Article
 Evaluating quality of life
 in adults with profound
 learning difficulties resettled
 from hospital to supported
 living in the community
 David Sines
 Buckinghamshire New University, UK
 Elaine Hogard
 Lakehead University, Canada
 Roger Ellis
 University of Ulster, UK, University of Chester, UK, and Buckinghamshire
 New University, UK
 Abstract
 This article describes a longitudinal evaluation of the quality of life of service users with profound
 learning difficulties who were resettled from hospital accommodation to supported housing in the
 community. The Trident approach was used for the design of the evaluation with data gathered
 regarding outcomes, process and stakeholder perspectives. Using a specially designed tool, quality
 of life was measured in seven domains for 39 service users in the hospital as a base line and at six
 months, twelve months and eighteen months in supported housing. A statistically significant
 improvement in quality of life overall and in each of the seven domains was found. Questionnaire
 surveys of parents/next of kin and support staff confirmed these findings as did a number of
 qualitative case studies.
 Keywords
 evaluation, learning disability, quality of life, resettlement, supported housing
 Date accepted: 10/9/2012.
 Corresponding author:
 David Sines, Buckinghamshire New University, 106 Oxford Road, Uxbridge UB8 1NA, UK.
 Email: david.sines@bucks.ac.uk
 Journal of Intellectual Disabilities
 16(4) 247?263
 ? The Author(s) 2012
 Reprints and permission:
 sagepub.co.uk/journalsPermissions.nav
 DOI: 10.1177/1744629512463840
 jid.sagepub.com
Introduction
 This article evaluates the quality of life (QofL) changes experienced by 39 adults with profound
 learning difficulties who were resettled from Orchard Hill Hospital to supported living in the
 community and reports the results. A longitudinal, repeated audit measure was applied on four
 occasions over 18 months using a specially developed instrument together with qualitative
 vignettes of individual lives and surveys of stakeholder views.
 Brief reviews on the international movement towards deinstitutionalization for those with
 severe learning disability and the evidence regarding its effects are provided. The concept of QofL
 and its measurement is briefly reviewed in its application to those with learning disability. The
 evaluation is then described with regard to design and methods, the instrument used, results,
 discussion and conclusions.
 International policy and practice in deinstitutionalization
 Deinstitutionalization, that is the replacement of institutions by services in the community, has
 been described as probably the most important change in the policy and pattern of service pro-
 vision for learning disability in the last 50 years (Mansell, 1996). From early initiatives in the UK
 (Tizard, 1960) and Sweden (Grunewald, 1997), the policy has been pursued, notably in North
 America, Scandinavia, Britain and Australia. Mansell (2006) describes the process as beginning
 typically with the replacement of hospital institutions with relatively large residential homes.
 These larger models are then superseded by group homes in which seven or eight people needing
 high levels of support live together with the help of a resident staff. Beadle-Brown et al. (2007)
 report the process of closing hospital-based provision as nearly complete in England, Scandinavia,
 the United States, Australia and Ireland, just beginning in many countries in eastern Europe, and,
 with a vague hope, further afield in, for example, Taiwan.
 Orchard hill resettlement
 The resettlement of service users from Orchard Hill Hospital to supported living in the community was
 a difficult and protracted process. This reflected conflict between the strongly held views regarding the
 best support for the service users, the consequences of a national scandal regarding abusive practices
 and the transfer of responsibility for commissioning from health care to social services.
 A decision to close Orchard Hill was taken by the Health Authority in 1999. A significant cause
 of delay was the challenges brought by parents and relatives of three patients, who were supported
 by the specialists in the Hospital. Ultimately, the challenges were unsuccessful, and agreement was
 reached to close the hospital.
 In 2007, the Health Care Commission inspection into the Sutton and Merton Primary Care
 Trust?s (PCT) Learning Disability Services identified the institutional abuse arising out of the
 models of care in operation and recommended that the people be moved to new community
 placements, based on assessed need, as soon as possible but no later than 2009.
 The number of long-term service users at Orchard Hill had reduced from 189 in 1990/1991 to 89
 by January 2007. This fall in numbers had been brought about mainly by people being resettled
 into community settings, and partly by the death of few people. The remaining service users at
 Orchard Hill were generally young, between 30 and 50 years old, and most of them had complex
 social care and health needs, whilst older and more able people have been relocated previously.
 248 Journal of Intellectual Disabilities 16(4)
The rate of resettlements into the community settings accelerated, and there were 80 people
 remaining by the end of April 2007. During 2007?2008, 79 of the 99 patients were transferred to
 other accommodation.
 In August 2008, a communique? announced that the Orchard Hill, the last long-stay hospital in
 England for people with learning disabilities, was due to close at the end of April 2009, which was
 5 years after the original government deadline for all long-stay hospitals to close.
