The relationship between building design and residents’ quality of life in extra care housing schemes

Health & Place 21 (2013) 52–64

Contents lists available at SciVerse ScienceDirect

Health & Place journal homepage: www.elsevier.com/locate/healthplace

The relationship between building design and residents’ quality of life in extra care housing schemes Alison Orrell a,n, Kevin McKee b, Judith Torrington c, Sarah Barnes d, Robin Darton e, Ann Netten e, Alan Lewis f a

Rehabilitation and Assistive Technology Research Group, ScHARR, University of Sheffield, Room 106, The Innovation Centre, Portobello, Sheffield S1 4DP, UK School of Health and Social Studies, Dalarna University, 791 88 Falun, Sweden c Derwent House, 7 Red Lion Row, Grindleford, Hope Valley S32 2JJ, UK d School of Health and Related Research, University of Sheffield, Regent Court, 30 Regent Street, Sheffield S1 4DA, UK e PSSRU, School of Social Policy, Sociology and Social Research, Cornwallis Building, University of Kent, Canterbury, Kent CT2 7NF, UK f Manchester Architecture Research Centre, School of Environment and Development, University of Manchester, 1.11 Humanities Bridgeford Street Building, Oxford Road, Manchester M13 9PL, UK b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 March 2011 Received in revised form 24 October 2012 Accepted 2 December 2012 Available online 24 January 2013

Well-designed housing is recognised as being an important factor in promoting a good quality of life. Specialised housing models incorporating care services, such as extra care housing (ECH) schemes are seen as enabling older people to maintain a good quality of life despite increasing health problems that can accompany ageing. Despite the variation in ECH building design little is known about the impact of ECH building design on the quality of life of building users. The evaluation of older people’s living environments (EVOLVE) study collected cross-sectional data on building design and quality of life in 23 ECH schemes in England, UK. Residents’ quality of life was assessed using the schedule for the evaluation of individual quality of life-direct weighting (SEIQoL-DW) and on the four domains of control, autonomy, self-realisation and pleasure on the CASP-19. Building design was measured on 12 user-related domains by means of a new tool; the EVOLVE tool. Using multilevel linear regression, significant associations were found between several aspects of building design and quality of life. Furthermore, there was evidence that the relationship between building design and quality of life was partly mediated by the dependency of participants and scheme size (number of living units). Our findings suggest that good quality building design in ECH can support the quality of life of residents, but that designing features that support the needs of both relatively independent and frail users is problematic, with the needs of highly dependent users not currently supported as well as could be hoped by ECH schemes. & 2013 Elsevier Ltd. All rights reserved.

Keywords: Building design Housing Extra care housing Older people Quality of life

1. Introduction Well-designed housing is recognised as being an important factor in promoting a good quality of life (QoL) (Evans et al., 2002; Parker et al., 2004; Wahl et al., 2009). Good quality housing is also seen as being instrumental in fulfilling the health and social care agendas for older people in the United Kingdom (UK), i.e., preventing or delaying the need for care (Department of Health, 2001). Moreover, good building design should contribute positively to making housing better for people to live in. The UK National Strategy for Housing in an

n

Corresponding author. Tel.: þ44 11 422 228272. E-mail addresses: A.Orrell@sheffield.ac.uk (A. Orrell), [email protected] (K. McKee), J.M.Torrington@sheffield.ac.uk (J. Torrington), S.Barnes@sheffield.ac.uk (S. Barnes), [email protected] (R. Darton), [email protected] (A. Netten), [email protected] (A. Lewis). 1353-8292/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.healthplace.2012.12.004

Ageing Society (Communities and Local Government, 2008) recommended that housing should support healthy, active and independent living in welcoming communities and be inclusive, attractive and sustainable for the ageing population. Specialised housing models incorporating care services, such as extra care housing (ECH) schemes, are seen as enabling older people to maintain a good QoL despite increasing health problems that can accompany ageing. Such schemes allow older people to live in their own flats or bungalows with a range of facilities and support designed to meet their needs (Department of Health, 2010). This is the first study to investigate the relationship between the design of ECH and QoL.

2. Background Between 1983 and 2008 people aged 85 years and over formed the population sub-group in the UK that proportionately increased

A. Orrell et al. / Health & Place 21 (2013) 52–64

the most, from 600,000 to 1.3 million (Office for National Statistics, 2009). Population projections suggest that by 2033 this number will have increased again to reach 3.2 million, accounting for 5% of the total population. Furthermore, the population of centenarians is projected to increase from 11,000 in 2008 to 80,000 in 2033, and the population of state pensionable age (65 years and over) is projected to increase by 23% from 11.8 million to 15.6 million (Office for National Statistics, 2009). For policy makers such population ageing has implications for spending on health and social care services, resulting in new policy paradigms such as, ‘‘active ageing’’ and ‘‘ageing in place’’, which are developed to enable older people to maintain their mobility and independence, thereby avoiding expensive and dependency-enhancing institutional care (Walker & Lowenstein, 2009). The physical, psychological and societal effects of ageing are well documented with older people at risk of being affected by comparative poverty, poor physical and/or mental health, chronic physical or sensory impairment and social isolation. Such factors impact on housing need in terms of the suitability of housing design, location and facilities to support an acceptable QoL for older people. Conceptually, QoL is perceived as a multidimensional construct containing domains of physical health, psychological wellbeing, social relationships and the physical environment (World Health Organisation Quality of Life (WHOQoL) Group, 1998). Ageing is often perceived as negatively affecting a person’s QoL but, when other factors are controlled for, the effects of age on QoL may disappear. For example, results from the first wave of the English longitudinal study of ageing investigating QoL suggest that mobility and independence in activities of daily living, living in a neighbourhood perceived as being good and having trusted relationships with friends and family are significant contributors to a good QoL, whereas a poor financial situation, depression, limitations in physical activities and having a chronic illness are associated with a poorer QoL (Netuveli et al., 2006). Furthermore, Wahl et al. (2009) suggest that the maintenance of independence in activities of daily living and QoL are related to the utilisation and optimisation of environmental resources, e.g., housing. Taken as a whole, the above research suggests that a good QoL can be sustained in later life given a supportive environment that reduces the impact of any functional limitations. The home environment is acknowledged as being the primary context for growing old (Scheidt and Windley, 2006; Wahl and Gitlin, 2007), with older people preferring to age in place for as long as possible (American Association of Retired Persons, 2003). As people age they spend more time in their home (Baltes et al., 1999; Wahl and Gitlin, 2007). Analysis of data from the English House Condition Survey suggests that people over the age of 65 years spend more than 80% of their time in their homes, increasing to more than 90% for people aged 85 years or over (Adams & White, 2006), whereas younger people spend more of their time away from the home due to work and social commitments (Gershuny et al., 2005). Spending so much time at home has the effect that the home becomes more important for creating meaning for the older person (Rubenstein and De Medeiros, 2004). Consequently, the home is more likely to enhance or to undermine the health and well-being of older people as environmental factors are not only related to negative health events, such as falls (Gitlin, 2003; Oswald and Wahl, 2004) or disability-related outcomes (Wahl et al., 2009), but also to positive health-related outcomes, such as independence in daily activities of living and subjective well-being (Evans et al., 2002; Oswald and Wahl, 2004; Oswald et al., 2007). The type of housing that people occupy as they age will therefore be important in fostering health and social well-being. ECH is a relatively recent form of housing that has been developed in the UK. Similar models are ’’Aanleunwonen’’ or ‘‘Wonen Plus Living’’ in the Netherlands and Germany, ‘‘Housing with Care’’ in Australia and Canada, and ‘‘Continuing Care

