and, therefore, avoiding delays in having to put these systems in place before returning home. Applying selection criteria was also identified as key to timely admissions to ESD (O). One team explained how they changed their model of service to improve responsiveness and further reduce hospital length of stay. Introducing two streams in the service allowed them to distinguish between mild to moderate patients with a straightforward discharge process and patients requiring more complex discharge planning. As a result, they managed to maintain the flow of patients from the hospital and free therapy time:

We honed our model to that under 15 days, over 15 days because prior to that, we all saw – all the patients that came in were referred and ready to go home, but what we found was everybody was getting caught up in very complex discharge planning and what have you, that actually to respond and keep the length of stay down for the more mild to moderate, higher level patients, became quite difficult. [. . .] So to keep this flow going, we developed this, so that we’re able to respond much quicker and that some of the therapists are able to free themselves up and keep themselves more flexible and not have a full caseload of very complex people.

FGBS2

Patients with mild to moderate disability levels were also considered more likely to be safe at home (O). Ensuring the safety of patients who had more severe disability was an important concern. One team questioned whether or not they would be able to provide the right equipment to support these patients at home (C). Relying on relatives to support patients with severe disability was risky (C) and respondents highlighted the need to consider the willingness and readiness of patients’ spouses and families to assume caring responsibilities. One team also noted the potential implications of admitting patients with more severe disability on staff well-being (O). After a trialling period of accepting more complex strokes, staff observed that staff sickness rates increased dramatically. Staff working harder to maintain the quality of the service at the expense of their own health was suggested as a potential explanation:

They get home and it’s actually the carers that really struggle and they just can’t cope, and that’s when accidents happen and they end up back in hospital, or infections and that sort of thing come on quite quickly.

FGES1

The context of an ageing population called for the management of increasingly complex patients with multiple health-care needs (C). Cognition was mentioned as an important comorbidity that could influence outcomes; however, this was mediated by other factors, such as a supportive home environment, social isolation and mood (C). Adding cognition to the eligibility criteria, however, was not considered the way forward because it could lead to the exclusion of patients who could benefit from the service, despite their cognitive problems. Teams also discussed how medical advances, such as the introduction of thrombolysis and thrombectomy, successfully addressed physical impairment but left other potentially hidden issues, including cognition, fatigue and mood, creating a ‘new type’ of patients: the ‘walking wounded’. Given that the services’ eligibility criteria focused on physical function, it was common for teams to admit patients who had retained some motor ability while having very complex needs (M). For instance, it was suggested that thrombolysis patients tended to be of working age and required additional support, such as vocational rehabilitation. The teams felt the pressure to respond to the needs of these patients without, however, always having the resources to do so (O):

At the minute because of thrombolysis and thrombectomy, we’re starting to see some new trends [. . .]. So the thrombectomy, thrombolysis as well, we’re seeing that they’re coming out a bit shocked, you know, that they’re physically fine but there’s a lot of quite unusual cognitive deficits coming out of thrombectomy and so I think in that there’s probably going to be another shift again of the kinds of patients we’re seeing in working age people, as well. From a vocational rehab point of view, we’re not funded to provide vocational rehab but again, we’ll do it. We’ll support people in it and the numbers of working age patients are going up as well.

FGBS1

I feel that some bed managers don’t understand what is a stroke. They think that if you’re medically stable, you can go home. But apart from being medically stable, you should be rehabilitation stable as well. Cognitively, if the patient is not able to function, can’t remember how to make a cup of tea, turn off the faucet, take his or her medication, it’s not going to be safe. We had one patient who was sent home, medically stable but cognitively not. The following week, the patient was cooking at 1 o’clock in the morning and all the medications went all over the floor. We had to phone the GP and send her back to the hospital because she doesn’t have capacity with her safety. If patients will go back to the hospital, it’s a failed discharge.

FS1

Flexibility in applying eligibility criteria

The availability and quality of community rehabilitation services in the catchment area (C) was a key factor influencing the level of flexibility demonstrated by the teams in the use of eligibility criteria (Figure 11). In some sites, patients who did not meet the criteria for ESD could have to wait for weeks or months to be admitted to a community rehabilitation service. The linearity of stroke care pathways also meant that patients who could not handle the intensity of ESD (e.g. owing to infection or fatigue) at the point of discharge could lose the opportunity for specialist rehabilitation (C). In response, some ESD teams felt under pressure to use criteria flexibly and provide short-term support to patients who fell through the net or were deemed to be too complex for ESD (M). ES16 explained that, although their managers knew about these patients, they were not included in the official caseload nor reported on the SSNAP:

We may get somebody that, for example, has come out of a specialist neuro rehab unit who really should be going into community therapy, they won’t; they’ll come to ESD. Or somebody that’s needing less-intensive therapy who under different circumstances would go through community therapy but because they’re going to wait for 18 weeks they won’t, they’ll come through ESD. So we take anybody that’s had a new stroke and has got rehabilitation potential.

DS14

The difficulty at the moment is you might have someone who, in the early stages, it takes a long time to rehab but say in 3 months time they start to make some progress, but unfortunately they’re then out of the stroke pathway that we have, so it’s then very difficult to think, well this patient wasn’t ready at that time but now they may benefit from a period of rehabilitation and whether there’s a pathway that we can bring them back in when maybe rehab is more beneficial for them.

FGES1

FIGURE 11.

Flexibility in applying eligibility criteria CMOs.

Within the context of a fragmented pathway (C), it was not uncommon for hospital staff to upgrade a patient to make them eligible for ESD or delay a patient’s discharge until they met the selection criteria (M). According to ESD staff, these practices led to preventable hospital re-admissions or hindered ESD services’ potential to achieve a demonstrable reduction in hospital length of stay (O):

Some of the patients are, sort of heavier, more kind of on the cusp of whether we would accept them or not, and sometimes they will tell us that they hold patients in, because they think that they can get them to a level where they’re suitable for ESD. Because then they know that they’re getting really good rehab maybe.

CS1&2

As noted by one of the team leads, accepting patients with more severe disability would call for a more integrated stroke care pathway in which health-care and social services co-ordinate to support these patients at home:

If you had the resources could be extended to more severe patients?

Big question. You definitely have to have your care packages to deal with that.

It would require more integrated systems?

I definitely think so. It couldn’t just be picked up by health, so there would have to be very clear integrated working with people that are going to be ring fenced for that, and that’s the difficulty, isn’t it? Why would you take someone who was more dependent home, when you are going to have to rely on many more agencies to make sure that they get out there in a timely manner. You’ve got an ambulance crew that are prepared to take someone into the house because we have huge issues with transferring patients from the ambulance, to get them in to their homes again. So, that’s not just a decision that can be made by an ESD team providing it. It’s also about whether ambulance would support it, whether social services would support that, whether the carer’s want it.

ES1

In response to acute trust pressures, three teams saw higher-intensity patients in care homes to teach the staff positioning, mouth care and communication. This was considered important because care home staff had a high turnover and needed this support to effectively handle these patients. Although the service’s input would be short term, there was an impact on caseload and resources because double visits could be required. The aim was to prevent re-admissions and later complications rather than achieve improvements on disability:

However, if we’ve got somebody that’s going into a nursing home or a care home, particularly if they’re really quite impaired so if they’re bed bound or if they’ve got a new communication impairment, we will put two people into the care home for a short period of time just to make sure that the care home know how to move somebody, and also so the family are aware, because there’s such a high turnover in the care home staff.

DS14

Admitting more complex patients also related to how flexible the services could be around extending patients’ length of stay to spread treatments over a longer period of time (C). Most services did not have the capacity to significantly extend patients’ length of stay and called for a combined service that would provide both ESD and community stroke team components, tailored around patients’ needs. One service (site B) operated a case management system that involved bursts of treatment followed by breaks and scheduled reviews. Patients would be seen for a long time but not with the same intensity, which allowed services to maximise capacity while considering the medium- and long-term patient needs:

I think it’s a huge issue, I think the fact that we have an ESD team and then we have a CST [community stroke team] is a really, really big issue. Because I think the fact that you don’t just get your 6 weeks and then you are discharged is just so important because we know everybody won’t rehabilitate in 6 weeks, no matter how much you try and how skilled your team is. So for me, to think about the kind of long, you know, the medium-term and long-term support is crucial for people because the problems that they encounter post stroke carry on very often.

BS19

Responsiveness

These CMOs concerned how teams responded to clinical guideline recommendations for provision of responsive rehabilitation, with treatment beginning within 24 hours of hospital discharge (Figure 12). Teams had different approaches to managing the tension between capacity issues (C) and meeting responsiveness targets. In some sites, patients could be held in the acute trust until a slot became available (M), which unnecessarily prolonged their hospital length of stay (O). When caseloads were full (C), some teams refused admissions completely (M) and patients were discharged to the care of a non-specialised rehabilitation service, for which waits could be long. Where other services did not exist, patients did not receive any rehabilitation at all (O):

The difficulty for us is when we have far too many people coming out and it’s not possible to get around to see everyone within 24 hours [. . .] I think from a point of view of trying to get the message through to commissioning, the business plan that we put forward was focusing on the impact of having a higher number of patients, meant that the length of stay was increased. So, if you really want this snappy service, where you get people through quickly, you’re not going to achieve that. If you’ve got far too many patients coming out there, you can’t get around to seeing them.

ES1

We couldn’t accept that patient, because that day we didn’t have slots available and that patient couldn’t stay in the hospital for any longer because they’d allocated a package of care for her. So that patient didn’t receive ESD when she really, really did need it.

FGCS1

FIGURE 12.

The CMOs on ESD service responsiveness.

When patients were deemed safe at home (C), patients’ timely discharge from the hospital would be prioritised even if they had to wait at home for an ESD slot to become available (M). This approach resulted in interruptions to patients’ rehabilitation at a critical point for their recovery time (O). In these cases, the focus of the team shifted towards managing patients’ expectations while they were waiting to be seen. Referrers did not always relay this information to patients and in-reaching to the hospital allowed ESD staff to ensure that patients knew what to expect. To minimise the break in rehabilitation provision, two ESD teams would ask their respective acute trust to send therapy plans for the assistants to start work before a formal therapeutic assessment could be completed in the community:

So at the point of referring, we’re saying how long, roughly, it’s going to take us to get in. So we say, is the patient safe to wait for the week and then if that is the case, generally we’re in there quicker, but it sets their expectation that the referrer is ensuring that they’re safe and that we’re not promising to be in quicker than we possibly can be.

FGDS1

The urgency to see all patients within 24 hours was questioned by some respondents (M), who felt that stroke survivors and their families could use more time at home to settle and identify their rehabilitation goals (O). In the context of capacity issues (C), some teams suggested that, in line with their overarching patient-centred rehabilitation approach, patients’ individual circumstances should dictate how quickly they were seen:

I think you’ve got those people on one end of the scale who haven’t processed that they’ve had the stroke really, don’t feel that they’ve got a lot of deficits. And when you go out very quick to say ‘I don’t need anything’, my feeling is if we were to leave them, they know we’re there, but leave them for a week or two to process some of that or just be in the background, I think they might identify things that are more of a problem. [. . .]. Does that person need to be seen in 24 hours over that person who’s got a therapy visit and should we cancel it? So we’ll make contact within that time. We’re working a bit more with the hospitals to say ‘can you let us know whether they’re that urgent really?’. In theory, based on the criteria, nobody should be coming out needing us to go that rapidly.

FS12

Quite often they may have been discharged the night before so we’re seeing them on the morning. All they’ve done is go home, probably not even unpacked and they’ve gone to bed and then they’ve woken up and they’re seeing a therapist. Some of the feedback I’ve had personally is that ‘Well, I haven’t had a chance to try anything so I can’t really say how I’m managing because I only came out of hospital last night’.

AS16

Adopting a flexible approach in how and who makes the first contact with patients (M) was used strategically to improve responsiveness (O). Telephone contact was used by some teams at weekends to complete initial holistic assessments, enabling them to extend the service to 7 days. One service trained experienced RAs to complete initial holistic assessments, which helped the service to work across a large rural geographical area and meet their responsiveness targets:

You need the therapy in there as quickly as possible but at the same time we need to make sure that it’s a safe discharge. So I would say there’s benefits to all the staff doing it, but some of the band 3’s and 4’s, their initial assessments are far superior really, making sure they’ve covered everything. I would get distracted by the OT part of it and lose your focus slightly. So, yeah, they would see them at the weekends as well so if we know someone’s coming on a Friday, that’s booked in for Saturday.

FS12

Rehabilitation intensity

The NICE clinical guideline6 suggests that an ESD team should provide rehabilitation and care at the same intensity as they would if the person were to remain on a stroke unit, that is 45 minutes of each appropriate therapy every day (Figure 13). Against a background of staff shortages and reduced resources, teams were struggling to strike a balance between responding to an expanding list of referrals and offering the intensity required to achieve improvements in patient outcomes (C). For instance, staff at site D reported that they could offer the recommended intensity to a maximum of 19 patients, but there were 30 patients in their caseload at the time of the interviews:

It would be lovely to be able to offer the intensity that each patient needs and deserves. I think at the moment we offer the best service that we possibly can, and I think patients really acknowledge and are grateful for that [. . .]. We either follow the guidelines and see a few patients or we see more patients but we’re really struggling to get the outcomes. So where is the balance?