 Advantages and disadvantages of resettlement
 In a review of published studies on the effects of deinstutionalization, Watts et al. (2006) found
 both advantages and risks for service users following a move out of the hospital and their review
 confirmed the seminal work of Emerson and Hatton (1994) in which mixed effects were also
 found. In the most recently published review, Chowdhury and Benson (2011) found that relocation
 had a generally positive impact on the QofL of service users. However, many of the studies
 identified that improvements were most prominent shortly after the move and plateaued after a
 year. In addition, service users continued to have low levels of community integration, and there
 was some evidence that health care might not be adequately met in the community. A limited use
 of standardized measures and also a lack in subjective assessments of QofL were found. More
 detailed studies of factors that are critical in affecting QofL were advocated.
 QofL and its measurement
 QofL is not a simple construct. It covers relatively objective external factors, such as the physical
 environment, and also less tangible features, such as satisfaction with that environment, and even
 more complex feelings, such as autonomy and choice. Given this complexity, measurement also
 poses problems, and these are exacerbated when those whose QofL is being studied have learning
 disabilities. Usually, a key indicator in QofL measures is the views of those concerned. If these
 views are not accessible, a proxy view has to be created through close observation of behaviour and
 circumstances.
 The concept of QofL and the means to measure it have been the focus of much recent research
 (Cummins, 1996; Skea, 2008). Skea (2008) identifies three areas of concern in the measurement of
 QofL in service users with learning disability. These include conceptual issues regarding the nature
 of QofL, methodological problems in equating subjective and objective measures as well as
 challenges in linking research with service utility. Hatton (1998) offers a radical critique of QofL
 as a concept and basis for assessment and concludes that the subjective measurements should be
 abandoned as posing insuperable methodological problems and that ?the quality of life approach,
 which claims to liberate people from a medical model, may paradoxically serve to extend the
 license of services to exert control over all facets of a person?s life?.
 Verdugo et al. (2005) suggested that ideally the measurement of QofL should be characterized by:
 1. recognition of its multidimensional nature, involving core domains and indicators;
 2. the use of both subjective and objective measures;
 3. the use of multivariate research designs to evaluate the ways in which both the personal
 characteristics and the environmental factors relate to the assessed QofL of the person;
 4. a systems perspective that captures environmental influences at a micro-, meso- and macro-
 level and
 5. the involvement, if possible, of those with learning disabilities in the design and implemen-
 tation of QofL assessment, research and evaluation.
 Sines et al. 249
In compiling and structuring instruments, there is a good agreement within the literature that the
 measurement of QofL should be based on both domains and indicators (Felce, 2007; Verdugo
 et al., 2005). Domains are the areas considered significant in QofL and indicators are the obser-
 vable features within these domains that can be measured. Thus, domains are the subdivisions of
 the overall notion of QofL and each domain provides a focus on a particular aspect of personal
 well-being. Taken together, the agreed domains signify the range over which the QofL construct
 extends.
 The 11 most common domains used within the QofL assessments are summarised by Verdugo
 et al. (2005) as follows:
  Physical well-being,
  material well-being,
  interpersonal relations and social inclusion,
  personal development,
  self-determination,
  emotional well-being,
  rights,
  environment,
  family relationships,
  recreation and leisure activities and
  safety/security.
 Felce (2007) in an elision of Verdugo suggested six domains, which are:
  Physical well-being,
  material well-being,
  social well-being: interpersonal relationships and community involvement,
  productive well-being: personal development, choice and occupation,
  emotional well-being and
  civic well-being.
 Cummins (1997) covered similar but not identical ground with six domains, such as:
  safety/health,
  material well-being,
  intimacy: close relationships with family and friends,
  place in society: activities with people outside the home,
  productivity: achievement in life and
  emotional well-being.
 He renames physical well-being as safety/health, focuses social well-being as intimacy on one
 hand and place in society on the other and, presumably, subsumes civic well-being in place in
 society.
 Schalock et al. (2004) again covered similar ground but with some additional suggestions:
  physical well-being,
  material well-being,
 250 Journal of Intellectual Disabilities 16(4)
 interpersonal relationships,
  social inclusion,
  personal development,
  self-determination,
  emotional well-being and
  rights.
 His significant additions include notions of rights and self?determination, whilst personal devel-
 opment replaces productivity and achievement.
 Indicators are the measurable elements within a domain that are taken as evidence that the
 person?s well-being reaches an appropriate standard.
 Verdugo et al. (2005) emphasizes the pervasive distinction between subjective and objective
 measures. In principle, objective indicators are accessible to public observation, whereas sub-
 jective indicators are accessible only to the concerned individuals or have to be inferred publicly
 from more objective measures. Objective measures have involved the use of audit tools, ques-
 tionnaires and interviews alongside other approaches, such as ethnographic studies, direct beha-
 vioural observation, discrepancy analysis and multidimensional scales (Keith, 2001).
 Nine tools were reviewed in detail and included, which are as follows:
  Evaluation of Quality of Life Instrument (Nota et al., 2006),
  The Life Experiences Checklist (Ager, 1998),
  Lifestyle Satisfaction Survey (Krapu et al., 2006),
  Index of Participation in Domestic Life (Stancliffe et al., 2010),
  Life Satisfaction Index (White, 2007),
  The Comprehensive Quality of Life Scale ? Intellectual/Cognitive Disability (Cummins,
 1997),
  Quality of Life Questionnaire (Schalock and Keith, 1993),
  The Mood, Interest and Pleasure Questionnaire (Ross and Oliver, 2003) and
  Measure of Independence and Choice (Grierson, 2006).
