Reablement services for people at risk of needing social care: the MoRe mixed-methods evaluation

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 13/01/17. The contractual start date was in October 2014. The final report began editorial review in November 2017 and was accepted for publication in September 2018. 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.

Declared competing interests of authors

none

Disclaimers

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

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Beresford et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

Background

Increased life expectancy, brought about mainly by improvements in health care, presents a number of health and care policy challenges. Increased rates of hospital admissions due to long-term health problems or frailty, subsequent delays in discharge from hospital, and growing demands for social care caused by heightened rates and levels of dependency have placed significant demands on services.

The 1990s saw the development of models of care to address these challenges, which, collectively, came to be designated as ‘intermediate care’. 1,2 This concept became formally recognised and defined in the National Service Framework for Older People,3 published in 2001:

. . . a new layer of care, between primary and specialist services . . . to help prevent unnecessary hospital admission, support early discharge and reduce or delay the need for long-term residential care.

p. 13. 3

© Crown copyright. Contains public sector information licensed under the Open Government Licence v3.0

In response to the National Service Framework, and supported by £900M investment from the UK government, new models of care, or practices, emerged.

The health-care sector saw the development of admission avoidance and supported early discharge schemes, typically described, or defined, as intermediate care. Importantly, some – but not all – supported early discharge schemes specifically sought to restore an individual’s ability to look after themselves, perhaps independently, in their own homes.

Similarly, local authorities (LAs) began to develop interventions for individuals who presented concerns in terms of their ability to continue to stay well and live independently, or at least remain in their homes with low levels of support. 4 Importantly, these latter developments were informed by a challenge, issued by the Department of Health and Social Care and directed at LAs, to develop approaches to care that reduced dependency on services and supported individuals to ‘make most use of their own capacity and potential’. 5 These twin levers saw the emergence of services across the country that shared similar features: short-term and intensive support delivered in the home with a focus on regaining, or preventing the decline of, daily living skills and social participation. By the early 2000s this approach had gained considerable traction and government support. 5,6 The term ‘reablement’ was used to describe the approach and significant levels of investment made in the development of such provision. In some countries, a similar shift in, approach to providing social care support was also taking place; some also referred to this new approach as ‘reablement’. Others, for example the USA, Australia and New Zealand, used the term ‘restorative care’. 7

The challenge of defining reablement

The past 15 or so years have seen changing and inconsistent use of the terms ‘intermediate care’ and ‘reablement’, both within policy8 and as applied to specific health and social care provision. In terms of the latter, the criteria, or ‘labels’, applied to different government funding initiatives (typically directed via health to encourage integrated planning) go some way to accounting for this.

In the context of this study, it is not useful or necessary to recount these changes in great detail. However, this lack of a shared definition was something that had to be explicitly addressed when developing the bid for funding and the study protocol. Indeed, the lack of an agreed description of reablement using a standard intervention framework9 continues to present significant challenges to those seeking to review existing evidence and those conducting primary research. 10

Defining reablement for the purposes of the study

This study was commissioned in mid-2014, and arose from a commissioning call issued by the National Institute for Health Research (NIHR) in early 2013. From the outset, it was essential that a clear definition of reablement was established that was relevant and meaningful in terms of current policy and practice, and made reference to the concepts of intermediate care and rehabilitation. Therefore, an analysis of recent evaluative literature11,12 and current policy and practice guidance documents7,13 was carried out. It was clear from this exercise that two key characteristics distinguish intermediate care and reablement from other health and care services:

1. The objectives of intermediate care/reablement. These are – acute admission avoidance at the point of clinical need for acute care; early supported discharge after acute admission; longer-term avoidance of unplanned hospital admission; reduction in the use of home-care services; and avoidance of admission to long-term care.

2. The time-limited nature of the service offered (usually up to a maximum of 6 weeks). This is the key defining characteristic that distinguishes intermediate care or reablement from, for example, generic rehabilitation services.

A further characteristic emerged as distinguishing reablement from intermediate care, namely its restorative, self-care approach. In other words, a reablement service is about enabling people to regain or retain self-care function for themselves, rather than providing input that replaces that function (e.g. reablement teaches people how to cook for themselves again, rather than providing meals on wheels). Table 1 sets out these distinctions, and overlaps, between reablement and intermediate care.

The study’s definition of reablement

Drawing on the work referred to above, the following definition of reablement was used.

• Intervention objective:

  • to support people to regain or maintain independence in their daily lives.

• Intervention approach:

  • to restore previous self-care skills and abilities (or relearn them in new ways) that enable people to be as independent as possible in the everyday activities that make up their daily lives (e.g. cleaning the house, shopping, or bathing and dressing themselves) rather than having someone (e.g. an informal or formal carer) do things ‘to’ them or ‘for’ them. 7,8 The provision of equipment may be used to support this

  • individualised and goals-focused.

• Population:

  • individuals returning home from hospital or other inpatient care setting following an acute episode

  • individuals in whom there is evidence of declining independence or ability to cope with everyday living.

• Nature of intervention delivery:

  • intensive

  • time-limited (up to 6 weeks)

  • goals-focused

  • delivered in the usual place of residence.

This definition aligns with current policy and practice guidance. 14,15

Subsequent developments in policy and definitions of reablement

While the study was under way, the Care Act 201415 – heralded as the most significant reform to social care in over 60 years – became law. Full implementation of the Act is ongoing, but phase 1 implementation had significant implications in terms of the perceived role of reablement within the wider portfolio of social care provision.

First, reablement was presented as one of the core interventions that can delay, or reduce, demands for care services and keep individuals living independently in their own homes. To this end, LAs are now required to consider providing reablement before, or alongside, carrying out a needs assessment. Furthermore, this approach should now be considered both for individuals not previously known to adult social care and for existing users. Second, reablement was presented as an intervention falling under the umbrella term of ‘intermediate care’, and was specifically identified as having the function of helping individuals leave hospital in a timely way and regain their independence. The need for integrated working with health services to deliver this was made explicit.

Together, these elements of the Care Act 2014 – and the continued growing and significant concerns within the NHS about the delayed discharge of older patients16 – have seen the emergence and adoption of ‘discharge to assess’ pathways, with these pathways including discharge to social care for assessment and reablement. 17

The 2017 National Institute for Health and Care Excellence (NICE) guidance, Intermediate Care Including Reablement14 makes clear that, as an intervention, reablement is regarded as one of the four core elements of ‘intermediate care’, the other three being (health) crisis response and bed- and home-based intermediate (health) care. It is important to note that the guidance continues to represent reablement as being distinctive in its restorative approach and focus on supporting independence in self-care and everyday life skills, with social care leading on its delivery. The guidance also stresses the need for integrated working within the context of providing intermediate care.

These significant changes in health and social care policy and guidance had two important implications for this study. First, it meant that changes to services and service developments were happening during the study period (2014–17), and are ongoing. Second, it signals that, as an intervention, reablement is highly topical and likely to remain a core aspect of meeting the health and social care needs of older people. As a result, the findings from this study are highly relevant and timely.

Existing evidence on reablement

The 2017 NICE guideline14 offers a useful review of existing evidence. It concludes that the quality of existing evidence on the effectiveness of reablement is not as high as for other forms of intermediate care (e.g. home- or bed-based intermediate care). To date, there have been just three comparative evaluations, of variable quality, that have used randomisation,18 none of which is UK-based. Among another four non-randomised comparative evaluations,19–22 two were carried out in England, both of which reported in 2010. 19,20 On the basis of this set of evidence, the following conclusion was drawn:

There is a moderate amount of moderate quality evidence that reablement is more effective when compared with conventional home care.

NICE Guideline 74, p. 137. 14 © NICE 2017 Intermediate care including reablement. Available from www.nice.org.uk/guidance/ng74 All rights reserved. Subject to Notice of rights NICE guidance is prepared for the National Health Service in England. All NICE guidance is subject to regular review and may be updated or withdrawn. NICE accepts no responsibility for the use of its content in this product/publication

It was noted that the evidence was most consistent and positive with respect to care needs or impact on service use. Similarly, overall, findings related to quality of life and ability to carry out activities of daily living (ADL) suggest that reablement is an effective intervention, although this was not found in a low-quality trial. 11,12 NICE also reports that it did not find any evidence on cost-effectiveness that was relevant to the UK.

A separate set of evidence concerns issues of the design of reablement services. Here, low- to moderate-quality evidence indicates that access to particular specialisms (e.g. physiotherapy, occupational therapy) may influence the effectiveness of reablement. 19,23,24 In addition, there is reasonable evidence to suggest that the skills of reablement workers may have an impact on outcomes,24,25 for example being able to judge the timing and degree of support to offer a service user when carrying out a task and increasing service users’ confidence and motivation. Two studies conclude that reablement services should include the ability to respond to users’ goals that concern their social or leisure lives, which, for some users, will be a high priority or have the greatest impact on their lives. 25,26 Finally, there is some indication that service models should incorporate the ability to be flexible and responsive to user needs and progress. 24 However, this may have implications for service users’ views and experiences of the service. 23,24

Some existing studies have investigated issues related to the characteristics of the service user. There is reasonable evidence that individual motivation may have an impact on the effectiveness of reablement. 24,25 More specifically, there is weak- to moderate-quality evidence, primarily based on practitioner views, that individuals with end-of-life care needs or complex needs should not be referred to reablement services. 19,20,27 No studies have investigated the effectiveness of reablement in people living with dementia.

Finally, in terms of service users’ and family members’ views and experiences, the key issue identified by existing research is the potential lack of understanding of the objectives of reablement, and the fundamental difference of approach compared with home care. 23,24,26,27

A number of other systematic reviews of reablement were published in 2016/17. 10,28–30 Their inclusion criteria (in terms of study design and quality) vary and, as a result, the conclusions drawn differ somewhat. However, all note the pressing need for further research as a result of the core place given to reablement within health and social care policy, particularly with respect to older people and the investment in services delivering this intervention.

Evidence gaps

Studies into the effectiveness of reablement per se are beginning to be reported, although, as noted above, more are required. A further gap in evidence concerns the way in which to deliver reablement. Since the early days of ‘reablement’ in the late 1990s, different localities and sectors have developed different service models by which to deliver this intervention. 7,14,31 Thus, in addition to evaluating the impact of reablement with a no-intervention comparator group, a complementary stream of research is required that looks at different approaches, or service models, to providing reablement. In addition, although there is some evidence on factors that may have an impact on intervention outcomes (e.g. service user characteristics such as motivation/engagement, reason for referral to reablement, comorbidities and living circumstances), clearly, further work in this area is required and particularly with respect to factors that are amenable to intervention. Finally, and also highlighted in the 2016 NICE guidance,32 increasing the evidence base on the costs and cost-effectiveness of reablement and providing reablement to people with dementia are both important priorities.

Study aims and objectives

This study was funded by NIHR in response to a commissioned call33 that sought research proposals addressing the following overarching questions: how effective are reablement services in enhancing self-care and independence in the population they are designed to cover; and how are they best delivered?

When responding to this call, we chose to focus on some of the specific evidence gaps discussed in the previous section. Table 2 presents the study aims and associated objectives as set out in the study protocol. 34

Thus, the focus of the study was not the effectiveness per se of reablement. Rather, its aim was to evaluate and compare existing reablement service models and to also conduct a smaller, parallel piece of work focused on reablement for groups for whom adjustments to a generic model of provision may be required (e.g. young adults, people with dementia).

Structure of the report

The report comprises 10 chapters. Chapter 2 provides a high-level overview of the study and reports deviations from the original protocol. Chapter 3 reports on work package (WP) 1, the national survey of reablement services and the reablement service models derived from that work. Chapter 4 describes the selection and recruitment of research sites for WP2 and provides a description of the characteristics of each site. In Chapter 5 we report the outcomes evaluation (WP2a). Chapter 6 describes one element of the process evaluation (WP2b), namely the interviews with reablement staff. In Chapter 7 we report the second element of WP2b: the user perspective. The economic evaluation (WP2c) is reported in Chapter 8. Chapter 9 turns to WP3, the qualitative study of practitioner views and experiences regarding reabling people with dementia. Chapter 10 discusses the implications of the study findings for health and social care, and recommendations for future research.

Introduction

This chapter provides a broad overview of the design of the study and any deviations from the protocol. A detailed description of objectives, design and methods are described in the chapters reporting the various aspects of the study. The original protocol for the study has been published. 35 However, deviations from this protocol were required, all of which were discussed with the project’s Study Steering Committee (SSC) and approved by NIHR.

Study design and structure

The study comprised three WPs:

• WP1 – mapping reablement services and developing a typology of reablement service models.

• WP2 – a mixed-methods comparative evaluation of up to four reablement service models, as identified in WP1, investigating outcomes, costs, cost-effectiveness, and service user and practitioner experiences.

• WP3 – an investigation into specialist reablement services/practice approaches, and the rationale for adjustment made to generic provision.

A survey was used to map reablement services (WP1). An observational study design was used to generate quantitative and qualitative evidence regarding the outcomes, effectiveness and cost-effectiveness of different reablement service models and the factors affecting individual outcomes (WP2). Qualitative methods elucidated descriptions of practice and practitioners’ views regarding the provision of reablement to people with dementia. WP1 preceded WP2 and WP3. WP2 and WP3 ran concurrently. The study was conducted between October 2014 and November 2017.