(DS7)

FIGURE 13.

The CMOs on rehabilitation intensity.

In this context, guideline recommendations on intensity were treated with scepticism. Most respondents thought that they were logistically unrealistic or even inappropriate for their patient group and highlighted the need to provide rehabilitation intensity tailored to people’s needs (M). It was suggested that patients received a bespoke service while services made an efficient use of resources as a result (O). According to respondents, patient feedback often substantiated this view, with patients requesting the time and space to ‘feel at home’ and resume some of their activities:

And we do all the other referrals, like psychology, or if they need extra equipment, that can all get started and you can support them because they may be thinking about their finances and, you know, returning to work and feeling very depressed, and actually you couldn’t then really get started straight away and say ‘I know you’re depressed and you’re thinking about your money and there’s loads of things wrong with the house and the family, but we’re going to have to get stuck in because we’ve got this limited period’.

FGAS1

Some of the feedback we had from people was that sometimes it got a bit too much because they might have reablement coming in three or four times a day to support them with their care, and then we were going in three times a day and it was like, ‘Oh, I’m at home’. [. . .] They were at home. They wanted to be pottering and doing things that they wanted to be doing.

DS14

Gaps in the local stroke care pathway and slow response from other services (C), such as GPs, resulted in staff having to work with many different hats on to try and cater for patients’ needs (M). Addressing unmet information needs was a common issue, as was ensuring that patients received the right medication, which could take hours of work for staff to clarify. Although rehabilitation intensity was diluted (O), staff saw it as part of their role to support their patients with reaching out to other services, especially when their physical and cognitive issues meant that they relied on ESD services for help:

The other end of therapy is sorting out things with them, like an emergency, if something happened to them. We don’t just stay there 45 minutes. Sometimes we stay there for 1 hour and 30 minutes because of other issues. Sometimes a patient’s partner or wife has had an emotional breakdown, and we talk to them and phone the Stroke Association [London, UK] to provide carer support. Those things that we’re dealing, it’s not just the therapy. It’s about the whole, because our approach is holistic.

FS1

Discharge letters themselves can be quite confusing, so I had to sit and study a discharge letter the other week that said a certain drug had to be stopped. I’ve phoned up the consultants in the past just to get clarification so that the patient knows what they’re doing.

FGES2

The impact of care arrangements on rehabilitation provision was stressed by respondents (C). They explained that if the care package was not working well and the amount of support that the family received was inadequate, then the priority of the therapy could become addressing personal care needs rather than providing active rehabilitation. Teams were required to liaise and co-ordinate with multiple services to provide emergency cover for patients until a care package was put in place. They were often put under pressure to problem solve beyond their services’ remit (M), moving the focus away from hands-on therapy and diluting the intensity of rehabilitation (O):

If a patient is just staying in hospital, got a bed, taking up a bed and they can go home but we’re waiting for care, we can sometimes offer the morning. We’ll offer like a morning call.

So instead of waiting for the social care to sort things out you can . . .

We can be that stop-gap.

FGFS1

All of the services sought strategies to enhance the intensity of rehabilitation practice without further stretching their capacity (C). To achieve this, they offered relevant training to family members and care providers, implemented telerehabilitation and promoted a self-management ethos (M). They envisaged to promote rehabilitation practice between sessions and empower patients and carers to effectively manage their recovery after discharge from the service (O):

We have a whole load of iPads [Apple Inc., Cupertino, CA, USA] that we can leave there and they can work, they love doing that. So that gives them the opportunity to practise between sessions.

DS7

We also recruit some of the family members as well.

They might help with therapy, so just similar to how a rehab assistant would do if there’s a relative that can help.

Do you show them how to do things?

Yes, because sometimes they don’t want somebody else coming in. They say they’ll just do it themselves if we show them.

FGBS1

In one site, staff were discouraged from seeing patients daily for 6 weeks; it was thought that this intensity could create psychological dependency and result in people feeling abandoned when discharged. In most sites, visits were the most intense initially and then tailed off, with patients being directed towards self-care in preparation for discharge:

I think the other reason why it’s highly performing is it was set up so that people could have frequent, quite intense input, and that’s something we’ve moved away from a bit. So, people aren’t seen every day like they used to be, but we’ve moved away from it very consciously – we’ve not done it because we were forced to, because we can’t fill vacancies, or the demands are too high. We’ve done it because it’s the right thing to do for the patients and some of the outcomes are better if you give people that self-efficacy to do things themselves. In other services, that self-care element has been forced upon them rather than choosing it. So, I think that’s helped us to be a highly performing service.

FS4

Capacity issues often meant that services relied on staff ‘walking the extra mile’ and maintaining momentum despite the difficulties (M) to meet intensity targets. To quote one of the respondents, teams strived to ‘offer a champagne service on beer money’ (FGCS2). Although their hard work helped them to meet patients’ needs, it also meant that services could become victims of their own success because senior managers and commissioners could maintain an unrealistic picture of the service’s capacity. As a result, resources remained unchanged and both rehabilitation intensity and staff’s well-being were compromised (O). One team had started documenting unmet visits, hoping to collect the evidence to make a case about the need for more staff. Although it was frustrating to accept that they did not meet their intensity targets, they also felt that it was necessary to acknowledge that working overtime to keep up with demand was not sustainable (site E):

I love my laptop, it comes home with me. I can do referrals, but I do it when the children are in bed. [. . .]. So time at work becomes so precious and so little because it’s consumed, so all those other tasks, we probably give a lot of hours in our own time without realising it – which is fine. We work for the NHS so I think it comes with the package.

AS17

We’ve been over our capacity for a while now and to the commissioners, I can imagine they’re sitting there and going, ‘Yeah, but you’re managing’. [. . .]. I can spread out a patient to have one visit every 3 weeks and they only get visited twice but it still looks good on paper that we’ve got that many people going through the service. But then – you know, ‘Who’s [patient name]? I’ve only seen her once’.

And the thing is once they’ve been discharged with us because they’ve got their timeline– they’re having to put more money into community for them to pick up what we couldn’t achieve, so at the end of the day it’s costing them more money because they’re having to get another service involved.

Or re-admissions.

FGDS2

Team composition

Respondents endorsed recommendations by clinical guidelines and the ESD consensus that ESD is delivered by a fully resourced and well-co-ordinated MDT. Particular emphasis was placed on the role of nurses, social workers and RAs, and relevant CMOs are presented in the following sections. Configurations relevant to other disciplines are presented and discussed in Appendix 4.

Access to nurses

With ESD services being the first point of call post hospital, staff felt tasked with ensuring that patients were doing well medically (C) (Figure 14). Having a stroke-specialist nurse in the team (M) allowed them to respond in a timely fashion to patients’ medication queries and address information gaps around medical issues (O). As noted by respondents, nurses were in a better position to identify when liaison with a stroke consultant or the GP was required:

We sometimes get patients that are really struggling with administering diabetes med[ication]s, insulin and things, and we’ve had to involve the nurses in that. I had a patient last week that had got a really high pulse rate [. . .] so it was a conversation with the stroke nurse, who then spoke to the stroke consultant and had a conversation with the GP to tweak some of the medication to try and bring that down. So there was a bit of MDT work with that.

CS3&4

FIGURE 14.

Access to nurses CMOs.

In one site, nurses had assumed the role of the ‘gatekeeper’ of the service by conducting the in-reaching and initial screening of the patients (C). Their thorough medical assessment (M) helped ensure that patients who fit the criteria were admitted to the service as opposed to patients with increased risk of re-admission (O). Respondents also noted that registered nurses had a more holistic approach (M) with their initial assessment of patients, touching effectively on sensitive issues, such as mood and incontinence:

I think that’s part of the in-reaching role is to make sure that they are medically OK. So possibly one of the strengths is two nurses who do the in-reaching; we might be a little bit more forensic in terms of the medical side than maybe our therapy colleagues. So we’ll be looking at blood pressure, continence, medication, those sorts of things maybe and whether they’ve had all the correct investigations and sort of maybe a little bit more detail than maybe the therapists might be looking – have a different angle on it.

ES16

In addition to providing advice and reassurance to the MDT around medical issues (O), nurses’ presence in the team was perceived as a factor contributing to the development of an interdisciplinary skillset across the team, facilitating in this way the provision of holistic care and improving the efficiency of activities (O). In one site, participants described how the lack of a nurse in their team meant that they had to spend a lot of their time trying to secure nursing input from external services, such as district nurses, and the support that they would receive would not always be stroke specific.

A frequently raised issue was that nurses’ contribution was not captured by SSNAP, leading to a big loss of information on ESD input:

The nurses spend an awful lot of time with our patients and none of their data is captured on SSNAP. I know how hard the team works and it’s just so, so frustrating when you see – because we get data sheets out from our SystmOne and our information team whenever the patient is discharged, and you see 2000, 3000 minutes and you think – and that’s effectively worthless because we can’t record it anywhere.

ES17

Access to rehabilitation assistants

Respondents agreed that under the supervision of specialist staff, RAs had a key role in freeing time for therapy staff and maximising the service capacity and cost-effectiveness (O) (Figure 15). As noted by two teams, the success of this process depended on effective delegation (M). This related to registered staff’s experience and ability to effectively manage increased workload. Less-experienced staff could be risk averse or lacked the confidence to judge the right time to hand over therapy tasks to RAs (C); as a result, patients’ length of stay in ESD could increase:

The level of caution, or that ability to let a person do something takes longer, so the length of stay takes longer. That person who isn’t as experienced and isn’t on the higher banding, tends to – you can see that their length of stay with the patient is longer and they need more support. It’s fine, you can pass that to the rehab assistant now and you can take on a new patient.

BS20

FIGURE 15.

The CMOs concerning access to RAs.

Therapy input was shared between registered staff and RAs based on patient need (M). One team lead stressed the need to empower experienced RAs to handle difficult situations, escalating when they judged necessary, and encourage them to bring in new ideas that would benefit their patients (M). Face-to-face daily communication with registered therapists at their office base (C) was thought to allow RAs to address any concerns and receive guidance and reassurance (M):

To be honest, if patients really needed a physiotherapist every single day for the whole of the 6 weeks, that would be hard to do. But actually, we’ve upskilled our assistants to a level now where once we’ve got our plan and we’re happy with that plan and to deploy them, and they can go. It depends on their need.

FS2

Research assistants’ daily contact with patients could ensure that simple care activities, such as washing and dressing, were carried out in a rehabilitative way as part of their therapy plan (O). Having generic RAs with an interdisciplinary skillset in the team was thought to further promote a holistic and interdisciplinary rehabilitation ethos:

I think having generic assistants helps because they are, they do cover all disciplines. So they are by their nature kind of able to go in and do different things and not be very set on just having one side of things.

ES3

Access to a social worker

Delays in social care arrangements could prolong hospital length of stay or lead to unsafe discharges (C) (Figure 16). ESD services admitting complex and more dependent stroke cases were the ones mainly affected. The problem was accentuated when effective communication channels between the ESD services and the referring teams were not in place and, as a result, ESD teams were not informed about the delays in a timely way (C). Slots were unnecessarily reserved, limiting the capacity of the ESD service for new patients. Having a social worker embedded in the pathway and liaising with both the acute and the ESD services (M) streamlined discharge, as it permitted the allocation of targeted admission slots, ensuring that the patient was seen soon after their return home (O). One team contrasted between referring wards with and referring wards without social workers:

Ward [X] have their own social worker, so they’re much more precise with what date their patients are discharged from the hospital, and that makes it much easier for them to refer a week early, and line it up with one of our dischargers and the process is really smooth.

And [Hospital X] don’t have that, so we ask them to give us 3 days’ notice but very often they get a last minute notice, the care package has changed, or it’s suddenly arrived, and patients will go out without them ringing us and it makes it a little bit more difficult. [. . .]. The main problem is the late notice discharges. If we get a lot of them, it makes it really challenging for us.

CS1 &2

FIGURE 16.

The CMOs concerning access to a social worker.

Following discharge home, ESD services could be called to address issues arising from delays in social care input to prevent hospital re-admissions. However, as noted by a social worker, it would be difficult for them to provide a quick and effective response to premature and unsafe discharges if they were not already familiar with the cases. They highlighted the importance of the timely involvement of a dedicated social worker who would follow patients and their families in their journey from the stroke unit into the community (M). Having access to the same electronic note systems (i.e. SystmOne) was emphasised (C), especially in the context of the peripatetic nature of the ESD team. Being able to draw the right information would allow the social worker to identify areas of concern, develop a personalised and detailed care plan and observe patients’ progress (O):

We went to visit and she told us that morning and then when we saw her walking, we realised she wasn’t going to be able manage, so we arranged for care but then care wasn’t available that night, so we did it between the team this morning. We’ve prevented hospital admissions, we’ve prevented them having to into respite and things like that, just by doing that one extra thing. By me doing that lady’s tea when I put her to bed, and then the band 7 went and got her up It was just, we all had to one little thing out of our normal day, but it stopped that lady going into hospital.