 From these, four tools were selected, including items that were clearly relevant to the chosen
 domains, and had demonstrated validity and reliability. Each contained objective measures.
 The first was The Comprehensive Quality of Life Scale ? Intellectual/Cognitive Disability of
 Cummins (1997). Items are scored by subjective and objective means. Subjective questions are
 asked to service users or their proxy respondents. Objective scoring looks at publicly identifiable
 indicators of the item. No correlation was found between the subjective and the objective measures
 in this tool, and when available, the proxy measures were not found to correlate with service user
 responses. However, a correlation of objective data gathered across all the measures was found.
 Such results led the researchers to conclude that the objective responses may be ?the only feasible
 way forward? when service user views are not obtainable or if there are questions surrounding the
 use of proxies (Perry, 2002).
 The second tool was the Quality of Life Questionnaire of Schalock and Keith (1993). This
 questionnaire includes 40 criterion-referenced items and focuses on four QofL domains: satisfac-
 tion, competence/productivity, empowerment/independence and social belonging/community
 integration. The scale has been well evaluated by researchers (Plimley, 2007).
 Sines et al. 251
The third tool was The Mood, Interest and Pleasure Questionnaire of Ross and Oliver (2003).
 This is based on an informed observer questionnaire and purports to measure levels of mood,
 interest and pleasure, by asking informants to rate operationally defined observable behaviours. It
 aims to access the important but relatively hidden personal experience and feelings of the
 population.
 The fourth tool was the objective measure developed by Grierson in his unpublished MSc
 dissertation (Grierson, 2006) to assess choice and independence in service users with learning
 disabilities.
 Design and methods
 The design of the evaluation was structured using the trident programme evaluation method (Ellis
 and Hogard, 2006) that focuses evaluation questions and data gathering on outcomes, process and
 stakeholder perspectives.
 The principal intended outcome for the resettlement was to achieve an improvement in the QofL
 of the service users. The evaluation used a design of repeated measure (Frison and Pocock, 1992),
 where the same measure is used with the same subjects at different points of time. This design, with
 data collected in a longitudinal time frame, allows the assessment of change over a period of time.
 It also allows two different conditions for the subjects to be compared. In this study, a comparison
 was made between the results when the subjects were in Orchard Hill Hospital and when they were
 in supported housing in the community. The measure was used on three occasions after resettle-
 ment, that is, at 6, 12 and 18 months.
 Another indication of quality of life outcomes was provided in the stories or vignettes that the
 care staff was encouraged to produce, regarding individual service users.
 Third, the views of care staff and parents/next of kin regarding the QofL of the service users
 were solicited using questionnaires based on the domains of the audit tool.
 The audit instrument was developed in accord with the following four main design principles:
  It should reflect, through its domains, the priorities for social care in Our Health, Our Care, Our
 Say (Department of Health, 2006).
  It should be valid and reliable, and to achieve this, use, where ever possible, the items from
 established measures in relationship to its domains.
  It should be based on objective rather than subjective indicators.
  It should be practicable and feasible for administration by care staff with a close
 knowledge of the service users and verification by social needs staff, managers, advocates
 and relatives.
 The instrument was developed with the support of a steering group consisting of senior social
 services staff, nominated experts, advocates, parents and next of kin of service users as well as the
 evaluators. Given the severe learning disabilities of the service users, it was not practicable to have
 a service user member, but the advocates on the group maintained a robust and committed position
 on behalf of the service users.
 The instrument was piloted with a group of support staff in a different facility, and this led to the
 minor modification of certain items to improve intelligibility and reliability.
 The validity of the instrument was addressed in several ways. First, a number of items
 were derived from instruments whose validity and reliability had already been demonstrated.
 252 Journal of Intellectual Disabilities 16(4)
Second, the instrument had high-face validity as judged by the members of steering group and
 care staff. Third, the instrument had good construct validity in that it reflected the priorities of
 DHSS guidelines. Fourth, the concurrent validity of the instrument was demonstrated by the cor-
 relation of its results with qualitative vignettes and surveys of the views of care staff and relatives.
 Seven domains were chosen for the instrument reflecting the literature and DHSS guidelines,
 including ?valuing people now?; and ?our health, our care, our say?.
 Domain 1: quality and location of housing;
 Domain 2: care planning and governance;
 Domain 3: physical well-being;
 Domain 4: social interaction and leisure activities;
 Domain 5: autonomy and choice;
 Domain 6: relationships;
 Domain 7: psychological well-being.
 In deciding the indicators and the questions for these domains, items were drawn, particularly,
 from The Comprehensive Quality of Life Scale ? Intellectual/Cognitive Disability (Cummins,
 1997), The Mood, Interest and Pleasure Questionnaire (Ross and Oliver, 2003) and the work of
 Grierson (1996) on choice and autonomy. In addition, the questionnaire benefited from the
 comments and suggestions of members of the evaluation steering group.
