Long title: Complementarity of welfare provision in the 'mixed economy' of care for carers of people with dementia: a longitudinal study

Transcription

1 Long title: Complementarity of welfare provision in the 'mixed economy' of care for carers of people with dementia: a longitudinal study Short title: Mixed Economy for CAre in DementiA (MECADA) Project Reference: RES Awarded to: Georgina Charlesworth 1, Paul Higgs 1 and Fiona Poland 2 Institutions: University College London 1 and University of East Anglia 2 Full report Background The concept of a mixed economy of care is one in which elements of planned care provision are shared with elements derived from more free market sources in which user and consumer choice may also feature. In relation to care in the UK such a mixed economy of care is seen as sharing of responsibility between state agencies, voluntary organisations and private and commercial organisations to deliver local support and care services. People with dementia have considerable care needs and their family carers often provide intensive care over an extended period of time. Assessment of family carers in their own right has come to the fore in the UK over the past decade since the introduction of the Carers Acts and National Strategy for Carers. This emphasis on the need for care for carers has coincided with the development of local markets in care with social service departments being encouraged to provide fewer direct care services themselves and increase commissioning of independent service providers (Ware et al., 2001). Despite the increased focus on carers needs and the development of a mixed economy of care, the interplay between providers of care for carers had not received attention. It was the aim of this research to study the interplay between different types of welfare provision using post-carers Act data on a sample of family carers of people with dementia to identify whether or not there was evidence for welfare pluralism (similar kinds of service being offered by different actors), substitution/'crowding out' (as one service increases, another is 'pushed out') crowding in (presence of 2 or more services 1

2 being provided in parallel, with once provider potentially acting as either a magnet or bridge to subsequent services or form of support) or complementarity (activities differ between actors, & are unaffected by each other). Data for this project was drawn from the Befriending and Costs of Caring (BECCA) project (Charlesworth et al, 2008; The BECCA trial was a multisite randomised controlled cost-effectiveness trial of befriending for family carers of people with dementia, funded by the Health Technology Assessment Primary Research programme. A total of 236 family carers of people with dementia were assessed and randomised into the trial, and over 80% of participants were followed up for 2 years post-baseline assessment. The trial was approved by the Eastern Multi Regional Ethics Committee (MREC), and by the 5 Local Ethical Research Committees (LRECs) in Norfolk and Suffolk, and by the Barking and Havering LREC. All participants gave informed consent for their anonymised responses to be pooled with others for analysis and dissemination. Based on existing literature relating to service receipt by older people and family carers, we hypothesised that: There would be evidence of both welfare pluralism (where there is duplication of support services by different providers) and complementarity (where different providers provide different services e.g. emotional vs instrumental) Carers would receive more emotional support provision (e.g. support groups) from the voluntary sector, & more instrumental support provision from statutory sector (including support from private providers commissioned by Local Authorities) There would be evidence that contact with voluntary sector services acts as a bridge to use of statutory services Receipt of welfare provision will be influenced by gender and kinship of the carer to the care-recipient with dementia, but not by measures of burden or social isolation. support provision from family and friends would diminish over time. 2

3 Objectives The specific objectives of the project were to: 1) review existing research and policy literature on welfare provision for carers of people with dementia, and the relationships between formal and informal care in this population 2) describe the quantity (intensity and duration) of instrumental and emotional support received from statutory, voluntary and 'informal' (family and friends) sources, and investigate the relationships between carer burden, social isolation, kinship and welfare provision (a) cross-sectionally and (b) longitudinally using the Befriending and Costs of Caring (BECCA) dataset 3) present findings in the context of theories of service use and service interrelationships, and to describe the implications for policy 4) disseminate finding to appropriate groups Each of the 4 objectives has been addressed. Methods and findings for objectives 1 and 2 are presented in turn, followed by a discussion of the findings in the context of previous work (objective 3) and a consideration of the implications for policy (also objective 3). Objective 4 (dissemination) is described in the sections of the report on activities and outputs. Objective 1 A range of literatures was searched, including academic and social policy. Identified documents highlighted the need to consider social support networks and short breaks. Both support networks and respite breaks formed part of the following analysis, and can therefore be considered in the context of existing research and policy. Objective 2 This objective involved the secondary analysis of data from the Befriending and Costs of Caring (BECCA) dataset, describing informal support, the support receipt from statutory, private and third /voluntary sectors, and the cross-sectional and longitudinal relationships between carers characteristics and welfare provision. 3