FGES1

I said this could have been an unsafe discharge and then people expect me to jump into things like that and I would be very reluctant to do that, because I do not know if it’s a mental health issue or what issue, if the person is not known to us.

CMS15

Seven-day service

These CMOs concern participants’ responses to recommendations about the provision of a 7-day service (Figure 17). Most teams did not provide a 7-day service, mainly because of capacity issues (C). Concerns were voiced that 7-day working reduced therapists’ time during the week, which, in turn, impeded communication with the MDT or other services and could negatively affect responsiveness. Where the service was provided, this was done either with reduced therapist input or with non-registered staff doing the home visits or providing telephone cover (C). Running the service without the presence of experienced therapists, however, meant that non-registered staff could be put under pressure to handle complex cases without the required support and expert input (M). Regarding the benefits of covering weekends, opinions were divided. Some respondents considered 7-day working as instrumental to the provision of responsive ESD (O), especially for patients discharged on a Friday or over the weekend:

I think it’s a good idea to have a 7-day service because what you’re doing is that early supported discharge and that can be any day of the week. If you weren’t functioning on a Saturday or Sunday, I think you would hold up discharges. But that’s a very basic – it’s absolutely got to happen. I think there will be more therapists that work on Saturdays and Sundays, but we need to have more staff to do it. And more money in the budget to pay them weekend rate.

FS2

Originally, when we were first up we did cover weekends as well on a rota, but we found that it wasn’t a very useful thing of our time, really.

Why was that?

I think it’s hard to, as a team, communicate. If you needed to speak to someone, sometimes in the week they weren’t there and then you’d have to wait for them, so it was kind of delaying things.

FGES1

FIGURE 17.

The CMOs on provision of weekly service > 5 days.

Weekend visits were discussed as an opportunity to meet family members and capture important information regarding patients’ adjustment at home. Seven-day working allowed for flexibility on when to see patients, however, it was also acknowledged that not everyone wished to be seen at weekends, especially the high-functioning patients. Respondents referred to patients’ feedback, which suggested that some of them found daily home visits by health-care professionals overwhelming, preferring some downtime and quality, uninterrupted time with family and friends (C). The need to consider and respect these requests was stressed (M) with the view to providing a patient-centred service, tailored to individual circumstances (O):

Sometimes the patients find it quite invasive having somebody from our team come every day, and then even sometimes two people will turn up at the same time, and then if they’ve got carers coming in as well, which they might have for a short amount of time, it can be really overwhelming, you know, this constant stream of people coming through your house. So I think sometimes they like a bit of a day off.

CS1&2

At the end of the day there’s very few patients that can tolerate seven visits per week from us with a therapy plan of an hour and a half. [. . .]. They often ask actually and say, ‘No, please don’t come’.

FS2

Multidisciplinary team working

Clinical guidelines and the ESD consensus have identified the delivery of ESD by a co-ordinated MDT that meet regularly and offer training opportunities as key to the delivery of an effective evidence-based service. Participants’ accounts further highlighted the importance of good multidisciplinary teamwork as a core component of a successful ESD service. They also identified processes and mechanisms through which effective multidisciplinary teamwork was achieved, which mainly included (1) formal and informal communication processes, such MDT meetings; (2) interdisciplinary working; and (3) collegiality and good leadership. CMOs concerning each of these mechanisms are described in separate sections below.

Team meetings

Regular MDT meetings with good attendance were thought to remediate breakdowns in team communication caused by increased workload and travelling (Figure 18). Despite the challenges of getting everyone physically together in times of high pressure (C), all teams agreed on the importance of safeguarding these meetings (M) as essential to achieving clear communication between staff members and ensuring that staff were well informed about patients’ progress (O). MDTs were forums for discussing goal-setting, sharing expertise and learning from each other. Complex cases could be approached collaboratively, drawing on the pool of experience in the team (M). Ensuring that every discipline was part of the decision-making process was seen as a key function of the MDT (M), serving a holistic approach to rehabilitation. Team leads viewed MDT meetings as a time investment that was necessary for their staff to ‘gel’ as a team, promote trusting relationships and avoid silo working (O). Staff members from all disciplines and bandings should feel empowered to actively participate and contribute ideas to these meetings:

That’s the power of the MDT, it’s not a single person seeing one person. It is a shared responsibility, it’s a shared kind of understanding across the team.

FGDS1

It’s the one time in the week when everybody should come together to discuss the patient’s needs and their plans. So primarily it’s to ensure that everybody is clear about what’s happening with that patient and if there are going to be changes to their stay with the service, perhaps we are going to bring their discharge date forward if they are progressing well and they are achieving their goals, then that is discussed with everybody involved, it’s not just one of us making that decision because most patients have more than one discipline going in to see them.

CS5

FIGURE 18.

The CMOs concerning team meetings.

In addition to formal meetings, most teams emphasised the value of opportunistic face-to-face interactions (M) not only for maintaining communication channels but also for promoting staff well-being (O). In one site, team leads talked about how they would have ad hoc face-to-face conversations with their therapists about treatment approaches and clinical reasoning. The nature of community working required staff to handle difficult and stressful situations in people’s homes (C). One team described how having an office base meant that they could ring in and the whole team in the office would get involved in problem-solving. Meeting colleagues at the end of the day helped staff combat feelings of isolation and loneliness that were related to being on the road most of the day, and presented opportunities for mutual support (O). It was also suggested that physical proximity helped socialising across bandings and professions, promoting an interdisciplinary ethos (O). According to a service manager, the lack of frequent face-to-face interactions in their site did not allow them to fully appreciate staff’s workload and slowed down their reflexes in relation to staff who struggled or underperformed:

I’m much more reliant on other members of the team coming to me and saying, ‘Something is not quite right with that member of staff’. Whether that’s around performance or well-being.

DS14

And because the office is so small we’re all sitting together, we’re always having lunch all together. There’s no allocated physio area or allocated OT area, or speech area.

FGCS1

You could work remotely. But a lot of what our handover is, is verbal. Observational, subjective, the handover that you give. Also, some of the therapeutic things that you’re doing, and you actually demonstrate what you – how you want that exercise performing. Or they want to come and show you what they’ve observed; you know, the posture that they’re sitting in or whatever. So, I don’t think that once a week would work. I think they need to be within that team environment more often than that. Not necessarily daily but maybe twice a week, three times a week, just being in that office. It’s important.

FS2

Interdisciplinary working

Respondents agreed that effective multidisciplinary working was enhanced by the blurring of professional divides and an interdisciplinary approach to the rehabilitation provision (M) (Figure 19). The challenges of community working (C) called for cross-boundary working to provide intensive input in an efficient way (O). Holistic goal-setting meant that staff needed to work closely together to help patients to consider and practice different aspects of activities (C). In the context of a time-limited service (C), quick adjustments in the therapy plan could be necessary to better meet patients’ needs (O), a process facilitated by the development of an interdisciplinary skillset across the team (M):

You play every role. You play an occupational therapist, you play a speech and language therapist, you play physio, don’t you, you don’t just go in and do one thing that is physio. You talk about how they’re feeling, what appetite, sleep is like, how they’re managing their everyday tasks, how their speech is. You cover all bases.

FGCS1

That’s where the MDT works, it’s just really natural and really in sync. The OT’s in this service they’ll have OK physio skills, and the physios will have OK OT skills. I think part of that as well is the therapy assistants are all generic. So, you’ve got the qualified therapists and the unregistered staff, and they’ll do all three elements of the treatment, whereas in some other services, you have OT assistants and physio assistants. So, you’re siloing your workforce the whole time.

FS4

FIGURE 19.

The CMOs concerning interdisciplinary working.

Training was used to nurture interdisciplinary working (M) because it helped staff understand each other’s work and feel comfortable with providing aspects of different roles (O). Some respondents noted that, in addition to the development of awareness of other disciplines, the development of an interdisciplinary skillset promoted trust and respect for all professions in the team (O). One team lead saw it as part of their role to nurture this climate; they thought that staff could only treat their patients with respect if they felt respected within their team:

I think if you’re in somebody’s house and there’s a medication issue or a mobility issue, as an occupational therapist, you can’t, in somebody’s house, say that’s not my job. I’ll go back and I’ll – you just deal with what you’ve got to deal with in that person’s house on the day.

You mentioned earlier that you do change hats a bit, so is it something – does it come through training or does it come through experience of learning to work with each other?

I think both, really. We get a level of training. You know, from a blood pressure point of view, the team lead will do a level of training with every new member of staff, but again because we talk a lot in the MDTs, there’s a lot of learning done in MDTs as well, from that point of view and we deal a lot with lifestyle, anxiety and fatigue, which really are something that we all deal with. It’s not profession specific.

FGBS1

One team introduced the term ‘nerapists’ as a reference to their nurses’ and therapists’ shared skillset and approach to rehabilitation. Despite this, they emphasised the need to strike a balance between specialism and interdisciplinarity (M) to avoid duplication and ensure that people are comfortable and confident with what they do without feeling ‘threatened’ by other disciplines (O):

A ‘nerapist’ is a nurse therapist. So they come and are like, ‘Oh my God, I can’t test urine, I can’t do a BP’. So everybody has to have the basics, nobody should freak out. A nurse shouldn’t freak out at a stick. An OT shouldn’t freak out about a blood pressure. [. . .] So the physios we’ve got to trust each other and we’ve got to respect that we can do those basics. The bigger the team has got and the busier we’ve got. I see we’re separating off into our houses again, which we need to watch. But it’s hard because people feel secure in their professional thing. [. . .]. Sometimes I look through a system and I’m like, ‘Well the OT has been out and talked about the bungalow, the physio’s been and talked about the bungalow, the nurse has talked about the bungalow. Stop with the bungalow, you know it’s a bungalow – just do your specialist bit’.

NRS4

Leadership

The importance of leadership in driving effective MDT working was emphasised in the interviews (Figure 20). Reflecting on the qualities of a good leader, respondents noted the importance of acknowledging the link between staff well-being and staff performance. This was particularly in the context of the impact that travelling and handling stressful situations in the community could have on staff well-being and job satisfaction (C). For one of the team leads, good teamwork was synonymous to ensuring that staff were happy and proactive:

It sounds corny but if you come in and your team look happy and they’re proactive and they’re getting out and doing what they need to do, then that’s good for me.

BS20

FIGURE 20.

The CMOs concerning leadership.

Supporting their staff in learning and upskilling (M) was thought to contribute to achieving high job satisfaction and maintaining the team’s momentum (O). Clinical leads and managers were praised by their teams for creating a safe environment for staff to reflect on and acknowledge their limitations, while providing the opportunities to address knowledge gaps through further training. Despite capacity issues (C), time for supervision and training was protected in most teams (M); it was a way to promote staff retention and well-being and cascade knowledge of evidence-based standards through the team (O):

How can we deliver quality, effective, treatment approach if half of my team don’t have the training? So that’s what I’ve always been asking my manager, and that’s the agreement. Training wise, my manager is very supportive with that. The trust is very supportive because we said we can’t see patients effectively if we don’t have knowledge and skills.

FS1

I think staff support must work because the retention here is very good [. . .]. The training aspect makes people feel that they are constantly learning and being able to manage things. And the supervision aspect of people being able to explore what they are doing with people, where the blocks are with people and how they are dealing with things themselves as well keeps the system quite efficient really and effective.

BS19

In addition to encouraging reflexivity and striving to develop and grow, good leadership was linked to enabling staff members to engage in the running and development of the team (M). It was suggested that people perform better when they have ownership over aspects of the team function (O). Part of the leaders’ role was to keep the team informed about developments, share their vision and invite ideas. When introducing changes, some team leads stressed the importance of securing the team’s buy in and support to take things forward. Successful improvements happened collectively, with staff working in partnership to develop and implement a plan of action:

What would you like to see from good leadership?

Well, it’s asking you for an idea or suggestion and then taking that on board if you suggest something and then kind of running with it or showing their appreciation if you suggested something. It’s not kind of dismissing it. You know, I’m valued as a band 3 as much as everybody else, no matter what band in the team. Our managers are fantastic.

FGCS2

As discussed in previous sections, capacity issues meant that staff were often required to work over time to provide the best care for their patients (C). Feeling respected and appreciated (O) by their colleagues and managers was highly valued by respondents. Some leaders took active steps to develop a non-hierarchical ‘culture of kindness’ and collegiality by adopting an open-door policy and encouraging simple day-to-day acts of appreciation and reciprocity (M). The aim was to bring out the best in each staff member so that everyone benefited from each other’s strengths. ‘Leading by example’ (M) was identified by some respondents as key in inspiring dedication and setting the tone for the relationships between team members and towards patients (O):

For them to see that you’re a good leader, that you’re doing what you ask them to do, you have to show them that you’re willing to do it as well, I think you have to empathise, patients wise and to your colleagues, but also instil in them the importance of ‘How would you feel if that was your relative?’. And also, appreciation of the staff. I think dedication as well. I always try to be honest and transparent. That’s my principle; you have to be honest. You can’t tell them, yes we will do that, and then we can’t do that. So, we are trying our best, we will do it, this is the plan.