Ethics considerations

Work package 1 was defined as a service audit by the Health Research Authority and did not require ethics approval. For WPs 2 and 3 ethics approval was obtained from the North East York Research Ethics Committee (REC), UK (REC reference 15/NE/0299). Substantial amendments arising from deviations from the original protocol were approved by this committee.

Public and service user involvement

We worked with public and service user representatives throughout the project. During the project design phase, we consulted our research unit’s longstanding Adults, Older People and Carers Consultation Group about the proposal we were developing. We also sought advice from professionals who represent voluntary sector organisations supporting people who use reablement services. These discussions highlighted several issues that informed our decision-making:

1. the need to track the use of health/community/social services and not just hospital services

2. the belief that in-house and outsourced services may be different, with in-house teams regarded as ‘care driven’ and contracted-out services perceived as ‘profit driven’

3. to compare, if possible, NHS and LA provision

4. the importance of exploring the ethos/philosophy used by service managers

5. the importance of examining the impact multiple impairments have on outcomes and how these are addressed by reablement services.

Throughout the project we updated, and sought feedback from, this group. This helped us to be certain that the aims of the project remained important to the public, as well as to service commissioners and providers, and that the outcome measures were relevant to service users.

The SSC included service users who had experience of reablement services and representatives of voluntary sector organisations supporting people who use reablement services. Throughout the project, the SSC provided advice on methods, research materials and project management. The service user and public representatives were contacted between meetings for feedback on recruitment and data collection materials. The SSC also provided a forum for the research team to discuss their initial research findings and consider their implications. Members were kept up to date between meetings with a quarterly newsletter.

Work package 1: mapping reablement services and developing a typology of reablement service models

The aims of WP1 were to generate ‘stand-alone’ evidence on reablement services in England and to develop a typology of reablement service models. This typology was then used to identify and select (up to) four service models for evaluation in WP2. It was also used to provide preliminary evidence and inform sampling decisions for WP3. Survey methodology, comprising a three-stage process, was used:

1. identification of all reablement services in England

2. identification of a key informant(s) in each reablement service

3. data collection from these key informants.

Work package 2: an evaluation of different models of providing a generic reablement service

The purpose of WP2 was to evaluate different reablement service models in terms of service user outcomes, the experiences of delivering and receiving reablement, and the relative costs and cost-effectiveness of the models. It comprised three elements:

• WP2a – outcomes evaluation

• WP2b – process evaluation

• WP2c – economic evaluation.

Samples for WP2b were drawn from the WP2a sample (service users) and WP2a research sites (staff). Service user-reported data for WP2c were collected within WP2a data collection processes.

We encountered significant issues with WP2 with respect to:

1. recruitment of research sites

2. study set-up in research sites

3. in some research sites throughput was much lower than expected, which significantly affected recruitment.

The first two issues caused considerable slippage in the project timetable. Given this, and coupled with the very slow rates of recruitment in two research sites, the decision was made not to extend the study until sample size requirements had been achieved across all research sites. As a result, the study closed early in three sites and recruitment was not started in the final site. Deviations from the protocol are detailed later in this section.

Work package 3: an investigation into specialist reablement services/practice approaches

The aim of WP3 was to investigate the organisation and delivery of reablement services to people with specialist needs.

The findings from WP1 indicated that the majority of specialist provision or practice concerned people with dementia. Therefore, with support from the SSC, and with agreement from NIHR, WP3 focused exclusively on this population.

This WP comprised:

• a case study of ‘adapted or extended practice’ within generic reablement services and of specialist provision across 10 case sites

• an investigation into the costs of such provision in each site.

Qualitative interviews with service leads and reablement workers were used to investigate the approaches, service structures, practices and experiences of providing reablement to people with dementia. Data on service costs were collected via a structured questionnaire administered during the interview with service leads. Data were collected from January to July 2016.

Introduction

The main purpose of WP1 was to provide a current picture of reablement provision in England. This included describing the organisational, structural and skill mix features of reablement services and the scope of reablement being delivered, and then looking at how these features affect intervention objectives, operating practices, referral routes, assessment tools and processes, outcome measurement and destination following discharge. It also sought to establish the costs of reablement.

Findings informed a number of elements of the evaluation WP (WP2), including selection of sites, decisions regarding the selection of explanatory variables within the outcomes analyses and the topics covered in the process evaluation. WP1 data were also used to identify the focus of, and thereby potential services to approach for, WP3.

Methods

Findings

The survey findings are reported in three sections:

• the core characteristics of reablement services

• service provision and delivery

• the costs of reablement.

Data not presented in the text are provided in Appendix 2.

Summary

Reablement services varied in terms of their organisational base (LA vs. NHS vs. integrated service); over half were LA services. Just over half were ‘standalone’ services, as opposed to integrated within wider intermediate care provision in the locality. The latter was more likely to be observed if the service was based in the NHS. Most services were delivered by in-house teams. Where aspects of the service were outsourced, this was most likely to be the hands-on delivery of the intervention, as opposed to the assessment processes.

Services could be grouped according to the types of staff working in the service. Not all services had OTs, and LA-based services were most likely to have them. NHS or integrated services presented the most multidisciplinary profile. It appeared that workers in some services had both home care and reablement clients. The scope of reablement input varied between services. Some did not work on aspects of functioning external to the home environment: we labelled this ‘functional reablement’. ‘Comprehensive reablement’, which included reabling with respect to getting out and about outside the home and re-engaging with social activities and friends, was, however, the more common approach. The primary objective of services, and particularly among LA services, was to restore everyday living skills. Multidisciplinary teams, which tend to be based in the NHS, were most likely to report ‘preventing hospital admission’ as an objective.

Services varied in terms of the clients they worked with. Over one-third offered reablement to all referrals for home care received by their LA. Just over one-quarter accepted self-referrals. The majority of services reported that they accepted a wide range of people. The most commonly reported exclusion criteria were individuals requiring end-of-life care and for whom there was no evidence of reablement potential. Over one-quarter of services reported that cognitive impairment and/or dementia could be a reason to refuse a referral.

The typical duration of reablement reported was 4–6 weeks. Services delivering comprehensive reablement were more likely to report a longer duration than those delivering functional reablement. Most services reported that they were able to extend the period an individual remained in the service, but this was predominantly attributed to delays in arrangements for longer-term care. Charging policies for extended involvement varied. LA-based services were most likely to charge.

A small minority of services used standardised outcome measures. Professional judgement with respect to achievement of reablement goals appeared to be the predominant approach to assessing progress. Almost two-thirds of services reported that, on discharge from reablement, only a minority were transferred to home-care services. Services providing functional reablement were least likely to report this as the typical outcome at discharge.

The section of the survey on costs was poorly completed. When sufficient data were available, the average cost of reablement per case was calculated at £1728 (excluding six highly atypical cases). There was relatively little difference in cost between services providing functional reablement and those providing comprehensive reablement.

Introduction

This chapter reports the selection and recruitment of services to WP2, each representing one of the service models identified by WP1. Each service is then described in comparison with the others. These descriptions are based on data collected during WP1, information gathered at meetings with services during site recruitment and study set-up, documentary material, and interviews with staff and commissioners within the process evaluation (WP2c).

Identification and recruitment of the case sites

Work package 2 was a mixed-methods evaluation of up to four reablement services, each representing a different model of delivering reablement, using an observational study design. Identification of these service models was a key output of the national survey of reablement services (WP1) as reported in the previous chapter. Two concepts emerged from our analysis of the survey data that best served to distinguish reablement services: the type of reablement input, and staffing (see Chapter 3, The core characteristics of reablement services).

In terms of the type of reablement input, the survey identified three types: ‘functional’, ‘comprehensive’ and ‘social’. (The last was very unusual and not relevant to WP2 as it is restricted to mental health settings.) The majority of services (60.8%) represented in our survey reported that they delivered ‘comprehensive reablement’. ‘Comprehensive reablement’ fully adhered to what, at the time of the study, was understood to properly constitute a reablement intervention,15,39 and this remains the case. 7 Thus, for the evaluation phase, we sought only to recruit services delivering ‘comprehensive reablement’.

The second concept that best differentiated services was staffing. Four ‘types’ emerged:

1. reablement with occupational therapy

2. home-care reablement

3. multidisciplinary reablement

4. reablement workers.

For the evaluation, therefore, we sought to represent services that were delivering comprehensive reablement but differed in terms of staffing.

Outputs of the WP1 cluster analysis of staffing typologies were used to identify services to approach regarding participation in WP2. Services closest to the centre of each cluster (in other words, best representing each staffing typology) were approached first regarding their participation in the evaluation. If the invitation was declined we moved on to the next ‘closest’ service. This continued until the involvement of a service was secured. Hereafter, we refer to these services as our research sites.

Unfortunately, significant delays experienced by the project led to the decision to close, rather than extend, the study. This meant that recruitment did not open in the research site that would have represented the ‘reablement worker’ typology.

The work package 2 research sites

The checklist of features for describing complex interventions developed by Dorling et al. 40 was used to determine the range of information reported on research sites. This information is presented in the following sections. We note that the level of detail is constrained by the need to ensure the anonymity of the research sites. The information presented is correct for the time when recruitment to WP2 was open and the intervention was being delivered to study participants.

Summary

This chapter describes the services that acted as research sites for the evaluation WP (WP2). It demonstrates the range of service and practice characteristics about which services may differ, many of which were explored in the outcomes and economic evaluation (WP2a and 2c). It also reveals that different service models may operate within a single commissioning locality. These findings highlight the importance of understanding a wide range of service and practice characteristics when conducting evaluations of complex interventions. Finally, we note that all of these services had changed in some way between the national survey and WP2 commencing (a period of around < 1 year), reflecting, we believe, changes associated with the implementation of the Care Act 2014. 15

Introduction

This chapter reports the outcomes evaluation (WP2a). The difficulties encountered during the execution of WP2 (see Chapter 2, Work package 2: an evaluation of different models of providing a generic reablement service), and its impact on sample sizes achieved across research sites, mean that the findings should be considered as indicative. However, they are important given the dearth of existing evidence, particularly in terms of the exploration of the impact on service user, intervention and service characteristics on outcomes, and findings regarding the different outcomes assessed.

Objectives and study design

The objectives of WP2a, revised in response to lower than expected sample sizes in research sites, were:

• to describe changes in outcomes between entry into and discharge from reablement

• to describe outcomes at 6 months post discharge, compared with outcomes at discharge

• to explore whether outcomes at discharge are associated with individual and/or service/intervention delivery characteristics.

Study design

An observational study of a cohort of individuals receiving reablement from one of three reablement services across England in which outcomes were tracked from entry into the service, at discharge and at 6 months post discharge.

Methods

Study recruitment and retention

Recruitment, taking place over a 6-month period (between autumn 2016 and late spring 2017), was a three-stage process.

1. Assessor seeks consent to contact: during the assessment of a new referral, carried out within 1–3 days of referral, the assessor briefly introduced the study, handed over the study information sheet and requested permission for the LSO to contact the service user by telephone to discuss participation (‘consent to contact’). Research sites were instructed that ‘consent to contact’ should be sought from all eligible referrals. The outcome of that discussion was recorded on a pro forma, and, for those not agreeing to ‘consent to contact’, the reason for this decision was recorded.

2. LSO contacts service user: service users agreeing to be contacted by the research team were contacted by telephone by the LSO. During the call, further explanation was given regarding study participation and, if agreed, arrangements were made for a home visit.

3. Home visit by the LSO: during the home visit, consent was taken and T0 data collection was undertaken.

Figure 1 provides an overview of the recruitment process and outcomes. Recruitment materials are available (see Report Supplementary Material 2). A discussion of our experiences of recruitment to and retention within the study, and their implications, can be found in Appendix 5.

FIGURE 1.

Recruitment process and retention: T0 to T1.

Recruitment

In total, 498 individuals across the three research sites were invited to give ‘consent to contact’. Over half (n = 276, 55.3%) agreed. Of these, 198 (71.7%) agreed to a home visit once they had been contacted by the LSO. A total of 186 individuals (representing 67.4% of those who gave ‘consent to contact’ and 37.3% of those invited to give ‘consent to contact’) were recruited to the study and had baseline data collection completed.

Retention at T1

The reablement service notified the local study team when a study participant was being discharged. This prompted the LSO to recontact the study participant to request a home visit to conduct T1 data collection. Unfortunately, in two research sites, on 34 occasions the service failed to notify the local study team about a discharge (18.3% of the T0 sample). These individuals were retained for T2. Among the remaining study participants, retention at T1 was good (84.2%). Thus, T1 data collection was achieved for 128 out of 186 study participants. Figure 2 provides an overview of attrition and retention between discharge and T2.

FIGURE 2.

Retention: T1 to T2.

Retention at T2

In total, 162 individuals were retained for follow-up 6 months after discharge from the reablement service. Unfortunately, 46 individuals could not be followed up because of the closure of the study. However, we attempted T2 data collection with the remaining 116 study participants. In the original study protocol we had planned to use home visits again to collect T2 data. In two sites, owing to study closure, local study teams were no longer in place. As a result, we administered the research instruments by post. In the third site (site C), for those recruited early in the study timeline, it was possible for the LSO to attempt to collect T2 data via home visits (n = 21). A T2 sample size of 64 out of 116 was achieved (55.2%). Retention was higher when home visits were used to collect data than when postal administration was used (91% vs. 51.8%). In total, the number of study participants for whom we had complete T0, T1 and T2 data was 52 out of 186.