4 Methods Participants Participants were carers of people with dementia who were providing 20 or more hours per week care for a person with a primary progressive dementia living in the community when recruited to the BECCA trial. Recruitment for the trial took place through primary, secondary and voluntary sector care in Norfolk and Suffolk (East Anglia, UK), and the London Borough of Havering, with some self-referrals following media publicity or word of mouth. Of the 236 participants, 64% were female, 67% were spouses and 80% were co-resident with the person with dementia. The mean age of the carer was 68 years (sd 11.4) and the mean duration of care provision was 4.1 years (sd 3.8). The mean age of the care-recipients was 78 years (sd 8.7). Measures Information was collected through semi-structured interviews using psychometric questionnaires and a resource use / service receipt data collection tool. The subset of data analysed for the current study included selected demographic characteristics, level of carer burden (Carers Assessment of Difficulties Index; Nolan & Grant, 1992), and use of day centres and sitting services, homecare and extended respite. Where such services had been used, information had also been collected on the service provider (voluntary, statutory or private sector). Support from family and friends was assessed in three ways, namely perceived support (Multidimensional Social Support Scale; Zimet et al, 1988), received support ( Do any relatives, friends or neighbours currently provide daily and/or weekly help or support?(regular support); Have you had any help or support from family, friends or neighbours on an occasional basis? (occasional support), and support networks. The assessment of support networks was the Practitioner Assessment of Network Type (PANT; Wenger, 1994) consists of 8 items/questions items assessing frequency of contact with family, friends and neighbours, geographical proximity to family, and information on local social ties (participation in social clubs, groups, religious meetings). Responses are used to classify individuals into one of five network types (Family Dependent, Locally Integrated, Local Self-Contained, Wider Community Focused, Private Restricted), or classified as Inconclusive. The characteristics of each of these networks are summarised in Charlesworth et al. (2007) (a nominated output). The instrument has been found to correlate highly with a range of demographic variables, 4

5 level of service use and response to interventions. PANT items were also used to construct a Social Isolation Index akin to that of Scharf & Smith (2004). The Index provided a measure of Social Isolation ranging from zero not isolated to 3 extremely isolated. Data Analyses Data was analysed using SPSS version 12.0 for Windows (2005). The phases of analysis were: dataset reorganisation, analysis of missing variables, use of descriptive statistics and finally cross-tabulations (with Chi 2 statistic) and comparisons of means (with t or F-tests). Levels of missing data were small in comparison to the sample size. Dataset reorganisation Received support: Participants receiving regular and occasional support were classified as receiving regular support. Service provider: the original BECCA dataset had separate labels for NHS and local authority provision. These were joined to form a single statutory sector variable. Sitting services: Information on Sitting Services had been recorded variously under the categories of homecare receipt and carer services. A new sitting services variable was created with data extracted from home care and carers services variables. Respite Breaks: Sitting service and Daycare variables were combined to create the Respite Break variable, reflecting the function of daycare from the perspective of most carers. Emotional care : in the protocol we had expressed an intention to compare emotional and instrumental support. However, inspection of the data within the category carers services revealed limitation in the quality of data available. There was often no indication of service provider and the focus of interventions not clear. Therefore the emotional vs instrumental support service comparison was not carried out. 5