53

Retirement Communities’’ and ‘‘Assisted Living’’ in the USA. ECH has been developed to provide higher levels of support than sheltered housing. Residents have self-contained apartments or bungalows in developments that include a wider range of communal facilities than are normally provided in sheltered housing schemes, and personal care based on assessment of needs is available. The expectation is that for many residents ECH will be a home for life and will reduce the need for residential or institutional modes of care. Between 2004 and 2010 the UK government provided £227 million total capital funding from the Department of Health’s Extra Care Housing Fund Initiative to stimulate developments and partnerships between social services departments, housing authorities, care providers and the private sector and social housing developers (Darton et al., 2012). There is considerable variation in size and typology across schemes, encompassing villages with several 100 units to small scale developments and remodelled schemes. The scale of the development determines a number of factors. Larger schemes, such as retirement villages with more than 100 dwellings offer economies of scale, allowing for more extensive communal areas and a wider range of non-care facilities and activities, e.g., fully equipped gymnasia and spas, restaurants and activity specific workshops (Croucher, 2006). Smaller schemes are easier to site and simpler to plan because they have fewer facilities (Callaghan et al., 2009), and may use communal facilities for more than one type of activity, e.g., a restaurant can also serve as part of the communal activities area. In some instances existing sheltered housing schemes have been remodelled as ECH schemes. Remodelling typically involves ‘‘cutting and carving’’ of existing sheltered housing in order to provide larger individual units and communal spaces, renewed services and finishes and to bring buildings into line with current regulatory requirements (Wilkes, 2007). Despite the variation in the building design of ECH, the underlying aims of such schemes are to maintain or improve independence in daily activities, reduce social isolation and improve the QoL for residents. To date, however, there is a lack of research that has examined the relationship between building design and the QoL of older people living in ECH schemes. Furthermore, until now, there has not been a suitable tool available to capture and to measure building design in ECH as the building design in this context refers to a variety of building attributes ranging from the use of space (macro level design) to individual design elements, e.g., type of tap (micro level design). The aims of this study were therefore twofold: (1) to produce an evidence-based building evaluation tool that would be suitable for use in assessing housing developments designed for older people and (2), to use this tool to explore the relationship between building design and the QoL of older people living in ECH schemes. The EVOLVE (Evaluation of Older People’s Living Environments) tool (Lewis et al., 2010a) was developed by the research team from literature reviews, policy guidelines, reviews of recent buildings, design guidance, the results of evaluations of building surveys, quality indicators, focus groups with ECH scheme residents and their relatives, and expert consultations. This development work is reported elsewhere (Barnes et al., 2012; Lewis et al., 2010a, b). In this paper we report the key findings of analyses which investigated the relationships between the micro and macro aspects of building design using the EVOLVE tool and the QoL of older people living in ECH schemes.

3. Method 3.1. Design A cross-sectional survey-based design was used.

54

A. Orrell et al. / Health & Place 21 (2013) 52–64

3.2. Sample

3.5. Outcome measures—Quality of life

A stratified sample of 35 extra care schemes from the public sector (housing association ownership) and the private sector (private, profit-making ownership), representing three different categories of ECH models (small new build, remodelled and extra care villages) was identified from the Elderly Accommodation Counsel (2007) and from data from the evaluation of ECH carried out by the Personal Social Services Research Unit (PSSRU) at the University of Kent (Netten et al., 2011). Housing providers and housing managers were contacted for permission to recruit their ECH scheme into the study. Standard sample size calculations indicated that 128 participants would be needed to detect a medium sized effect ( ¼0.15) with a ¼0.05 and b ¼0.80 on one of the proposed quality of life measures (Cohen, 1998). As individual participants sampled within each scheme may share a similar QoL, a larger sample size of 192 participants was calculated to take into account the clustered structure of the data (Ukoumunne et al., 1999). To allow for missing data the goal was to recruit 10 residents from 25 extra care schemes. Of the 35 schemes sampled and approached to participate in the study, 12 declined to take part. The majority of these were from the private sector, and three schemes refused to participate due to concerns about the ‘over-researching’ of residents. The recruited sample of 23 schemes (3 from the private sector) was located across the country: 11 from the north of England, 5 from the Midlands and 7 from the south of England. Five of the schemes were remodelled former sheltered housing, 6 were extra care villages and 12 were small, new-build schemes. The smallest scheme was comprised of 30 living units and the largest contained 246 living units.

Quality of life was assessed using two separate instruments: the schedule for the evaluation of individual quality of life-direct weighting (SEIQoL-DW) (Hickey et al., 1996; Browne et al., 1997) and the CASP-19 (Hyde et al., 2003). Instruments for measuring quality of life can be classified as either generic or diseasespecific, and can be further classified into those that include individualised values or those that use standardised items (de Leval, 1999; Fayers and Machin, 2000). The SEIQoL-DW is a three stage, interview-based, individualised generic measure of perceived QoL which derives a single QoL score from an individual’s selection and rating of the areas of their life that are most important to them. The SEIQoL-DW has been used with older people in the general population (Mountain et al., 2004; Seymour et al., 2008), frail older people (McKee et al., 2002), and in older people receiving renal replacement therapy (McKee et al., 2005). Respondents are asked to: (1) nominate five areas of their lives that they feel are important in determining their quality of life; (2) rate their current level of satisfaction in each quality of life area on a visual-analogue scale from 0 to 100 and; (3) indicate the relative weight of importance that they place on each quality of life area by using a mechanical device that rotates five discs to produce a pie chart. A SEIQoL-DW score is calculated by multiplying the levels and weights to produce a number ranging from 0 to 100, with higher scores indicating higher quality of life. There are, however, several limitations in administering the SEIQoL-DW to older people. It is time consuming to complete (mean time approximately 38 min) and impaired vision, a lack of manual dexterity, fatigue and confusion impact on the ability of older individuals to complete the second and third stages of the tool (McKee et al., 2002; Mountain et al., 2004). Hence, we modified the second and third stages of the SEIQoL-DW so that participants were only asked to choose and to rank five QoL areas, and to say how satisfied they were with each QoL area on a four point Likert scale (‘‘very satisfied’’, ‘‘satisfied’’, ‘‘dissatisfied’’ or, ‘‘very dissatisfied’’). QoL scores were calculated by summing the multiples of the ranks and satisfaction weights to produce a number ranging from 25 to 100, with higher scores representing higher QoL. QoL data was also collected on the four domains of control, autonomy, self-realisation and pleasure using the CASP-19 (Hyde et al., 2003). The CASP-19 is a generic standardised QoL measure consisting of 19 items which cover the above four theoretical life domains. It has been used with older adults in the English longitudinal study of ageing (Marmot et al., 2003) and the 11th wave of the British Household Panel Survey (Taylor, 2003). Each life domain on the CASP-19 consists of 4 or 5-items presented as statements. For example, ‘‘I feel free to plan for the future’’ (Control), ‘‘I can do the things I want to’’ (Autonomy), ‘‘I choose to do things I’ve never done before’’ (Self-realisation) and ‘‘I enjoy being in the company of others’’ (Pleasure). Respondents assess each statement with regards to their own feelings about life on a four point Likert scale (‘‘this often applies to me’’, ‘‘this sometimes applies to me’’, ‘‘this does not often apply to me’’ or, ‘‘this never applies to me’’). Several of the items are reverse coded for scoring and the scale scores are summed to produce an index of QoL. The possible QoL scores range from 0 to 57, with a higher score indicating a good QoL (Hyde et al., 2003; Wiggins et al., 2008).

3.3. Participants Individual scheme managers were contacted for permission to access their scheme to collect data via the EVOLVE tool and to approach potential participants for interview. Scheme managers were asked to distribute an invitation letter and an information sheet about the study to all of the residents that they deemed capable of taking part in the study. Residents were ineligible if they had severe cognitive impairment or were too frail to complete a lengthy questionnaire-based interview. Eligible residents were asked to inform their scheme manager if they were willing to take part in the study and were asked again for confirmation of their willingness to participate in the study by the researchers prior to being interviewed. A total of 164 residents within the recruited 23 extra care schemes consented to participate in the study and 163 participants produced complete data.

3.4. Materials A questionnaire was developed containing standardised instruments and individual items measuring resident demographic and background characteristics, such as age, gender, length of stay, current type of tenure, satisfaction with accommodation, and quality of life. The EVOLVE tool1 was used to evaluate the quality of building design in the ECH schemes. The main components of the questionnaire and EVOLVE tool are described below. 1 The EVOLVE tool, with instruction for its use, is available to download at http://www.dhcarenetworks.org.uk/IndependentLivingChoices/Housing/Topics/.

3.6. Building design measures—EVOLVE and its domains The EVOLVE tool consists of a comprehensive list of building features (items) classified into 13 ‘‘building user-related domains of interest’’ (see Appendix 1). Individual building features were judged by experts, non-experts and members of the research