Loss to study through death or permanent move to residential care

According to the information available to us, very few participants were lost to the study because they had died (n = 10) or permanently moved to residential care (n = 1). We would note, however, that these may also be the reasons why we failed to re-establish contact with some study participants at T1 and T2.

Analytical plan

The statistical software Stata^® 14.2 (StataCorp, College Station, TX, USA) was used.

Results

Summary

The smaller than expected sample size, particularly in two research sites, means that the data presented should be considered as exploratory.

Improvements on all outcomes – health-related quality of life (EQ-5D-5L), social care-related quality of life (ASCOT-SCT4), practitioner-reported functioning (Barthel Index) and self-reported functioning (NEADL scale) – were observed at discharge from reablement (T1) compared with entry into the service (T0). For the subsample for whom data were collected at both time points, improvements in health-related quality of life, social care-related quality of life and practitioner-reported functioning were statistically significant.

Outcomes had further improved at 6 months post discharge, but the improvements in health-related and social care-related quality of life were not statistically significant. However, the improvement in self-reported functioning was statistically significant, a contrast with the degree of improvement at discharge, which was not significant. Barthel Index scores were not available at this time point. Using the GHQ-12, we also looked at mental health outcomes at 6 months’ follow-up. A (non-significant) improvement in the mean GHQ-12 score was observed between T1 and T2.

A set of regression analyses explored, in a preliminary way, the association between service- and individual-level characteristics on outcomes at discharge from reablement. There was some evidence of a positive association between intervention engagement, better mental health and having sufficient money and health-related quality of life at discharge. There was weaker evidence of a positive association between this outcome and sharing the home with others and self-reported functioning at entry into reablement. There was also weak evidence of an association with two service characteristics: receiving reablement from a single team (vs. multiteam) and from in-house (vs. outsourced) reablement workers.

In terms of social care-related quality of life, there was some evidence of association between this outcome at discharge and a number of individual (e.g. sufficient money, referral reason, mental health at entry, engagement with reablement) and service characteristics (e.g. intervention fidelity, duration of reablement). A weak association between this outcome and informal carer involvement and in-house versus outsourced reablement workers was also found.

With respect to practitioner-reported functioning at discharge, there was some evidence that the following individual characteristics were associated with more positive outcomes: having sufficient money, being referred for reablement to support return home from hospital (as opposed to support remaining at home) and engagement with reablement. There was also some evidence of a positive association between self-reported functioning at discharge and health-related quality of life at entry into the service and practitioner-reported functioning. There was also some evidence that informal carer involvement was negatively associated with this outcome. In addition, there was weak evidence that living alone and better mental health at entry into the service were associated with better practitioner-reported functioning at discharge. In terms of service characteristics, there was some evidence that practitioner-reported functioning at discharge was positively associated with receiving reablement from a single team. There was also weak evidence that receiving reablement from in-house, as opposed to outsourced, reablement workers was positively associated with practitioner-reported functioning at discharge.

Finally, in terms of self-reported functioning at discharge, there was some evidence of a positive association with engagement with reablement. There were a number of other individual characteristics for which there was weak evidence of a positive association with self-functioning at discharge. These were health-related quality of life and practitioner-reported functioning at entry into the service, living alone, having sufficient money, being referred to support a return home from hospital and being younger. There was some evidence of an association between self-functioning at discharge and two service characteristics: the reablement service being formed of a single team and having in-house reablement workers.

Introduction

This chapter concerns the experiences of staff directly involved with delivering reablement, both those in reablement assessment teams and those who do the hands-on reablement work with service users. Investigating staff views was one element of the process evaluation (WP2b), the other being research with service users and family members. The purpose of this aspect was both to understand the experience of providing and delivering reablement and to elicit views regarding the outcomes of reablement and the factors that have an impact on outcomes.

Methods

Focus groups were held with assessors in each site and with reablement workers in two sites (the third site, an outsourced service, declined to participate). Data collection took place between April and July 2017.

Findings

We report the findings under five key themes:

1. the impact of the Care Act 201415 on reablement services and practice

2. the impact, and suitability, of reablement as an intervention

3. user and family understanding of reablement

4. individual and contextual factors having an impact on reablement outcomes

5. practice and delivery issues.

Summary

This chapter has reported on findings from focus groups with staff working in the reablement services acting as research sites for the study’s evaluation WP (WP2). Interviewees included those with an assessment and review role and reablement workers. Staff based in-house and outsourced teams were represented. The interviews explored experiences of delivering reablement from a number of perspectives.

Both assessment staff and reablement workers discussed the impact of the Care Act 2014,15 and also NHS discharge-to-assess policies, on the characteristics of their caseloads, with reabling to full independence no longer the predominant outcome. At the same time, reablement was still regarded as a relevant and effective intervention for individuals with support needs associated with staying in their own homes, or returning home following an inpatient episode.

A key impact of merging assessment and reablement services was the significant increase in the number of referrals, which some services were struggling to manage. This appeared to be exacerbated by a lack of sufficient home-care provision (typically outsourced) or a refusal by providers to accept referrals for onward transfer of individuals needing ongoing support. This had the impact of slowing the service throughput.

There was a strong and consistent belief of the superiority of reablement over ‘traditional home care’. The skills and knowledge of staff to adopt a reabling approach were regarded as vital to the success of the intervention. In addition, staff believed that the frequent social contact resulting from home visits by reablement workers served to reconnect, and rekindle, service users’ interest in everyday life.

The issue of user and family understanding of reablement was raised across all focus groups. A lack of understanding of the difference between reablement and home care was regarded as a barrier to the effectiveness of reablement and distracted staff time away from delivering the intervention. Staff reported that they had been unsuccessful in finding ways to address this issue.

Aside from the issue of user, or family, understanding of the intervention, a number of other factors were identified as having an impact on reablement outcomes. These included service user characteristics (e.g. personality, problem-solving skills, mental health, cognitive impairment, the presence of a partner/spouse). Housing was also identified as a potential barrier to success, as installation of aids or minor modifications to the home could be difficult to achieve if the individual rented their home.

A number of aspects of service delivery and practice were also identified as having an impact on the effectiveness of reablement. Flexibility in the timing and duration of visits, and an expectation that this would reduce over the course of reablement, was noted. However, some staff believed that this was not sufficiently incentivised in contracts with outsourced providers. Adequate monitoring of an individual’s progress and supervision of reablement workers was regarded as critical and more difficult to achieve in services where assessors and workers were based in different teams. There were mixed views about the optimum number of workers who should be assigned to a case: advantages of having one or two or, alternatively, multiple workers were both articulated. Finally, access to specialist expertise, particularly physiotherapy and mental health, was regarded as important, but difficult, if not impossible, to achieve in a timely way.

Introduction

This chapter reports the second element of WP2b, the study’s process evaluation of reablement, which involved interviews with service users and family members.

Methods

Findings

We present the findings from the interviews with service users using a ‘chronological’ framework: presenting views and experiences encountered during an individual’s journey through reablement and life since being discharged. Although we tested and looked for discernible differences in respondents’ accounts between services and sites, none were apparent. Thus, findings are presented generically and concern experiences of five different reablement services: one each from sites A and B, and three from site C. Findings from the interviews with family members are situated towards the end of this account.

Key factors affecting outcomes

Key factors perceived as having an impact on outcomes were:

• the client–reablement worker relationship

• workers’ reablement skills

• users’ confidence in the worker

• the duration of visits

• a reluctance to accept the need for support

• records of progress.

Family members’ accounts of the reablement intervention

In this section we present two case studies of reablement from the perspectives of a family member: first, a daughter and, second, a husband.

The end of the intervention

The majority of interviewees described themselves as feeling ready for reablement to end; some even reported asking for it to stop as soon as they felt that they were managing independently. However, some were encouraged to remain in the service to take full advantage of the ‘free’ support being offered. In addition, a few interviewees reported that reablement had ended unexpectedly with little or no notice. Typically, this was bound up in their understanding that the intervention was of a fixed duration of 6 weeks, suggesting an incomplete understanding of reablement as an intervention that seeks to support recovery of skills, with discharge being linked to this.

Some interviewees had moved on to home-care packages following discharge from reablement. For those eligible for statutory support, this had been a seamless transfer with the same agency now providing home care. Others had had to source and fund their continuing care. Interviewees reported that the reablement service had provided them with information about local care agencies, but a few people had experienced difficulties finding an agency with the capacity to take on their care.

Feelings were mixed when reablement came to an end. Those who were able to function independently again were pleased that they no longer had to plan their day around workers’ visits:

It became more of a nuisance to me ‘cos I have things to do here.

S1

Oh thank goodness, to be honest, yeah, I can get on with me life now [laughs] because I was invited out and I thought I, I can’t go out ‘cos those carers are coming.

C1

Those who still needed help with washing and dressing or preparing a meal were thankful when ongoing care was found.

On the other hand, many of those who had enjoyed the company of the reablement workers and grown close to their workers described feelings of sadness and loss that they would not be seeing them again:

Oh well I was sorry to see them go, not because I didn’t feel better . . . you get to know people a little bit and I thought oh they won’t be coming in, you know.

A1

I looked forward to them coming and I missed them when they went because they’d been very . . . friendly . . . you know, they, they didn’t stay long, obviously, ‘cos they’ve got other people to go and see to, but they always did make 5 minutes’ time for chat.

N1

For some, their feelings were tempered by gratitude that reablement had been made available to them, and by a desire not to divert valuable resources away from those in greater need:

I didn’t want to be that person that was stopping someone else coming out of hospital, I wasn’t that bad.

M1

. . . I wanted them to go because I felt that there was other people needing it more than me [laughs] type of thing, you know.

P1

Interviewees’ current situations

At the time of interview (between 2 and 21 weeks post discharge) most interviewees were managing without home-care support. Specialist equipment (e.g. dressing aids, walking sticks/frames, wheeled trolleys and grab rails), most of which had been brought in during reablement, played an important role in this. As noted above, a few had moved on to long-term care packages.

Many, however, described relying on family, friends and neighbours for help with shopping, cooking and other household tasks and for providing regular social contact and emotional support. A few also paid privately for help with housework or gardening or for meals-on-wheels services.

Views about service improvements

Interviewees were generally very positive about the reablement service they had received. Not unexpectedly, therefore, there were few suggestions about ways to improve the service. The suggestions that were made can be organised into three themes:

1. workforce issues

2. information at the start of the intervention

3. scope of reablement.

The majority of suggestions centred on the number of workers within a service and the need to bolster the pool of reablement workers. More workers, it was felt, would help to address the issue of the unpredictable nature of the timing of visits and visits that were rushed because of pressures of work. In terms of the pool of workers, it was suggested that more should be done to recruit men into the workforce. Indeed, one-third of male interviewees said that they preferred the company and attention of male reablement workers to that of female ones.

In a few instances, there was some evidence of confusion about when the reablement service would start following discharge from hospital. For example, a few reported that they had expected a reablement worker to call the day they had been discharged from hospital and had felt let down that the worker did not arrive until the following day or later.

The final suggested area of service improvement centred on a desire for the intervention to be used with respect to activities outside the home, for example going for a walk or catching the bus to the local shops and services. Some interviewees believed that help with these activities was just as necessary in enabling people to remain living independently in their own homes. Interviewees expressing these opinions said that it would be helpful if reablement services could be tailored to their individual needs and priorities.

Summary

This chapter reported the experiences of individuals who had received reablement. The experiences of the two family members we were able to recruit to the project, a spouse and a daughter, were also described.

The majority of interviewees, at least at the start of reablement, did not fully understand the nature of this intervention, regarding it instead as home care. The exception to this was individuals who had received reablement in the past. Some aspects of dissatisfaction with the service they had received can be attributed to this misunderstanding. Aside from this, the main source of dissatisfaction concerned unpredictable visit times and ‘bedtime’ visits happening too early in the evening.

Overall, the great majority of interviews reported positive outcomes in terms of the core objective of reablement, namely independence. Those not reporting these gains all had chronically poor health. That said, they had valued the support they had received. Four (connected) aspects of practice were identified as having supported reablement outcomes: first, the provision of personalised and enabling assistance; second, advice on alternative ways to carry out tasks and activities; third, building confidence; and, finally, interviewees spoke of the comfort of knowing that someone was aware of them and their situation and would be visiting them. It is important to note that few could recall goal-setting and it was not clear from service users’ accounts whether or not they had been involved in identifying intervention objectives. An additional outcome experienced during the intervention was companionship, and some described this as the most valued outcome.

Six factors were identified as having an impact on outcomes: the service user–worker relationship, workers’ reablement skills, service user’s confidence in the worker, duration of home visits, willingness or reluctance to accept support, and being able to review records of progress. Continuity of workers was not identified by interviewees as important to the outcomes achieved. Furthermore, those being supported by a number of reablement staff did not usually report this as a problem or a source of dissatisfaction with the service.

Overview

This chapter reports WP2c, which examined the resource use and costs associated with the package of care used by individuals receiving reablement. In addition to the reablement service, it considers the resource use and costs falling on the health care, social care and voluntary sectors, as well as users’ out-of-pocket costs and any use of informal care.

