An investigation of Reablement or restorative homecare interventions and outcome effects: A systematic review of randomised control trials

Abstract The effect of Reablement, a multi‐faceted intervention is unclear, specifically, which interventions improve outcomes. This Systematic Review evaluates randomised controlled trials (RCTs) describing Reablement investigating the population, interventions, who delivered them, the effect and sustainability of outcomes. Database search from inception to August 2021 included AMED, ASSIA, BNI, CINHALL, EMBASE, HMIC, MEDLINE, PUBMED, PsycINFO, Google Scholar, Web of Science, Clinicaltrials.gov. Two researchers undertook data collection and quality assessment, following the PRISMA (2020) statement. They measured effect by changed primary or secondary outcomes: no ongoing service, functional ability, quality of life and mobility. The reviewers reported the analysis narratively, due to heterogeneity of outcome measures, strengthened by the SWiM reporting guideline. The search criteria resulted in eight international studies, five studies had a risk of bias limitations in either design or method. Ongoing service requirement decreased in five studies, with improved effect at 3 months shown in studies with occupational therapist involvement. Functional ability increased statistically in four studies at 3 months. Increase in quality of life was statistically significant in three studies, at 6 and 7 months. None of the studies reported a statistically significant improvement in functional mobility. Reablement is effective in the context of Health and Social Care. The outcomes were sustained at 3 months, with less sustainability at 6 months. There was no statistical result for the professional role regarding assessment, delivery and evaluation of interventions, and further research is justified.


| INTRODUC TI ON
Reablement is restorative home care supporting individuals physically, socially and psychologically to regain the health, skills and independence required for daily living (Clotworthy et al., 2021). As defined by an International Delphi Study, "Reablement is a person-centred, holistic approach that aims to enhance an individual's physical, and or other functioning, to increase or maintain their independence in meaningful activities of daily living at their place of residence, and to reduce their need for long-term services" (Metzelthin et al., 2020 p11).
Reablement, valued for its potential to decrease demand for home care (Cochrane et al., 2016), facilitate hospital discharge, and its costeffectiveness (Francis et al., 2011;Kjerstad & Tuntland, 2016;Tuntland et al., 2017). People who benefit most have mild or moderate frailties or live alone . Furthermore, Reablement has no impact on the burden of informal caregivers compared to standard care (Senior et al., 2014).
The ethos of Reablement is that the carer works 'with' the person to achieve goals, rather than doing the activity 'for' the person (Metzelthin et al., 2020;Tew et al., 2014). The ideal Reablement harnesses strengths and has a highly functional and social connectivity focus (Doh et al., 2020), with meaningful and achievable goals, focusing on what matters to the person (Social Care Institute of Excellence (SCIE, 2020). (Metzelthin et al., 2020;Whitehead et al., 2018), and can be time-limited (Clotworthy 2021). Past systematic reviews indicated a need to evidence the effectiveness of Reablement interventions (Boniface et al., 2013;Legg et al., 2015). Reablement interventions are usually, but not exclusively, provided by Occupational Therapists to enable self-care activities: Activity analysis, motivational coaching, practising skills, risk enablement, compensatory techniques, assistive technology, equipment and adaptations (SCIE, 2020;Whitehead et al., 2018;Zingmark et al., 2020). Activity analysis is a unique Occupational Therapist skill (Thomas, 2012).  found that mobility was a key priority for particpants goals regardless of their health condition.

Reablement involves multi-faceted interventions
The academic literature is unclear on the specific professional roles involved in Reablement (Metzelthin et al., 2020;Pettersson & Iwarsson, 2017). The most successful Reablement according to SCIE (2020) has occupational therapy input. Royal College of Occupational Therapists (RCOT, 2019) argue that Occupational Therapists are best placed to deliver specialist and complex Reablement interventions and to supervise or train others due to the scope of their professional training.
Reablement service delivery models vary depending on organisational constructs.  conducted a UK survey reporting 53% of Reablement services were delivered by the Local Authority, and 17% included Occupational Therapists.

| Search strategy
The research question was based on the PICO framework, population, intervention, control and outcomes (Thomas et al, 2022).

The reviewers used the Preferred Reporting Items For Systematic
What is known about this topic?