 Wherever possible, questions were constructed with a scale of possible responses with one to be
 ticked by the respondent. Scoring was then based on 5 for the most favourable and 1 for the least
 favourable condition in the choices. The final version of the scale included 50 questions with,
 wherever possible, quantified alternatives to be ticked.
 Given that the audit tool could not be completed by the individual service users, consideration
 was given to various alternatives, including completion by a carer, an advocate, a parent or a next
 of kin. It was decided that a primary carer would be in the best position to answer the questions in
 the tool, provided they were in regular sustained contact with the individual. The issue then became
 ensuring the validity and reliability of the responses of the carer. It was decided that each com-
 pleted audit would be checked by both the manager of the carer and by an independent social
 services needs assessor. Opportunities were also provided for an advocate and relative to sign off
 the audit, if they wished.
 Training was provided in the completion of the tool for two senior carers who were then
 responsible for training the primary carers. The tool was piloted through two audits: first of 10
 residents in a comparable provision and then of 29 residents in the Orchard Hill programme. The
 results of these pilots were discussed at the Steering Group to produce the final version which is
 copyrighted as the Quality of Life Questionnaire V.3 for Assessing Quality of Life in Individuals
 with Learning Disabilities (Ellis et al., 2008).
 The audit of the QofL of 39 residents was undertaken on four occasions:
  whilst resident at Orchard Hill,
  6 months after resettlement,
  12 months after resettlement and
  18 months after resettlement.
 The initial audit was undertaken in November 2008 and provided a baseline against which the
 subsequent audits could be compared. Whilst recognizing that this retrospective method is open to
 Sines et al. 253
criticism, we are satisfied by its validity and reliability through the triangulation of the judgments
 involved. Steps were taken to ensure that the audits were completed from January 2006 by carers
 who were familiar with the residents and their circumstances in Orchard Hill. All of the audits,
 including the retrospective audit, were carried out by front line staff with the support from their
 service managers; advocates and carers were involved wherever possible. The retrospective audits
 were further quality assured by senior staff who had worked at Orchard Hill since 2005.
 Subsequently, the audit tool was modified in the light of experience and review by the steering
 group, and the final version of the tool was used for the retrospective audit and the subsequent
 audits. The 6-, 12- and 18-month audits were dated from the baseline of November 2008 and were
 thus completed for May 2009, November 2009 and May 2010, respectively.
 The responsibility for the completion of the audits rested with the new care home providers, all
 of whom were subject to overview by the Head of Provision for Learning Disabilities and Mental
 Health Services. Problems were not reported in the completion of the audits and on the few
 occasions when there were missing sheets this was quickly rectified.
 In addition to the quantitative data gathered using the instrument, qualitative data were gathered
 through inviting care staff to produce vignettes of success stories in the lives of service users.
 Respondents were encouraged to describe the behaviours, which they believed to represent the
 improvement in the QofL of service users following their resettlement.
 The process of resettlement was explored through interviews with senior staff in the responsible
 social services department. The process followed was described by Hogard (2007). The interviews
 began by asking for a chronological account of the resettlement process. Respondents were then
 asked to identify particular successes and, if any, failures in the process. They were then asked to
 identify lessons learned. A process description was subsequently produced to identify key features.
 This description was then verified by the respondents and is included below.
 Stakeholder perspectives were sampled using two questionnaires designed to replicate the seven
 domains of the audit tool and to invite direct comparisons of life for the service users in Orchard
 Hill and supported accommodation. Similar questionnaires were devised for support staff and
 parents/next of kin. They covered the seven domains of the audit tool and asked whether they
 believed that the life of the service users had improved, deteriorated or remained much the same
 following resettlement.
 The service users
 All the 39 service users who were resettled from Orchard Hill Hospital to supported housing (34) or
 residential accommodation (5) had profound learning disabilities and many had additional disabilities,
 including autism, Down?s syndrome, epilepsy and various physical disabilities, including sensory
 impairment. The genders were almost equal with 20 females and 19 males. The average age was
 51 years with a range from 41 to 68 years and a SD of 5.68. Of the 39 service users, the majority (31)
 was described as having no spoken language with seven having severely limited language and one very
 limited language. Most had lived in Orchard Hill Hospital since early childhood until middle age.
 Results
 In this section, the results of the four audits completed for these service users are presented together
 with a brief summary of the ?good news? vignettes and the surveys of care staff and parents/
 relatives. A summary description of the process of resettlement is given.
 254 Journal of Intellectual Disabilities 16(4)
T
 ab
 le
 1.