6 Results of cross-sectional analyses Of the 226 carers in the cross-sectional analysis for whom support receipt data was available, 69 (31%) were not receiving any informal support; 59 (26)% received such support occasionally (less than weekly) and 48 (44%) received it regularly. For those receiving regular support, family/friends replaced the carer for a mean of 2 hours per week (s.d.=4), and assisted for 8 hours per week (s.d.=19). Of the 234 carers for whom service receipt data was available, 117 (50%) care-recipients attended daycare, homecare was in place for 61 (26%), 42 (18%) carers used a sitting service and 56 (24%) care-recipients had spent time in respite accommodation in the 6 months prior to baseline assessment. In total 156 (67%) carers had made use of at least 1 of these 4 services. In other words, the receipt of regular (weekly or daily) support from family, friends or neighbours was more common in the sample than was use of any one service, but overall service receipt was more common than informal support. Table 1: Cross-tablulation of support received from family, friends and neighbours (regular, occasional or none) with service receipt (receipt of 1 or more services vs no service use) (n=226*) Regular support (n=98) Occasional (n=59) None (n=69) n(%) n(%) n(%) Services 1 or 70 (71) 40 (68) 40 (58) more (n=150) None (n=76) 28 (29) 19 (32) 29 (42) Chi 2 = 3.58 (p=.187) * Missing support receipt data for 10 carers Table 1 shows a cross-tablulation of receipt of carers informal care (regular, occasional or none) and receipt of services (1 or more vs none). Almost three quarters (71%) of carers receiving regular informal support were also in receipt of 1 or more service 6

7 whereas only 58% of those without informal support used services. This would be in keeping with the concept of crowding in. However, there was no significant relationship between the receipt of services and the receipt of family support (chi 2 =3.36, p=.187). Having considered the relationship between formal and informal support, we turn to the characteristics of carers associated with each of the formal (table 2) and informal (table 3) care categories. Although more spouses use services than non-spouses, this is due to the high number of spouse carers overall. It is the non-spouses and their relatives with dementia who are significantly more likely to use services. Users of services also report having cared for significantly longer than do non-users of services and to report higher burden. There were no differences between those who did and not use services in terms of age, gender, co-residency with the person with dementia, social isolation or perceived social support. 7

8 Table 2: Characteristics of carers who do, and do not, use services at baseline (n= 234*) Use of at least 1 No use of services service (n=156) (n=78) n(%) n(%) 2 (p-value) Gender female 102 (65) 50 (64) male 54 (35) 28 (36).038 (.846) Kinship spouse 96 (62) 62 (80) non-spouse 60 (38) 16 (20) (.006) Co-resident Yes 135 (87) 68(87) No 21 (13) 10 (13).019 (.892) Social isolation* Not isolated 52 (34) 22 (29) Low isolation 45 (29) 25 (32) Moderate 41 (27) 24 (31) High 15 (10) 6 (8) (.754) Mean (sd) Mean (sd) t-test (p-value) Age 67.1 (11.9) 68.8 (10.7) (.101) Duration of care (4.3) 3.2 (2.3) 3.1 (.002) MSPSS (9.63) (9.37) (.45) CADI (9.97) (8.85) 6.707(.00) ** 2 cases missing service use data; *4 of the 124 missing data for social isolation MSPSS = Multidimensional Scale of Perceived Social Support; CADI = Carers Assessment of Difficulties Index; + Equal variance could not be assumed (Levene) 8

9 Table 3: Characteristics of carers who do, and do not, receive support from family, friends or neighbours at baseline (n= 234) Regular Occasional None (n=98) (n=59) (n=69) n(%) n(%) n(%) 2 (p-value) Gender female 66 (67) 39 (66) 39 (57) male 32 (33) 20 (34) 30 (43) 2.25 (.325) Kinship spouse 54 (55) 51 (86) 50 (72) non-spouse 44 (45) 8 (14) 19 (28) (.000) Co-residence Yes 76 (78) 57 (97) 63 (91) No 22 (22) 2 (3) 6 (9) (.001) Social isolation Not isolated 41 (42) 19 (32) 11 (16) Low isolation 27 (28) 16 (27) 25 (36) Moderate 21 (21) 20 (34) 22 (32) High 8 (8) 4 (7) 9 (13) (.028) Mean (sd) Mean (sd) Mean (sd) F (p-value) Age 66.4 (11.6) 70.8 (9.9) 68.1 (12.4) 2.71 (.068) Duration of care 3.89 (2.91) 4.38 (4.67) 4.46 (4.42).509 (.602) MSPSS (8.58) (8.68) (9.73) (.000) CADI (11.27) (9.50) (10.58).360 (.698) * cases missing for SI ** cases missing service use MSPSS = Multidimensional Scale of Perceived Social Support CADI = Carers Assessment of Difficulties Index 9