Methods

Results

Duration and intensity of reablement

Figure 3 shows the duration and intensity of reablement.

FIGURE 3.

Duration and intensity of reablement.

Planned duration and intensity of reablement

Information on the planned duration of reablement was available for 184 out of 186 (99%) study participants. For the majority of participants, reablement was planned to last 6 weeks (n = 170; 91%). The number of planned reablement sessions was recorded for all study participants but one. Most study participants had two sessions per day, which is equivalent to 14 sessions per week (n = 83; 45%). Over one-third had one session per day, equivalent to seven sessions per week (n = 66; 35%). Some study participants had a more irregular pattern, with 1–6 sessions per week (n = 12; 6%); a similar number of participants had three sessions per day (n = 12; 6%). Eight study participants (4%) had more than three sessions per day.

Actual duration of reablement

Information on the actual duration of reablement was available for 175 (94%) study participants, and was calculated as the time between baseline and discharge. The median duration of reablement was 4 weeks (mean 3.9 weeks; minimum 0.6 weeks, maximum 6.4 weeks). Information on the actual number of sessions was not recorded.

Summary

The analysis presented in this chapter is the first, to our knowledge, to explore the feasibility of collecting and using a broad set of costs and outcomes data in the context of reablement services in England. We undertook a comprehensive descriptive and exploratory analysis using regression analysis, to investigate the determinants of costs during reablement given the information available at baseline. This was informed by the first ever, to our knowledge, systematic review of the methods applied for the economic evaluation of reablement. 36

The tool developed to collect information on resource use and costs falling on health care, social care, voluntary services, private costs (out-of-pocket) and informal care time (SCPQ) appears to have performed well, with response rates (overall questionnaire and per question) generally high. We would note that study participants could choose to have the questionnaire administered by the LSO as a structured interview, or could self-complete it with support from the LSO. Thus, self-completion without support and/or no choice over mode of administration has not been tested.

The planned duration of reablement was, on average, 6 weeks and the actual duration of reablement was 4 weeks. Most study participants were planned to have one or two reablement sessions per day.

Services most used before receiving reablement and up to 6 months after discharge were hospital, community health care and social care. Few major house adaptations, mostly in the period before reablement, were reported. Some minor house adaptations and equipment items were reported. Most study participants received help from family and friends. A downwards trend in resource use from the period before reablement to 6 months post discharge was observed.

Public sector costs, falling on health and social care, constituted the largest cost category. Of the public sector costs, costs of hospitalisations with overnight stays were the largest cost item at all time points. Out-of-pocket costs were generally low, at a maximum of £3 per week on average. In the other costs category, the major cost item was informal care, followed by major house adaptations and equipment. From a public sector perspective, the key cost items were hospital overnight stays and social care.

Study participants who received reablement from outsourced providers were associated with higher hospital costs and lower social-care costs than those who received reablement from an in-house service.

Referrals to reablement because of a musculoskeletal problem or a fall, or because of an infection, were associated with lower costs of community health care during reablement. Furthermore, arthritis and chronic respiratory conditions were associated with higher costs, whereas cardiovascular conditions were associated with lower costs.

In terms of demographic characteristics, older study participants were associated with higher costs of community health care; by contrast, male participants were associated with lower costs. Study participants who rented their home were associated with higher costs of social care. Finally, study participants who reported having sufficient money to live on were associated with higher costs of community health care.

Finally, unit costs were sourced from national reference materials. 52,53 We were unable to find unit costs from some services, and had to generalise from similar services. We were explicit about the unit costs we used and our costing assumptions. Our extensive reporting of the use of services allows future studies, such as modelling studies, to use these data and apply other unit costs. However, the sample size limited the tests that could be undertaken, and the inferences that it is possible to draw. Specifically, we were unable to evaluate the cost-effectiveness of service models or to make robust inferences about which characteristics and factors are predictors of costs.

Introduction

There is a growing population of people who have need for reablement, which may challenge generic services and/or practice. In particular, the number of people in the UK living with dementia is expected to double in the next 30 years. 56 Evidence from previous UK studies indicates that people with complex needs (e.g. dementia, sight loss, learning disabilities) may be ineligible for support from generic reablement services, or may not experience the same level of benefit as ‘typical’ users. 20 Indeed, our own survey of services (WP1) (see Chapter 3, Service provision and delivery) found that over one-quarter of services (28%) reported specifically excluding people with dementia. Furthermore, people with dementia may also be more likely to be susceptible to commonly reported exclusion criteria, including not having reablement potential (29%), having longer-term care needs (17%) and a lack of engagement (see Chapter 3, Service provision and delivery).

However, there is growing interest in reablement for people with dementia. 7 The recent NICE guideline14 notes that people with dementia may be excluded from reablement services because they are perceived as unlikely to benefit from the intervention. NICE’s position, however, is that people with dementia should not be excluded from reablement on the basis of their diagnosis, stating:

Consider reablement for people living with dementia, to support them to maintain and improve their independence and wellbeing.

NICE Guideline 74, p. 9. 14 © NICE 2017 Intermediate care including reablement. Available from www.nice.org.uk/guidance/ng74 All rights reserved. Subject to Notice of rights NICE guidance is prepared for the National Health Service in England. All NICE guidance is subject to regular review and may be updated or withdrawn. NICE accepts no responsibility for the use of its content in this product/publication

The guideline also draws attention to the virtual absence of any evidence on the effectiveness and cost-effectiveness of reablement for people with dementia, and the views and experiences of people with dementia, their families and professionals (e.g. managers and front-line practitioners) in terms of service delivery, practice and outcomes. Indeed, with respect to the topic of views and experience, no studies were identified by NICE in the evidence synthesis work carried out to support and inform the guideline’s development. Addressing these evidence gaps is presented as one of the research recommendations arising from the development of the guideline.

Our research questions, originally formulated in 2013, address some of these gaps in evidence and are as follows:

• How are generic reablement services adapting their service model or practice to accommodate the needs of ‘specialist’ groups? What is the rationale for these adjustments?

• What has been their learning experience as they have developed this provision?

• What are the barriers to the delivery of reablement and achievement of positive outcomes for these groups in the context of generic reablement services?

• How do specialist services approach and deliver reablement? How does it differ from generic provision?

• What are the barriers to the delivery of reablement and achievement of positive outcomes for groups with specialist needs within specialist services? Do these differ from those for people with low-level needs using generic services?

• What do specialist reablement approaches or services cost?

• What specific ingredients of service and support arrangements might best support the quality of life and independence of people with specialist needs?

• What outcomes might be appropriate to assess the success of these services in a future evaluation?

This workstream was led by Parveneh Rabiee. She conducted all data collection and led on the qualitative data analysis. All members of the research team were involved in the development of data collection tools, data analysis and interpretation. WP3 was conducted during January–July 2016.

Study design and methods

Work package 3 comprised a case study of adapted or extended practice within generic reablement services for people with dementia and dementia-specific provision. In terms of the generic services, we sought to include services reporting to have a ‘dementia protocol’ and those without such a protocol. Our assumption was that having a protocol might indicate more specialist or systematised approaches to providing reablement to service users with dementia. Data were gathered from service leads and reablement workers. We also sought to collect data on the costs of dementia-specific provision or practices. Qualitative and quantitative (costs only) methods were used.

The inclusion criteria were:

• The service/practice had existed in their present form for at least 1 year, thereby ensuring that the sample would comprise practitioners who had some degree of experience of working with people with dementia and that the practice was reasonably well established.

• The service had a relatively large proportion of people with dementia within their caseload; again, the rationale for this was that interviewees would have a depth of cumulative experience.

A dementia reablement service

One of the services recruited to this WP was a service describing itself as a ‘dementia reablement service’. During interviews with staff, it became clear that the service was outside the remit of this study. We do, however, offer a brief report on this service because it is an interesting example of implementing the reablement approach for a different purpose, as is also the case for reablement for people with mental health difficulties.

Generic reablement services and people with dementia

The remainder of this chapter reports our findings with respect to the experiences and practices of generic reablement services in relation to working with service users with dementia, and how they have responded to any additional needs and challenges.

Summary

Interviews with service leads and front-line practitioners drawn from eight generic reablement services and with extended experience of reabling people with dementia yielded rich data on reabling people with dementia.

Interviewees agreed that reablement is an appropriate intervention for people with dementia. However, the stage of the dementia will influence the focus of reabling work and intervention objectives. In addition, the expectation that reablement will eliminate the need for ongoing service involvement may not be appropriate, particularly in terms of expectations around long-term outcomes. Service leads did not always feel that commissioners understood the argument for this. Interviewees also believed that adopting a reablement approach within an extended assessment of social-care needs was an effective approach and supported service user acceptance of involvement of services should this be required.

Front-line practitioners offered detailed descriptions of the ways in which reabling people with dementia may involve additional areas of work, or increase the time and attention needed in these areas. These included fostering engagement, supporting reorientation to the home environment, re-establishing routines, connecting into local support services, working with family members and, at each visit, needing to re-establish the therapeutic relationship and finding a means to corroborate that self-care or other activities had been undertaken.

Interviews explored whether or not service policies and practices were adjusted to accommodate the additional or unique needs of people with dementia who are being reabled. None had formal processes or protocols related to the delivery of reablement to people with dementia. Although all interviewees agreed that longer home visits were typically required, only two services did this. Similarly, only a minority of services routinely provided reablement for longer than the usual 6-week period, or restricted the number of reablement workers assigned to a case.

Although access to generic training on working with people with dementia was reported, none of the reablement workers reported that they had received training on reabling people with dementia. Staff working in outsourced services were most concerned about the lack of training they had received.

Finally, in selecting case sites for this WP, a specialist dementia reablement service was identified that worked with people soon after their diagnosis. Its remit excluded it from our study, but a short description of the service was included in this chapter as an illustration of applying a reablement approach in a different context.

Introduction

This chapter considers the strength and limitations of this study before moving on to discuss the study findings, placing them in the context, where possible, of existing evidence.

Strengths and limitations

To our knowledge, the national survey of reablement services (WP1) generated the most comprehensive and detailed snapshot of reablement services ever conducted. It demonstrated that organisational and practice characteristics differentiate services and it also generated another source of evidence on the costs of providing reablement. The detailed descriptions of the organisation and practice of the services acting as research sites in WP2 are a rich case study of the variability of the way reablement services are organised and delivered.

The evaluation WP (WP2) encountered a number of challenges. The recruitment of research sites and study set-up took much longer than anticipated, and severely affected the progress of the study along the original timeline. A second key issue was that the rate of referrals and/or throughput of two research sites significantly affected the levels of recruitment to the study. As a result, the decision was taken to close the study rather than extend and/or add other research sites, which would have been required to achieve all of the study objectives. The implication of this was that our data set for the outcomes and economic evaluation was much reduced. Furthermore, it was not possible to collect T2 (6-month follow-up) data from all study participants. Thus, it is appropriate to conceive the outcomes and economic evaluations as both pilot and feasibility studies, rather than the full and robust study that was intended. Despite these difficulties, however, the recruitment rate (> 37%) was good, given the nature of the population and the study design. Retention to the study was good (> 80% at T1 and > 90% at T2, when, as per the protocol, home visits were used to collect data). In addition, for the first time, the association between a wide range of individual and service characteristics, some of which are amenable to change, have been subject to an initial exploration.

The process evaluation (WP2b) was less affected by these difficulties. Service users across all three research sites were recruited, although the recruitment of family members to interview about their experiences proved extremely difficult. Rich data were generated from interviews with service users and staff; both make a significant contribution to a rather limited existing evidence base.

Our investigation into staff’s views about reabling people with dementia (WP3) recruited one fewer service than was planned and we did not achieve total samples of service leads and front-line staff. However, again, the data generated were rich and, we believe, are an important addition to the very limited evidence base on this topic.

The limitations of the outcomes and economic evaluations places considerable restraint on the extent to which implications for policy and practice can be drawn. That said, we do believe that the study has strengths and makes an important and valuable contribution to the existing evidence base. They lie in the following areas: understanding of current service design, structures and practices with respect to the delivery of reablement; research design and research methods (including implementing research in non-NHS settings); and informing the future research agenda. Although it is not possible to draw firm implications for health and social care practice from our findings, some comments, based on both our qualitative and quantitative findings, are possible. Chapter 11 sets out these contributions and reflections regarding service organisation, delivery and practice.

The organisation and practice of reablement in England

The first WP of this project (WP1) (see Chapter 3) comprised a national survey of reablement services in England carried out in mid-2015. Surveys of services are, by definition, snapshots of provision at a particular moment. Findings, therefore, have to be read in that light.

Our findings also have to be understood in relation to the decisions we made while determining our sampling framework and selecting services to approach. Most important of these was the focus on services that aimed to help people regain everyday living skills, in their usual place of residence, for a defined period of around 6 weeks, and for which no charge was made. This definition is close to the original conception of ‘social’ reablement as envisaged by its originators in the Department of Health and Social Care. 4,57 It is also close to the definition of reablement used since 2013 in the annual National Audit of Intermediate Care (NAIC). The 2014/15 NAIC (which covers the period of our survey) received responses from only 44 reablement services. 5 To the best of our knowledge, then, our survey provides the first large-scale, detailed and comprehensive picture of reablement services in England.