 Su
 m
 m
 ar
 y
 o
 ft
 he
 re
 su
 lts
 at
 fo
 ur
 au
 di
 tp
 oi
 nt
 s
 N
 am
 e
 Qu
 ali
 ty
 an
 d
 lo
 ca
 tio
 n
 o
 fh
 ou
 sin
 g
 C
 ar
 e
 pl
 an
 ni
 ng
 an
 d
 go
 ve
 rn
 an
 ce
 Ph
 ys
 ica
 lw
 e
 ll-
 be
 in
 g
 So
 cia
 li
 nt
 er
 ac
 tio
 n
 an
 d
 le
 isu
 re
 ac
 t
 Au
 to
 no
 m
 y
 an
 d
 ch
 oi
 ce
 Re
 lat
 io
 ns
 hi
 ps
 Ps
 yc
 ho
 lo
 gic
 al
 w
 e
 ll-
 be
 in
 g
 Qu
 ali
 ty
 o
 flif
 e
 Re
 tr
 os
 pe
 ct
 to
 Jan
 ua
 ry
 20
 06
 M
 ax
 im
 um
 30
 45
 35
 74
 65
 30
 25
 30
 4
 M
 ea
 n
 17
 .3
 3
 0.
 08
 17
 .2
 8
 21
 .9
 2
 24
 .2
 6
 8.
 28
 17
 .2
 6
 10
 6.
 28
 SD
 5.
 21
 0.
 48
 3.
 62
 8.
 26
 11
 .2
 4
 6.
 26
 2.
 55
 25
 .0
 2
 Pe
 rc
 en
 ta
 ge
 58
 2
 57
 30
 37
 28
 69
 35
 6
 m
 o
 n
 th
 s
 M
 ea
 n
 25
 .5
 8
 39
 .0
 8
 19
 .1
 5
 32
 .1
 8
 53
 .8
 7
 12
 .7
 4
 19
 .8
 7
 19
 2.
 26
 SD
 1.
 94
 12
 .0
 6
 5.
 36
 13
 .5
 3
 6.
 52
 6.
 56
 3.
 47
 40
 .9
 6
 Pe
 rc
 en
 ta
 ge
 85
 76
 55
 43
 83
 42
 79
 63
 12
 m
 o
 n
 th
 s
 M
 ea
 n
 24
 .5
 8
 39
 .9
 2
 20
 .3
 3
 28
 .5
 1
 54
 .2
 8
 10
 .9
 2
 20
 .3
 1
 19
 2.
 95
 SD
 3.
 53
 4.
 62
 4.
 10
 12
 .7
 2
 6.
 21
 6.
 17
 2.
 69
 27
 .4
 6
 Pe
 rc
 en
 ta
 ge
 82
 78
 58
 39
 84
 36
 81
 .2
 4
 63
 18
 m
 o
 n
 th
 s
 M
 ea
 n
 28
 .1
 1
 36
 .2
 6
 20
 .6
 32
 .5
 9
 56
 .6
 2
 13
 .6
 2
 20
 .0
 3
 20
 6.
 95
 SD
 2.
 53
 5.
 28
 3.
 54
 .
 10
 .3
 1
 6.
 74
 6.
 78
 2.
 42
 24
 .2
 4
 Pe
 rc
 en
 ta
 ge
 93
 .7
 80
 .6
 58
 .9
 44
 87
 45
 .4
 81
 .2
 68
 255
QofL audit
 For each of the four audits, Table 1 shows the overall results for the group of 39 residents and the
 results for each of the seven domains that made up the QofL tool. Overall and for each domain, an
 average (mean) score is shown together with the maximum possible score. The achieved score is
 also shown as a percentage of the maximum possible score. The SD of each score is shown, this
 being an indicator of the spread or dispersion of scores.
 Comparisons were made and tested for statistical significance for an overall QofL score for all
 the 39 residents at the retrospective baseline and at 18 months. Comparisons were also made for
 each of the seven domains. A t test was used to test for significant difference between baseline and
 18-month audits, and a significant difference was found for overall QofL and for each of the
 domains. A t test calculates the probability of the difference in scores occurring by chance as
 opposed to indicating a significant difference. The differences from the overall QofL comparison
 and the comparisons for the seven domains were all significant at <0.0001 level. The two-tailed p
 value in each case was <0.0001. By conventional criteria, this difference is considered to be highly
 statistically significant with only a 1 in 10,000 chance that it occurred by chance.
 A comparison was also made with the five service users who were resettled in a care home and
 with the 34 who were resettled into supported living accommodation. No significant difference
 was found.
 The purpose of the audit tool was to detect differences between QofL at Orchard Hill and in
 supported accommodation in the community. The simplest indicator of QofL for individuals and
 hence the group is the overall QofL score out of a possible 304. This score is arrived at by summing
 the scores in the seven domains for each completed audit. Figure 1 shows an average (mean)
 calculated from all the 39 completed audits. The SD for the scores is shown and this gives an
 indication of the spread of scores between the 39 individuals. Finally, the mean is computed as a
 percentage of the maximum score possible in each domain and overall.
 The most significant single finding is that this average QofL score was 106.28 for residents
 when at Orchard Hill rising to 192.26 after 6 months in supported accommodation, a level which is
 maintained for 12 months (192.95) and which then rises to 206.95 after 18 months. This represents
 an improvement, in percentage of possible score, from 35% to 68%. This is shown in Figure 1 as
 the percentage of possible maximum at each audit. On the basis of these results, it is clear that the
 overall QofL for the 39 residents improves after their resettlement into supported accommodation.