FS1

Summary of programme theory 1 findings

Despite differences in contexts and models of service operation and contextual influences, all teams identified core evidence-based components as central to the intervention effectiveness. Most respondents agreed that adhering to selection criteria regarding disability levels helped the services to manage capacity and promote responsiveness. However, the lack of or slow response of community rehabilitation services in the local pathways put ESD and/or referring teams under pressure to use eligibility criteria flexibly to respond to the needs of more complex patients. Most services were unable to offer the therapeutic intensity required to address the needs of patients with more severe disability, and attempting to do so could dilute the intensity of service for the rest of the patients. Research and clinical recommendations around intensity and responsiveness were treated with scepticism, and respondents called for flexibility in their implementation to account for capacity issues and to respond to patients’ individual needs and circumstances. Involving carers in the rehabilitation plan, promoting a self-management ethos and implementing telerehabilitation approaches were thought to increase practise between sessions and improve adherence. Mechanisms thought to streamline discharge and help teams to meet their responsiveness targets included having access to a social worker, adopting a flexible approach in terms of who conducts the initial assessment and establishing ongoing communication and a trusting relationship with hospital staff. The role of RAs in facilitating the delivery of an intensive and responsive service was highlighted, as was the need for appropriate supervision and upskilling. Effective MDT working with frequent formal and informal meetings was key to the services’ successful operation. Key mechanisms to enhancing MDT working were identified, and included interdisciplinary working and leadership.

Programme theory 2: ‘core evidence-based components of ESD will operate differently in urban versus rural settings’

Investigating how participating services achieved their targets unveiled subtle but important differences between urban and rural settings, especially regarding the delivery of an intensive service tailored to patients’ needs. Staff interviews supported the programme theory; however, they also highlighted the need to consider service capacity and the time spent travelling to further disentangle the influence of geography. Teams covering a large catchment area experienced similar challenges to rural services, although to a lesser extent. Findings are presented in two sections: (1) the first section describes how rurality and other important contextual influences interacted with staff’s efforts to offer an equitable and intensive service across their catchment areas; and (2) the second section focuses on travel time as a prevalent contextual challenge and how teams adapted MDT structures and processes to mitigate its impact.

Providing an equitable service

The time required to reach their patients (C) was identified by respondents as an important contextual influence, which could affect the number of patients they could see per day and the intensity of service that they could offer, especially against a background of reduced resources and stretched capacity (C) (Figure 21). Although all teams perceived travel times to be an issue that had to be considered and managed, the impact was more pronounced in services covering rural and/or large catchment areas, for which teams spent a big part of their day on the road, leaving less time for hands-on therapy. In rural areas, further delays were caused by poor road infrastructure and poor connectivity for mobility phones and satellite navigation systems. When coupled with capacity issues, achieving the recommended intensity targets was especially challenging for services covering rural/large geographical areas (C). Some participants from these sites voiced concerns about whether or not patients living at distal points of the catchment area received the same rehabilitation intensity as those living closer to the service:

I think if you live up on the north you’re probably not going to get as many visits as if you live around the corner. And it’s very hard to be equitable when you’re looking at patients and the geography and the number of visits they need and the staff you’ve got to make sure – so of course, if they live around the corner I would say, I don’t think we’ve ever actually done a piece of work looking at that, but I would say anecdotally that probably if you live up on the north you may not get as intense a service.

ES16

FIGURE 21.

The CMOs concerning an equitable service.

To try and cater for patients based on their needs and not where they lived (O), rurally based teams would often have to work overtime or cancel training activities, ‘putting their patients first’ (FGES1) (M), to increase their capacity. They also spread the number of visits permitted by their capacity evenly across their patients to give the opportunity to those living remotely to access the same amount of therapy as those living closer to the service (M). However, when capacity was low (C), this strategy could mean that all patients on the caseload received a less-intensive service over a longer period of time (O). Staff at one of the rural sites explained that, with intensity being diluted, the service had to be prolonged to allow patients the time to achieve benefits before discharge. This in turn had a knock-on effect on patients’ flow from the hospital and the service’s capacity (O):

It really is about what patients’ need is and that’s where you get this conflict, because actually, they would benefit from this, but the reality is we can only give this. Then the outcome is that they’re sat there longer than they should be and we’re not seeing them as quickly and not progressing as quickly and they’re getting bored because they’ve got exercise plans which are too easy and they’re not feeling challenged.

ES1

In rural and/or large geographical areas (C), a lot of time and effort was invested in timetabling and planning home visits. The same was reported in a site with a high level of deprivation and a recent increase in patients with a history of alcohol abuse (C). Staff visited in pairs for safety or to obtain a holistic assessment while saving on mileage and costs. However, the complexity and resources required to organise these visits (M) often had an impact on the time available for hands-on therapy (O), especially when the team did not include a dedicated administrator:

[Distance] cuts down the number of patients that you see if you are sending two staff out at a time and we do have a few complex ones like that.

FGSF1

If anyone’s got a history of alcoholism we have to go out in twos because we just can’t predict it. That then makes it so much more complicated because we all manage our own diaries and just getting the right person at the right time to be able to double up with you is difficult.

AS1

Respondents from services covering rural/large geographical areas (C) noted that the complexity of their timetabling allowed little flexibility when arranging home visits for patients living remotely (M). This could mean reduced opportunities for rehabilitation provision tailored to patients’ needs (O), for instance arranging shorter and more frequent visits to manage patients’ fatigue. It was also suggested that patients living closer to the service might benefit from cancellations and opportunistic visits, or even experience a more responsive service (O). One team explained that when capacity was low and patients had to be put on a waiting list, patients living closer to the location of staff members would be prioritised once a slot became available:

It does make a difference, though, because if there’s someone just down the road and someone’s had a cancellation and they’re in the office, they could quickly nip out to see them in their afternoon where they’ve got a slot, whereas if they’re right at the top of north of the county then they can’t just nip out to see them.

FGES1

We’ve got people that are on the real outer borders. And if there’s nobody else out there, I’m just thinking the border which might take you over an hour to get to, if you’ve got nobody else out in that direction, I suspect the clinicians are thinking carefully about how frequently they need to go out there. I’d like to think it’s not impacting on – but I think that would be – that’s kind of human nature really, isn’t it?

DS14

All of the teams were set to encourage functionally oriented rehabilitation practice. In urban and inner-city areas, goal-setting could involve using local facilities, such as catching public transport or visiting a shop or café. Some teams organised on-site rehabilitation groups to augment practice and achieve the recommended intensity. However, patients living remotely, especially those living in rural areas, had a disadvantage in terms of being able to access such services because they often relied on poor or non-existent public transport, while fatigue and physical disabilities made travelling even less desirable (C). If support with transport could not be arranged by the service, patients often refused participation (M) and, therefore, missed out on opportunities for rehabilitation practice and socialisation (O):

And what would be the reasons why they couldn’t attend?

Some decide it’s too far too travel so we only have this location to work from so some live in the peripheries of the county so they would travel for an hour, you know, they don’t want to do that.

Do they just miss out then?

Yes.

There is no alternative that will replace that opportunity?

No, the resources in that workshop we can’t take out of the workshop so the equipment we can’t take from there to the patient’s home even though some of it is mobile. Patients that attend that group are the ones that have transport or want to come.

CS5

I have got a patient this afternoon who is very remote and that’s one of the things we are going to be talking about, would you like to be socialising more, to do that would you like to travel? So we would be mindful of it and offer it and see if we could link in with social services who can provide transport so that those patients could potentially get picked up, taken to a group and brought back. I think, we are kind of aware of possible things that effect their rehabilitation and try and offer things around it.

ES18

Managing the impact of travel time

Although services covering rural and/or large geographical areas were particularly affected by the time spent on the road, all teams took active steps to mitigate the impact of travelling on the operation of their services. The following two sections describe CMOs related to services’ efforts to reduce travel times and use resources efficiently (O) through (1) the adaptation and management of MDT structures and processes (M) (see Adapting multidisciplinary team structures and processes), and (2) flexible working arrangements (M) (see Flexible working arrangements).

Adapting multidisciplinary team structures and processes

The CMOs described in this section concerned how teams adjusted the way that their MDTs were organised and operated (M) to mitigate the impact of travel time on the operation of the service (O) and staff’s well-being (O) (Figure 22). All of the teams organised patients geographically and most sites implemented ‘patch working’: dividing their MDTs into subteams that covered different geographical areas (M). This strategy was thought to allow services to meet responsiveness targets and deliver the greatest intensity of service possible. In one site, the MDT was split into five subteams; it was envisaged that this would ensure that patients across the catchment area had equal access to senior therapists:

Because we work geographically we do reduce the amount of travelling we have to do and we can plan our own diary so we have that flexibility to say, ‘OK. I’m seeing that person, they live here. So I’ll see that person this time and then I’ll come back to here, or I’ll finish where I can just go straight home so I’ll do all of these at the end of my day’. So we have that flexibility with our diaries.

AS16

FIGURE 22.

Adapting MDT structures and processes CMOs.

Site C divided the weekly MDT meetings into two patches to make the best use of staff’s time, reducing the length of the main MDT meeting. In addition to the weekly MDT and wider meetings, some teams had introduced separate patch meetings to plan the visits for the following week. During these meetings, staff would collectively do their timetabling for the week to ensure that every patient was seen daily and that there were no overlaps. This process was thought to encourage communication between staff members and allowed subteams to discuss cases in more detail:

We’ve just fairly recently introduced patch meetings, so that we can just sort diaries for the week, so just making sure that patients get a visit on each day, and nobody clashing with anyone else. We just do it together, standing in front of the roll with their diaries, and go, right who’s going what day, and then they can say, they’re off on annual leave and can you cover that day instead of that day and switch it all round.

CS3&4

Good planning and co-ordination across subteams was key. When the service capacity was stretched, subteams would work flexibly and support each other to keep caseloads evenly spread and allocate resources where they were most needed (M). The latter was not always easy to achieve, especially in the context of wide/rural geographical catchment areas; the complexity associated with travelling in these areas made it difficult to plan the capacity of each subteam. As discussed in one of the focus groups (FGDS1), patches tended to continually shift as the caseload from each geographical area fluctuated. Despite the challenges, patch-working was thought to allow the teams to respond faster to changing circumstances:

Because you realistically see one in the morning and one in the afternoon. And there are times that if they all happen to live within a 5-mile radius, which has happened, then you can see four of them quite comfortably in a day. And it’s really variable to then say OK yes we can have this number of patients or this capacity because it’s not really a straight forward figure.

ES3

From my point of view, the patch-working is essential in a team, when you cover a huge area because it allows you that flexibility to sort of say right, OK, we’re not getting as many referrals from [area X] so you shift your patch up a little bit and we can work from there.

FGDS1

In sites that covered large geographical areas with high levels of rurality, when possible, staff treated patients who were based near their homes. However, despite the steps taken to reduce the burden of travelling, many respondents from across the six sites talked about how spending a big part of their day on the road affected their work–life balance and their physical and mental well-being (C). Respondents also felt that the way that the national rates for mileage and car wear and tear were calculated placed staff who travelled a lot, and especially those in lower bandings, at a disadvantage. These issues affected staff morale and led to occasional resistance to work in areas that required a lot of travelling (C). Team leads tried to manage the situation by having staff or subteams rotate in covering geographical patches (M) to share the burden of travelling and promote job satisfaction (O):

People will sit in the area that’s closest to their home generally speaking. So, we kind of allow half an hour travel per patient, as an average.

DS14

It’s quite a long way and we have had to ask the other existing support workers to cover that area. There is some of them that are saying, I won’t do it and we have had to try and say to that person, everybody is going to have to take a turn doing that area. One of the new ladies that we have recruited will have to go there permanently and she has been told that at interview.

CS5

To maintain continuity of care despite staff rotations (O), one team explained that, once staff were assigned to a patient, they would stay with them until they were discharged (M):

Once you’ve got that patient in your team, we’ll be their stroke nurse for the whole journey that they’re with us; so they know your stroke nurse, you can ring her on this number, and the same with the therapists. Generally, they’ll see the same people, so once they’ve got to know us all and they’ve got to meet us, they’ve got a bit of continuity of care, so we get to know the patients really well.

FGES1

Flexible working arrangements

Where good information technology (IT) facilities were in place (C), flexible working arrangements that allowed staff to return to the office only when necessary (M) could further help to manage travel time and optimise resources (O) (Figure 23). Staff were encouraged to start and finish their day with patients rather than in the office, which, for some teams, meant that they could be visiting the base only once per week:

If I had a space in the morning and my first patient was down south, I might come here just to maybe get some resources or do my e-mails quickly, but I certainly wouldn’t come here if a patient wasn’t on the way. And I might use it as a base to come back for lunch or something, but only if I was close otherwise it’s a layby or wherever.