A. Orrell et al. / Health & Place 21 (2013) 52–64

team using the Delphi process and the nominal group technique (Pill, 1971; Jones and Hunter, 1995; Hallowell and Gambatese, 2010) to be important to the QoL of building users in ECH schemes. Six of the domains relate to the Universal Needs for anyone regardless of age (comfort & control, dignity, personal care, personal realisation, socialising within scheme, connecting with wider community) and seven domains represent Support for Impairments Associated with Older Age (accessibility, dementia support, physical support, sensory support, safety, security, working care [support for care staff]). Individual building features are also cross classified according to the spaces where they appear, e.g., within the participants’ living units, communal facilities, circulation spaces, staff and services areas or broader aspects of the building’s site and location. Building design was measured on all domains except the dementia support domain using the EVOLVE tool. We omitted the dementia support domain because residents with dementia were excluded from participating in this study. For the purpose of this study an item score of one was recorded if the item was present in all of the applicable building spaces and a score of zero recorded if the item was absent from any of the applicable building spaces. The resulting matrix allows relative building scores for each domain to be calculated using the following formula: (number of items present/total number of items)  100 (c.f. Lewis et al., 2010b). Relative building scores for each domain can thus range from 0 to 100, with higher scores indicating that all or most of the pertinent items for good building design in that domain are present. 3.7. Individual level (level 1) mediator—dependency We focused on dependency, i.e., ability to perform everyday tasks, as a potential mediator of the relationship between building design and QoL at the individual level. Participants’ dependency levels were measured on the 5-item Barthel index (Hobart and Thompson, 2001). This is a short form of the 10-item Barthel index (Mahoney and Barthel, 1965) which captures activities of daily living (ADL), and is suitable for use in clinical trials, epidemiology studies and audit as it has strong agreement with the original 10-item index (Hobart and Thompson, 2001). The 5items on the shortened index relate to: transfer, bathing, toilet use, stairs and mobility. Scores range from 0 to 20, with a low score indicating greater dependency. Dependency was represented as a dummy variable coded 1 for high dependency and 0 for low dependency. Individuals who scored 12 or less on the 5item Barthel index were categorised as being highly dependent for self-care or required significant help, whereas individuals who scored more than 12 were categorised as being largely independent for self-care or required incidental help. 3.8. Individual level (level 1) controls We controlled for key sociodemographic factors that could account for the relationship between building design and QoL, namely age, gender and type of tenure. Gender and tenure were represented as dummy binary variables coded 1 for males and ownership of ECH and 0 for females and renting of ECH, respectively.

55

3.10. Procedure Data collection took place between March 2009 and December 2009. The resident questionnaire was administered via face-toface interviews with participants. The interviews were held in participants’ homes and lasted on average for 60 min. The EVOLVE tool was administered in parallel to the interviews, covering the participant’s living unit and the scheme’s other spaces. Additional building data were collected from architectural plans and consisted of the ratio of private to public areas within the scheme, the net area of the participants’ living units, the travel distance from the participants’ living units to the nearest lift, and the number of bedrooms in the living unit. Recordings were also made of environmental performance in terms of daylighting, artificial lighting and heating. 3.11. Data analysis Preliminary data analyses were performed to determine normality of data, homogeneity of variance and independence of the data using histograms and frequency distributions, box plots and scatter plots. Baseline characteristics were descriptively summarized by using the median and inter-quartile range (IQR), or frequencies and percentages. The age difference between males and females was analysed using an independent t-test. A Pearson product-moment correlation coefficient was calculated for the relationship between QoL scores on the SEIQoL-DW and CASP-19. Standard content analysis procedures were employed to group the QoL dimensions nominated by participants when completing the SEIQoL-DW into generic categories (Smith and Davies, 2010). Due to the clustered structure of the data, multi-level linear regression methods (Goldstein, 1995) were applied to examine the effect of building design on QoL. Residual plots, Cook’s Distance, Mahalanobis Distance and Centred Leverage Value were used to check the model assumptions of linearity, normality and constant variance, and to examine the effects of outlying values. All data were analysed using SPSS version 19.0 and the individual variables were group mean centred (Peugh, 2010). Statistical significance for the multi-level linear regression was assessed with a two-tailed a o0.10. This level of alpha was chosen due to the fact that the attained sample size was lower than desired. With an underpowered analysis, the risk of Type II error is inflated, and the alpha was set so as to counter this problem. One way analyses of variance (ANOVA) using the Student–Newman–Keuls post hoc test were performed to determine if there were differences in the aggregate summary measures for each building type, i.e., small new build, remodelled and extra care village. 3.12. Ethics Ethical approval for this study was obtained from The University of Sheffield Research Ethics Committee. Written informed consent was obtained from all participating schemes and participants.

4. Results

3.9. Building level (level 2) mediator

4.1. Descriptive results for schemes and EVOLVE domains

We focused on scheme size, i.e., the number of living units within the extra care scheme, as a potential mediator of the relationship between building design and QoL at the building level. Size was represented as a dummy variable coded 1 for a large scheme (more than 75 living units) and 0 for a small scheme (less than 75 living units).

Table 1 summarises the scheme characteristics and EVOLVE domain scores for the three types of scheme. Inspection of Table 1 shows that the highest domain scores were for security (69 to 80/100) and personal care (65–70/100) and that the lowest domain scores across schemes were for sensory support (32–35/100), physical support (41–44/100) and comfort & control (31–42/100). Variation in domain

56

A. Orrell et al. / Health & Place 21 (2013) 52–64

Table 1 Building characteristics and domain scores. Characteristic Number of schemes Location Urban Suburban Rural Tenure Social renting Leasehold/joint ownership Mixed Number of living units Number of rooms in living unit Number of bedrooms in living unit Area of living unit (m2) Distance of living unit from lift (m) Number of communal facilities Domain scorea Universal needs Personal care Socialising within scheme Connection with wider community Dignity Personal realisation Comfort & control Support for older age Security Working care Safety Accessibility Physical support Sensory support a

Small new build number (%)

Remodelled number (%)

Extra care village number (%)

12 (52)

5 (22)

6 (26)

2 (16) 5 (42) 5 (42)

2 (40) 3 (60) 0 (0)

0 (0) 4 (67) 2 (33)

6 0 6 Median 42 5 1 53

(50) (0) (50) (Q1, Q3) (39, 50) (4, 5) (1, 1) (48 62)

5 (100) (0) (0) Median (Q1, Q3) 39 (37, 40) 5 (4, 5) 1 (1,1) 42 (41, 47)

1 3 2 Median 217 6 2 62

(17) (50) (33) (Q1, Q3) (78, 225) (3, 6) (1, 2) (43, 85)

23 (13, 42) 10 (8, 13)

23 (8, 31) 8 (8, 10)

13 (4, 45) 16 (14, 18)

65 60 65 58 43 31

(60, (25, (38, (52, (42, (28,

69) 75) 69) 64) 45) 40)

65 50 62 40 41 40

(60, (25, (38, (40, (38, (35,

65) 75) 69) 44) 43) 40)

70 75 33 53 52 42

(55, (50, (31, (38, (43, (30,

71) 100) 54) 57) 53) 48)

69 47 58 53 44 35

(65, (44, (56, (48, (43, (31,

80) 60) 60) 57) 54) 38)

80 67 52 50 43 32

(76, (57, (51, (49, (40, (31,

82) 73) 52) 58) 44) 36)

72 67 53 53 41 34

(53, (55, (52, (52, (40, (29,

72) 71) 54) 60) 48) 36)

Range from ‘0’ lowest domain score to ‘100’ highest domain score. Q1, Q3 are the lowest and upper quartile figures.

scores between scheme types was notable on several domains. Within the category of Universal Needs, connecting with wider community had a notably lower score in extra care villages than small new build and remodelled schemes, while extra care villages had a higher domain score relative to the other two types for socialising within the scheme domain. Remodelled schemes had low scores on the dignity domain relative to the other two scheme types; small new build schemes had a particularly low score on the comfort & control domain, while extra care villages had the highest scores for personal realisation. Within the category of Support for Old Age, remodelled schemes scored highest for security, while small new build schemes had notably low scores on working care. The other four domains within this category demonstrated less variation by scheme type. 4.2. Descriptive and bivariate results for scheme residents Descriptive data for participants are shown in Table 2. More women (n¼106, 65%) than men (n¼57, 35%) took part in the study. There was no significant difference in age between male and female participants, t[1 6 1] ¼ 1.12; p¼.265 (95% CI  4.56, 1.26). The median number of months living in the scheme was 19, and the main form of tenure was social renting (n¼128, 79%). A substantial minority of residents used a wheelchair (n¼67, 41%), and the overall median dependency on the 5-item Barthel was 13. The majority of participants reported having a high QoL (median scores on SEIQoL-DW and CASP-19 were 92 and 39, respectively), and 94% (n¼153) of residents were either satisfied or very satisfied with their accommodation. Table 3 presents the frequencies and rank order of the various dimensions of life that participants nominated as being important to their QoL on the SEIQoL-DW. The most frequently nominated dimensions were ‘‘relationships’’, ‘‘maintaining independence’’ and

‘‘being active’’, whereas the least frequently nominated dimensions were ‘‘financial matters’’, ‘‘religion’’ and ‘‘psychological health’’. Aspects of building design and the physical environment were reflected in the nomination of ‘‘home environment’’ and ‘‘external environment’’ as dimensions important for QoL, respectively, the 6th and 7th most frequently nominated dimensions. 4.3. Multivariate analyses As there was a significant relationship between QoL scores on the SEIQoL-DW and CASP-19 (¼.45, po001), the CASP-19 was selected as the outcome measure for determining the effect of building design on resident QoL. The CASP-19 was selected in preference to the SEIQoL-DW as it demonstrated greater variance and less skew. Table 4 presents the results of the multilevel regression models of the effect of building design (EVOLVE domain scores) on QoL (CASP-19 score). The domains safety, accessibility and working care were significantly negatively associated with QoL; while the domain security was significantly positively associated with QoL. The first column of Table 4 (Model A) reports the results from the unconditional (intercept only) model. In all of these models 9% of the variability in QoL was associated with differences between the schemes. Furthermore, there were significant differences in domain scores across scheme type for safety (F¼21.64 df¼2160 p¼ o0.01), accessibility (F¼4.46 df¼2160 p¼0.01) working care (F¼16.52 df¼2160 po0.01) and security (F¼10.83 df¼2160 po0.01). Post hoc analyses revealed that the score on the accessibility domain was significantly higher for the extra care villages (55) compared to the small new build (52) and remodelled schemes (53), but was significantly lower (64) on the security domain compared to the small new build (71) and remodelled schemes (75). In contrast post hoc analyses disclosed that the score on the