At the time of the survey, the organisational base of over half of services was the LA. Such services were less likely to be part of any wider intermediate care provision within the locality and were typically presented as a separate, or standalone, service. A small minority of reablement services were based in the NHS, although there were others that were based in integrated NHS/LA organisations, or organisations jointly funded by the NHS and LA. In both cases, these services were more likely to be part of wider intermediate care provision.

The majority of reablement services did not contract out any aspect of the provision to another agency. If they did, it was most likely to be the delivery of reablement support: that is, the ‘hands-on’ work of delivering the intervention through home visits.

Overall, reablement services shared the common objectives of helping people regain everyday skills, reducing the need for ongoing home (social) care and preventing longer than necessary stays in hospital. Preventing admission to residential long-term care was less often reported (by around half) and was closely associated with certain types of organisational base and structure. In terms of services’ primary objectives, most identified this to be either help with regaining everyday skills or reducing the need for ongoing home (social) care. Two services, both of which were NHS based, identified preventing hospital admission during acute illness as the primary objective of their service.

Around two-thirds of reablement services reported that they worked to reable individuals across a wide range of domains of everyday living activities that allow people to live as independently as possible. We defined these types of service as ‘comprehensive reablement’. However, one-third of services were more restricted in the way they implemented the reablement approach and we labelled these types of service as ‘functional reablement’. An additional distinct cluster of reablement input was also identified, which we referred to as ‘social reablement’. It differed considerably from ‘functional’ and ‘comprehensive’ reablement in its service characteristics and aspects of service provision and delivery. Both our survey and other literature suggest that ‘social’ reablement is most likely to be delivered by mental health services. 58,59

How services assessed service users and measured progress towards reablement varied, with most apparently reliant on non-standardised measures or professional judgement to assess their effectiveness. A surprisingly high proportion of respondents were not able to report the usual outcome of service users after discharge. Nonetheless, where this information was provided, the usual outcome most often reported was discharge without involvement of care services. This is broadly in line with results from the 2015 NAIC. 5

Overall, therefore, this survey of reablement services revealed a common objective of restoring, as much as possible, an individual’s functioning and ability to live independently. There was, however, heterogeneity in terms of where services were located, the degree to which reablement was integrated into wider intermediate care provision, the staffing of services and the extent to which services were being outsourced. In addition, not all services reported delivering comprehensive reablement, focusing instead only on functioning with respect to ADL. Finally, it was clear that the routine use of standardised measures to assess and monitor outcomes was rare.

This broad understanding of reablement services is supplemented by the much more detailed picture of the three reablement services that constituted the research sites for the evaluation WP (WP2). Chapter 4 described and compared these services and it reinforced the notion of diversity that exists in reablement service models. Thus, even within just three services, varying approaches to, or models of, providing reablement are captured in terms of both organisational and service characteristics and practice. It also highlighted that, within a single site, a number of different delivery models may be operating. Importantly, no consistent patterns of association of service characteristics were observed, something sometimes also found in the analysis of the national survey. Indeed, it was a recognition of this variety in terms of service delivery and practice that stimulated NIHR’s commissioned research call on this topic, and the questions posed.

The costs of providing reablement

The national survey of reablement services (WP1) also collected data on the costs of reablement and service caseload. The average cost of reablement was £1455. However, 61% of services did not provide information on costs, and, among those that did, there was wide variation in the answers from £20 to £3333 per case. When services with a cost per case of under £500 were excluded, the average cost per case rose to £1728. Both of these figures are within the ranges reported in the literature. Using data from a previous evaluation of reablement,20 the 2015 Unit Costs of Health and Social Care52 reports a cost per case of £2096. The 2015 NAIC59 reported a cost per case of £1484. This provides some reassurance that most of survey respondents provided accurate information.

However, collecting provided information on costs did present issues, in terms of both non-response and apparently implausible answers. We can hypothesise that respondents found the questions difficult to understand or answer, or that they did not have access to the required information. Furthermore, some answers yielded such low costs being calculated (e.g. £20 per case) that errors in survey completion may have also occurred. Overall, our experiences suggest that future attempts to collect data on the costs of reablement services should consider targeting at least some questions to finance departments, which are more likely to understand the nature of the information required and have the relevant information to hand. Furthermore, to highlight errors to the respondent, electronic surveys could be designed so the implied cost per case is calculated automatically and made visible to the respondent.

The outcomes of reablement

The outcomes evaluation was reported in Chapter 5. Given issues with sample size, particularly in two research sites, it was not possible to compare outcomes between sites, and only descriptive analyses of the total sample were therefore reported.

Improvements on all outcomes [i.e. health-related quality of life (EQ-5D-5L), social care-related quality of life (ASCOT-SCT4), practitioner-reported functioning (Barthel Index) and self-reported functioning (NEADL scale)] were observed. For the subsample for whom data were collected at both time points, improvements in health-related quality of life (index score and VAS), social care-related quality of life and practitioner-reported functioning were statistically significant.

Outcomes (excluding the Barthel Index, which was not collected at this time point) had further improved at 6 months post discharge. Compared with the mean improvement in EQ-5D-5L and ASCOT-SCT4 index scores between T0 and T1, the difference between T1 and T2 scores was smaller and not statistically significant. However, for the NEADL scale, size of improvement in mean score was greater between T1 and T2 than between T0 and T1 and, unlike T0 to T1, it was statistically significant. Using the GHQ-12, we also looked at mental health outcomes at the 6-month follow-up. A (non-significant) improvement in mean GHQ-12 score was observed between T1 and T2.

One of the objectives of the study was to compare outcomes between the research sites, each site representing a different model of providing reablement. This was not possible because of inadequate sample sizes. However, our findings still make a useful and important contribution. First, they suggest that a number of outcome measures should be used to capture the impacts of reablement. Second, the inclusion of a measure of mental well-being, an outcome not usually included in evaluations of reablement, should be considered. This is particularly pertinent given concerns about mental well-being, social isolation and ageing. 60 Third, they indicate the validity and value of capturing longer-term outcomes. Fourth, and related to the previous point, they suggest that the chronology of the impact of reablement may vary between outcomes.

Factors associated with outcomes of reablement

Reablement is a complex intervention61 with no fixed protocol; rather, its core approach is to tailor the intervention to each individual while remaining within the overall paradigm of a restorative, intensive, time-limited intervention. 62 In addition, a number of external and person-centred factors are likely to moderate or mediate intervention effectiveness. 62 These issues were clearly reflected in NIHR’s commissioning brief for this study. 33 Emphasis was given to the need for evidence on service, intervention and individual characteristics that have an impact on intervention effectiveness.

A core objective of the study, therefore, was to explore the impact of a range of characteristics on reablement outcomes. We used regression analysis to conduct initial, exploratory analyses; the sample size limited the complexity of modelling work. However, we would argue that the findings are highly useful given that the existing evidence is extremely limited. 14 We now consider the findings (described in Chapter 5) with regard to each characteristic, placing it, where possible, in the context of existing evidence.

Overall, our findings support the notion that reablement should be offered to individuals regardless of age and sex. We have not identified any studies that offer counter-evidence. We used the question ‘Do you feel that you have enough money to live on?’ as a proxy indicator of perceived financial situation. The conclusion from our analysis is that having sufficient money may affect reablement outcomes. However, the nature of this association is likely to be complex, with a number of different issues at play, including, for example, housing quality and suitability63 and general physical health. An additional explanation that has begun to be explored within rehabilitation64 is the potential impact of service users’ socioeconomic status on therapist attitudes and decision-making.

We found evidence that not having family or friends as informal carers was associated with greater improvement in some outcomes (e.g. ASCOT-SCT4, Barthel Index). Similarly, there was weak evidence that living alone was associated with a greater improvement in professional and self-reported functioning (Barthel Index and NEADL scale, respectively). However, in terms of health-related quality of life, living alone was negatively associated with positive outcomes. Overall, these findings accord with a randomised controlled trial of reablement versus usual care11 in which having a carer was related to receiving ongoing care at 3 and 12 months post intervention in both arms of the study. The wider literature offers support for the notion that living alone may affect aspects of health status in a positive way. For example, a recently published systematic review65 of adherence to exercise programmes for older people found that living alone is associated with higher rates of adherence.

Our analysis indicated that the association between referral reason and outcomes may vary according to type of outcome. In terms of social care-related quality of life, there was some evidence that ‘return home’ referrals do better than ‘remain at home’ referrals. However, the opposite was observed with respect to self-reported functioning (NEADL scale). Findings such as these reiterate the importance of assessing a number of outcomes, as the concepts captured differ.

We found that functional ability at referral into the service may have an impact on quality-of-life outcomes and functional outcomes at discharge. This aligns with findings from a randomised controlled trial of reablement. 11 We note, however, that even small reductions in dependency can have an impact on levels of need for ongoing care. Related to this, and equally important, is the evidence that maintaining some degree of independence is highly valued by older people. 66 In terms of physical comorbidities, we found no evidence of an association between this variable and health and social care quality of life at discharge from reablement, and some evidence of possible association with functional outcomes. It is useful to refer to evidence from previous studies that ongoing health issues may affect reablement outcomes, as any gains in functioning and independence may be regarded as likely to be short-lived. 14

Two previous studies24,25 report practitioners’ and/or service users’ beliefs that service users’ motivation is an important factor in successful reablement. Another67 describes service users saying that they wanted to feel that they were ‘working with’ their reablement worker, which gives a sense of active, collaborative involvement. These studies, and practice guidance, identify motivating and increasing the confidence of service users to be key aspects of the reablement worker role. 7,14,24,25,67 We have not identified any previous evaluative research exploring this concept: the lack of a measure may account for this. This study has generated, in an initial and exploratory way, some evidence (and across all outcomes) of the association between engagement and outcomes. A 2015 conceptual review68 concluded that engagement is a multidimensional construct comprising both a co-constructed process and a patient state. The authors argue that this highlights the practitioner’s role in supporting engagement, and note that work to support engagement is likely to require an individualised approach by sufficiently skilled practitioners.

Our analyses also indicate that mental health at entry into reablement was associated with health and social care quality of life at discharge and also, although a weaker association was found, functioning. We have not identified any previous evaluative studies that have explored the association between mental health and reablement outcomes. The insufficient sample size means that we have not been able to run any multiple regression analyses; these would have allowed us to explore, for example, the relationship between mental health and intervention engagement. Intuitively, this would seem an important relationship to understand. Certainly, as reported in Chapter 6, practitioners believed that mental health has an impact on the effectiveness of reablement. For this reason, staff identified the need for access to mental health expertise and input within reablement provision.

We were also able to test, in a preliminary and exploratory way, the association between some service- or intervention-specific characteristics and reablement outcomes. Aside from some relevant qualitative evidence of user and practitioner views on the importance of access to physiotherapy and occupational therapy, and ‘joined-up’ provision with the service and between other services,19,23,24 to our knowledge this is the first time such work has been carried out.

The characteristics we tested were:

• duration of the intervention

• reablement workers’ adherence to reablement principles

• whether the service comprised a single team or separate teams of assessment and reablement worker staff

• whether reablement workers were located in-house or outsourced from a private provider.

In terms of functional outcomes, there was some evidence to indicate that being looked after by a single team was associated with better outcomes. There was also weak evidence that a single team model is associated with better health-related quality of life at discharge.

In addition, there was some evidence that intervention fidelity and duration of reablement are associated with outcomes, but only with respect to social care-related quality of life.

Finally, there was some or weak evidence from the regression analyses that in-house reablement workers, as opposed to workers from an outsourced team, are associated with more positive outcomes at discharge across all outcome measures.

The wider literature on outsourcing of adult social care has raised questions about the quality of care (e.g. Hudson69). Our interviews with staff offer some further reflections (see Chapter 6, Practice and delivery issues). They noted that outsourced providers may not be incentivised, and nor is it in their ‘business’ interests, to seek to reduce levels of input with a service user. This is contrary to practice within in-house teams, where this as a clear objective. Second are issues of training and skills; a number of interviewees believed that there is greater ‘churn’ within care agencies than among in-house workers, and thus the cumulative experience and skills of delivering reablement may be less evident. Third, outsourcing arrangements may place any ongoing monitoring and review by the assessment team, and supervision and support to reablement workers, at, at best, arm’s length. That said, this may also be the case where separate in-house services deliver assessment and ‘hands-on’ reablement.

Again, we note that we can only hypothesise about the possible mechanisms at play and the connections between some of these service characteristics, some of which may have become clearer had we been able to conduct multiple regression analyses.

Finally, we remind readers that the association between outcomes and other service characteristics could not be explored because the sample sizes were too small for some categories. These were as follows: whether or not an OT was integral to the team; location of the service within an integrated health and social care organisation versus a single agency; and the policy regarding the number of reablement workers allocated per case. Based on our interviews with service users and professionals working in reablement (reported in Chapters 6 and 7, respectively), and the wider literature on issues such as continuity of care and skill mix/access to skills within the reablement service,14,19,23,24 we would argue for the importance of including these characteristics in future research, as well as those for which we have been able to conduct some initial exploratory analyses.

The impact of reablement on resource use

An economic evaluation (WP2c), running in parallel with the outcomes evaluation (WP2a) and reported in Chapter 8, explored the impacts of reablement on resource use and costs. The findings from this aspect of the study are limited in a similar way to those of the outcomes evaluation, and, therefore, conclusions should be circumspect.