10 As for table 2 (formal service use), it is the non-spouses in table 3 (informal support) who are proportionately more likely to receive support. In contrast to table 2, levels of social isolation and perceived support differ significantly between the levels of informal support. Receipt of informal support is associated with greater perceived support whereas perceived social support is unrelated to receipt of services. In addition level of service use, but not quantity of family support is associated with level of carer burden. The received support measure does not take into account the social network available to a carer. Two carers may receive similar levels of support, but for one that could represent a fraction of the potential support whereas for another it could be at ceiling. Using data on social networks takes into account the broader availability (or otherwise) of informal support. Table 4: Day care use at baseline, by carer network type (n=216 * ) Use (N=108) n (%) Non use (N=108) n (%) Family dependent 22 (20) 19 (18) Locally integrated 31 (29) 16 (15) Local self-contained 25 (23) 21 (19) Wider community-focused 7 (7) 21 (19) Private restricted 23 (21) 31 (29) * 20 missing 10

11 Table 5: Respite care use at baseline, by carer network type (n=214 * ) Use (n=52) n(%) Non use (n=162) n(%) Family dependent 14 (27) 26 (16) Locally integrated 15 (28) 33 (20) Local self-contained 12 (23) 34 (21) Wider community-focused 1 (2) 26 (16) Private restricted 10 (19) 43 (27) * 22 missing There was a statistically significant relationship between network type and use of day care (table 4; chi 2 (4)=13.54, p=.009) and also between network type and use of respite care and respite (table5; chi 2 (4)=10.90, p=.028). Carers in locally integrated networks were more likely to make use of respite and daycare whereas carers in private restricted network less likely to make use of day care or respite care. This is again indicative of a crowding in of support where those with the higher levels of available support from family, friends and/or neighbours are also likely to use services. Conversely those with the least access to informal support are also most likely to be non-users of services. Having considered the (lack of) relationship between services and family support for carers, we now turn to look at the different providers of services. An implicit expectation was that there would be evidence of a mixed economy, with services provided by a range of sectors. Figure 1 demonstrates the plurality of welfare providers for any one service-type. However, it is only in Home Care that more than one sector is meaningfully involved. The statutory sector was the predominant provider of daycare, and the voluntary sector the main provider of sitting services. 11

12 Figure 1: Number of carers in receipt of services at baseline by provider (statutory, voluntary, private) Number of carers Statutory Voluntary Private 20 0 Daycare Homecare Sitting Figure 2 takes into account the 3 study localities and illustrates issues of localisation in commissioning. For example, home care was most commonly provided by the statutory sector in Norfolk, and by the private sector in Suffolk; sitting services were most commonly provided by the voluntary sector, especially in Norfolk. Caution is required when interpreting the data for three reasons. Firstly, numbers are small for all categories other than for statutory sector daycare. Secondly the characteristics of the person with dementia are not considered, and some specialist services may only be available within a particular sector in any one locality. Thirdly, the categorisation to sector was based on the carers belief of the service provider. Commissioning and joint funding (part-payment by families and local authorities) may have led to mis-perceptions of the provider by the carer. For example, although most sitting services in Havering are provided through CrossRoads, there was no recorded voluntary sector provision of sitting services in this population. This may be due to the particular organisation of services in the locality where health, special and voluntary sector services were all based in the same building. It may therefore have been particularly difficult for carers to identify the service provider. In spite of these notes of caution, figure x nonetheless demonstrates the influence of localisation in the mixed economy of welfare provision. 12