Reablement is a multi-faceted complex intervention.
What does this paper add?
1. Reablement interventions are heterogenic, in terms of the timescale of delivery, and dose, and this influences outcomes.
2. Interventions delivered during Reablement can reduce the need for ongoing home care and improve activities of daily living and quality of life at three to four months, with sustainability beyond 6 months.
3. Professional role for assessment, delivery or evaluation of Reablement progress did not have a statistical consequence on outcome effect, and therefore cannot be generalised, this justifies further research.
Reviews and Meta-analyses (PRISMA) (Page et al. 2021)  Two reviewers independently undertook the search, critically appraising the studies for eligibility using a recognised tool (Critical Apporaisal Skills Programe (CASP, 2021), and disagreements were resolved through reflection, careful re-examining of the data and discussion. The full search results with inclusion and exclusion reasoning is available in Appendix S1. All completed data collection forms and other study information are available through correspondence with the author.
The synthesis method involved a three-stage process. Firstly, the interventions and comparisons were established in the protocol.
Secondly, the reviewers completed the data extraction process establishing the characteristics of each study, tabulating each outcome (Appendices S2 and S3) enabling comparison. The 'intention to treat' outcome effect was used to evidence assignment to the intervention, missing data reported in the risk of bias analysis .
The reviewers used the revised Rob2 risk of bias framework tool (Cochrane, 2021) to establish the internal validity of the studies.The tool has signalling questions with detailed explanations and an embedded answer algorithm to determine the level of concern about issues that are likely to affect the rability to draw reliable conclusions from the study. The risk domains are the randomisation process; deviations from intended interventions; missing outcome data; outcome measurement; and selection of the reported results. In addition, the reviewers used the supplementary questions for cluster RCT trials (Eldridge et al., 2020).
The Rob2 criterion for overall risk of bias is based on a combination of the embedded algorithm and assessor judgement, the assessor's decision has the final influence on risk weighting. Higgins et al. (2020) define the criteria as follows: Low risk of bias: all domains were judged low risk.
Some concerns: at least one domain had a concern, but not a high risk of bias in any domain.
High risk of bias: either the high risk of bias in one domain or some concerns in multiple domains.
Finally, stage 3 pulled the strands of the analyisis together to deterime the quality. Two reviewers discussed and determined each risk decision, including the direction and strength, giving certainty of bias for each study, strengthening the link between study design and resulting intervention effects (Sterne et al., 2019). The cumulative evidence synthesis was coherently summarised to determine that the true effect lies within a particular range or side of a threshold, using the definitions established by the Grade, Recommendations, Assessment, Development and Evaluation (GRADE) working group (Hultcrantz et al., 2017).
The reviewers strengthened the reporting of the synthesis by using SWiM , a framework designed to improve narrative reporting of the analysis process, supported by visual data to describe the range, distribution and effects (Robertson-Malt, 2014).

| Declaration of sources of funding
The primary author is funded for a PhD study by the National

Institute for Health Research Applied Research Collaboration East
Midlands, United Kingdom.

| RE SULTS
The reviewers found nine journal articles, reporting eight studies from the literature search strategy, presented using the PRISMA (2020) flow diagram for systematic reviews, Figure 1  Two published journal articles Parsons et al., 2013), reported the same registered RCT, both are included in the analysis to capture the risks associated with the secondary outcome measure documented in Parsons et al. (2013). Reasons for other studies' exclusion are available in Appendix S1.

| Population
The mean age of 1777 participants included in the studies reviewed was 80.35 years. All had difficulty completing activities of daily living at home, requiring a home care service.
Age, over 65 years, was an inclusion criterion in five studies. In addition, Parsons et al. (2012Parsons et al. ( , 2013 included over 55 years old from Māori or Pacific Islander ethnicity, and Tuntland et al. (2015) and Whitehead et al. (2016) included adults over 18 years old. All studies included more females than males ( Table 1). People who lived alone, reported in six studies (Table 1), showed a higher incidence in the intervention group for all studies, except Burton et al. (2013).