DS7

You tend to have to be quite resourceful in your day as well. Some of us use the live SystmOne, which is where we do our notes. So we might choose to do notes in between visits to help speed things—

So you don’t have to travel back to the base and do your notes.

We’ve got the new mobile app that we download everything onto before we leave the office, then you can do your notes and you can access it.

FGES2

FIGURE 23.

Flexible working CMOs.

Remote working helped to reduce travel time, but it could also mean that there were less opportunities for interaction and for the exchange of information, expertise and emotional support (O), all of which were deemed by respondents as particularly important in community working. Teams highlighted the value of developing and sharing good local knowledge on a team level (C) to maximise efficiency. These exchanges were put at risk by a lack of face-to-face communication between team members. Although technology facilitated flexible working arrangements, it did not address all of the issues arising from the lack of face-to-face interactions. As noted by respondents, one had to be an excellent communicator to ensure that staff and patients knew when patients were seen and by whom merely through mobile telephones and texting:

Parts of the city where you can’t park because you’ll be clamped, things like that. Don’t park there because you think you can park there but you’ll get a ticket. This is where you need to park. All that knowledge you do build up as a team.

ES1

I think there’s a really strong argument for actually having physical contact with people. We can see if a member of staff is struggling. You can just pick up on a whole load more if you’re actually seeing the team. And it is hard, you know, it’s not easy working out in the community without coming back and touching base I think you lose that sense of team. And yes, we’re very good at using e-mail on our phones but it’s not the same. And I know that members of the team who don’t attend MDT because of their working structure, they battle. They don’t like not being a part of it, they want to come back into it. And we do telephone handovers but nothing replaces actually being in MDT, actually seeing each other otherwise it’s really lonely being out there [laughs] all week. So just having that one contact I think is really valuable.

DS7

Summary of programme theory 2 findings

The findings suggested that rurality, especially when coupled with capacity issues, could influence (1) the intensity of rehabilitation provision, (2) teams’ flexibility to adjust the frequency and duration of the visits to people’s needs, and (3) patients’ access to facilities that would promote rehabilitation practice and socialisation. Staff responded by working overtime or extending the length of the service, which, however, could have a knock-on effect on service capacity and patient flow from the hospital. The time required to travel to patients was a key contextual influence, which meant that services covering large catchment areas were also affected, albeit to a lesser extent than rural services. Therefore, the results support the initial programme theory but also highlight the importance of considering travel times and service capacity in interpreting the intended and unintended outcomes. To mitigate the impact of travel times on their operation, services organised their MDTs around local geography and encouraged flexible working arrangements. When good IT facilities were in place, returning to the office only when necessary contributed to the efficient use of resources; however, unless frequent face-to-face communication was safeguarded, working remotely minimised the opportunities for the exchange of information, expertise and emotional support between staff members.

Programme theory 3: ‘the quality of communication processes between key stakeholders and services in the local stroke care pathway will influence the provision of a seamless, patient-centred pathway’

The initial programme theory concerning communication was developed during the preliminary analysis of staff interviews while data collection was ongoing. It suggested that the quality of the communication processes between key stakeholders and services in the pathway would influence the implementation of an evidence-based service. As data collection and analysis progressed, new insights required this programme theory to be revised to better reflect the staff accounts. The refined programme theory focused on patient experience and the provision of a seamless stroke care pathway as the key areas influenced by the quality of communications between services and stakeholders. Findings are structured in three sections corresponding to CMOs concerning communication between ESD and (1) patients and carers, (2) stroke units and (3) services across the stroke care pathway.

Communication with patients and carers

Unrealistic expectations of ESD were common, and often reflected gaps in information provision prior to discharge or even referrers’ lack of clarity around the intensity and length of the service (C) (Figure 24). All teams recognised the need for timely, clear and constant information provision to both patients and families (M). This was important for addressing concerns about returning home and misconceptions about their recovery (O):

I don’t think referrers know us as well, so they probably don’t know exactly what we can provide. So in terms of patient expectations sometimes you get ‘Well I was told I was going to be seen every day’ or they’re not happy with being discharged at 6 weeks or they’re just in a bad place because they’ve just had a stroke and they’re low in mood.

FGES1

FIGURE 24.

The CMOs on communication between ESD teams, patients and carers.

Families occasionally held cultural expectations that therapists would ‘do it for’ the patient rather than patients taking responsibility for their own rehabilitation and recovery (C). Being honest and realistic in their communication (M) was identified as key in managing expectations and improving adherence to rehabilitation (O). Striking a balance in relation to information provision was also considered necessary to ensure that patients were not overwhelmed. Some teams discussed how they developed their information material to consider stroke-related cognitive and language difficulties (C). In response to patient feedback, one service adapted their discharge letters (M) to make them more accessible (O):

I think they do come with some expectations and the people that in-reach on the ward will go and introduce themselves to the patient and give them a bit of an explanation anyway. So, they do have some idea of what’s likely to happen. But I think they’re often still quite surprised at how much we see them and how much they get and how much they improve. I don’t think they expect to get as much as they do actually. But some people want lots of therapy and some people don’t, and we try to have that negotiation with them earlier on. I think sometimes people want to keep us a bit longer but that’s always going to be the case isn’t it, no matter how long we see people for.

FS3

According to respondents, patients would often feedback that their discharge from ESD came too soon; this observation was common across all sites (C). As noted by a team lead, discharge from ESD services should be actively managed to avoid leaving patients and their families with a sense of ‘abandonment’ and to avoid an abrupt ending of the service. Timely communication (M) was key to effectively preparing patients practically and emotionally for their discharge from ESD services (O):

The ending bit used to be a problem about saying to patients, ‘We think – there’s that awful thing that’s used – you’ve maximised your rehabilitation capacity now so we’re going to say goodbye’. It’s not very helpful for patients to hear that, as in I’m going to have a meeting with the patient today to tell them I’m not going back next week. You should have done it long before that. The patient should be aware and ready that this the process they’re on.

BS20

Carers were acknowledged to be a vital component of the rehabilitation journey (C). Establishing and maintaining communication with the carers (M) was required to manage the delicate family dynamics that come as part of home-based rehabilitation and to help carers to feel engaged in the rehabilitation process (O). Staff worked to acknowledge carers’ presence and role in the rehabilitation plan, honour their own wishes and emotions, and provide support. Early goal-setting that involved the carers allowed patients and their families to have a clear understanding of what they were working towards and start the process of adjustment. This, in turn, helped to engage everybody in optimising the amount of time spent on functional therapy practice between formal therapy sessions:

Occasionally, the carer might find it as an intrusion because you’re going in someone’s house and it’s a very different dynamic to being in hospital. [. . .] They come as a package quite often I think, your family dynamics. But I will go out and talk to the carer. So, I’ll actually say ‘I’ve come out today but perhaps we can go and have a cup of tea just to see how you are as well’. Some of that is about acknowledging that they’re part of the picture, but they’re also going through a lot. It’s about giving them the forum to talk through those things as well. Some people didn’t like each other in the first place as well, which is always very difficult. So, you get people dynamics where they were living very separate lives to start with. [. . .] The big thing is to acknowledge that person that’s living with somebody has something to say. Nobody is a passive recipient of anything.

B20

Communication between early supported discharge services and stroke units

Good communication with teams from the referring stroke units (M) was described as a key enabling factor to a patient’s smooth discharge process (O) (Figure 25). Respondents emphasised the need to be proactive, especially in the early days of the services’ development, as referrers may not fully understand the screening and referral process nor appreciate the quality of the service that patients would receive at home (C). In one site, the service manager had engaged in a lot of preparatory work to establish good communication with referrers, visiting the stroke unit daily for about 1 year before the service’s full operation. Communication and rapport-building efforts were a vehicle to educating about the service’s role and remit, shifting attitudinal barriers and promoting a trusting relationship with the referring team (O):

Developing communication and rapport and I suppose trust as well, communication and trust. Because if the therapists don’t trust the community team then they are reticent to let that patient out so you have to develop that and that’s developed over time.

FGBS2

FIGURE 25.

The CMOs on communication between ESD teams and stroke units.

Four teams in-reached to the acute trust and perceived this process as a central feature of their services’ models, enhancing relationships and formal and informal communication (C). Two teams were embedded in the acute trust, and attended ward rounds, were involved in assessment and assisted with discharge planning. Not everyone agreed on the benefits of in-reaching. Some teams had trialled and dismissed in-reach as a model, deciding that it was not an efficient use of their resources. To mitigate the lack of direct communication about patients, they had put other processes in place. These processes included designing their own referral forms to provide detailed information about patients, which, however, were often poorly completed by the referring teams; as a result, ESD teams would have to either screen all referrals and telephone for missing information or search the trust’s databases. They also made periodic visits to the stroke unit to refresh ward staff on their role and the referral criteria, a process further necessitated by the high turnover of acute staff:

I think the communication’s strong and because we pitch up at their MDTs frequently, we can ask the questions and prompt that because obviously when there’s a turnover of staff at the hospital end or our end, we do need to keep revisiting the criteria, so yes we do that. I think trying to get night-time charts from nursing staff and things, so that we know we can help form our risk assessment. So it’s a very MDT approach, asking the nurses for their thoughts on the ward as well about the patient because they see them throughout the night. Often their behaviour and their needs can be quite different.

Do you see yourselves then as – do you see them as part of the team?

Yes, and I think that’s another reason why the speed of getting the ESD patients out has become better over time, because they trust us.

I think as you say, we have a quite trusting relationship, I think it works both ways. They know they can send people home earlier than maybe if they were elsewhere because of us.

They know us and we know them, so we work with them.

FGBS2

We did trial in-reaching for a short period of time but it just didn’t quite work and we didn’t feel it was a good use of our time, sitting in a meeting where you’re discussing lots and lots of people and then we’re just coming back with a list of names when actually, we could be out there treating patients. Sometimes the dates that they propose weren’t necessarily the date they were discharged either.

AS16

In one site, senior management encouraged meetings between the managers of ESD and stroke unit teams (M); it was envisaged that if they could understand each other’s pressures and needs they would be more supportive and collaborative (O). Other commonly practised communication processes included staff rotations, where teams were employed by the same organisation, and hosting visits by acute trust staff to shadow ESD team members where they were not from the same organisation. Respondents agreed that cross-boundary working (M) promoted an understanding of the ESD role and assisted in achieving appropriate and timely referrals (O):

What I say to them is ‘Oh, I’m from ESD, do you know what we do?’. And I give them a leaflet and I say, ‘Come and spend some time with us’, but it’s about good relationships, communicating. I usually try and grab any new nurse that I don’t know and sometimes if they’ve got a good mentor the mentor will send them down. So yeah I always ask people to come and be here.

ES4

There were differing opinions of the value of physical proximity with the stroke unit (C). For one ESD service, co-location with the acute team meant face-to-face interactions, shared IT and training events, or even supporting each other when at overcapacity. By contrast, communication with other referring hospitals was more strenuous because, other than speaking over the telephone, they did not have many opportunities to build up rapport in the same way:

I think one of the things that really helps is that the stroke rehab unit and the early supported discharge team are effectively the same team. You know, we’re the same organisation and we’re on the same site, so it’s much more seamless than other referral routes. We communicate quite closely with the ward team here. Some of the staff have worked in both environments, which helps with that as well. [. . .].

Another advantage with some of the staff who are on the ward, who have worked in ESD, they appreciate what it’s like to work in ESD, so when they’re making their referrals to us, they consider that. You’re not just doing it to this other team, that you have no idea of their working processes.

Yes, and also when I’m with board round, I’ve got both hats on. ‘What about this patient, what about that patient?’. You get a really good handover. So it’s those personal relationships that we’ve built up as well, that make a difference.

FGDS1

Communication across the stroke care pathway

Community-based care was characterised as one of the riskiest parts of the patient’s journey through the pathway owing to the difficulty in predicting how families or systems are going to cope or manage (Figure 26). ESD service staff often assumed a case worker/key worker role, liaising with other services (as for instance GP and social care) and signposting patients to ongoing services (C). Teams reported that they had to contact multiple organisations both statutory and voluntary and much time could be spent in identifying which service was appropriate based on the patients’ GP and their postcode. At the same time, the frequent restructuring and decommissioning of services was not communicated effectively, leading to staff not knowing what was available for their patients (C). Meetings across the entire pathway, as regular service management meetings, quality improvement meetings or whole pathway service redesign meetings, promoted clarity on the availability and role of each service and awareness of the pressure points in the pathway (M). Communication was streamlined as a result, leading to smoother and safer handover between services (O):

If you have a very clear pathway from acute to the community, where everyone knows their role and doing that, then less time is wasted and it’s a more streamlined process, and timely, and also safer, hopefully, as well. Everyone takes part in developing the pathway, which is I think what happened here.

FGBS2

FIGURE 26.

The CMOs concerning communication across the stroke care pathway.