13 Figure 2: Number of carers in receipt of each service type by both provider type (statutory, voluntary and private) and location (Norfolk, Suffolk, Havering) Statutory Voluntary Private Statutory Voluntary Private Number of carers Statutory Voluntary Private Havering Suffolk Norfolk Daycare Homecare Sitting Objective 2b: Longitudinal analyses Participants Of the 236 carers who undertook baseline assessment, 46 carers had withdrawn from the study by 2 year follow-up, most commonly due to ill-health. Ninety-four carers were no longer caring for their relative at home (38 care-recipients in residential care; 56 carerecipients had died). The total number of participants who continued to care for the person with dementia at home was 96. The mean age of the longitudinal sample at baseline was 67 years (sd 10.6) and the duration of care was 4.4 years (sd 4.1). Approaching two thirds (64%) were female and 76% were spouses. The mean age of the care-recipient with dementia was 75 years of age (sd 9.2). 13

14 Results of longitudinal analyses Details of informal support were available for 93 of the 96 carers in the longitudinal sample. The proportion of carers in each support category (none, occasional, regular) at baseline was similar in this group to the larger sample in the cross-sectional analysis. At follow-up, one in 6 carers received no support (n=16, 17%), one in 3 (n=34, 34%) had occasional support and almost half received regular support (n=45, 48%). The proportion of carers receiving no support reduced over time whereas the proportion in both the occasional and regular categories increased. This is contrary to the hypothesis that family support would decrease over time. However, the mean total number of hours per week respite provided for carers by family, friends or neighbours was only 1 hours per week (sd=3.17), and the number of hours assistance per week was 4.3 (sd 19.8). Figure 3:The number of carers in the longitudinal sample who were in each received support category at baseline and follow-up Baseline Follow -up None Occasional Regular 14

15 Table 6: Baseline characteristics of carers who do, and do not, receive support from family, friends or neighbours at 2 year follow-up (n= 93 * ) Regular (n=45) Occasional (n=32) None (n=16) n(%) n(%) n(%) 2 (p-value) Gender female 28 (62) 21 (66) 9 (56) male 17 (38) 11 (34) 7 (44).400 (.819) Kinship spouse 28 (62) 30 (94) 13 (81) non-spouse 17 (38) 2 (6) 3 (19) (.005) Co-residence Yes No Social isolation Not isolated 18 (40) 9 (28) 2 (13) Low isolation 11 (24) 10 (31) 8 (50) Moderate 11 (24) 12 (38) 5 (31) High 4 (9) 1 (3) 1 (6) + Mean (sd) Mean (sd) Mean (sd) F (p-value) Age 65.2 (11.3) 70.2 (8.74) 66.0 (11.7) 2.26 (.111) Duration 4.1 (2.8) 4.2 (3.1) 6.2 (7.6) 1.66 (.197) MSPSS 46.7 (9.2) 44.9 (9.98) (6.9) 9.30 (.000) CADI 52.5 (9.56) 51.4 (11.6) 59.8 (14.0) 3.34 (.040) * 3 cases missing for Received Support (regular, occasional, non) variable MSPSS = Multidimensional Scale of Perceived Social Support CADI = Carers Assessment of Difficulties Index + Statistical analyses not performed due to small numbers in high proportion of cells. 15

16 Baseline variables kinship, perceived support and carer burden were all significantly associated with informal support at follow-up. Carers not in receipt of family care at follow-up had reported the highest levels of carer burden at baseline. The analysis of carer characteristics associated with service use / non-use was not repeated for the longitudinal data as so few carers and their care-recipients were no longer in receipt of any formal services. In parallel with increased number of carers in receipt of informal support over two years, receipt of services also increased (see figure 4). Figure 4: Comparison of the number of carers in the longitudinal analysis either not using services, or using 1, 2 or 3 of respite breaks, homecare and/or extended respite Baseline Follow-up No. service t ypes The pattern of service providers associated with each type of services remained similar over time, but the major player (s) for each type of service became more prominent over time (figure 5). 16