| Intervention
The intervention, role, components, dose, comparator and service delivery timescale are documented in Table 2.
All service models had an element of Reablement home care.
However, the service delivery varied, described as including a care coordinator assessment with home carers following a Reablement plan (King et al., 2012;Parsons et al., 2012Parsons et al., , 2013, and an assessment followed by face-to-face monitoring or review of the person's Reablement progress (Burton et al., 2013;Hattori et al., 2019;Sheffield et al., 2013;Tuntland et al., 2015;Whitehead et al., 2016).
Where a study reported a review or monitoring of progress, visits ranged from three to five ( Table 2).
The intervention components focused on a complex range of strategies to increase motivation, ability and mobility ( Table 2).
Functional ability is a descriptive term used for how people complete Activities of Daily Living (ADL): washing, dressing, meal preparation, independent living skills, shopping, and housework.
Whereas functional mobility describes the ability to sit to stand, balance and walk.

| Control
In all studies, the control group received standard home care.

| Outcomes
The outcome measures are tabulated in Appendix S2. Two studies use the dichotomous measure of no ongoing home care service as a primary measure, taken from administrative databases (Hattori et al., 2019;Lewin et al., 2013); three studies report on this outcome as an incidental finding (King et al., 2012;Sheffield et al., 2013;Whitehead et al., 2016).  3.4.1 | What are the interventions, when are they delivered and by whom?
All studies delivered Reablement interventions with individuals in their homes except Hattori et al. (2019) who delivered some educational aspects to participants in a community group setting.
The interventions investigated were multi-faceted, with each study focusing on different combinations ( Table 2)  The issue was the diverse range of tools used to measure, some valid and reliable, others not (Appendix S2) and the consequence this had on results.

| Are there any long-term benefits?
Five studies reported a descreased need for ongoing home care post-intervention ( Table 3). The lack of reported data for between group difference and transparency of method, affected the comparison of studies that reported an incidental effect.
Where the need for ongoing home care was a primary outcome measure, effect is presented using odds ratio with 95% confidence intervals, showing a statistical significance ( Figure 2).
Next, (Table 4) shows the long-term benefits of functional ability.
There is an increase in effect post-intervention reported in four studies, with a slight decrease after 6 months (Whitehead et al., 2016), and 9 months (Tuntland et al., 2015) (Figure 3).
Comparing each study reporting a positive effect, Sheffield et al. (2013) measures with the Functional Independence Measure (Mackintosh, 2009), reporting statistical effect as a p value but with no further between-group comparable data. Next, King et al. (2012) use Nottingham Extended Activities of Daily Living (NEADL) (Nouri & Lincoln, 1987), measuring the between-arm mean difference in score, from baseline to 7 months as 0.3 (−1.4 to 2.1), p = 0.71.
Next, taking change in the quality of life, four studies evidenced improvement ( Table 5) Tuntland et al. (2015) used COOP/Wonka score (Kinnersley et al., 1995), whereas other studies used the Short-Form SF-36 health survey (Ware et al., 2000), reporting effect based on the total score for SF-36, combining the physical and mental aspect.
Comparisons between studies measuring quality of life using SF-36 ( Figure 4) show a similar difference in effect for the SF-36 physical score at 6-7 months; King et al. (2012) and Parsons et al. (2012) used a similar sample size. Whitehead et al. (2016) report less effect is reported at both 3 and 6 months with a smaller sample.
Lastly, three studies report the positive effect on functional mobility, Table 6. Burton et al. (2013), did not achieve their primary outcome using a composite of valid and reliable outcome measures, aiming to evidence the statistical effect of LIFE interventions in a summary variable at 2 months, ( Table 6). the studies using this measure were unable to produce a statistically significant short-or long-term result for changes to functional mobility.