 As stated above, there was no significant difference in the QofL scores for the five service users in
 a care home and the 34 in supported living.
 It should also be noted that all the seven domains show improvement, although some signifi-
 cantly show more improvement than the others. Table 2 shows, in rank order, the improvement in
 percentage terms (achieved score out of possible score) for each domain.
 Care planning and governance showed the most dramatic improvement reaching a high per-
 centage from a very low base line. This reflects the fact that the standards now set by the social
 services for care planning, which were not in place and being met at Orchard Hill, are now
 established and being followed (Figure 2).
 Autonomy and choice showed the next largest improvement, which is exactly what might have
 been hoped for when residents moved from the relatively controlled environment of hospital to the
 more autonomous and personalized environment of the supported housing. At a descriptive and
 qualitative level, there are dramatic differences between the restrictive regime at Orchard Hill
 described in the commission report and the more independent and individual life for the service
 256 Journal of Intellectual Disabilities 16(4)
users in the supported housing. This is reflected in the range of choice detected by the audit tool
 (Figure 3).
 As would have been expected, the quality and location of housing also improves significantly in
 the new purpose built housing (Figure 4).
 The two key social domains, relationships and social interaction and leisure, show improve-
 ment, but are still at a relatively low percentage (<50%) of what might be achieved. These are both
 areas that would benefit from further attention to maximize the potential advantages of living in the
 community. A risk identified in the literature is that sheltered housing might actually restrict social
 interaction compared with the range of interactions available in an institutionalized setting with the
 number of service users and care staff (Figures 5 and 6).
 68%
 0%
 20%
 40%
 60%
 80%
 100%
 63% 63%
 35%
 Base 6 months 12 months 18 months
 Quality of life
 Figure 1. Overall quality of life score shown as a percentage of maximum possible.
 Table 2. Percentage of improvement for seven domains in rank order
 Domain Baseline (%) 18 months (%) Improvement (%)
 Care planning and governance 2 80.6 78
 Autonomy and choice 37 87 50
 Quality and location of housing 58 93.7 35.7
 Relationships 28 45.4 17.4
 Social interaction and leisure 30 44 14
 Psychological well-being 69 81.2 12.2
 Physical well-being 57 58.9 1.9
 2%
 76% 78% 81%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Care planning and governance 
Figure 2. Percentage of possible maximum for care planning and governance.
 Sines et al. 257
37%
 83% 84% 87%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Autonomy and choice
 Figure 3. Percentage of possible maximum for autonomy and choice.
 58%
 85% 82%
 93.70%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Quality and location of housing
 Figure 4. Percentage of possible maximum for quality and location of housing.
 28%
 42% 36%
 45.40%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Relationships
 Figure 5. Percentage of possible maximum for relationships.
 30%
 43% 39% 44%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Social interaction and leisure activities 
Figure 6. Percentage of possible maximum for social interaction and leisure activities.
 258 Journal of Intellectual Disabilities 16(4)
Psychological well-being, which was relatively high in Orchard Hill, has nevertheless improved
 and maintained a fairly steady level since the beginning of the resettlement. This domain was
 measured through observations of behaviour and appearance from which feelings could be inferred
 (Figure 7).
 Physical well-being has shown the smallest improvement, but this probably reflects both a good
 standard of physical care at Orchard Hill and fairly intractable conditions in the individuals. Since
 there were fears that a move from hospital to community might lead to a deterioration in physical
 and medical care, the maintenance of this standards was significant (Figure 8).
 As a qualitative complement to the QofL audit, a number of ?stories? of individual service
 users were provided by care staff who were familiar with service users in Orchard Hill and
 their supported housing. A total of 18 such stories were collected; the excerpts given below
 provide heartening evidence of the improvement in their lives and behaviour through
 supported housing, personalization and choice. They are in striking contrast with the
 descriptions of patients? lives in the Healthcare Commission Inspection Report (2007) on
 Orchard Hill.
 The tenant was not only making a choice but it was being listened to and acted upon.
 After a very few weeks in her new flat she looked relaxed, and greeted me quietly and appropriately.
 She is obviously enjoying having her own space when she needs it and maybe this is helping her to be
 more sociable at other times.
 69%
 79% 81.24% 81.20%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Psychological well being
 Figure 7. Percentage of maximum possible for psychological well-being.
 57% 55% 58% 58.90%
 0%
 20%
 40%
 60%
 80%
 100%
 Base 6 months 12 months 18 months
 Physical well being
 Figure 8. Percentage of possible maximum for physical well-being.
 Sines et al. 259
One of the people we support, whilst selecting a lounge suite for their new home, went straight over to a
 suite unprompted and sat in the arm chair smiling and happy.
 One person we support has purchased an MP3 player and they now enjoy relaxing in their own home
 and listening to music they have selected.