17 Figure 5: Bar charts showing service use by sector (statutory, voluntary and private) at baseline and follow-up for the N=96 carers in the longitudinal analysis Baseline Follow-up Daycare Home care Brief 'at home' respite Residential respite Stat Vol Private N ca 25 rer s Daycare Home care Brief 'at home' respite Residential respite Stat Vol Private The final two tables in this report look at the relationship between social networks and the type of service use for the longitudinal sample (tables 6 and 7). In both tables the social network categorisation from baseline. The most common service configurations at baseline were no service use (n=42) followed by use of respite breaks only (that is use of sitting services and day care for the person with dementia). After 2 years, use of respite breaks was the most common configuration (n=25), followed by a combination of home care and respite breaks (n=22). Although carers in all network types increased their service use over time, this was particularly marked for carers in the Locally Integrated network type. Carers in the Wider Community Focussed network type showed least increase in service use. Numbers were too small for formal statistical analysis, however, there were qualitatively different patterns of service use associated with the various network categories. The overall pattern is again of those carers with the best links locally also being the ones with the highest levels of service use. 17

18 Table 6: N (%) of carers in the longitudinal analysis in each social network who were using respite breaks, home care and/or extended respite at baseline (N=92) Family dependent (n=15) Locally integrated (n=22) Local selfcontained (n=12) Wider communityfocused Private restricted (n=21) Inconclusive (n=7) (n=15) None (n=43) 5 (33) 6 (27) 4 (33) 10 (67) 14 (67) 4 () Respite breaks 3 (20) 8 (36) 2 (17) 3 (20) 2 (10) 2 () only (n=20) Home care only - 2 (9) 2 (17) - 1 (5) (n=5) Extended respite only (n=0) Home care & 2 (13) 2 (9) 2 (17) 1 (7) 1 (5) 1 () Respite breaks (n=9) Home care & (5) Extended Respite (n=1) Respite breaks & 4 (27) 1 (5) 1 (8) 1 (7) 1 (5) Extended respite (n=8) Respite breaks, Home Care and Extended Respite (n=6) 1 (7) 3 (14) 1 (8) - 1 (5) 18

19 Table7: N (%) of carers in the longitudinal analysis in each baseline social network who were using respite breaks, home care and/or extended respite at follow-up (N=92) None (n=14) Respite breaks only (n=25) Home care only (n=5) Extended respite only (n=1) Home care & Respite breaks (n=22) Home care & Extended Respite (n=0) Respite breaks & Extended respite (n=10) Respite breaks, Home Care and Extended Respite (n=15) Family dependent (n=15) Locally integrated (n=22) Local selfcontained (n=12) Wider communityfocused (n=15) Private restricted (n=21) Inconclusive (n=7) - 2 (9) 1 (8) 5 (33) 4 (19) 2 (29) 5 (33) 3 (14) - 8 (53) 6 (29) 3 (43) 1 (7) 1 (5) 1 (8) - 1 (5) 1 (14) (5) - 4 (27) 6 (27) 7 (58) 1 (7) 4 (19) (27) 3 (14) 1 (8) - 2 (10) - 1 (7) 7(32) 2(17) 1(7) 3 (14) 1 (14) 19

20 Objective 3: Discussion of findings In keeping with international comparisons, there was evidence for accumulation of help from statutory and family supporters ( crowding in ) (Kundemund & Rein, 1999; Motel- Klingebiel et al., 2005). As with studies of interaction between voluntary and statutory provision for family carers, there was no evidence for substitution theory (Dahlberg, 2004, 2005, 2006). Taking the results as a whole, none of the existing models of interplay between support providers provides an adequate description of the patterns found in the BECCA dataset of care provision for family carers of people with dementia. The results reflect the influence of local circumstances and local markets on patterns of social care. Providers are increasingly informed by commissioning processes that reduce the scope for innovation in service provision and increasingly blurs distinctions between sectors (Baines et al., 2008). Implications for policy The mixed economy can be seen as a way of sharing different costs while also offering more variety and choice to service users in relation to meeting their needs. Choice may be considered an advantage from a free market perspective, but can be perceived as overwhelmingly complex form the perspective of carers. The challenges to households within the mixed economy of care have previously been raised (Baldock and Ungerson, 1994). Testament to the complexity of care managing are the difficulties that older people faced when given the opportunity to hold Individual Budgets (DH, 2005) in the recent pilots (Glendinning et al. 2008). active engagement with the mixed economy depends less on income, needs or access to information but more on participants views on what is appropriate 'consumer behaviour'. Activities Methods and findings were discussed with two consultation groups during the course of the project. The first, an academic advisory group, was convened by Paul Higgs at UCL and assisted with considering data in the context of research, theory and policy. The second, a consumer reference group, was convened by Fiona Poland at the University of East Anglia and drew on the experience of service users and direct service providers. All 3 grantholders, and the project researcher, have all presented at UK or international 20