| Risk of bias
Overall, four studies scored low risk of bias, three studies had some risk of bias, and two studies had a high risk of bias (Table 7).
The reviewers considered each risk domain for each outcome. In Parsons et al. (2012), the randomisation was a concern because there was little detail on the process reported, and the reviewers took the study on face value and assumed integrity, assessing as low risk overall. Whitehead et al. (2016) used valid and reliable standardised outcome measures. Howevever the reviewers were concerned about missed data or misinterpretaion of data, the study did not identify this limitation. Despite blinding the occupational therapy assessor, having unblinded participants and home care staff meant the assessor could have guessed the allocation group. When the reviewers considered collection of outcome data was face-to-face in the participant's home, and immediately entered into a database, the reviewers agreed a judgement that any bias through missed data or misinterpretation of data was reduced, and reduce the overall risk of bias score.
The reviewers upheld their concerns about the five RCTs with an overall risk of bias. In Sheffield et al. (2013) it was in the missing outcome domain, due to a high attrition rate explained as 'age ineligibility', those participants who dropped out had higher dependency scores on the Functional Independence Measure. Burton et al. (2013) deviated from intended interventions and outcome measurement.   Given that the TARGET intervention focused on person-centred goal setting and an ADL outcome measure is used at baseline, this raised a concern about reporting bias based on favourable results.

| Certainty of evidence
The reviewers quality assured the RCTs to determine certainty of effect, based on the assessment of five domains: risk of bias, inconsistency, indirectness, imprecision and publication bias (Schünemann et al., 2020), There were five studies with risk of bias limitations in either design or execution of the method (Table 7).
There was no inconsistency in any study serious enough to downgrade the evidence. In all studies reviewed, the evidence directly answered the review question, and there were no concerns about indirectness. In all the studies, the reported results were precise, and the quality of evidence was upheld (Appendix S3).
There was a probability of publication bias in Parsons et al. (2013), both reviewers agreed the reporting bias was not serious enough to downgrade the quality of evidence.
The reviewers considered the magnitude of the effect, in the case of Lewin et al. (2013)

| DISCUSS ION
The purpose of this systematic review, was to identify, describe, and critically appraise the interventions delivered and their effects on outcomes of Reablement using robust methodology; eight RCTs met the eligibility criteria. The reviewers found the data was methodologically heterogenic.
The search criteria for this review were based on evidenced based reasoning that Reablement interventions should continue until people have maximised their abilities and reached their person-centered goals; time-limited but not time-restricted (Clotworthy et al., 2021;SCIE, 2020;Doh, 2020). Unlike previous reviews of RCTs, the reviewers placed no restrictions on intervention delivery timescale in the eligibility criteria.
This review challenges a historical systematic review of RCTs that found no evidence of effect for Reablement delivered in six weeks (Legg et al., 2015), aim to achieve a particular outcome (Higginson & Carr, 2001).
Furthermore, the method for calculating the SF-36 was unclear in both King et al. (2012) and Parsons et al. (2012). A scoping review of studies reporting a total quality of life score using SF-36, evidenced 129 (75.0%) of the 172 studies did not specify the method for calculating the SF-36 total score (Lins & Carvalho, 2016). Despite a positive effect on the requirement for ongoing services in the studies with occupational therapy involvement at three months (Sheffield et al., 2013;Tuntland et al., 2015;Whitehead et al., 2016) and at 4 months Hattori et al. (2019), it wasn`t possible to unequivocally determine whether professional role influenced the difference in outcome due to heterogenetic data. The only UK study had a small sample n=30 (Whitehead et al., 2016), limiting the results for an otherwise methodologically sound study. The primary outcome was to explore feasibility of design, and whilst the study evidenced the number of visits by role, it did not explore whether frequency of Occupational Therapist visits optimised the outcomes. Pettersson & Iwarsson, (2017) identified in their literature review that there was a lack of definition of interventions and professional roles, and this remains a problem for evaluating Reablement intervention outcomes in relation to economic effectiveness and quality assurance. Reablement services deliver interventions for people with varying degrees of complexity and need, and the cost of regulated specialist professionals is greater than unqualified workers, therefore the staffing role and responsibility should be clear in any research examining the effectiveness of Reablement outcomes.
Three studies described Reablement workers competency to deliver interventions, their supervision, and training (King et al., 2012;Tuntland et al., 2015;Hattori et al., 2019). They do not specify how these can influence or assure better outcomes (Sims-Gould et al., 2017). This lack of detail evidenced a need for further research on Reablement workers competency, training, and supervision to enable a greater understanding of Reablement as a complex intervention. Furthermore, occupational therapists have a role in training Reablement workers to operate in an enabling way (Dibsdall, 2021).