A. Orrell et al. / Health & Place 21 (2013) 52–64

Table 2 Participant characteristics and quality of life outcomes. Characteristic

N (%)

Sample

163

Gender Male Female Age (years)  gender Male Female Marital status Single Married/cohabiting Divorced Widowed Time living in scheme (months) Tenure Social renting Leasehold/joint ownership Mobility Wheelchair user No mobility aid Mobility  Tenure Wheelchair user social housing No mobility aid social housing Wheelchair user leasehold/ joint ownership No mobility aid leasehold/ joint ownership Dependencya Satisfaction with living place Very satisfied Satisfied Neither satisfied nor dissatisfied Dissatisfied Quality of life score (SEIQoL-DW)b Quality of life score (CASP-19)c

Median

(Q1, Q3)

79 80

(71, 83) (75, 86)

19

(9, 37)

13

(9, 20)

92 39

(82, 100) (33, 48)

57 (35) 106 (65)

15 42 19 87

(9) (26) (12) (53)

128 (79) 35 (21) 67 (41) 96 (59) 61 (48) 67 (52) 6 (17) 29 (83)

109 (67) 44 (27) 6 (4) 4 (2)

a

5-item Barthel index, range from ‘0’ maximum dependency to ‘20’ independent. SEIQoL-DW, range from ‘25’ lowest quality of life to ‘100’ highest quality of life. c CASP-19, range from ‘0’ lowest quality of life to ‘57’ highest quality of life. Q1, Q3 are the lower and upper quartile figures. b

Table 3 Quality of life dimensions nominated by participants on SEIQoL-DW. QoL dimension

Rank

Frequency

Relationships (carers/family/friends) Maintaining independence Being active Socialising Physical health Home environment External environment Relaxing and indulging Financial matters Religion Psychological health

1 2 3 4 5 6 7 8 9 10 11

187 167 115 67 59 56 38 33 19 14 13

safety domain was significantly higher for the new build schemes (57) compared to the remodelled (51) and extra care village schemes (52), but was significantly lower (49) on the working care domain compared to the village (58) and remodelled schemes (62). The domains comfort & control, dignity, personal care, personal realisation, connecting with the wider community, socialising within the scheme, physical support and sensory support were not significantly associated with QoL (data not reported in Table 4) and more elaborate models were not developed for these domains. The second column (Model B) adds the individual control variables (age, gender, tenure) that could account for the relationship between

57

building design (domains) and QoL. Of the control variables, only tenure was significant in the models for security and working care. For no domain, however, was Model B a significant improvement over Model A. This is shown by the p value associated with the change in  2LL. The third column (Model C) adds the mediator dependency which was hypothesised to mediate the relationship between building design and QoL at the individual level. We found that dependency had a significant negative association with QoL across all domains (higher levels of dependency were associated with lower QoL). Model C was a significant improvement across all four domains. The fourth column (Model D) adds the mediator scheme size, which was hypothesised to mediate the relationship between building design and QoL at the building level. We found that scheme size had a significant positive association with QoL, and Model D was a significant improvement across all four domains. The fifth column (Model E) adds the cross-level interaction of dependency  scheme size. We found that this interaction term was significant in the models for safety and accessibility, where the term had a negative beta weight (high dependency in large schemes was associated with a lower QoL). Model E, however, was a significant improvement only for the safety domain. Consequently this interaction was broken down for the safety domain by performing separate multilevel models on dependency (low dependency and high dependency) and size (small scheme and large scheme). These models (see Table 5 and Table 6) were the same as the main model but excluded the main effects and interaction terms for dependency and size, respectively. Table 5 presents the results for dependency. The first column (Model A) reports the results for the unconditional (intercept only) model. The domain safety approached significance in predicting a negative QoL in participants with low dependency needs, but did not predict a difference in QoL in participants with high dependency needs. The second column (Model B) adds the individual control variables (age, gender, tenure) that could account for the relationship between safety and QoL. The individual controls did not explain any of the effects of building design on QoL. The third column (Model C) adds the mediator scheme size. We found that scheme size did predict a positive QoL for participants with low dependency needs but did not predict a difference in QoL for residents with high dependency needs. Table 6 presents the results for scheme size. The first column (Model A) reports the results for the unconditional (intercept only) model. The domain safety significantly predicted a negative QoL in participants living in a small scheme, but did not predict a difference in QoL in participants living in a large scheme. The second column (Model B) adds the individual control variables (age, gender, tenure) that could account for the relationship between safety and QoL. The control variables did not explain any of the effects of building design on QoL. The third column (Model C) adds the mediator dependency. We found that dependency did significantly predict a negative QoL for residents living in a large scheme, but did not predict a difference in QoL for residents living in a small scheme.

5. Discussion 5.1. Main findings In our study, which is the first to examine the effect of building design on residents’ QoL in ECH, certain aspects of design were associated with QoL. Notably, elements of design related to accessibility, safety and working care were negatively associated with residents’ QoL, while those related to security were positively associated with QoL. Across all these domains, the relationship between building design and resident QoL was mediated (at the

58

A. Orrell et al. / Health & Place 21 (2013) 52–64

Table 4 Effects of building design features (domains), controls and mediators on QoL score on CASP-19. Domain safety

Fixed effects Intercept Individual level 1 factors Age Gender Tenure Individual mediators Dependency Scheme level 2 factors Size of scheme Interaction Dependency  size Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p

Fixed effects Intercept Individual level 1 factors Age Gender Tenure Individual mediators Dependency Scheme level 2 factors Size of scheme Interaction Dependency  size Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p

Fixed effects Intercept Individual level 1 factors Age Gender Tenure Individual mediators Dependency Scheme level 2 factors Size of scheme Interaction Dependency  size Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p

Fixed effects Intercept Individual level 1 factors Age Gender Tenure Individual mediators Dependency

Model A unconditional Model B (SE)

Model Bþ individual controls B (SE)

Model C þdependency B (SE)

Model D þsize of scheme B (SE)

Model E þ interaction B (SE)

 1.03 (0.47)nn

 0.93 (0.47)nn

 0.87 (0.46)

 0.87 (0.47)

 1.03 (0.43)nn

0.02 (0.09) 0.58 (1.68) 5.30 (3.47)

0.01 (0.09) 0.06 (1.66) 4.21 (3.44)

0.01 (0.09) 0.05 (1.68) 4.19 (3.48)

0.03 (0.09) -0.03 (1.66) 2.70 (3.50)

 3.82 (1.61)nn

 3.90 (1.63)nn

 2.60 (1.71)

nn

4.94 (2.02)

5.11 (1.97)nn  10.54 (4.85)nn

9.08 (6.63)

9.30 (6.62)

9.56 (6.54)

3.22 (5.54)

2.83 (5.26)

0.09

0.10

0.10

0.04

1206.67

p¼ o 0.05nn

1204.02 2.65 3 p 40.05

1198.48 5.54 1 p o 0.05

1193.95 4.53 1 po 0.05

1189.30 4.65 1 p o 0.05

 0.75 (0.40)n

 0.61 (0.41)

 0.44 (0.41)

 0.44 (0.41)

 0.51 (0.41)

0.01 (0.10) 0.43 (1.69) 5.15 (3.54)

0.00 (0.09)  0.03 (1.68) 4.36 (3.50)

0.00 (0.10)  0.03 (1.70) 4.35 (3.54)

0.02 (0.09)  0.12 (1.68) 3.03 (3.57)

 3.65 (1.65)nn

 3.73 (1.67)nn

 2.53 (1.76)

4.96 (2.01)nn

5.12 (1.97)nn  9.47 (4.86)nn

8.95 (6.62)

9.14 (6.62)

9.36 (6.84)

2.95 (5.54)

2.55 (5.29)

0.09 1207.93

po 0.10n

0.09 1205.63 2.30 3 p 40.05

0.10 1200.82 4.81 1 p o 0.05

0.02 1196.24 4.58 1 po 0.05

1192.49 3.75 1 p 40.05

 0.38 (0.20)n

 0.41 (0.20)nn

 0.36 (0.20)