To support the economic evaluation, early in this project, a systematic review of the methods applied for the economic evaluation of reablement interventions was undertaken. 36 We identified five areas of methods uncertainty common to some, if not all, of the studies. There was a lack of clarity about which outcome measure(s) it was appropriate to use, as reported in four studies. 20,70–72 There was uncertainty in the cost savings achieved from reduced hospital stays. 72,73 Six studies20,29,71,74–78 discussed the uncertainty around the impact on carers. Six studies12,20,72,74,75,79 discussed the potential benefits of subgroup analysis to determine how the individual’s characteristics affect the costs and outcomes, and to tailor reablement to the individual’s needs. Two studies20,72 suggested that more research is required on the cost-effectiveness of different service models. The findings from this review supported the objectives and design of the economic evaluation. They also provided useful information regarding the methods and scope of data collection.

Overall, the findings from the economic evaluation suggest a downwards trend in resource use from the period before reablement to 6 months post discharge. This may reflect the study participants’ recovery after a period of ill health. However, it may be confounded by the losses to follow-up if the study participants who did not take part in the follow-up were those with the greater use of services. The frequent use of informal care suggests that this was an important source of support for study participants.

From a public sector perspective, the key cost items were hospital overnight stays and social care. Given the magnitude of the cost of hospital overnight stays, future studies should aim to collect this information in detail in order to cost it accurately. In addition, costs of major home adaptations and equipment should also be collected if this is funded by the NHS or the LA, given their magnitude. From a wider perspective, informal care time should be considered.

It is difficult to place our findings on the costs of reablement in context, given the paucity of previous research on reablement. Only one previous UK study reports this;20 in it, the health-care costs during reablement were mostly driven by hospitalisations. By contrast, in our study, hospitalisations during reablement were rare. The differences may be related to the different characteristics of the two study populations and the different study methodologies. Specifically, the study sample in Glendinning et al. 20 was much larger (n = 697), had a worse health-related quality of life at baseline, a greater proportion of study participants were categorised as ‘return home’ and a greater proportion were living alone. The duration of reablement was also longer at 8 weeks (vs. the 4 weeks in our study).

Factors affecting the costs of reablement

The type of provider, the reason for referral to the reablement service, comorbidities and the demographic and economic characteristics of service users were associated with higher costs.

Thus, study participants who received reablement from outsourced providers were associated with higher hospital costs and lower social-care costs than those who received reablement from an in-house service.

In addition, study participants who were referred to reablement because of a musculoskeletal problem or a fall, or because of an infection, were associated with lower costs of community health care during reablement. Furthermore, study participants with arthritis and chronic respiratory conditions were associated with higher costs, whereas cardiovascular conditions were associated with lower costs.

In terms of demographic characteristics, older study participants were associated with higher costs of community health care; by contrast, male participants were associated with lower costs. Study participants who rented their home were associated with higher costs of social care. Finally, study participants who reported having sufficient money to live on were associated with higher costs of community health care.

This analysis was conducted on a subset of the study sample (n = 86) for whom there were complete data on all variables of interest and, therefore, its predictive power should be interpreted with caution. Large sample sizes are usually required to predict health-care costs, given the high variability of costs. 80 We cannot exclude the possibility that these associations are spurious, given the limitations of the analysis, which we highlight below.

The results suggest that the type of provider (outsourced vs. in-house) may have an effect on costs associated with the use of health and social care services. Both in the main analysis and in the sensitivity analysis, outsourced providers were associated with higher hospital costs and lower social-care costs, controlling for other characteristics. This may reflect differences in the case mix that outsourced providers tend to receive that we were unable to control for, or differences in the way that reablement is delivered and its implications for service use.

The results also suggest that there may be some income inequality in the provision of hospital and community health-care services. These costs were higher for study participants who resided in more affluent areas or who perceived that they were in a better financial situation. Income inequality in the provision of health-care services in the UK has been reported in the literature. 81 By contrast, costs of social care were higher for those renting privately than for those who own their home. This is consistent with previous research. 20

At the individual level, we found that variation in costs was mostly related to the health of study participants, reflected in their EQ-5D-5L score, health problems leading to reablement or comorbidities. Therefore, we suggest that this information is collected, not only in future research but also as part of routine data collection by reablement services.

The experience of delivering reablement

To date there have been very few studies of practitioners’ views and experiences of delivering reablement, particularly in terms of intervention effectiveness and factors which support or act as barriers to this. Thus, some findings reported in Chapter 6 of this report [which form part of the project’s process evaluation (WP2b)] concern issues or topics not previously explored or identified.

Views on the importance of service user motivation – and the need, sometimes, for skilled workers to achieve this – was consistently expressed by staff taking part in WP2b (which have been previously described). 24 Staff participating in this study emphasised the importance of flexibility in the timing and duration of visits, and the difficulties and negative impacts of working with ‘resistant families’. Both this and another study27 also report views on the importance of access to specialist knowledge and expertise, with occupational therapy and physiotherapy, as well as community nursing teams, being variously mentioned. Finally, two previous studies24,27 also report reablement workers’ accounts of finding their work very satisfying.

The issue of a lack of awareness and understanding of the reablement approach, among service users and family members, was a dominant issue across interviews and something observed in previous research with staff. 24 It was perceived as affecting intervention effectiveness because of its impact on engagement with reablement and the therapeutic relationship.

The findings from our initial and exploratory modelling analyses of the impact of individual and service/intervention characteristics on outcomes [see Chapter 5, Social care related quality of life (Adult Social Care Outcomes Toolkit – self-completed)], and our tentative interpretations of them, certainly chime with the themes and issues that emerged from our interviews with reablement staff, which serve to offer further explanatory value.

The service user experience

Before this study, there was relatively little research on the perspectives of service users and family carers on reablement services. It was a topic area that was noted in the 2017 NICE guidance14 as needing to be incorporated into future research.

Some of the issues relating to the practicalities of service delivery that interviewees describe have been previously reported, for example dissatisfaction with unpredictable timings of reablement workers’ visits and experiences of rushed visits by reablement workers that may undermine reablement objectives. 23 Interestingly, previous studies23,67,82 have reported service users’ preferences for consistency in the reablement workers who visited them; this was something we did not find and, as reported in Chapter 6 (see Practice and delivery issues), it was not something that all staff agreed was helpful.

Despite some areas of dissatisfaction with the practicalities of service delivery, as have been reported in previous studies, our interviewees were typically extremely positive about their experience of receiving reablement and the outcomes it achieved. 20,26,27 Furthermore, the role of reablement workers in motivating service users, encouraging them to take responsibility and giving them confidence, and the impact on reablement outcomes, is something that other qualitative studies with service users have also reported. 25,83

Perhaps one of the most important issues emerging from these data is a lack of complete understanding about the approach and objectives of reablement. This resulted in individuals feeling dissatisfied or confused about the support they were receiving, and for a minority this remained their dominant view of the service. In addition, there was evidence that this misunderstanding of the role of the reablement worker, and their own role in the intervention, had an impact on the extent to which the service users engaged with the intervention. Another indicator of this issue is that there was inconsistency between interviewees as to whether or not they had understood, and co-developed, the objectives of their reablement. This issue has been reported in previous studies,23,26 although in another study27 service users were described as having a good understanding of reablement and its objectives. This is potentially significant in that it shows that it is possible for service users, and family members, to have a good understanding of the intervention.

Finally, we highlight the relational element to reablement that emerged from these data. Regular home visits by, typically, a positive and friendly workforce were often enjoyable, something that previous research has also reported. 67 For some service users, the visits addressed an issue of loneliness and social isolation. Importantly, our data reveal that the nature, or quality, of relationships with their reablement workers was identified as having an impact on service users’ engagement with the intervention. This accords with other work that argues that the worker/therapist–user/patient relationship plays a significant role in securing engagement with an intervention. 68

Reabling people with dementia

Existing research evidence on reabling people with dementia is extremely limited. At the same time, there is concern that people with dementia may be excluded from reablement. 7,14 Given the growing proportion of older people with dementia, and the implications, or associated risks, of the diagnosis in terms of need for support from health and social care services, it is not surprising that the NICE guidance includes, in its research recommendations, a call for evaluations of the effectiveness and cost-effectiveness of reablement for people with dementia and research to secure the views and experiences of key stakeholders. A standalone WP (WP3), reported in Chapter 9, comprising interviews with service leads and front-line staff explored their views and experiences of providing reablement to people with dementia.

Interviewees were in agreement that the dual objectives of ‘assessment and reablement’ services, which have formed the new ‘location’ for the delivery of reablement since the Care Act 2014,15 worked well for people with dementia. It was regarded as a positive and effective approach to needs assessment. People with dementia were also consistently viewed as having the potential to be reabled, particularly those with mild or moderate dementia. However, expectations about the extent to which everyday living skills/independence were restored, and the period of time over which these would be sustained, needed to differ from those for people without this diagnosis. There was strong consensus that a diagnosis of dementia should not, on its own, preclude offering reablement.

Another difference between reabling people with and reabling those without dementia identified by interviewees was that a greater emphasis was likely to be placed on working on, and supporting positive changes in, social outcomes (e.g. social networks, activities), both for the person with dementia and for family members.

For all of these reasons, interviewees were clear that adjustments should be made regarding the perceived objectives, and indicators of success, of reablement for people with dementia. This position accords with that set out in a 2017 ‘concept paper’ on reablement in dementia, which drew on both the existing evidence base and the views of strategic/service leads, practitioners and academics. 84

In terms of the practice of reablement, interviewees highlighted some distinctions and issues. Interviewees noted that greater investment had to be made in ‘pre-reablement’ work: getting the individual ‘ready’ for reablement. This included fostering engagement, re-establishing daily living routines and, for those returning home from hospital, settling into and reorienting to the home environment. In terms of ongoing adjustments to practice, practitioners spoke about, potentially, a need to re-establish their relationship with the user at each visit. When working with service users who lived alone, it was not always possible to reliably establish which tasks or activities had or had not been done (e.g. having meals, taking medication). Finally, work with family members was often presented as part of the intervention, although, as reported in Chapter 6, family members’ input was not always helpful or supportive of reablement objectives.

Interviewees also described the adaptations to their usual practice that may be needed. First, adjustments to communication were likely to be required, or helpful. This included simplifying language, repeating instructions and using visual cues. Second, there was agreement that it was likely that the duration and intensity (i.e. length of visits) of the intervention should, ideally, be extended, and that the number of reablement workers assigned to a case should be restricted. These adjustments were being formally implemented in only a minority of services represented in this study (although we would note that this gives no indication of a wider adoption of such approaches).

A final area of findings concerned staff training. Interviewees, and specifically those who were reablement workers, felt that the training they had received on dementia was inadequate.

Desiring to place these findings in some sort of context, and given the lack of existing evidence on staff’s views and experiences of reabling people with dementia, we have looked at existing evidence regarding the physical rehabilitation of people with dementia following hip fracture. There are certainly parallels in terms of findings. For example, a systematic review of the effectiveness of hip-fracture rehabilitation in people with dementia concluded that, for people with mild or moderate dementia, they may make similar gains in function and mobility to those of people without dementia. 85 A 2016 systematic review86 of ‘current best practice’ for rehabilitation interventions of older individuals with cognitive impairment post hip fracture concludes that implementing rehabilitation interventions with this population is feasible. It also notes that innovative approaches to support engagement with the intervention may be required. Finally, an Australian-based qualitative study of health professionals87 reported a strong consensus among interviewees that access to rehabilitation should be determined by an individual’s ability, or willingness, to engage with the intervention and not assessed solely on cognitive impairment. Linked to this, interviewees in this study reported that some professionals assumed that people with dementia have limited capability to benefit from rehabilitation.

Contribution to understanding reablement services in England

The timing of this study within a changing policy landscape (in particular, the Care Act 201415 and its subsequent implementation) had both positive and negative impacts. It meant that the study was being carried out at a time of enormous change within social care for older people and intermediate care provision. Specifically, the notion that all (new) referrals for home care/support should undergo reablement alongside a comprehensive assessment process led to considerable changes in the organisation of services and the perceived role, or objectives, of reablement.

Although these changes, being implemented while we were seeking to recruit services to the study and during study set-up, created difficulties for us, they do mean that the study has generated early and detailed evidence in three localities on the ways in which LAs have responded to the Care Act and wider work on integrating intermediate care provision. Importantly, our research sites were quite diverse, with a range of differences in terms of their characteristics. We have been able, albeit sometimes only in a preliminary way, to explore the impacts of these characteristics on outcomes, user experience and practitioner experience.

In addition, this study makes a significant contribution to our understanding of the ways in which reablement services are responding to people with dementia. The findings from our national survey indicate that some reablement services may still exclude people with dementia. Our qualitative research with reablement staff has generated rich evidence on the way services are responding and, where possible, the development of practices and strategies to support engagement and positive outcomes. However, we also report concerns about the lack of training and a perceived lack of recognition on the part of commissioners to the value of reablement for this population and also, within commissioning arrangements, the adaptations that may be required to maximise its impact.

Contribution to research design and methods

In meeting the health and social care needs of the elderly population, specifically those at risk of (increased) use of social care or unsuccessful discharge from hospital following an acute episode, reablement continues to be key to the policy and practice response in England and elsewhere. Given the paucity of existing evidence, evaluative research on reablement is, and will continue to be, needed. The findings from this study are an important and useful resource for such research in terms of issues of both feasibility and study implementation.