| S TRENG TH S AND LIMITATI ON S OF THE IN CLUDED S TUD IE S
The intervention effect was positive in all studies. Howevever, the outcome measures were diverse, and comparison of heterogeneous outcome data would be misleading. This limited the extent the reviewere could accurately compare these RCTs.
Overall, five studies had a risk of bias limitations in either design or method (Table 7), two large sample RCTs were judged high risk overall. Parsons et al., 2013). While the larger samples strengthen study results, these limitations affected the reviewers confidence in the findings.

| Strengths and limitations of the review method
The reviewers limited the search criteria to peer-reviewed RCTs, this excluded other relevant experimental studies, creating a narrower review (Schünemann et al., 2020), limiting the findings. The reviewers agreed to include a feasibility RCT (Whitehead et al., 2016) and prospective RCT  as they met the search eligibility criteria. The search terms were not extensive and this was a limitation. The reviewers excluded studies with participants not in receipt of a home care service, a limiting factor as informal care givers deliver home care for cultural or financial reasons.
The synthesis method clearly identified data from the reviewed studies was heterogenic, the variety of outcome measures used gave no scope for meta-analysis, subgroup analysis or meta-regression, limiting the extent of comparison. The SWiM reporting guideline  strengthened the narrative reporting method. The use of GRADE (Hultcrantz et al., 2017) addressed the subjectiveness of the reviewers analyisis of the quality of the studies, strengthening the assessment of certainty in the effect.
The method gave an opportunity to consider confidence in effects, quality, similarities, the impact of bias and applicability of the findings to the research question. The reviewers used the method to assess whether the effectiveness of the interventions was sensitive to clinical or methodological heterogeneity and whether the intervention effect itself was enough to eliminate any risk in the bias domains; this sensitivity analysis ensures confidence in the intervention effect, strengthening the study .

| Can the findings be generalised?
The intervention delivery, amount and type of interventions described in all studies could not be generalised due to the methodological heterogeneity of the data. The reviewers evidenced external validity in two studies with unrestricted age samples (Tuntland et al., 2015;Whitehead et al., 2016). The only health condition exclusions, dementia and palliative care suggest that Reablement has efficacy in a diverse range of conditions. Internationally, Health and Social Care organisations, policy drivers, workforce mix, and population culture vary, limiting generalisability.
It was not possible to determine whether a professional role has a statistical consequence that could be generalised. A definitive large sample UK trial aiming to determine the effect of Reablement interventions and the professional role administering them is required. The ethical and methodological considerations of RCTs are complex to resolve in a Social Care setting: randomising to intervention; blinding the researcher, professional completing the assessment and review, Reablement workers delivering the intervention and the participant receiving the intervention, meaning that it is most appropriate to use a cohort study method.

| CON CLUS IONS
The reviewers found diversity in the outcome measure indicating future research should establish an agreed outcome measure for Reablement, including need for ongoing home care, functional ability, mobility, and quality of life, to evidence Reablements effect.
The results show the need for an ongoing home care service decreased in five studies Table 3: Figure 2, with improved effect at three months shown in studies with occupational therapist involvement. Functional ability increased statistically in four studies at three months, Table 2. An increase in quality of life was statistically significant in three studies at six and seven months, table 5. None of the studies reported a statistically significant improvement in functional mobility, Table 6. The outcome of Reablement was most beneficial at three to four months, with some sustainability beyond six months (Tuntland et al., 2015;Whitehead et al., 2016).
Studies with Occupational Therapist's involvement showed a greater effect on outcomes, this was not statistically significant-Therefore, the effect of professional role delivering Reabelement could not be generalised. Care. A large sample UK trial aiming to determine the effect of Reablement interventions, and who is best placed to deliver them, is necessary.

ACK N OWLED G EM ENTS
The primary author Cate Bennett is an Occupational Therapist and would like to acknowledge her academic supervisors, employer Nottinghamshire County Council Adult Care, and the National

Institute for Health Research (NIHR) Applied Research Collaboration
East Midlands (ARC EM) who fund this Phd Study.

CO N FLI C T O F I NTE R E S T
There is no conflict of interest, the views expressed are those of the author and not necessarily those of the NIHR, Nottinghamshire County Council or the Department of Health and Social Care.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.