 0.36 (0.20)

 0.36 (0.20)

0.03 (0.09) 0.38 (1.68) 6.79 (3.44)nn

0.01 (0.09)  0.10 (1.67) 5.63 (3.42)

0.01 (0.09)  0.11 (1.68) 5.61 (3.46)

0.03 (0.09)  0.18 (1.67) 4.50 (3.49)

 3.67 (1.62)nn

 3.74 (1.64)nn

 2.69 (1.72)

nn

4.94 (2.01)

5.09 (1.97)nn  8.74 (4.81)

8.81 (6.58)

9.15 (6.58)

9.39 (6.50)

3.03 (5.51)

2.64 (5.28)

0.09 1207.78

po 0.10n

0.09 1203.66 4.12 3 p 40.05

0.10 1198.62 5.04 1 p o 0.05

0.03 1194.06 4.56 1 po 0.05

1190.80 3.26 1 p 40.05

0.67 (0.38)n

0.74 (0.37)nn

0.68 (0.37)

0.68 (0.37)

0.69 (0.37)

0.02 (0.09) 0.73 (1.68) 6.77 (3.45)nn

0.00 (0.09) 0.20 (1.67) 5.59 (3.42)

0.01 (0.09) 0.20 (1.68) 5.58 (3.46)

0.02 (0.09) 0.13 (1.67) 4.46 (3.48)

 3.78 (1.61)nn

 3.86 (1.63)nn

 2.78 (1.72)

A. Orrell et al. / Health & Place 21 (2013) 52–64

59

Table 4 (continued ) Domain safety

Scheme level 2 factors Size of scheme Interaction Dependency  size Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p

Model A unconditional Model B (SE)

Model Bþ individual controls B (SE)

Model C þdependency B (SE)

Model Dþ size of scheme B (SE)

Model E þ interaction B (SE)

4.94 (2.02)nn

5.09 (1.98)nn  8.95 (4.81)

8.92 (6.62)

9.29 (6.62)

9.54 (6.54)

3.20 (5.54)

2.80 (5.30)

0.09 1208.31

0.11 1204.00 4.31 3 p 40.05

0.10 1198.56 5.44 1 p ¼ o 0.05

0.04 1194.06 4.5 1 p o0.05

1190.64 3.42 1 p4 0.05

po 0.10n

p o 0.01.

nnn

n

p o 0.10. p o0.05.

nn

Table 5 Effects of the safety domain, controls and mediators on QoL score on CASP-19 of residents with low and high dependency needs. Domain safety

Fixed effect Intercept Individual level 1 factors Age Gender Tenure Scheme level 2 factors Size of scheme Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p

Low dependency

High dependency

Model A unconditional B (SE)

Model B þindividual controls B (SE)

Model C þ size B (SE)

Model A unconditional B (SE)

Model B þindividual controls B (SE)

Model Cþ size B (SE)

 1.04 (0.56)n

 0.97 (0.57)

 1.07 (0.57)

 0.91 (0.78)

 0.79 (0.75)

 0.90 (0.77)

 0.08 (0.13) 1.34 (1.96) 1.38 (3.86)

 0.06 (0.13) 1.44 (1.97) 0.97 (3.86)

0.15 (0.13)  4.25 (2.95) 10.91 (7.34)

0.17 (0.13)  4.21 (2.94) 8.91 (7.90)

5.97 (2.50)nn 17.57 (10.67)

16.83 (10.49)

7.38 (9.31)

0.19

0.18

0.09

733.77

732.63 1.14 3 p 40.05

728.19 4.44 1 p o 0.05

p o 0.10n

 2.69 (4.01)

459.72

p 40.05

455.47 4.25 3 p 40.05

455.02 0.45 1 p 40.05

p o 0.01.

nnn

n

p o 0.10. p o0.05.

nn

level of the individual) by dependency, which had a negative effect on QoL; by (at the building level) scheme size, which had a positive effect on QoL; or, in the case of the safety domain, by an interaction of these two factors. Other aspects of building design connected to, for example, physical support, sensory support, personal realisation, comfort & control, and dignity, were not found to be associated with residents’ quality of life. There was considerable variation across domains in how well the ECH schemes performed. Schemes performed well in terms of security and personal care, but less well in terms of comfort & control and sensory support. There was also considerable variation in domain scores across scheme types with, for example, small new build schemes relative to the other two types performing well on dignity but poorly on working care; remodelled schemes performing relatively well on security but relatively poorly on dignity; and extra care villages performing relatively well on socialising within the scheme and relatively poorly on connecting with wider community. This variation in scores across domains and

by domain across scheme types suggest that the EVOLVE tool is sensitive to the diverse nature of building design in this form of housing for older people. The majority of residents in all of the schemes reported a good QoL, with the median score on the SEIQoL-DW, for example, approaching the ceiling score of the instrument (92/100). On this instrument, the three most commonly cited dimensions important for QoL were ‘‘relationships’’, ‘‘maintaining independence’’, and ‘‘being active’’. ‘‘Home environment’’ and ‘‘external environment’’ were the 6th and 7th most commonly cited dimensions. Residents reported a high level of satisfaction with their accommodation, with 94% very satisfied or satisfied. 5.2. Support for ageing and QoL The intention of ECH is to enable older people to live independently in appropriate housing for as long as possible, to reduce social isolation and improve QoL (Communities and

60

A. Orrell et al. / Health & Place 21 (2013) 52–64

Table 6 Effects of the safety domain, controls and mediators on QoL score on CASP-19 of residents living in small and large ECH schemes. Domain safety

Fixed effect Intercept Individual level 1 factors Age Gender Tenure Individual mediator Dependency Variance components Intercept (scheme level) ICC  2LL X2Change dfChange p n

Large scheme

Small scheme

Model A unconditional B (SE)

Model B þindividual controls B (SE)

Model C þdependency B (SE)

Model A unconditional B (SE)

Model B þ individual controls B (SE)

Model C þ dependency B (SE)

0.40 (0.92)

0.77 (0.84)

0.46 (0.80)

 1.55 (0.54)nnn

 1.58 (0.54)nnn

 1.48 (0.54)nnn

0.10 (0.27) 3.71 (3.31) 11.69 (4.25)nn

0.22 (0.26) 2.92 (3.10) 8.65 (4.18)nn

0.04 (0.10)  0.31 (1.87)  0.81 (4.96)

0.02 (0.10)  0.65 (1.87)  1.49 (4.93)

 9.18 (4.10)nn 33.36 (33.16)

38.67 (33.29)

34.82 (29.72)

0.27

0.37

0.27

255.54

247.02 8.52 3 po 0.05

242.35 4.67 1 p o0.05

p 40.10

 2.68 (1.72)

939.57

p o0.01nnn

939.32 0.25 3 p 40.05

936.93 2.39 1 p 40.05

po 0.10. nn

p o0.05. po 0.01.

nnn

Local Government, 2008). Contrary to expectations, therefore, low scores were obtained across schemes for the EVOLVE domains physical support and sensory support, and to an extent, accessibility, suggesting that there may be inadequate provision of some of the individual building design features considered to be important for supporting ageing in ECH, particularly for people with high dependency needs. Indeed, Callaghan et al. (2009), in their report investigating the development of social well-being in new ECH schemes, suggest that extra care villages may suit the more able and active older population. Our findings, however, suggest little difference across scheme types in terms of scores on accessibility, physical support and sensory support. Wahl et al. (2009) have shown that the magnitude of accessibility problems in the home environment is strongly associated with dependence in activities in daily living. In our study, however, accessibility was negatively associated with QoL, although this association did not remain significant once other factors such as dependency and scheme size were accounted for. The negative association between accessibility and QoL was mirrored by the negative associations between safety and working care and QoL. One interpretation of this finding is that where buildings perform highly on these domains, the resulting effect on residents is of a living environment that is over-institutional in appearance and not home-like. For example, the living units in ECH schemes have relatively large bathrooms and purpose built assistive bathrooms are available within all schemes, which support the working environment for care staff. Buildings that signify to their users that they are designed for older people (such as where there are many building elements that reflect the needs of disabled or frail users) can have a negative impact on residents via the activation of stereotypes of age and ageing, which have been demonstrated to impair people’s functioning and cognitive performance (Levy, 2000; Stein et al., 2002). As the majority of participants in our sample were relatively mobile (59% required no mobility aid), it could be argued that building features that support older age will not impact highly on their satisfaction with their living environment or their QoL, as such features are not expressly meeting their own needs as users. Indeed, following the earlier argument, the support for old age features that do exist