First and foremost, the findings from this study, in terms of both the complexity of reablement provision and our (early) findings on the individual-, intervention- and service-level factors that may affect intervention effectiveness, point strongly to the value of observational studies within any body of work evaluating reablement.

Second, the findings offer insight into the individual-, service- and intervention-level characteristics that should be considered in future evaluative research.

Third, the study provides evidence, based on practitioner views, on the outcome dimensions that evaluations of the effectiveness of reablement for people with dementia should consider.

Fourth, when regarded as a feasibility study, it offers the following:

• data on the performance of, and effect sizes for, four outcome measures – two quality-of-life measures and two measures of functional ability (at discharge and 6 months post discharge)

• preliminary data on mental health outcomes

• data on recruitment, retention and attrition rates

• approaches to embedding research staff within LA services to support recruitment and timely data collection

• information, of relevance to research teams and research funders, that may help to pre-empt the difficulties and challenges of implementing a complex study in LA settings

• initial data on the effectiveness of alternative modes of administering research instruments.

Finally, the study has contributed in terms of the development of three new research instruments that we believe are highly relevant to future evaluations of reablement.

First, two new measures were created: the HRERS-RV and the ERPC. To our knowledge, no equivalent measures exist. We have noted that both would be improved by additional work; but, even in their current state, they offer a means by which user engagement in reablement and intervention adherence can be captured. Our initial findings suggest that these are, potentially, important variables in terms of their impact on user outcomes. Crucially, both are variables that are amenable to intervention by reablement services. Second, a questionnaire (the SCPQ) to assess resource use was developed and performed well.

Informing health and social care practice: some considerations

Although noting the time that has passed (and policy changes that have occurred) since the survey of reablement services (WP1) was carried out, it is valid and important to particularly highlight some findings. First, the evidence of ‘two tiers’ of reablement provision (comprehensive vs. functional) is of concern. Also concerning is that, based on evidence from our process evaluation, services that purport to provide comprehensive reablement are not necessarily doing so. Comprehensive reablement, which adheres to the original concept of reablement and pays attention to a holistic restoration of everyday living beyond simply basic functional skills, demands more from service providers, including having a knowledge of, for example, local social groups and daytime activities and being able to work with service users outside their home (should that be required). Our process evaluation suggests that this is an important issue. The lack of a more holistic approach was one of the few areas of improvement raised by service users we interviewed. Furthermore, a failure to do this meant that service users’ priorities were not necessarily attended to. From the interviews with professionals working in reablement services, workers’ knowledge of local social/activity provision was regarded as an active ingredient of reablement, and this was also stressed by staff who participated in WP3, during which we explored the issue of reabling people with dementia. These findings certainly align with wider evidence that attending to individuals’ social capital is a core aspect of preventing physical and mental health problems. 88 This is, perhaps, particularly pertinent for older people who are at an increased risk of social isolation and loneliness. 89

A second issue to particularly highlight from our survey of reablement services is that monitoring of outcomes through the use of a standardised measure is extremely unusual. Our experiences of using the Barthel Index is that, with training, it is feasible to integrate this into the assessment and review process. In addition, we encountered no difficulties using the NEADL scale, a self-report measure of (extended) ADL functioning.

The difficulties and challenges that this study encountered meant that an insufficient sample size was recruited for the evaluation WP. As a result, the findings from the outcomes and economic evaluation should be treated with caution. Indeed, with respect to these elements of the study, we would suggest regarding these as initial, exploratory evidence. It is, therefore, not appropriate to use these findings to draw implications for health and social care practice.

However, we do note that in our exploratory analyses of factors that may moderate or mediate the effectiveness of reablement, or explain differences in reablement outcomes between research sites, are user, service and intervention characteristics that are amenable to change in terms of service structure and practices.

In terms of individual characteristics, these include mental health and informal carer involvement. There are no indications that characteristics such as age or presence of comorbidities affect intervention outcomes. Engagement with reablement was the factor with the strongest and most consistent evidence regarding impact on outcomes. Although we have classified engagement as an individual characteristic, recent thinking stresses the role of the professional in engendering and maintaining engagement. Findings from our process evaluation fully align with that position. This also revealed that service user and family member (mis)understandings of reablement could be a barrier to engagement and that staff skills vary. We would suggest that these are two areas that service providers and commissioners may want to explore.

To our knowledge, this study is the first to explore the possible impact of organisation and delivery factors on outcomes. The findings are tentative and uncertain. However, intervention fidelity, duration of reablement, service structure (single vs. separate assessment and delivery teams), and in-house versus outsourced reablement worker teams were found to have some degree of (statistical) association with at least one outcome. The findings from the process evaluation, particularly the interviews with staff working in reablement services, both support and offer explanations for these findings. However, the significance of these associations in terms of practice or commissioning decisions could not be established.

Furthermore, this study incorporated collecting qualitative evidence from a range of stakeholders. Although this evidence is limited to three reablement services, and the data are incomplete in terms of full representation of all stakeholder groups across all research sites, our qualitative evidence consistently supports, and offers some degree of explanation for, our exploratory quantitative analyses.

Thus, although not able to draw practice implications from our study findings, we strongly advocate for the need for further research to ensure that commissioning, strategic and practice decision-making is evidence informed.

Finally, we draw attention to the findings from the final WP (WP3), which looked at reabling people with dementia. There was strong and consistent support from participants in this qualitative study that a diagnosis of dementia per se should not preclude someone from being offered reablement. However, interviewees consistently agreed that aspects of service delivery and practices may need to be adjusted or adapted. Within this, interviewees agreed that all staff working within the reablement pathway should be adequately trained in dementia. Although there is not necessarily yet the evidence to offer detailed guidance on practice, it should be possible to devise training that considers the implications of a dementia diagnosis on delivering reablement.

Recommendations for future research

The findings from this study indicate that securing robust evidence with regard to the following topic areas would be highly relevant and useful to commissioners, managers, front-line staff and policy-makers.

• A large-scale, multisite, mixed-methods outcomes and economic evaluation of reablement. We would suggest an observational study design that allows a robust and thorough investigation of the impacts of service, intervention and individual characteristics on reablement outcomes, and includes people with dementia:

  • We recommend that such a study involves a larger number of services than were used for the current study.

  • We also note the importance of sufficient resources being made available to services acting as research sites.

  • Nesting a local study team within the service appears to work well. The success of the study in terms of retention and data quality indicates the value of using home visits for data collection.

• Furthermore, to properly execute such a study, the following additional and preliminary pieces of work are recommended:

  • Identification of service and intervention characteristics deemed important and relevant to commissioners, strategic leads and service managers. These should then be represented in the research sites.

  • Further psychometric evaluation, and, if required, development, of the HRERS-RV and the ERPC is recommended.

  • Through consultation with stakeholders, via a systematic review and, if necessary, primary research, the identification and testing of a relevant set of reablement outcome indicators for people with dementia.

  • Exploration of using routinely collected data on service use (e.g. Hospital Episode Statistics) as part of the outcomes and resource use data set.

• The development/identification and evaluation of feasible interventions, delivered prior to/on the point of entry into reablement, to improve understanding and expectations of reablement, both on the part of potential service users and family members.

• The identification and evaluation of strategies and practices to support engagement with reablement. We would suggest a systematic review to start, which seeks to synthesise evidence across a wider scope of interventions including rehabilitation. Such a review, and any ongoing work, should include people with dementia.

• A systematic review, again synthesising evidence from reablement and also other relevant interventions, of practices and technologies that support the understanding, assimilation and remembering of instructions, or stages, of executing tasks, routines and daily activities among people with dementia.

Contributions of authors

Bryony Beresford (Professor, Health and Care Services Research) was a co-applicant and principal investigator (May–November 2017). She oversaw the final stages of the project and the outcomes and process evaluations within WP2. She contributed to all WPs and led the writing of, or contributed to the chapters for, the final project report.

Rachel Mann (Research Fellow) was involved in the WP1 data collection and analysis. She supported/led study set-up and day-to-day management of research sites and data collection for WP2 and supported/led the supervision of local study teams. She contributed to the process evaluation (WP2b) data collection and analysis, the analysis of service characteristics and report writing. She supported the collection of costs data and contributed to all other aspects of the project.

Gillian Parker (Professor, Social Policy Research) was a co-applicant. She oversaw the WP1 data collection and the analysis and reporting of this element of the project. She contributed to all other elements of the project.

Mona Kanaan (Senior Research Fellow, Statistics) was a co-applicant. She provided statistical expertise to the project, led on the outcomes data analysis and reporting (WP2a) and supervised other staff involved.

Rita Faria (Senior Research Fellow, Health Economics) was a co-applicant. She worked on the costs aspects of the WP1 survey and the economics evaluation (WP2c), leading on data analysis, writing and the systematic review. She also led the analysis and reporting of ASCOT-SCT4 outcomes data (WP2a).

Parvaneh Rabiee (Senior Research Fellow, Health and Social Care Services Research) was a co-applicant. She led WP3, collected all data and conducted data analysis and report writing. She contributed to the data collection for WP2b (process evaluation) staff focus groups.

Helen Weatherly (Senior Research Fellow, Health Economics) was a co-applicant. She oversaw the health economics evaluation (WP2c) and supervised junior health economists working on the project.

Susan Clarke (Research Fellow) conducted the data collection for and data analysis of service user and family member interviews within WP2b, and also supported the focus groups with staff within this WP. She assisted and supported the production of the final report.

Emese Mayhew (Research Fellow) led on data management and cleaning of outcomes data (WP2a), conducted data analysis for the outcomes evaluation and report writing, and the psychometric testing of the HRERS-RV and the ERPC. She assisted and supported the production of the final report.

Ana Duarte (Research Fellow) contributed to the health economics evaluation (WP2c) in terms of data management and cleaning, analysis and writing.

Alison Laver-Fawcett (Senior Lecturer, Occupational Therapy) provided occupational therapy expertise and advice to all aspects of the project, and was directly involved in the development of the HRERS-RV and the ERPC used in WP2a.

Fiona Aspinal (Senior Research Fellow, Health and Social Care Services Research) was lead applicant and principal investigator (October 2014–May 2017). She oversaw the delivery of the project until May 2017 and contributed to all WPs. She supervised junior staff and local study teams and led on delivering WP1.

Publications

Faria R, Kiss N, Aspinal FJ, Harden M, Weatherly HLA. Economic evaluation of social care interventions: lessons drawn from a systematic review of the methods used to evaluate reablement. Health Econ Outcome Res Open Access 2015;1:1.

Faria R, Weatherly HLA, Kiss N, Manca A, Parker GM, Beresford BA, et al. The challenges in evaluating the cost-effectiveness of complex interventions. Value Health 2015;18:A727–8.

Mann R, Beresford B, Parker G, Rabiee P, Weatherly H, Faria R, et al. Models of reablement evaluation (MoRE): a study protocol of a quasi-experimental mixed methods evaluation of reablement services in England. BMC Health Serv Res 2016;15:375.

Mayhew E, Beresford B, Laver-Fawcett A, Aspinal F, Mann R, Bechtold K, Kanaan M. The Hopkins Rehabilitation Engagement Rating Scale – Reablement Version (HRERS-RV): development and psychometric properties [published online ahead of print December 12 2018]. Health Soc Care 2018.

Data-sharing statement

Available data have been included in appendices and supplementary material. Any queries or data requests should be submitted to the corresponding author for consideration. Access to available anonymised data may be granted following review.

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.

Background

The HRERS48 was developed in the USA to measure engagement with physical/functional rehabilitation interventions. It is a five-item, practitioner-completed scale and has good psychometric properties. In developing the measure, and based on reviews of existing research into patients’ participation in rehabilitation interventions, the authors of the scale defined the construct of ‘engagement’ as comprising five dimensions: attendance, ability to participate affected by cognitive impairments or low mood, positive attitude, acknowledgement/acceptance of need, and active participation. The HRERS comprises five items, each representing a different dimension (Table 33).

A six-point response format (i.e. never, seldom, some of the time, most of the time, nearly always and always) is used by clinicians to report the frequency, or consistency, with which they observed each dimension.

Methods

Cognitive interviews were undertaken with 10 reablement workers based in two services. All except one completed this process with respect to two service users: this allowed us to explore whether or not the statements and response options were distinct enough to show differences between service users. A staged process was used, with revisions being made to the HRERS-RV during the process. Thus, a set of interviews were conducted, findings were shared and discussed within the research team and modifications were agreed. It was then subject to further cognitive interviews and so forth.

Testing the psychometric properties of the Hopkins Rehabilitation Engagement Rating Scale – Reablement Version

Total scores and distribution of scores across items

A total score was calculated by summing the scores on each of the five items (maximum range 0–30). A higher score represents greater engagement with the intervention. Figure 4 presents HRERS-RV total scores for the sample. The distribution is skewed to the left, demonstrating a ceiling effect (see Figure 4). The majority of the participants (< 74%) scored highly on the instrument (total score of ≥ 25).

FIGURE 4.

The HRERS-RV: distribution of total scores.

Looking at scores on an item by item basis, there was a pattern of high scores across all the dimensions regarded as contributing to the construct of engagement (Figure 5).