21 conferences on aspects of the project. An end of project dissemination event was held at UCL that was advertised through the British Society for Gerontology listings. Outputs All project outputs have been submitted to the Information Centre via the ESRC Society Today website. The 11 project outputs include: 2 journal publications, 1 book chapter, 4 conference presentations, the 2 consumer consultation meetings, academic advisory group event and a project dissemination event. The two nominated outputs are a conference poster on the relationships between social networks and service use cross-sectionally (Higgs et al., 2007), and a journal article on the social networks/service use relationship longitudinally (Charlesworth et al., 2007). Further outputs are planned, as described in the End of Award form. 21

22 Impacts Prior to this work, the pattern of welfare provision for carers of people with dementia had received little attention. Findings are of direct relevance and use to policy makers, funders of carers services, and the third sector. There may be actual benefits for carers in terms of impact on awareness of different aspects of service provision, the importance of taking into account carers social networks, and considering the potential for knowledge on social networks to inform planning for future services. Findings have also informed an application to the Wellcome Trust, led by Professor Martin Orrell (UCL) and involving Paul Higgs and Georgina Charlesworth. The links made with actual and potential members of the academic advisory and consumer reference groups are being maintained to carry forward to current and future work, including the NIHR funded SHIELD project (Support at Home: Interventions to Enhance Life in Dementia) led by Prof Martin Orrell and involving both Georgina Charlesworth and Fiona Poland. Future Research Priorities It is intended that this work form the basis of a future programme of research in which further issues of support for family carers are addressed. For example, to study the impact of welfare provision on carer health and well-being, and the value of carer support on the wellbeing of carers as they move through transitions on the course of caring (e.g. carer recipient moving into residential or nursing care; bereavement). It is also hoped that the study will encourage more attention being focused on the evolving nature of marketised social care provision on models of social policy both in relation to national and more localised circumstances. Word count:

23 References Aneshensel, C.S., Pearlin, L.I., Mullan, J.T., Zarit, S.H. & Whitlach, C.J. (1995) Profiles in Caregiving: the unexpected career. Academic Press: London. Baines S., Wilson R. & Hardill I. (2008) "'They Just Don't Understand Us!': Learning and Reflection from Commissioning Relationships in the Mixed Economy of Care" at: NCVO/VSSN Researching the Voluntary Sector Conference, University of Warwick. Warwick, UK. Baines, S., Lie, M. & Wheelock, J. (2006) Volunteering, self help and citizenship in later life. A collaborative research project by Age Concern Newcastle and the University of Newcastle upon Tyne. Newcastle: University of Newcastle and Age Concern. Baldock, J. & Ungerson, C. (1994) Becoming consumers of community care: households within the mixed economy of welfare. Joseph Rowntree Foundation in conjunction with Community Care. Charlesworth, G., Tzimoula, X., Higgs, P. & Poland, F. (2007) Social networks, befriending and support for family carers of people with dementia. Quality in Aging - Policy, practice and research, 8(2), Dahlberg, L (2004) Substitution in statutory and voluntary support for relatives of older people. International Journal of Social Welfare, 13, Dahlberg, L. (2005) Interaction between voluntary and statutory social service provision in Sweden: a matter of welfare pluralism, substitution or complementarity? Social policy and Administration, 39, Dahlberg, L. (2006) The complementarity norm: service provision by the welfare state and voluntary organisations in Sweden. Health and Social Care in the Community, 14, 4, Department of Health (1996) Carers (Recognition and Services) Act 1995: Policy Guidance. Department of Health, London Department of Health (1999)Caring about Carers: a National Strategy for Carers. Department of Health, London Department of Health (2005) Independence, Well-being and Choice: Our vision for the future of Adult Social Care London: Department of Health Department of Health (2007) Our NHS, Our Future: NHS next stage review (interim report London: Department of Health 23