may, for such residents, have a detrimental effect on well-being and self-esteem. Nevertheless, it is clearly important that housing for older people is an environment that supports independence. Findings of studies by Netuveli et al. (2006) and Molzahn et al. (2010) indicate that maintaining independence in activities of daily living is associated with a good QoL in older people. This is echoed in our study by the frequency with which residents cited ‘‘maintaining independence’’ and ‘‘being active’’ as important for QoL, and by the fact that dependency was associated with low QoL when added to all our domain score models. Yet there is a balance to be struck between supporting independence and manifesting an institutional atmosphere. Previous research investigating building design and QoL in residential and nursing homes for older people using the Sheffield Care Environment Assessment Matrix (SCEAM) demonstrated a negative association between safety and low dependency residents’ QoL (Parker et al., 2004). The authors suggested that care homes adopted a risk-averse environment to comply with health and safety legislation. For example, residents’ independence and ability to move freely inside and outside the home was compromised by the use of fire-protection devices that kept doors, which required considerable force to reopen them, closed. Parker et al. (2004) concluded that alternative and available solutions should be incorporated into the design of buildings for older people to reduce this conflict between safety and independence. Although the ECH schemes in our sample had incorporated newer safety features and assistive technologies into their building design in order to address the issues of safety and independence, such as automatic fire door closers which allow fire doors to remain open, our findings reflect those of Parker et al. (2004). We found a negative association between buildings’ scores on the safety domain and residents’ QoL, with additional variance in residents’ QoL accounted for by the interaction of resident dependency and scheme size. This finding highlights the difficulty and conflicts in incorporating good building design in ECH which maximises provision for safety and health whilst maintaining independence for older people. Good security within extra care schemes is a key issue for residents (c.f. Barnes et al., 2012). In this study, scores on the

A. Orrell et al. / Health & Place 21 (2013) 52–64

security domain were found to positively predict residents’ QoL, and scores for security within our sample of schemes was high. At first, such a finding might be thought to counter the previous argument about the negative effect of an ‘institutional’ environment on QoL. ECH residents, however, have the ability to lock their own front doors, and internal security within the schemes is evidenced by the provision of progressive privacy (Fletcher et al., 1999) incorporated within the building design as well as the use of physical barriers, such as doors, lifts, corridors and separate wings to keep public and private areas separate. As such, security is realised in a relatively covert manner, and rather than provoking a feeling of institutionalisation, the knowledge that one lives in a secure environment (both in terms of the scheme’s design and location) could have a beneficial psychological effect on residents. Further work, however, is required to determine what influences whether residents experience levels of security positively or negatively. 5.3. Universal needs and QoL Social interactions are important for older people’s QoL (Netuveli et al., 2006; Bowling and Gabriel, 2007; Street et al., 2007; Seymour et al., 2008; Golden et al., 2009). Indeed, in a national study of old age, Bowling and Gabriel (2007) reported that the main attributes important to respondents for their QoL included social relationships and social activities. Furthermore, Golden et al. (2009) reported that when age, sex, depression, cognitive impairment and disability were controlled for, social engagement was associated with a better QoL in older people and Thomas (2010) suggests that it is often better for the well-being of older adults to give rather than to receive social support. In our study, ‘‘relationships’’ was the factor most frequently cited by participants as important to their QoL, with ‘‘being active’’ the third most commonly cited factor. With respect to the EVOLVE domain scores attained by the schemes, whilst participants’ ability to socialise within schemes with other residents, friends and family members appeared to be well supported, especially within the village schemes, connecting with wider community was substantially lower in the villages than in the smaller new build and remodelled schemes. This differentiation in domain scores across scheme categories may, in part, be explained by the number of facilities available on site in villages for residents. Furthermore, extra care villages are more likely to make their facilities available to non-residents than new build and remodelled schemes (Croucher, 2006). Hence residents in village schemes do not have to leave the site in order to experience and to maintain social relationships and social activities. Scheme locality may also impact on whether people choose to socialise within the scheme or outside in the wider community. In our study none of the village schemes was in an urban location, whereas two of the small new build and two of the remodelled schemes were in urban localities. This may reflect the greater flexibility of smaller schemes. Thus villages may have to trade off location for size. For some older people living in schemes in rural areas, the cost and limited transport services available for accessing services and social activities located in urban and/or regional centres may restrict people’s ability to connect with the wider community (Manthorpe et al., 2008), thereby promoting and encouraging the utilisation of the facilities available within the scheme. Extra care villages also performed better than the other two scheme types on the personal realisation domain. This may be due to the overall median size of the living units in the extra care villages being greater than those of the other scheme types, providing more flexibility for personal expression in furniture,

61

decoration, and so on. The living units also had more bedrooms, which residents could utilise for activities of their choice. Schemes scored well on personal care, reasonably well on dignity, and poorly on comfort & control. ‘‘Personal care’’ and ‘‘dignity’’ are currently high profile topics within policy and practice debates on the status of care for older people, with the suggestion from some actors that such concepts should be used to drive up the standard of care (Tadd et al., 2011). As such, the provision in housing design of environmental support for achieving care standards of this type is important. Comfort and control is afforded in housing through the provision of operable windows, blinds, air-conditioning and large and easy to operate technology interfaces (van Hoof et al., 2008). Thermal comfort, described as ‘‘the state of mind, which expresses satisfaction with the thermal environment’’ (ASHRAE, 2004), or rather a lack of it, is known to restrict the health and well-being of older people and can induce problematic behaviour in people with dementia (van Hoof et al., 2010). Hence, good ECH building design should give residents the opportunity to adjust the thermal environment inside their own living units to suit themselves. Thirdly, despite, however, the apparent importance for residents’ QoL (and indeed generically for QoL across the population) of such issues as personal realisation, personal care, dignity, and comfort and control, our findings suggest that scheme scores on these and other domains within the ‘universal needs’ category were not in themselves associated with resident QoL. Perhaps this finding is to be expected. If the environment is conceptualised as a ‘‘distal’’ factor for QoL, ‘‘proximal’’ factors such as mood/mental health, social activity and relationships and health and mobility would all be theoretically expected to have a more profound impact on QoL than the built environment (Lawton, 1983; Lawton et al., 1995). This is reflected in our findings, which showed that the vast majority (94%) of our sample was either ‘‘very satisfied’’ or ‘‘satisfied’’ with their living place, despite low scheme scores on the Universal Needs domains of comfort & control and personal realisation, and also on the Support for Older Age domains of physical support and sensory support. This apparent contradiction may reflect reluctance among older people to report dissatisfaction (Bowling, 2002). A similar picture emerged from the focus group discussions held with residents in the early stages of the project where residents expressed overall satisfaction with the extra care lifestyle, but made considerable criticisms of the building designs in detail (Barnes et al., 2012). Inflated levels of satisfaction with accommodation might also be a reflection of some participants having lived in poor quality housing before moving into ECH, as older people in the UK are the age group most likely to occupy non-decent homes (Donald, 2009). In our sample, however, the home environment and the external environment ranked 6th and 7th, respectively, with regards to their importance for individual QoL. These findings suggest that there may be little contradiction between having relatively high levels of QoL and yet at the same time occupying a somewhat less than excellent living environment; the environment is important, but not as important as other factors. In the context of ECH schemes, these other factors might be, for example, the quality of care provided by staff, the nature of one’s relationship with other residents, and one’s own health (reflected in the importance of dependency for QoL in our findings). The role of the environment should not be downplayed, however, as scheme size was found to be a very significant factor in the relationship between building design domains and residents’ QoL, with larger scheme size associated with higher QoL. It is possible that the effect of scheme size on QoL operates partly through certain features, such as individual living unit size, number of communal facilities etc. There may, however, be advantages to living in larger schemes in terms of variables that

62

A. Orrell et al. / Health & Place 21 (2013) 52–64

are more difficult to measure, such as the atmosphere created by a larger-scale community and an environment that is more selfsufficient and self-contained. It is worth noting, for example, that on all but one of the Universal Needs domains, the larger scheme type, the extra care village, scored better than the other two scheme types.

5.4. Study limitations The main limitations of this study are fourfold. First, the results of this study cannot be easily generalized to the population of older people in ECH because our sample consisted of self-selected participants who were initially selected by the scheme managers and this may have introduced bias. Second, due to a smaller than intended sample size, the power of our study was reduced and the risk of a type II error increased, whereby important relationships between building design and the QoL of people living in ECH are dismissed because of a lack of statistical significance. This problem was countered by accepting a higher than usual alpha for determining statistical significance, but this also leads to an accompanying risk of inflated Type I error. Third, this was a cross-sectional rather than a longitudinal study. Consequently, we were only able to capture a snapshot of people’s QoL on the day that they were interviewed and were only able to observe associations rather than distinguish causal effects. Finally, the use of broad QoL indicators as an outcome variable may have been inappropriate. Previous research has shown that limited effect sizes are obtained with regards to the association between personal and environmental characteristics (Spokane, 1985) and that this may be attributable to the use of unsuitable outcomes where the association may only have a small role to play (Holland, 1987). Rather than measuring global QoL, other outcome measures theoretically linked to the physical environment, such as social activity, social wellbeing and loneliness, may have been more sensitive to the impact of building design on the individual.