 There is another service user whose outbursts of self-injury have decreased quite significantly after
 moving out of Orchard Hill.
 Spontaneously, she took herself to the front of the crowd that had gathered for the opening event, and
 made a speech, welcoming everyone to her new home, her new flat.
 The most liberating aspect of his move for him is that he has a garden where he can enjoy the sun, he
 has his own flat and belongings, can eat what he wants when he wants and he can walk into Cheam
 Village whenever he wants. He loves getting out and about and just smiles so much now.
 The process part of the evaluation focussed on the approach and procurement procedures of
 the council and these bear careful study to identify how a resettlement associated with a clear
 excellent improvement in QofL can be achieved so cost effectively. Fundamental to their approach
 was the design of a new model of supported living based on the individual needs of people as
 identified in their statutory assessment of needs. The council commissioned a model of social care
 in supported living in clusters of one bed flats where service users/citizens could gain new skills for
 more independent living. Economies of scale were achieved by sharing staff across day and night
 time support. Three care providers had already been procured by the PCT, and Sutton Council
 agreed an exemption from tendering to facilitate the transition from health to social care. The three
 social care providers had agreed to take NHS staff on TUPE (protected [Transfer of Undertakings
 (Protection of Employment) Regulations]) transfer and costed care plans were agreed on this basis
 with a new supported living service specification and hourly rate negotiation. The supported living
 service specification used the Reach (??Reach Standards of Supported Living?? promulgated by
 Advance Support) standards to ensure personalization of care.
 Stakeholder perspectives were solicited for the evaluation from parents/next of kin and care
 staff. To a large extent, their views supported the audit findings, although both the groups
 expressed concern about friendships and social relationships that achieved relatively less (although
 still significant) improvement when compared with other domains in the QofL audit. This, we
 believe, is a major area for further development. The response rate of both the groups in the surveys
 was disappointing, which suggests that these are early days for them to have formed confident
 views and that they should be surveyed regularly as the resettlement matures.
 Discussion and conclusions
 The design of this evaluation research was based on the well-tried trident method (Ellis and
 Hogard, 2006) in which the focus of research questions and data gathering is on outcomes, process
 and stakeholder perspectives. This approach has integrated a number of data streams to provide a
 distinctively comprehensive evaluation of the resettlement compared with other reported evalua-
 tions (Chowdhury and Benson, 2011). The research combined a relatively objective audit with qua-
 litative descriptions and questionnaire surveys of stakeholder views. Each of these methods has
 strengths and weaknesses but their combination provides a degree of complementarity.
 260 Journal of Intellectual Disabilities 16(4)
With regard to outcomes, the main part of the study was the completion of the specially devised
 QofL audit at four points in time to assess whether the major outcome of an improvement in QofL
 for service users appeared to have been achieved. At the conclusion of this part of the research, it
 appeared clear from the QofL audit data that there was an appreciable improvement in QofL for all
 the 39 residents in their new supported accommodation and care home compared with Orchard
 Hill. These audit results were confirmed in both the qualitative vignettes of service users and the
 perceptions of care staff and parents/next of kin.
 Comparing the baseline audit with the one obtained after 18 months shows highly significant results
 for QofL as a whole and for the seven domains through which it was measured. There are differences in
 the level of improvement between different domains and this would merit further study.
 The research design cannot of course indicate unambiguous cause and effect between
 accommodation and QofL, but it is a not unreasonable conclusion that the resettlement from
 hospital to supported housing appears to be associated with this improvement in QofL.
 The improvement in QofL found in this research should be considered in relationship with the
 costs of the supported accommodation, which compared more than favourably with costs else-
 where and for other forms of provision.
 The results are consistent with those found in other studies in both the immediate improvement
 following resettlement and the relative plateauing over an 18-month period. The risks associated
 with resettlement were not apparent at this stage but would merit further auditing over a longer
 period. Whilst all the domains showed improvement, the relatively smaller improvement for social
 relationships suggests the need for further study, and this reflects the findings summarized in
 Chowdhury and Benson (2011) in which the community relationships had showed least improve-
 ment after resettlement.
 The concentration on outcomes and on the process of commissioning in this evaluation study
 did not address the process whereby the outcomes are being achieved beyond a description of
 accommodation and support. Further study would be desirable to identify key variables associated
 with QofL outcomes.
 After this study, Sutton Council has commissioned continued monitoring of the effectiveness,
 impact and acceptability of the relocation. In particular, this will include repeating the QofL audit,
 monitoring provision against known risk factors, developing more refined audit in key social areas
 and continuing surveys of stakeholder satisfaction. It is the intention that the project should be used
 as a basis for devising sustainable internal monitoring procedures.
 Funding
 This research was supported by a grant from the London Borough of Sutton.
 References
 Ager A (1998) The BILD Life Experiences Checklist. London: British Institute of Learning Disabilities.
 Beadle-Brown J, Mansell J and Kozma A (2007) Deinstitutionalization in intellectual disabilities. Current
 Opinion in Psychiatry 20: 437?442.