24 Department of Health (2008) Making personal budgets work for older people: developing experience. London: Department of Health Glendinning, C., Challis, D., Fernandez, J.-L, Jacobs, S., Jones, K, Knapp, M., Manthorpe, J., Moran, N., Netten, A., Stevens, M. & Wilberforce, M. (2008) Evaluation of the Individual Budgets Pilot Programme Final Report: Social Policy Research Unit, University of York Higgs, P., Tzimoula, X., Poland, F. & Charlesworth, G. (2007) To those who have more is given ; Network type and service use in England. Poster Presentation at the Gerontological Society of America 60th Annual Scientific meeting, 16th-20th November, San Francisco. HM Government (2007) Putting People First: A shared vision and commitment to the transformation of Adult Social Care London: Department of Health nes%20%20et%20al.pdf Kunemund, H. & Rein, M. (1999) There is more to receiving than needing: theoretical arguments and empirical explorations of crowding in and crowding out. Ageing and Society, 19, Motel-Klingebiel, A., Tesch-Roemer, C. & von Kondratowitz, H-J. (2005) Welfare states do not crowd out the family: evidence for mixed responsibility from comparative analysis. Ageing & Society, 25, Nolan, M.R., Grant, G.J. (1992) Regular respite: an evaluation of a hospital rota bed scheme for elderly people: London:Age Concern. Paxton, W., Pearce, N., Unwin, J. & Molyneux, P. (2005). The voluntary sector delivering public services: Transfer or transformation? York, Joseph Rowntree Foundation. Pickard, L. (2001) Carer Break or Carer-blind? Policies for Informal Carers in the UK. Social Policy & Administration 35 (4), Poland, F., Tzimoula, X., Higgs, P. & Charlesworth, G. (2007) Longitudinal changes in social networks and carer support for family carers of people with dementia. Paper presentation at Social Network Conference, 14th July, Queen Mary University, London. Scharf, T. & Smith, A.E. (2004) Older people in urban neighbourhoods: addressing the risk of social exclusion in later life. In: C Phillipson & G Allan (Eds). Social Networks and Social Exclusion: Sociological and Policy Perspectives. Aldershot: Ashgate Publishing Ltd. 24

25 Ware, P., Matosevic, T., Forder, J., Hardy, B., Kendall, J., Knapp, K., and Wistow, G. (2001) Movement and Change: Independent Sector domiciliary care providers between 1995 and Health and Social Care in the Community 9(6), Ware, T., Matosevic, T., Hardy, B., Knapp, M., Kendall, J., and Forder, J., (2003) Commissioning Care Services for Older People in England: the View from Care Managers, Users and Carers. Ageing and Society 23, Weiner, K., Stewart, K., Hughes, J., Challis, D., Darton, R. (2002) Care Management Arrangements for Older People in England: Key areas of Variation in a National Study. Ageing and Society 22, Wenger, G.C. (1994) Support Networks and Dementia. International Journal of Geriatric Psychiatry Wenger, G.C. (1994) Support Networks of Older People: A Guide for Practitioners. Centre for Social Policy Research and Development, University of Wales, Bangor. Wenger, G.C. (1996) Social networks and gerontology. Reviews in Clinical Gerontology Wenger, G.C. (1997) Social networks and the prediction of elderly people at risk. Aging and Mental Health 1(4) Wenger, GC. (1991) A network typology: from theory to practice. Journal of Aging Studies, 5 (2): Zimet GD, Dahlem NW, Zimet SG, Farley GK. The Multidimensional Scale of Perceived Social Support. J Pers Assess :