6. Conclusion This is the first study to investigate the relationship between building design of ECH and QoL. The EVOLVE tool provided a detailed description of the ECH schemes in our sample, and was sensitive to variation within and across scheme types. The QoL of residents in our sample was found to be good, with high levels of satisfaction expressed, despite scores on the EVOLVE domains suggesting that the schemes in the study could be better designed to meet the needs of older people, and to promote such universal needs as choice and control and personal realisation. Where the environment was found to be related to resident QoL, high scores on the security domain suggest that the covert forms of security found in ECH promotes well-being, while the negative effect of high scores on domains such as accessibility, safety and working care, suggest the designers of ECH must strive to build in features that support frail older users in such a way that an institutional atmosphere is not created, and that signifiers of ‘‘old age’’ and frailty are kept to a minimum. The detail offered by the EVOLVE tool in its description of the scheme environments thus offered a more sophisticated understanding of the relationship between the environment and resident QoL than would a tool that provided a global or unidimension assessment. Our participants did nominate the environment as important to their QoL, but less often than such dimensions as relationships, maintaining independence, and being active. Thus, the personal attribute of dependency was overall more influential on residents’ QoL than the environment in our analyses, mediating the effect of the EVOLVE domain scores. Nevertheless, the importance of scheme size as a correlate of QoL in our analyses underlines the power of the environment to create a foundation for how older residents of ECH can experience a good quality of late life. Overall, therefore, our results suggest that the design of ECH can support and promote social capital and care, which is in keeping with the UK National Strategy for Housing in an Ageing Society (2008).

Table A1 Domain Universal needs Comfort and control Dignity Personal care Personal realisation Connecting with wider community Socialising within scheme Support for older age Accessibility

Physical support Sensory support Dementia supporta Safety Security Working care

a

Description

Extent to which an individual can control their internal environment. Examples of this are; (1) dwellings have a manual thermostat and (2) all radiators have an individual thermostatic temperature control Extent to which the building allows other people to respect personal space and privacy in personal care. Examples of this are: (1) dwellings have a front door bell and (2) assisted bathrooms are accessed via private rather than public spaces Extent to which the building allows residents to be independent in daily activities performed for self-care, such as feeding, bathing and toileting. Examples of this are: (1) dwellings have a shaver point in the bathroom and (2) there is a WC close to the communal facilities Extent to which the building allows people to undertake the activities of their choice within their home. Examples of this are: (1) the main bedroom can accommodate twin beds and (2) there are areas in the garden where residents can grow plants and vegetables. Is the way in which the scheme fits into the wider community? Examples of this are: (1) scheme is located within an existing residential neighbourhood and (2) the scheme is located within a 400 m travel distance of a public transport terminus Extent to which the building allows people to socialise in both public and private spaces. Examples of this are: (1) the activity room seating arrangement allows personal conversation and (2) the garden has area for activities, such as, bowling Extent to which the building allows people, especially wheelchair users, to independently move around various parts of the scheme including dwellings, communal facilities and circulation spaces. Examples of this are: (1) level thresholds and (2) there are at least two lifts. Is the way in which the building enables residents with physical disabilities to use the scheme successfully? Examples of this are: (1) handrails in corridors and (2) large cross type or lever handle taps the residents; bathrooms and kitchens Is the way in which the building enables residents with sensory impairments to use the scheme successfully? Examples of this are: (1) rooms have natural light and (2) use of non-reflective finishes on doors Extent to which the building supports the needs of people with dementia. Examples of this are: (1) clear way finding and (2) provision of reference views Extent to which the building provides a safe and healthy environment for residents. Examples of this are: (1) non-slip floors in the bathrooms and (2) handrails do not have protruding ends that can catch on clothing or handbags Extent to which the building provides a secure environment for residents. Examples of this are: (1) the front door to the resident’s dwelling has an entry phone system and (2) there is a security barrier between public and private areas Extent to which the building provides an environment that allows staff to deliver the highest level of care. Examples of this are: (1) there is more than 600 mm round the sides of the washbasin to allow a carer to be present and (2) there are low level shower enclosures in the bathrooms to prevent carers from getting wet

Domain omitted in the analysis in this paper.

A. Orrell et al. / Health & Place 21 (2013) 52–64

Issues remain to be resolved about whether such schemes can promote QoL in both relatively independent and frail older users, with the level of support for older age designed into current ECH schemes less than could be expected.

Acknowledgements This study was supported by a grant from the Engineering and Physical Sciences Research Council, UK and by our partners the Department of Health Housing Learning and Improving Network (LIN) and EAC (Elderly Accommodation Counsel). We would also like to express our gratitude to the residents and staff at the extra care schemes participating in this study and to Chris Parker for her help with the analysis and Jacquetta Holder for her advice.

Appendix A See Appendix Tabels A1. References Adams, S., White, K., 2006. Older People, Decent Homes and Fuel Poverty: An Analysis Based on the English House Condition Survey. Help the Aged, London. American Association of Retired Persons, 2003. These four walls. Americans 45 þ talk about home and community. /http://assets.aarp.orgS. ASHRAE, 2004. Ansi/ashrae standard 55-2004, thermal environmental conditions for human occupancy, Atlanta, GA, USA. American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. Baltes, M., Maas, I., Wilms, H-U., Borchelt, M., Little, T., 1999. Everyday competence in old and very old age: theoretical considerations and empirical findings. In: Baltes, K., Mayer, K. (Eds.), The Berlin Aging Study. Cambridge University Press, Cambridge, pp. 384–402. Barnes, S., Torrington, J., Darton, R., Lewis, A., McKee, K., Netten, A., Orrell, A., 2012. Does the design of extra care housing meet the needs of residents? A focus group study. Ageing & Society 32, 1193–1214. Bowling, A., 2002. An ‘‘inverse satisfaction law’’? Why don’t older people criticise health services?. Journal of Epidemiology and Community Health 56, 482. Bowling, A., Gabriel, Z., 2007. Lay theories of quality of life in older age. Ageing & Society 27, 827–848. Browne, J., O’Boyle, C., McGee, H., McDonald, N., Joyce, C., 1997. Development of a direct weighting procedure for quality of life domains. Quality of Life Research 6, 301–309. Callaghan, L., Netten, A., Darton, R., 2009. The development of social well-being in new extra care housing schemes. /www.jrf.org.uk/sites/files/jrf/care-andwell-being-full.pdfS. Cohen, J., 1998. Statistical power analysis for the behavioral sciences, second ed. Lawrence Erlbaum Associates, Publishers, Hillsdale, NJ. Communities and Local Government, 2008. Lifetime homes, lifetime neighbourhoods: a national strategy for housing in an ageing society. /www.commu nities.gov.ukS. Croucher, K., 2006. Making the Case for Retirement Villages. Joseph Rowntree Foundation, York. ¨ Darton, R., Baumker, T., Callaghan, L., Holder, J., Netten, A., Towers, A., 2012. The characteristics of residents in extra care housing and care homes in England. Health and Social Care in the Community 20, 87–96. de Leval, N., 1999. Quality of life and depression: symmetry concepts. Quality of Life Research 8, 283–291. Department of Health, 2001. National service framework for older people. Department of Health, London. /http://www.doh.gov.uk/nsf/olderpeople. htmS. Department of Health, 2010. A vision for adult social care: capable communities and active citizens. Department of Health, London. /http://www.dh.gov/en/ Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/ DH_121508S. Donald, I.P., 2009. Housing and health care for older people. Age and Ageing 38, 364–367. Elderly Accommodation Counsel, 2007. Guide to Extra Care Retirement Housing in England. EAC, London. Evans, G., Kantrowitz, E., Eshelman, P., 2002. Housing quality and psychological well-being among the elderly population. Journal of Gerontology B: Psychological Sciences 57, 381–383. Fayers, P., Machin, D., 2000. Quality of Life. Assessment, Analysis and Interpretation. Wiley, Chichester. Fletcher, P., Riseborough, M., Humphries, J., Jenkins, C., Whittingham, J., 1999. Citizenship and Services in Older Age: The Strategic Role of Very Sheltered Housing. Beaconsfield: Housing 21.