Liaison between teams was hindered by a lack of joined-up or shared IT systems in many areas (C). Acute trusts and ESD services were rarely able to share electronic records, scan results or test results, which hampered the sharing of information about patients. Different therapies were sometimes employed by different organisations and, therefore, had separate electronic notes or even paper-based notes. In one team, SystmOne contained standardised templates across all four patches (C), which meant that teams could access each other’s caseloads and notes (M). This meant that they could cover for each other in busy periods (O):

I think SystmOne is really working well, because we can see what other people have put on the system, and providing we’re all on SystmOne we can see what the care plans are. So that doesn’t hinder if we’re working across team, that’s a fantastic facility. Where it falls down is when we’ve got patients coming in from say some of the hospitals and we can’t see what’s happened to the patient before they come to us. So particularly for speech therapy, I don’t know whether it happens in the other therapies quite so much, but they might be complex, they might’ve had a lot of assessment and we have no knowledge.

I think from a hospital point of view we have that frustration as well, and with the absolute kind of utopia would be, we’re all on the same system, so that you can see scans, if they’ve had a stroke prior to that and we can see what’s happened and where – that would be the absolute ideal, wouldn’t it, if we were all on the same system.

I mean to be fair we can access the care portal, but it’s still not the same as what we can access on SystmOne, so while we can see the big things on the care portal we can’t see some of the . . .

Therapy notes.

. . . details. We can’t access those.

FGFS2

Achieving and maintaining good communication with commissioners was identified as an important parameter in developing an ESD service and establishing its position in the stroke care pathway (O). One team described liaising with commissioners as a time-consuming and stressful exercise that could end fruitless as ‘they don’t speak the same language’; it resulted in ESD leads disengaging from the process because they felt that they had to prioritise their clinical work. As noted by a service manager, cultural differences between organisations reflected differences in their communication style (C); they thought that this characterised communication efforts between commissioners and clinicians and stressed the importance of ESD services tailoring their networking approach (M) to effectively reach out to different stakeholders (O):

I think there are very, very many competing priorities for funding other than commissioners, what you have to do is make sure that you get the ear of the commissioner, establish your relationships and networks and never miss an opportunity and that’s very much an individual thing.

CS6

Summary of programme theory 3 findings

Findings highlighted the importance of good communication in developing collaborative trusting relationships between key stakeholders and promoting the provision of a seamless patient-centred pathway. In their communications with service users, participants stressed the need to be honest and pragmatic and ensure the ongoing provision of tailored information to both patients and carers. Building rapport and a two-way communication with carers as well as patients was considered key to effectively managing expectations and engaging carers in the rehabilitation process. Creating opportunities for formal and informal communication with the stroke unit was identified as a mechanism driving the development of a trusting and collaborative relationship with hospital referrers, and contributed to improving the referral and discharge process. Establishing processes for regular cross-service communication emerged as a facilitatory mechanism to ESD teams’ efforts to liaise with multiple stakeholders across the pathway to ensure smooth and safe handovers for their patients. Against a background of cultural differences between organisations, the need to tailor communication style to engage stakeholders, such as commissioners, was important in securing their buy-in and ongoing support.

Discussion

Overall findings supported our programme theories by (1) indicating the value of core evidence-based components to successfully deliver the intervention in different settings; (2) demonstrating how features of geography and particularly the time required to travel to patients shaped the delivery of certain intervention components; and (3) demonstrating why and how effective communication across the pathway can influence the provision of a seamless and patient-centred pathway.

In line with our first programme theory, findings supported the need for the adoption of core evidence-based components in achieving effective ESD. Despite considerable variations in their model of operation, services identified key characteristics and processes that reflect research and clinical recommendations and, when in place, were believed to facilitate the delivery of evidence-based practice. These included the provision of an intensive and responsive service to eligible individuals by a co-ordinated MDT that met frequently. For instance, the application of eligibility criteria allowed services to manage capacity, promote responsiveness and ensure patient safety at home, although it was accepted that relying on physical disability scores to decide eligibility may not provide the full picture.

To deliver these components, however, services had to respond and manage countervailing conditions of specific contexts within which they operated. This reflects the Damschroder’s conceptualisation of a complex intervention as consisting of core active ingredients and an adaptable periphery that can be modified to account for contextual influences without compromising the intervention integrity. 14 For instance, services safeguarded their MDT meetings but adjusted the way that their MDT was organised, dividing the team into smaller subgroups corresponding to geographical locations with the view to reduce travel time and use resources efficiently.

Respondents recognised the need for intensive and responsive ESD; however, they disagreed with the prescriptive nature of the clinical guidelines,5 arguing that clinical judgement is required to ensure that individual patient circumstances are considered and the service best meets people’s needs. Participants did not recognise a dose-dependent relationship between intensity of practice and patient outcomes. They emphasised that the need to provide a holistic service addressing the multifaceted impact of stroke might take time away from hands-on practice. For instance, teams could be called to address mood problems or issues related to returning to work. The analysis also showed that intensity and responsiveness targets were sensitive to certain contextual characteristics, including the fragmentation of the local care pathway with lack of slow response of other services.

Service intensity and responsiveness were also influenced by the delicate balance between service capacity and the time that the team spent travelling to patients, as suggested by the analysis of the second programme theory. The impact was more prominent in rural services, but those covering large catchment areas were also affected. Distinguishing between remoteness and rurality103 might help to explain the impact of travelling times in these contexts and frame it in relation to geography. Barriers associated with rurality and travelling times stretched the capacity of these teams; when staffing levels were low, services often relied on staff commitment and hard work to sustain optimal levels of care. Roots et al. ’s104 study on rural rehabilitation practice in Canada found that OTs and physiotherapists ‘stretched their role’ beyond the normal expectation of the professions as a strategy to overcome resource shortages. In the study by Adams et al. ,105 workforce, including staffing levels, was identified as a key factor in delivering and sustaining a rural physiotherapy service.

Qualitative feedback included in the ESD consensus document13 also suggested that travelling times might reduce face-to-face contact and stretch service capacity. Supported by previous research, our findings suggest that the pressures posed by travel times and rurality make sustaining the quality of service provision a resource-intensive task. These teams may require higher staffing levels than similar-sized teams operating in urban contexts to meet their targets and deliver evidence-based care, which needs to be highlighted at a commissioning level and be taken into account in developing service specifications.

Good communication and the development of trusting, collaborative relationships was a consistent theme across the three programme theories. Findings related to the first programme theory focused on ‘inner setting’14 communication within the MDT. Although MDT meetings were recognised as the main forum for supporting communication and collective decision-making, the importance of informal face-to-face interactions was also stressed as key to exchanging expertise and practical and emotional support. The value of regular informal communication has been previously highlighted in the literature. 19,106 Waring et al. 19 found that informal interactions enable the development of trust and collaboration, and may contribute to improved hospital discharges. Episodes of opportunistic dialogue, facilitated by physical proximity, contributed to mutual learning and achieving interdisciplinary working in stroke rehabilitation settings, according to Clarke. 107 In this study, we found that efforts to reduce travelling times, through flexible working arrangements, could have a negative impact on face-to-face communication. Although the use of digital technologies supported communication and information exchange, they could not substitute face-to-face interactions. At a time when telerehabilitation approaches gain ground, our findings highlight the need to protect opportunistic communications and encourage face-to-face meetings.

Building on existing guidelines, this study examined staff’s perspectives on why and how certain evidence-based components influence outcomes, allowing a deeper exploration of certain elements of the intervention. With regard to team composition, the study clarified the role of RAs, for which agreement was not achieved on the consensus document. Based on our findings, a recommendation could be made for RAs being a core member of an ESD team. Findings also stress the need to address certain contextual elements that enhance the role of RAs, such as delegation from registered therapists and the need to empower and promote the autonomy of support staff through appropriate supervision and upskilling.

The access to nurses, psychologists, doctors and social workers offering stroke-specialist input was essential to promoting an interdisciplinary skillset and enabling a holistic approach to rehabilitation provision. The timely involvement of social workers who understand the needs of stroke survivors and who are embedded in the pathway was perceived as an important facilitatory mechanism, streamlining the discharge process and ensuring that stroke survivors received a tailored care plan. Social carers were called to play a boundary-spanning role,19 working across social care and health-care organisational boundaries to identify and address unmet needs in a timely fashion. The need for partnership working with social care has been a consistent finding in the ESD literature23,39,56,108 but the issue remains pertinent.

In addition to further supporting the value of certain MDT processes, such as formal meetings, the research design surfaced nuanced, qualitative features of MDT working, including interdisciplinary working and leadership. Across the six sites, respondents agreed on the need to invest in developing an interdisciplinary skillset and work ethos across the MDT. It was suggested that the nature of community working and the efficiency of activities required collaborative working and blurring of professional boundaries. As noted by Baillie et al. ,108 skills transfer permits a better management of resources in community working because it reduces the need to have more than one staff member at home visits. Baillie et al. 108 identified trust and confidence in colleagues’ expertise as a context facilitating the development of an interdisciplinary skillset. In this study, we found that interdisciplinary working further promoted trust and respect in other disciplines. Our conceptualisation of interdisciplinary working as a mechanism overlaps with the ‘role blurring’ construct identified by Sims et al. ,109 which enabled carryover of skills by other disciplines, enhancing the continuity of care and helping staff professional development. Interestingly, and in line with our findings, both Baillie et al. 108 and Sims et al. 109 stressed the need to protect expertise and specialism and ensure that staff did not feel professionally threatened.

The role of effective leadership in maintaining standards of care and protecting staff well-being and job satisfaction was highlighted in the findings. Key mechanisms for effective leadership included the provision of training and support, and the development of a non-hierarchical culture of respect and appreciation. Staff across disciplines and bandings were encouraged to participate in collective decision-making and were empowered to contribute to service improvements. In a study investigating rural physiotherapy service provision, Adams et al. 105 highlighted the importance of staff engagement in decision-making and the provision of professional development and upskilling as important workforce factors contributing to staff retention in rural areas. Motivation, empowerment and feeling supported and valued were key mechanisms of leadership identified by De Brún and McAuliffe,110 which were linked with staff and patient satisfaction and enhanced team working. A review of relevant evidence in health care by West et al. 111 stressed the importance of leadership in shaping organisational culture and influencing the success of service improvement efforts. Our results on the importance of MDT working resonate with realist research in stroke care by Harris et al. 112 Taken together, our findings further emphasise the need for services to invest in enhancing effective multidisciplinary teamwork. However, similar to Harris et al. ,112 our identified outcomes operated at a team and not at a patient level, which implies an indirect relationship between effective teamwork and patient outcomes. Future research could develop specific interventions targeted to elements of effective team working, focusing on clarifying their relationship with patient-level outcomes.

The complex and multifaceted impact of stroke requires that ESD services work across occupational and organisation boundaries, often championing patients’ needs, to ensure that they receive comprehensive, integrated care. As demonstrated by the analysis of programme theory 3, effective communication with other services was key but not easy to achieve, especially against a background of cultural divides and lack of integrated communication systems. The relationship with inpatient services was perceived to be a particularly important one in streamlining the discharge process and ensuring the appropriateness of referrals. Despite that most participating sites had a well-established position in the local pathways, ongoing efforts to clarify the ESD role and build a trusting relationship with referrers were considered necessary. The need to address knowledge gaps among hospital staff regarding the ESD remit and benefits has been stressed in the literature. 23,46,108,113 This study identified opportunities for formal and informal communication and cross-boundary working arrangements to be key mechanisms for building a collaborative relationship. In-reaching and physical proximity appeared to provide a facilitatory context, although not everyone shared this view.

Findings from across the three programme theories highlight the interdependency between the ESD services and the local stroke care pathways. The quality of communication and transitions across services had a direct impact on the operation of the service, including its ability to achieve its responsiveness and intensity targets. Supported by previous studies,13 our findings stress the importance of a more integrated and comprehensive stroke service provision. As noted by our participants, the lack of an integrated stroke care pathway was especially obvious when they attempted to address the needs of more complex patients who did not fall within the mild to moderate eligibility criteria. Integration may take different shapes or forms. 114 Although organisational integration may not be feasible, virtual integration may be achieved through the development of networks and alliances. 114 In this study, we found examples of services establishing cross-service communication processes to promote information-sharing and collaboration; these efforts need to be protected and further enhanced. Our findings also have implications for evaluating context in implementation research. Damschroder et al. ’s14 and Rogers et al. ’s115 conceptualisation of ‘Cosmopolitanism’ as an organisation’s networking efforts through, for instance, boundary-spanning roles, is consistent with our findings. Future studies should include representation of stakeholders from across the stroke care pathway to investigate these system-level factors in greater depth and identify further strategies to promote integration.

The RE methodology allowed us to address the complexity of ESD delivery in clinical practice and obtain a better understanding of how intervention resources and stakeholders’ reasoning interact with contextual conditions to generate intended and unintended outcomes. Findings reiterated the importance of context for successful implementation of large-scale interventions and highlighted contextual features and mechanisms that need to be either addressed or capitalised on to improve effectiveness. By distinguishing between intervention components, which need to be safeguarded and adaptable elements, this study could effectively inform customisation efforts and interventions aimed to maximise success in local contexts.