 Chowdhury M and Benson B (2011) Deinstitutionalisation and quality of life of individuals with intellectual
 disability: a review of the international literature. Journal of Policy and Practice in Intellectual Dis-
 abilities 8(4): 256?265.
 Cummins RA (1996) The domains of life satisfaction: an attempt to order the chaos. Social Indicators
 Research 38: 303?332.
 Sines et al. 261
Cummins RA (1997) Comprehensive Quality of Life Scale ? Intellectual Disability: ComQol-15. 5th ed.
 Melbourne: School of Psychology, Deakin University.
 Department of Health (2006) Our Health Our Care Our Say, London: Department of Health.
 Ellis R and Hogard E (2006) The trident: a three-pronged method for evaluating clinical, social and educa-
 tional innovations. Evaluation 12(3): 372?383.
 Ellis R, Hogard E and Watts V (2008) Quality of Life Questionnaire V.3 for Assessing Quality of Life in
 Individuals with Learning Disabilities. Chester, UK: SHEU, University of Chester Press.
 Emerson E and Hatton C (1994) The Impact of Relocation from Hospital to Community on the Quality of Life
 of People with Learning Disabilities. London, HMSO.
 Felce (2007) Quality of Life in Community Living Services: Issues of Objective and Subjective Measurement.
 Cardiff: Welsh Centre for Learning disabilities, Cardiff University.
 Frison L and Pocock SJ (1992) Repeated measures in clinical trials: Analysis using mean summary statistics
 and its implications for design. Statistics in Medicine 11(13): 1685?1704.
 Grierson A (2006) A stakeholder analysis of a quality of life for people with learning disabilities. MSc Thesis,
 South Bank University.
 Grunewald K (1997) Effect of social and educational policies on the number of persons with mild mental
 retardation in Sweden. Mental Retardation 35(3): 218?220.
 Hatton C (1998) Who?s quality of life is it anyway? Some problems with the emerging quality of life
 consensus. Mental Retardation 36(2): 104?115.
 Healthcare Commission (2007) Investigation into the Service for People with Learning Disabilities Provided
 by Sutton and Merton Primary Care Trust. London: Healthcare Commission.
 Hogard E (2007) Using Consultative Methods to Investigate Professional-Client Interaction as an Aspect of
 Process Evaluation. American Journal of Evaluation 28(3): 304?317.
 Keith KD (2001) International quality of life: current conceptual measurement and implementation issues.
 International Review of Research in Mental Retardation 24: 49?74.
 Krapu T, Meinke L, Kramer L, Friedman R and Voda J (2006) The life satisfaction survey (LSS): development
 of the instrument. In: Proceedings of the 4th international coach federation coaching research conference.
 Mansell J (1996) In: Mansell J and Ericsson K (eds) Deinstitutionalisation and Community Living: An
 International Perspective. London: Chapman and Hall.
 Mansell J (2006) Deinstitutionalisation and community living: Progress, problems and priorities. Journal of
 Intellectual & Developmental Disability 31: 65?76.
 Nota L, Soresi S and Perry J (2006) Quality of life in adults with an intellectual disability: the evaluation of
 quality of life instrument. Journal of Intellectual Disability Research 50(5): 371?385.
 Perry (2002) Subjective and objective quality of life: their interrelationship and determinants. ReFeR
 summary. Available at: http://www.refer.nhs.uk.
 Plimley LA (2007) A review of quality of life issues and people with autism spectrum disorders. British
 Journal of Learning Disabilities 35: 205?213.
 Ross E and Oliver C (2003) Preliminary analysis of the psychometric properties of the mood interest and
 pleasure questionnaire (MIPQ) for adults with severe and profound learning disabilities. British Journal of
 Clinical Psychology 42: 81?93.
 Schalock RL (2004) The concept of quality of life: what we know and do not know. Journal of International
 Disability Research 48(3): 203?216.
 Schalock R and Keith K (1993) Quality of life questionnaire. Mental Retardation 40(2): 157?165.
 Skea D (2008) Quality of life for adults with learning disabilities in private residential care: monitoring aspects
 of life experience over time. Mental Health and Learning Disabilities Research and Practice 5: 252?265.
 Stancliffe R, McVilly K, Radler G, Mountford L and Tomaszewski P (2010) Active support, participation and
 depression. Journal of Applied Research in Intellectual Disabilities 23(4): 312?321.
 Tizard J (1960) Residential care of mentally handicapped children. British Medical Journal 1: 1041?1046.
 Verdugo MA, Schalock RL, Keith KD and Stancliffe RJ (2005) Quality of life and its measurement: important
 principles and guidelines. Journal of Intellectual Disability Research 49:707?717.
 262 Journal of Intellectual Disabilities 16(4)
Watts V, Hogard E and Ellis R (2006) Resettlement of Individuals with Learning Disabilities into Community
 Care: Advantages and Risks SHEU Occasional Papers No 7. Chester, UK: University of Chester Press.
 White A (2007). A global projection of subjective well-being: a challenge to positive psychology? Psychtalk
 56: 17?20.
 Sines et al. 263