63

Gershuny, J., Lader, D., Short, S., 2005. The time use survey 2005: how we spend our time. /www.statistics.gov.uk/cci/article.asp?ID=1600S. Gitlin, L., 2003. Conducting research on home environments: lessons learned and new directions. The Gerontologist 43, 628–637. Golden, J., Conroy, R., Lawlor, B., 2009. Social support network structure in older people: underlying dimensions and association with psychological and physical health. Psychology, Health & Medicine 14, 280–290. Goldstein, H., 1995. Multilevel Statistical Models. Arnold, London. Hallowell, M.R., Gambatese, J.A., 2010. Qualitative research: application of the Delphi method to CEM research. Journal of Construction Engineering and Management 136, 99–107. Hickey, A., Bury, G., O’Boyle, C., Bradley, F., O’Kelly, F., Shannon, W., 1996. A new short form individual quality of life measure (SEIQoL-DW): application in a cohort of individuals with HIV/AIDS. British Medical Journal 313, 29–33. Hobart, J., Thompson, A., 2001. The five item Barthel index. Journal of Neurology, Neurosurgery and Psychiatry 71, 225–230. Holland, J.L., 1987. Some speculations about the investigation of personenvironment transactions. Journal of Vocational Behavior 31, 337–340. Hyde, M., Wiggins, R., Blane, D., Higgs, P., 2003. A measure of quality of life in early old age: the theory, development and properties of a needs satisfaction model (CASP-19). Aging and Mental Health 7, 86–94. Jones, J., Hunter, D., 1995. Qualitative research: consensus methods for medical and health services research. British Medical Journal 311, 376. Lawton, M.P., 1983. The varieties of wellbeing. Experimental Aging Research 9, 65–72. Lawton, M.P., DeVoe, M.R., Parmelee, P., 1995. Relationship of events and affect in the daily life of an elderly population. Psychology and Aging 10, 469–477. Levy, B., 2000. Handwriting as a reflection of aging self-stereotypes. Journal of Geriatric Psychiatry 33, 81–94. Lewis, A., Torrington, J., Barnes, S., Darton, R., Holder, J., McKee, K., Netten, A., Orrell, A., 2010a. EVOLVE: a tool for evaluating the design of older people’s housing. Housing Care and Support 13 (3), 36–41. Lewis, A., Torrington, J., Barnes, S., Darton, R., Holder, J., McKee, K., Netten, A., Orrell, A., 2010b. EVOLVE/http://www.dhcarenetworks.org.uk/IndependentLi vingChoices/Housing/Topics/S. Mahoney, F., Barthel, D., 1965. Functional evaluation: the Barthel index. Maryland State Medical Journal 14, 61–65. Manthorpe, J., Iliffe, S., Clough, R., Cornes, M., Bright, L., Moriarty, J., 2008. Older people researching social issues, 2008. Health and Social Care in the Community 16, 460–468. Marmot, M., Banks, J., Blundell, R., Lessof, C., Nazroo, J., 2003. Health, Wealth and Lifestyles of the Older Population in England: The 2002 English Longitudinal Study of Ageing. Institute of Fiscal Studies, London. McKee, K., Houston, D., Barnes, S., 2002. Methods for assessing quality of life and well-being in frail older people. Psychology & Health 17, 737–751. McKee, K., Parker, S., Elvish, J., Clubb, V., Nahas, M., Kendray, D., Creamer, N., 2005. The quality of life of older and younger people who receive renal replacement therapy. Ageing & Society 25, 903–923. Molzahn, A., Skevington, S., Kalfoss, M., Makaroff, K., 2010. The importance of facets of quality of life to older adults: an international investigation. Quality of Life Research 19, 293–298. Mountain, L., Campbell, S., Seymour, D., Primrose, W., Whyte, M., 2004. Assessment of individual quality of life using the SEIQoL-DW in older medical patients. Quarterly Journal of Medicine 97, 519–524. ¨ Netten, A., Darton, R., Baumker, T., Callaghan, L., 2011. Improving Housing with Care Choices for Older People: An Evaluation of Extra Care Housing. Personal Social Services Research Unit. University of Kent. Netuveli, G., Wiggins, R., Hildon, Z., Montgomery, S., Blane, D., 2006. Quality of life at older ages: evidence from the English longitudinal study of ageing (wave1). Journal of Epidemiology & Community Health 60, 357–363. Office for National Statistics National Population Projections 2009. /http://www. statistics.gov.uk/statbase/Product.asp?vInk=8519S. Oswald, F., Wahl, H-W., 2004. Housing and health in later life. Reviews of Environmental Health 19, 223–252. ¨ Oswald, F., Wahl, H-W., Schilling, O., Nygren, C., Fange, A., Sixsmith, A., Sixsmith, J., Sze´man, Z., Tomsone, S., Iwarsson, S., 2007. Relationships between housing and healthy aging in very old age. The Gerontologist 47, 96–107. Parker, C., Barnes, S., McKee, K., Morgan, K., Torrington, J., Tregenza, P., 2004. Quality of life and building design in residential and nursing homes for older people. Ageing & Society 24, 941–962. Peugh, J.L., 2010. A practical guide to multilevel modelling. Journal of School Psychology 48, 85–112. Pill, J., 1971. The Delphi method: substance, context, a critique and annotated bibliography. Socio-Economic Planning Science 5, 57–71. Rubenstein, R., De Medeiros, K., 2004. Ecology and the aging self. In: Wahl, H-W., Scheidt, R., Windley, P. (Eds.), Annual Review of Gerontology and Geriatrics. Aging in Context: Socio-Physical Environments. Springer, New York, pp. 59–84. Scheidt, R., Windley, P., 2006. Environmental gerontology: progress in the postLawton era. In: Birren, J., Schaie, K. (Eds.), Handbook of the Psychology of Aging. Elsevier, Amsterdam, pp. 105–125. Seymour, D., Starr, J., Fox, H., Lemmon, H., Deary, I., Prescott, G., Whalley, L., 2008. Quality of life and its correlates in octogenarians. Use of the SEIQoL-DW in wave 5 of the Aberdeen Birth Cohort 1921 Study (ABC1921). Quality of Life Research 17, 11–20.

64

A. Orrell et al. / Health & Place 21 (2013) 52–64

Smith, K., Davies, J., 2010. Qualitative data analysis. In: Dahlberg, L., McCaig, C. (Eds.), Practical Research and Evaluation: A Start-to-Finish Guide for Practitioners. SAGE, London, pp. 145–158. Spokane, A., 1985. A review of research on person-environment congruence in Holland’s theory of careers. Journal of Vocational Behavior 26, 306–343. Stein, R., Blanchard-Fields, C., Hertzog, C., 2002. The effects of age-stereotype priming on memory performance of older adults. Experimental Aging Research 28, 169–181. Street, D., Burge, S., Quadagno, J., Barrett, A., 2007. The salience of social relationships for resident well-being in assisted living. Journal of Gerontology B: Psychological Sciences 62, S129–134. Tadd, W., Hillman, A., Calan, S., Calan, M., Bayer, A., Read, S., 2011. Right place– wrong person: dignity in the acute care of older people. Quality in Aging and Older Adults 12, 33–43. Taylor, M., (Ed.) with Brice, J., Buck, N., Prentice-Lane, E., 2003. British Household Panel Survey User Manual. Volume A, Introduction. Technical Report and Appendices. UK Data Archive, University of Essex. Thomas, P., 2010. Is it better to give or to receive? Social support and well-being of older adults. Journal of Gerontology B: Psychological Sciences 65, 351–357. Ukoumunne, O., Gulliford, M., Chinn, S., et al., 1999. Methods for evaluating areawide and organisation-based interventions in health and health care: a systematic review. Health Technology Assessment. Winchester, England 3 (5) pp: iii-92.

van Hoof, J., Kort, H.S.M., Hensen, J.L.M., 2008. Thermal comfort and HVAC design for people with dementia. In: Proceedings of the IFA’s Ninth Global Conference on Ageing, Montreal: International Federation on Ageing, pp. 1–18. van Hoof, J., Kort, H.S.M., Hensen, J.L.M., Duijnstee, M.S.H., Rutten, P.G.S., 2010. Thermal comfort and the integrated design of homes for older people with dementia. Building and Environment 45, 358–379. Wahl, H-W., Gitlin, L., 2007. Environmental gerontology. In: Birren, J. (Ed.), Encyclopedia of Gerontology. Age, Aging, and the Aged. Elsevier, Oxford, pp. 494–501. Wahl, H-W., Schilling, O., Oswald, F., Iwarsson, S., 2009. The home environment and quality of life-related outcomes in advanced old age: findings of the ENABLE-AGE project. European Journal of Ageing 6, 101–111. Walker, A., Lowenstein, A., 2009. European perspectives on quality of life in old age. European Journal of Ageing 6, 61–66. Group, WHOQoL, 1998. The World Health Organisation quality of life assessment (WHOQoL): development and general psychometric properties. Social Science and Medicine 46, 1569–1585. Wiggins, R., Higgs, P., Hyde, M., Blane, D., 2008. Quality of life in the third age: key predictors of the CASP-19 measure. Ageing & Society 24, 693–708. Wilkes, D., 2007. Mini cost model: extra care housing. /http://www.building.co. uk/data/mini-cost-model-extra-care-houisng/3099666.articleS.