The analytical task of linking CMO chains is suggested to help unpack causation, increase the explanatory power of findings and improve the external validity of case studies. 8,20 Although the iterative analytic process required considerable investment of time and resources, we also found that the theory-informed approach and the application of the CMO heuristic facilitated the analysis of a very large and information-rich data set. The lack of prescriptive guidance on ‘how to’ configure CMOs makes the process an interpretative task that requires ongoing reflection and negotiation between researchers. In this study, having data independently analysed by two researchers helped to reduce ambiguity and increased the confidence in the trustworthiness of our findings. Disaggregating mechanisms between intervention resources and stakeholder reasoning was a particularly helpful approach, and we think that it allowed us to achieve a clearer delineation of mechanisms and contexts. We took a pragmatic position to configure our CMOs at a level of abstraction that would allow us to draw take-home messages relevant to clinical practice. However, we are conscious that this approach has led to a large number of CMOs and theoretical nuggets rather than an overarching higher-level theory. Although the exact replication of the identified CMOs in other settings is unlikely, the accumulation of insights from further testing of these programme theories in different geographical and organisational contexts could help to develop an evidence-informed theory base for the large-scale implementation of rehabilitation services in the community.

Conclusion

The findings supported our overarching programme theories and shed light on the key contextual features and underlying mechanisms driving the successful delivery of evidence-based ESD. The adoption of core evidence-based components was identified as central to the effectiveness of the intervention. Mechanisms facilitating patient discharge and ESD responsiveness included (1) the application of eligibility criteria, (2) having access to a social worker, (3) adopting a flexible approach in who conducts the initial assessment and (4) establishing ongoing communication with hospital staff. Rehabilitation assistants had a key role to play in supporting the delivery of an intensive and responsive service. Involving carers in the rehabilitation plan, promoting a self-management ethos and implementing telerehabilitation approaches may contribute to augmenting practise between sessions and improving adherence. Service intensity and responsiveness were influenced by the interaction between service capacity and the time spent travelling to patients. Rural services and those covering large catchment areas may require higher staffing levels than similar-sized teams operating in urban contexts to meet their targets and deliver evidence-based care. Across all settings, effective MDT working was perceived as key to ESD success. The need for good communication both within ESD and across services in the stroke care pathway was emphasised as a mechanism driving the development of collaborative trusting relationships between stakeholders and streamlining patient experience of care transitions.

Introduction

The previous chapter qualitatively explored how staff attitudes and behaviours influenced the way that stroke ESD services were configured and run. It also explored the way in which community-wide stroke pathway development and contingent structures and processes shaped service delivery. This chapter focuses on the views and lived experiences of stroke survivors and their families. The purpose is to provide additional depth of understanding on how the variation in operation and delivery of services, as described in Chapter 6, affect and are experienced by service users in the real world. The findings could help inform future service planning, including what may and may not add value from a service user’s perspective.

Methods

Five purposively selected ESD patients from each of the six ESD sites (n = 30) were invited to participate in semistructured interviews. Patients who were actively on caseloads were recruited by the ESD teams in consultation with the research team to ensure that the sample included patients with a variety of characteristics (e.g. sex, age and stroke severity). Interview guides were informed by the three programme theories described in Chapter 3. 116–118 However, a realist interviewing approach did not prove feasible because patient interviews required a more informal conversational style.

Results

Data were collected from 30 patients between November 2018 and November 2019. Interviews lasted between 24 and 70 minutes (mean 45 minutes, median 45 minutes), excluding consent time. All interviews were conducted in the patient’s home.

Discussion

Semistructured interviews with patients aimed to explore their views and lived experiences of receiving ESD services to help to augment the understanding of the three study programme theories relating to the adoption of evidence-based core components, geography and communications.

Patients were very satisfied with the ESD service, as has been found previously. 38,119 They also reported that it was the main conduit and catalyst for their recovery. Most patients had been told to expect a service after hospital discharge, but did not know or understand what the service was going to be. Several patients expected just to receive physiotherapy and, therefore, were delighted and surprised at the flexibility of the team to work holistically and support their psychological issues, as well as work towards meeting their physical needs. The teams’ ability to signpost and refer to wider community support services were also vital to assist with reintegration into family and community life. Patient expectations are important to manage because the evidence base suggests that positive expectations can increase patient engagement and motivation, and positively influence functional outcomes. 120,121

The responsiveness of the team in terms of the early first contact and visit was particularly important. It helped patients and families to feel safe in the transition from hospital, where they were heavily supervised, to home, where there was no immediate professional help and support. They particularly valued the team’s interventions on their behalf to sort out immediate problems and gaps in services. Previous studies have demonstrated that this transfer of care can be a traumatic time for patients and carers, and can increase stress and burden if not supported appropriately. 38,49,122,123

Patients especially appreciated that the service was delivered at home. This made the service more accessible and affordable, and meant that rehabilitation could be tailored to real lives, empowering patients to be more involved in their recovery. 49 These findings were replicated by Cobley et al. 38 in a study of ESD patients who reported that services delivered at home were more cost-effective and less demanding. It also enabled family members to be more involved in the delivery of therapy. Furthermore, research has shown that rehabilitation delivered in the home can improve high-level activities of daily living, such as walking and managing stairs, above that delivered in hospital settings,124 and other functional outcomes. 10,122,123

The person-centred approach afforded by the ESD teams with personalised goal-setting was a key component in their recovery by targeting what was most important to patients. This has been evidenced previously. 125 This helped with patient and carer motivation to extend therapy outside formal therapy sessions and helped to progress recovery at a more rapid rate. This included rapid assessments and the provision of access to aids and equipment, which was updated and changed as patients moved forward in their recovery. This is important because patient and family motivation and engagement are associated with good rehabilitation outcomes. 9 In addition, patients reported that this forward progression helped them to feel more positive and improved their mood.

Patients, overall, wanted to recover as quickly as possible, but recognised that ESD services were stretched and had to be shared out equally across the stroke pathway. They adapted their expectations of the service to the model of service in place in their area. They accepted that the number of visits and the length of service that they received was driven by the capacity in the team, and were happy to receive less than they ideally would like to make sure that those in greatest need also received a service.

Conclusion

Overall, patient experience of ESD was very positive. Patients reported that it was the main conduit and catalyst for their recovery. Early contact from the ESD team was critical to help patients and families in the transition from hospital to home. It was also the mechanism by which immediate problems were de-escalated and resolved, avoiding potential unplanned re-admission to hospital or transfer to residential care.

The delivery of the service at home enabled patients to engage and participate in their recovery more fully. This would not have been achievable if travel to a treating centre had been required. This was particularly relevant in rural and remote areas.

The person-centred, goal-oriented approach of the teams was highly regarded by patients. They reported that this optimised their engagement in the therapy and provided them with a locus of control, which they felt resulted in a more rapid and fuller recovery.

Patients accepted and were grateful for the dose and intensity of the therapy that they received. Although some would have preferred more frequent visits to speed-up their recovery, they were cognisant of the constraints that the teams were working under. Patients were satisfied with either the 5- or the 7-day service model, dependent on what was on offer to them.

Early supported discharge services delivered by experienced specialist stroke teams enhanced patient experience and optimised their potential recovery.

Programme theory 1: ‘The adoption of evidence-based core components is important for the intervention to be effective in practice’

A 17-item ESD consensus score was used successfully to quantitatively measure the adoption of core components of an ESD service model, for example team composition (core team and others), staff training, team meetings and service specificity (see Chapter 4).

Programme theory 2: ‘Core evidence-based components of ESD will operate differently in urban versus rural settings’

Chapters 4 and 6 provided evidence of variability in the types of ESD model adopted across both urban and rural sites. In Chapter 7, the CCA found that rural services were associated with higher costs per patient, which, in turn, were associated with adherence to evidence-based core components of the ESD model. This suggests that additional resources and costs are required for rural services to meet evidence-based criteria.

This was supported by findings in Chapter 8, in which the time required to travel to patients was highlighted. In the context of wide/rural geographical catchment areas, allocating resources was a challenge, particularly when there were capacity issues. Teams adjusted the way that their MDT meetings were organised and operated to respond to challenges posed by travelling. All teams organised patients geographically to improve efficiency and most sites implemented ‘patch working’, dividing their MDTs into subteams covering different geographical areas. It was through staff commitment and hard work that the impact of travelling times was mitigated, so that intensity and responsiveness targets could be met. Good planning and co-ordination across subteams was key.

Many respondents talked about how spending a big part of their day on the road affected their work–life balance and physical and mental well-being. Again, relating to this programme theory, the impact was more pronounced in teams covering rural and wide geographical areas.

Programme theory 3: ‘The quality of communication processes between key stakeholders and services will influence provision of a seamless stroke care pathway to patients’

All teams recognised the need for timely, clear and constant information provision to both patients and families. Being honest and realistic in their communication was identified as key in managing expectations and improving adherence to rehabilitation (see Chapter 8).

Findings from Chapter 9 showed that more could be undertaken to improve this, particularly with regard to the transition across hospital and community settings. Most patients had been told to expect a service after hospital discharge, but did not know or understand what the service was going to be. However, patient expectations were exceeded and, overall, patients were satisfied. The teams’ ability to signpost and refer to wider community support services was vital for patients to assist with reintegration into family and community life.

Timely communication was identified by staff as key to effectively preparing patients practically and emotionally for their discharge from ESD services. Staff reported that discharge from ESD services should be actively managed to avoid leaving patients and their families with a sense of ‘abandonment’ and to avoid an abrupt ending of the service. This stemmed from an observation common across all sites, which was that patients felt that their discharge from ESD services came too soon (see Chapter 8). This was confirmed by patients themselves, as reported in Chapter 9, in which more than half of the participants were concerned about what was going to happen after the initial 6 weeks of the service ended.

Good communication with teams from the referring stroke units was described as a key enabling factor to patients’ smooth discharge process. Communication and rapport-building efforts were a vehicle to educating about the service’s role and remit, shifting attitudinal barriers and promoting a trusting relationship with the referring team. In one site, senior management encouraged meetings between the managers of ESD and stroke unit teams; it was envisaged that if they could understand each other’s pressures and needs they would be more supportive and collaborative (see Chapter 8).

This aligns with findings from Chapter 4, in which attendance to the acute MDT meetings was a component of the ESD consensus score, which, overall, related to the responsiveness of ESD service delivery.

The implications of these findings for research and clinical practice are discussed in the next chapter.

Robustness of the results and limitations

We adopted a mixed-methods design that allowed us to draw information from multiple complementary sources using methodological guidance on the conduct of RE. 8

The study responded to the call for evaluation designs that consider the complexity of the large-scale service innovations in health care. 71 To achieve rigour in developing and reporting the study, the RE quality standards were applied. 37 Quantitative aspects of our methodological approach meant some deviation away from realist practice, including reporting standards. 21 To conduct robust quantitative analyses, context, mechanisms (resources)26 and outcomes were instead reflected in defined and measurable variables, the relationships between which were investigated with statistical modelling.

Implications for health care and policy

There were two findings of this study that were perhaps surprising based on previous research and require further comment in relation to health-care delivery and policy.

First, higher ESD consensus scores were not significantly associated with patient outcomes, as measured by the modified Rankin Scale. Although this could be interpreted as adopting evidence-based core components of ESD does not result in improved patient outcomes, we would emphasise caution at coming to this conclusion. Previous research has shown that ESD is associated with improved outcomes. 10,12 What our findings suggest is that the modified Rankin Scale is an insensitive measure of the effect of ESD on patient outcomes. This is in line with concerns from teams themselves about the reliability of the use of this score across the stroke care pathway. 76 We suggest routine use of other validated patient outcome measures, over time, to monitor the longer-term impact of ESD on patient recovery.

Our findings also showed that stroke patients who received ESD spent, on average, approximately 1-day longer in hospital. This is in contrast with previous clinical trials. 10 ESD patients were younger and more likely to have a lower modified Rankin Scale score at their final inpatient discharge. It was, therefore, important to compare their total inpatient length of stay with that of a group of patients matched for these (and other) important patient characteristics. It should also be emphasised that we used total inpatient length of stay data (combining different inpatient settings). Previous modelling appears to have been driven by assuming that patients with a higher modified Rankin Scale score receive ESD. 85 Although it is tempting to conclude that patients with more severe disabilities should be able to access ESD, previous research and our current findings suggest caution with this because without an appropriately resourced service, this may be detrimental to stroke patient recovery (and unsafe) over the longer term. 10

As reported in Chapter 5, the average length of hospital stay for stroke survivors overall has reduced dramatically over the last 20 years, from 40 days to 20 days, and, hence, all patients are being discharged much sooner. 80 Based on our qualitative findings, ESD patients spending, on average, 1-day longer in hospital could relate to the delays associated with the capacity of the ESD teams to accept patients (emphasising the importance of appropriate staffing). We also suggest that the additional day could be required to facilitate and co-ordinate the appropriate transfer of care to the ESD team (i.e. required administration) and to liaise with social care.

Here, we summarise other take-home messages with the aim to guide implementation and evaluation of ESD in practice.