FIGURE 5.

Distribution of responses across the five dimensions of the HRERS-RV. (a) Readiness; (b) acceptance; (c) attitude; (d) active; and (e) cognitive.

Construct validity

Factor analysis was performed to establish the construct validity of the HRERS-RV. An initial correlation matrix across the HRERS-RV variables confirmed that there was significant correlation between the five dimensions (Table 34).

TABLE 34 - Correlation matrix across HRERS-RV

*p < 0.001.

	Readiness	Attitude	Acceptance	Active	Cognitive
Readiness	1.0000
Attitude	0.7973*	1.0000
Acceptance	0.7022*	0.8098*	1.0000
Active	0.7961*	0.8988*	0.7104*	1.0000
Cognitive	0.3430*	0.3673*	0.3158*	0.3865*	1.0000

Bartlett’s test showed that the correlations between the five components were, overall, significantly different from zero (p < 0.001; n = 126). The determinant of the correlation matrix (0.018) indicated that there was no extreme multicollinearity across the variables, although they were not completely unrelated either. [The determinant of the correlation matrix is a measure of collinearity. In case of perfect correlation (singularity) the determinant is 0, and if all variables are completely unrelated, the determinant is 1. To be able to perform factor analysis, the determinant of the correlation matrix needs to be > 0.00001.] Iterated principal factor method was conducted using orthogonal varimax rotation. The Kaiser–Meyer–Olkin (KMO) measure verified the sampling adequacy for the analysis (KMO 0.82).

There was only one factor with an eigenvalue > 1 (eigenvalue = 3.38, after rotation = 2.89), explaining over 80% of the total variance. The (rotated) factor loadings ranged from 0.36 to 0.94. The scree plot (Figure 6) showed a clear inflexion after the first factor, justifying keeping one single factor in the analysis.

FIGURE 6.

Scree plot: HRERS-RV.

Table 35 shows the factor loadings for all five dimensions. Factor loadings show the degree to which a dimension is correlated to the underlying factor: ‘engagement with reablement practice’. Positive attitude and active participation in reablement have the highest factor loadings. Cognitive ability to take part in reablement has a factor loading of 0.36, which is below the threshold of 0.4 of significance. 92,93 Communalities represent the amount of variance in each variable that can be explained by the factor, whereas uniqueness shows the amount of variance that is not shared with other variables.

TABLE 35 - The HRERS-RV: factor analysis output

Loadings > 0.4 are considered significant.

HRERS-RV dimension	Rotated factor loading	Uniqueness	Communality
Ready to start session	0.76a	0.25	0.75
Cognitive ability to take part	0.36	0.81	0.18
Positive attitude	0.92a	0.03	0.98
Acceptance of the need for reablement	0.67a	0.25	0.75
Active participation in reablement	0.94a	0.07	0.93

Predictive validity

Finally, we explored associations between the HRERS-RV total score and the main outcome measures at T1: the EQ-5D-5L summary score, Barthel Index summary score and ASCOT-SCT4 index score (Figure 7).

FIGURE 7.

Associations between HRERS-RV score and main study outcomes. (a) ED-5D-5L; (b) Barthel Index; and (c) ASCOT-SCT4.

The HRERS-RV correlated weakly with the EQ-5D score (r = 0.263; p < 0.01) and the ASCOT score (r = 0.306; p < 0.01). It correlated more strongly with the Barthel Index measure of functional status (r = 0.454; p < 0.001).

Conclusions

The findings from our psychometric evaluation of the HRERS-RV raise some concerns but are, overall, positive. The relatively small sample size means that any interpretation should be cautious.

We would note the ceiling effect and question whether or not revisions to the response format may address this. Adding further items would mean a departure from the conceptual basis of this scale and its ‘parent version’ (HRERS). The tests of construct validity yielded good results, although we query whether item 2 should be retained or reworded. This was the most problematic item to adapt in terms of retaining a conceptual meaning true to the ‘parent version’. Finally, tests of predictive validity were in the direction expected, with strongest associations with functional ability outcomes at discharge from reablement.

On the basis of these results, we included the HRERS-RV within our suite of predictor variables. We refer the reader to Chapter 5, which reports that engagement in reablement, as measured by HRERS-RV, was found to be a significant predictor of health and functional outcomes. This supports the argument for including a measure of intervention engagement when evaluating reablement.

Further work to test, and perhaps to further refine, the HRERS-RV would appear to be worthwhile and important.

Development of the Experience of Reablement Practice Checklist

Two members of the research team (BB and A-LF), both of whom have experience of questionnaire development (one is also a senior OT with experience of delivering reablement), met to discuss and develop a set of five items that described the core components of the reablement approach. Each item was presented as a statement, with the respondent asked to report whether or not they agreed with a statement by answering ‘yes’ or ‘no’. After the meeting, the items were reviewed individually by the two researchers, after which a first full draft of the ERPC was agreed. This was then shared with the wider research team for comment. No further revisions were made at that stage. The draft version (Box 6) was then taken forward for testing via cognitive interview with 10 respondents, all of whom had recently used a reablement service and who were taking part in a wider pilot of the study’s research instruments.

Cognitive interviews with two respondents prompted some early revisions to the draft checklist. First, it was clear that they were struggling to know how to respond to the items that, at that stage, were presented as statements. They were heard reconstructing the statements into questions. It was also clear that they were not sure which ‘workers’ the checklist was referring to. It was also suggested that instructions about completion should be more detailed. Finally, in terms of layout, it was suggested that the checklist should be presented in portrait, rather than landscape, orientation, with response options given underneath.

Aside from retaining the items as statements, changes to the checklist were made in response to these observations and suggestions. Five further cognitive interviews indicated that we should reword the items into questions. We made this amendment and administered it, with a cognitive interview, to a further three respondents. We concluded from these cognitive interviews that respondents found it easy to complete; they had no suggestions about further improvements to the checklist. This version, therefore, became the final version of the ERPC (Box 7).

Psychometric properties

Factor analysis

The dichotomous nature of the response option meant that a polychoric correlation matrix was used to explore associations across the items constituting the checklist. We found significant associations across all five components. Effect sizes were largest between ‘aware of the overall purpose of reablement’ and ‘involved in setting goals’ (r = 0.943; p < 0.001; 95% CI 1.09 to 0.878), and between ‘aware of the overall purpose’ and ‘having specific goals to work towards’ (r = 0.726; p < 0.001; 95% CI 0.975 to 0.477). Exploratory factor analysis found one underlying factor with an eigenvalue of 2.549 (after orthogonal varimax rotation), explaining over 75% of the variance across the five components. The variables with the highest loadings on the factor are ‘being aware of overall purpose’ (0.92) and ‘being involved in own goal-setting’ (0.89) (Table 38).

TABLE 38 - The ERPC: results of factor analysis (n = 128)

Note

Factor loadings > 0.4 are considered significant.

Experience of reablement component	Rotated factor loading	Uniqueness	Communality
Overall purpose	0.65	0.45	0.55
Specific goals	0.92	0.06	0.94
Own goal-setting	0.89	0.08	0.92
Motivation	0.49	0.44	0.56
Feedback	0.48	0.41	0.59

A scree plot and factor loadings matrix (Figures 8 and 9) suggest that there is an additional underlying factor, albeit weaker, and with an eigenvalue of < 1.

FIGURE 8.

The EPRC factor analysis: scree plot.

FIGURE 9.

The ERPC factor analysis: factor loadings plot. Rotation = orthogonal varimax, method = principal factors.

‘Being motivated to be independent’ and ‘receiving feedback’ load slightly higher on this second factor (0.56 and 0.59, respectively). On the basis of this exploratory analysis, we can conclude that there seems to be a dominant factor explaining most of the variance across five checklist items, and this seems to represent the ‘practical’ aspects of reablement: ‘being aware of the overall purpose’, ‘setting specific goals’ and ‘involvement in goal setting’. There is a secondary factor representing the ‘psychological’ component of reablement: encouraging independence by motivating the client and providing feedback on progress. However, these hypotheses need to be tested further with a larger sample.

For the purpose of this study, therefore, we chose to use the ERPC as a categorical variable, taking two values: (1) ‘complete intervention adherence’ and (0) ‘partial intervention adherence’.

Conclusions

We attempted to create a simple checklist, completed by service users at discharge from reablement, that would capture the extent to which the intervention they had received adhered to the ethos and approach of reablement. In other words, a user-report measure of intervention adherence. Our testing of the checklist’s psychometric properties suggests that additional items and/or revisions to the response format should be considered and tested. Specifically, based on ERPC scores, two-thirds of study participants received reablement that adhered to the five components of reablement practice represented in the checklist. In addition, it should be considered whether or not further items should be created and tested that further distinguish the so-called ‘practical’ and ‘psychological’ aspects of reablement detected in our analyses.

However, for the purposes of this study, we were comfortable taking the ERPC forward as a predicator in our outcomes evaluation, not using a score but treating the score as a categorical variable: complete versus partial adherence.

Introduction

The commissioners of this research (the NIHR Health Services and Delivery Research programme) requested that the research team record its experiences of delivering the study, and reflect on implications for future research. This appendix reports this and considers the following in turn:

• research readiness and LAs

• working in a fluid service landscape

• the involvement of private providers in research

• ambiguity regarding eligibility of NIHR Clinical Research Network support.

Research readiness and local authorities

Working in a fluid service landscape

This study commenced just after one of the most significant reforms in social care policy in recent times. As described in Chapter 1, this stimulated changes in the approach to managing individuals referred to adult social care who potentially may need long-term packages or care, at home or in a residential setting. Specifically, a parallel process of assessment and reablement was advocated. Furthermore, reablement was clearly positioned as part of a locality’s intermediate care provision, with the expectation of an integrated approach to delivery. NICE’s 2017 guidance Intermediate Care and Reablement14 reiterated and fleshed out policy guidance associated with the Care Act 201415 and also the move within the NHS to a model of ‘discharge to assess’. As a result, the service landscape at the time the study was seeking to recruit research sites was quite fluid. This had an impact on the study in two ways. First, services keen and interested to take part were not able to because they anticipated a period of service reorganisation. Second, the work carried out as part of the WP1 survey to assign each reablement service to a particular type of service model was not necessarily accurate, either at the time of approaching services to take part in the evaluation (WP2) or by the time recruitment began. This, indeed, was our experience.

Participation of private providers in the study

All sites outsourced at least some of their reablement provision to private providers. In two sites we encountered difficulties securing their involvement in the research. As a result, in one site we were not able to collect, at discharge from reablement, reablement workers’ reports of user engagement in the intervention. In a second site, only two out of the five private providers agreed that the study could recruit individuals using their service. In all instances, non-involvement was attributed to the fact that involvement in research was not specified in the contract the provider had with the LA. We would note that this issue was also encountered at the survey stage (WP1), with some private providers declining to participate.

In addition, although they did not directly have an impact on the study, it is also useful to report that changes in private providers were observed during study set-up and recruitment, in terms of both providers losing or choosing to close contracts and new providers being commissioned.

Concluding comments

Integration, or joint working, between health and social care is regarded as an essential core feature of ensuring that older and disabled people live healthy lives and are as independent as possible. This means that research into these populations has, and will continue to, cut across the two sectors. This will, perhaps increasingly, require the active involvement of LAs in what might be regarded as health services research. The experiences of this study, and also reported by others, point to substantive differences in research awareness, capacity and infrastructure between LAs and the NHS. This has significant implications in terms of the efficiency of research and the ability to deliver, in a robust way, large and complex projects. It is for these reasons that the research team was encouraged by the funder to provide an account and commentary of the issues and challenges encountered in this study. The purpose of preparing this account is to support an awareness among the health and social care research community and research funders of the issues that may be encountered when engaging in social care research. It also raises the question of whether or not strategic investment in research capacity and infrastructure within LAs, as has been seen in the NHS, is required.

2016 versus 2017 EuroQol-5 Dimensions, five-level version, tariffs: index and EuroQol visual analogue scale scores

Scores for the two tariffs were highly correlated both at baseline and at discharge, with a correlation coefficient of 0.9988 and 0.9995, respectively. The mean difference between the two tariffs was approximately 0.014. The intraclass correlation coefficient estimate for the EQ-5D-5L 2017 tariff at discharge was 0.114 (standard error 0.116, 95% CI 0.0085 to 0.4898); using the 2016 tariff gave similar results (intraclass correlation coefficient estimate 0.116; standard error 0.117, 95% CI 0.0090 to 0.4914). Table 45 provides further summary statistics of the difference between the two tariffs at baseline and discharge.

FIGURE 10.

The T0 EQ-5D-5L index scores: 2016 and 2017 tariffs (site A).

FIGURE 11.

The T0 EQ-5D-5L index scores: 2016 and 2017 tariffs (site B).

FIGURE 12.

The T0 EQ-5D-5L index scores: 2016 and 2017 tariffs (site C).

FIGURE 13.

The T0 EQ-5D-5L index scores: 2016 and 2017 tariffs (total sample).

Outcomes data

FIGURE 14.

The EQ-5D-5L index scores: distributions at T0 and T1 (total samples).

FIGURE 15.

The EQ VAS scores: distributions at T0 and T1 (total samples).

FIGURE 16.

Distribution of EQ VAS scores: at T1 and T2 (total samples).