Recommendations for future research

Observational studies in Sweden suggest patient and caregiver benefits related to the provision of ESD in regular clinical practice, in line with our previous study findings from England. 86,87 Here, we showed the modified Rankin Scale to be of limited use in assessing ESD in terms of influencing patient outcomes. We suggest that further research using validated patient outcome measures (e.g. measuring activities of daily living, general health/mood and quality of life) at longer follow-up periods is required to enable the investigation of the impact of ESD on patients over the longer term. 10,12,87

The extension of national stroke registries with the inclusion of community stroke data, including measures of activities of daily living (e.g. Barthel index score or Nottingham Extended Activities of Daily Living score) and impairment, activity, participation and well-being (e.g. Therapy Outcome measure), could offer important opportunities to evaluate stroke service delivery beyond the hospital setting. 1,16

The evidence relating to the cost implications of ESD is mixed, with previous studies reporting reductions, increases and no differences in cost compared with treatment as usual. 10,79 An original attraction of ESD from a commissioning perspective was a reduction in length of hospital stay. 11 However, Teng et al. 101 postulated that any reductions in hospital length of stay associated with ESD translate into negligible cost savings because empty beds are then used by other patients. Our findings suggested that the benefits of ESD relate more to the provision of a responsive and intensive intervention when patients leave hospital, rather than shortening their overall length of hospital stay. The cost implications of integrating ESD and community stroke rehabilitation models and the consequences with regard to patient outcomes over the longer term need to be investigated.

We conclude that the adoption of defined evidence-based components are necessary for both urban and rural sites, and it is encouraging that we found examples of evidence-based ESD models in rural regions. Rural services were more costly but they also resourced larger numbers of staff on a per-patient basis, which ensured that they delivered an evidence-based service. We suggest evaluating these observed cost differences from a patient outcome perspective. More research is also required to identify what interventions could help services to manage logistical challenges associated with delivering effective home-based stroke rehabilitation in different geographical contexts.

Impact

We conducted virtual engagement events with representatives from each of the ESD services who had participated in this research to share our findings and discuss implications for practice. We facilitated sharing of perspectives on providing evidence-based community stroke care and encouraged staff to build on our recommendations and support each other as a community of practice. We also provided our stroke survivor participants with a feedback package, which included a lay summary of findings and a thank you for being involved.

In April 2020, Rebecca J Fisher was appointed the National Rehabilitation and Life After Stroke Workstream lead, with NHS England and NHS Improvement Clinical Policy Unit. This was in recognition of our team’s research in informing national commissioning guidance, clinical guidelines,133–136 and national and regional stroke care improvement. 137 Further details of national activities, informed by this research, can be found in Appendix 5.

Contributions of authors

Rebecca J Fisher (https://orcid.org/0000-0001-6866-6341) (Stroke Association Senior Lecturer, University of Nottingham) was the principal investigator and budget holder, and was involved in all aspects of the research.

Niki Chouliara (https://orcid.org/0000-0001-8294-0722) (Research Fellow, University of Nottingham) managed the day-to-day delivery of studies, led the use of realist methodology, conducted the rapid evidence synthesis, collected and analysed staff interview and focus group data, contributed to the interpretation of the results, wrote the report and disseminated the results.

Adrian Byrne (https://orcid.org/0000-0003-4887-2572) (Research Fellow, University of Nottingham) managed the day-to-day delivery of studies, led the design of the statistical plan, obtained data from SSNAP, undertook statistical analyses, conducted the CCA, managed stroke patient interviews, contributed to the interpretation of the results, wrote the report and disseminated the results.

Trudi Cameron (https://orcid.org/0000-0002-4273-3611) (Research Fellow, University of Nottingham) managed the day-to-day delivery of studies, collected and analysed staff interview and focus group data, analysed patient interview data, contributed to the interpretation of the results, wrote the report and disseminated the results.

Sarah Lewis (https://orcid.org/0000-0001-5308-6619) (Professor of Medical Statistics, University of Nottingham) contributed to the conceptualisation of the research, the development of the grant application, the development of the statistical plan, data analysis and interpretation, production of the report and disseminated the results.

Peter Langhorne (https://orcid.org/0000-0001-8185-2659) (Professor of Stroke Care/Geriatric Medicine, University of Glasgow) contributed to the conceptualisation of the research, the development of the grant application, the development of the statistical plan, the interpretation of results and the production of the report, and disseminated the results.

Thompson Robinson (https://orcid.org/0000-0003-2144-2468) (Professor of Stroke Medicine, University of Leicester) contributed to the conceptualisation of the research, the development of the grant application, the development of the statistical plan, the interpretation of results and the production of the report, and disseminated the results.

Justin Waring (https://orcid.org/0000-0003-1459-5896) (Professor of Organisational Sociology, University of Birmingham) contributed to the conceptualisation of the research, the development of the grant application, the interpretation of results and the production of the report, and disseminated the results.

Claudia Geue (https://orcid.org/0000-0003-2243-0733) (Health Economics Lecturer, University of Glasgow) contributed to the conceptualisation of the research and the development of the grant application, advised on cost–consequences analysis, contributed to the interpretation of results and the production of the report, and disseminated the results.

Lizz Paley (https://orcid.org/0000-0002-3879-5377) (Intelligence Manager, SSNAP at the time of the study) contributed to the preparation of the data, the interpretation of results and the production of the report, and disseminated the results.

Anthony Rudd (https://orcid.org/0000-0002-5738-2703) (Clinical Director of SSNAP at the time of the study) contributed to the development of the grant application, the interpretation of results and the production of the report, and disseminated the results.

Marion F Walker (https://orcid.org/0000-0002-3534-591X) (Professor in Stroke Rehabilitation, University of Nottingham) contributed to the conceptualisation of the research, the development of the grant application, the interpretation of results and the production of the report, and disseminated the results.

Publications

Fisher R, Chouliara N, Byrne A, Lewis S, Langhorne P, Robinson T, et al. What is the impact of large-scale implementation of stroke early supported discharge? A mixed methods realist evaluation study protocol. Implement Sci 2019;14:61.

Fisher R, Byrne A, Chouliara N, Lewis S, Paley L, Hoffman A, et al. Effectiveness of stroke early supported discharge: analysis from a National Stroke Registry. Circ Cardiovasc Qual Outcomes 2020;13:e006395.

Fisher R, Byrne A, Chouliara N, Lewis S, Paley L, Hoffman A, et al. Effect of stroke early supported discharge on length of hospital stay: analysis from a National Stroke Registry. BMJ Open 2021;11:e043480.

Data-sharing statement

This project involved secondary analysis of anonymised, routinely collected clinical data. The data were made available under a data-sharing agreement between the University of Nottingham, SSNAP and the HQIP. A condition of that agreement was that all the data would be destroyed at the end of the project. Therefore, we are unable to make them available to others. Further information can be obtained from the corresponding author.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data are vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care.

Search strategy (MEDLINE example)

Database: MEDLINE(R) (via Ovid).

Date of search: July 2018.

Date range searched: January 2000 to June 2018.

Publications were eligible for inclusion in the review if the following criteria were met.

Inclusion criteria

• Evidence published in peer-reviewed journals.

• Published in English.

• The title or abstract reported research on ESD/community stroke rehabilitation services.

• The title or abstract reported research on stroke survivors’ or carers’ experiences of receiving rehabilitation at home.

• The title or abstract reported research on stroke survivors’ and carers’ experiences in the early post discharge phase (up to 6 months post stroke).

• The title or abstract reported research on health-care professionals’ views on home-based rehabilitation.

• The results section described contextual factors that might facilitate or impede the implementation and delivery of home-based rehabilitation services.

Exclusion criteria

• No full-text paper available online or via The British Library.

• Publication did not report on factors influencing the process of discharge from hospital/transition to community rehabilitation/delivery of home rehabilitation.

Site A

Site B

Site C

Site D

Site E

Site F

Access to doctors: context, mechanisms and outcomes

Being well-supported medically was perceived by the teams with access to stroke consultants as an important component of their services’ success. Respondents valued the opportunity to access timely specialist medical input via telephone or e-mail when they were concerned about a patient (O). One team explained that liaising with a stroke consultant permitted the assessment of any missed cases and reintroduction to the stroke pathway (O). A stroke consultant noted that, although they did not have paid time in the ESD service, the development of a trusting relationship with the team over the years and their own experience and expertise in handling issues pertinent to ESD patients (M) accelerated the process of providing input and advice:

I don’t have any time in my job plan to provide input to this service. It’s usually people you can trust most because you know they’ve worked in it for a long time. You know if they’re asking you something which is a problem, then you can trust the information provided is going to be reasonably appropriate and you can – because we’re experienced, we’ve dealt with this a long time, most of the problems or some of the problems you’ve encountered before. So the answers are usually quite easy to provide advice on. So no it’s not an onerous problem. It could be very onerous if you start another service and they don’t know each other, and people are referring a lot of things and you don’t have experience. So in the beginning it might be very difficult to do and time consuming until people get used to it and build up those experiences.

BS8

In two sites, respondents contrasted between the medically and consultant-led ethos of hospital-based teams and their therapy-led hierarchical structure (C). They thought that these differences between the community and the acute context could influence doctors’ attitudes towards the service and their confidence that ESD would be the best discharge destination for their patients. Having an established, non-hierarchical relationship with stroke consultants (M) improved referrers’ understanding of the remit and quality of the service and helped to ensure that referred patients met ESD selection criteria (O). In the extract below, a therapist talked about how the consultant attached to their service respected their way of working and offered a much-needed objective perspective:

I think on the acute ward the consultant will have the final say, you may make a recommendation about a patient, you may think that a patient isn’t safe to go home for A, B, C reason but the consultant may still make a decision about that patient being discharged or something happening to that patient. Here, I feel like we are more – our opinions are more respected, and the consultant will always take on board our feedback and consider that in relation to the patient’s needs. And sometimes it’s quite good for somebody else to make that observation if you are in there with that patient you can’t always see perhaps that maybe they are not making as much progress as you would hope or would expect, so it’s good to have that objective viewpoint. On a ward, often the pace of the turnover, the nature of the role in which you are providing treatment is very different to ESD.

CS5

He [consultant] and I had some really robust conversations, because he was really struggling to see the benefit of assisted discharge at the time. The biggest area of discord was the fact that we didn’t put in a medical component. So it was a therapy-led service, and that caused quite a furore in the medic within our consultant body, I have to day; and hence some of our rigorous conversations.

FS13

Access to clinical psychologists: context, mechanisms and outcomes

Where available, access to clinical psychologists and neuropsychologists was perceived by staff as one of the strong aspects of their service model, enabling the provision of a holistic rehabilitation approach. For most patients, their return home was marked by the emergence of mood issues in response to the losses and restrictions brought on by stroke (C). Acknowledging the psychological impact of stroke for patients and their carers and providing timely intervention (M) was seen as a key factor influencing patients’ readiness and adherence to physical and cognitive rehabilitation interventions (O). The provision of mental health training and support to staff members was considered to be equally important (M). Respondents described how clinical psychologists taught them the basic skills that enabled them to effectively talk about adjustment, anxiety and depression (O):

We’ve had a lot of training from psychology about how to deal with difficult situations as well, so I think we do a lot of psychology ourselves, with the patients and a lot of the carers. Sometimes it’s not the patient that needs it. You sit with the relative because they’re the one that’s struggling to adjust.

FGBS1

The neuropsychologist we have in the team does training for everyone who starts with the team, and gets us ready to do distress checklists, the outcome measure for mood, and have those nice awkward conversations that sometimes you need to have with patients.

I remember when I was a band 5 physio if a patient in the hospital was crying you would just autopilot, I need to make you happy again, whereas there’s less of that approach and more of a, let’s have a conversation about this, and let’s get something written down, and then we can have a conversation with the neuropsychologist.

CS1&2

A commonly reported concern related to the time required to identify and address psychological issues, which could well exceed the short lifespan of the ESD service (C). The problem was accentuated when emotional struggles did not become apparent until a later stage of patients’ recovery. By the time that staff had the chance to build the rapport required to identify the problem and the referral to a psychologist was made, patients could be close to their discharge date. Where available, follow-up services could take over but they would not offer the required intensity to make a difference (C). In these cases, psychologists from two sites suggested that they could consider extending their intervention long after patients’ discharge from ESD (M). In this way, clinical psychologists attempted to respond to patients’ medium- and longer-term emotional needs; their input, however, would not be captured by SSNAP (O):

It’s very difficult to capture the psychologist, because SSNAP again, very naïve, they think the psychology is all done and dusted by the time they’re discharged from ESD and we fed back to SSNAP and said no, that’s not the case because a lot of the patients aren’t ready for psychology interventions until the end of ESD. So, there’s a whole raft of data that SSNAP are potentially missing because I don’t think we’ll ever get a true picture of the impact or the psychological and emotional problems of stroke, because of the way that data set is. It’s far too early. It’s not just the patient, sometimes it’s the carer’s.

ES1

What’s the longest you have seen a patient for?

I think the longest would be probably 3 years.

And that would be you personally or supervising other team members to provide that support?

That would probably be me personally, but it wouldn’t be weekly because we have got to have a throughput.

BS19