39

RESEARCH BRIEFING

Preventing loneliness and social isolation: interventions and outcomes

By Karen Windle, Jennifer Francis and Caroline Coomber

Key messages
Older people are particularly vulnerable to social isolation or loneliness owing to loss of friends and family, mobility or income. Social isolation and loneliness impact upon individuals quality of life and wellbeing, adversely affecting health and increasing their use of health and social care services. The interventions to tackle social isolation or loneliness include: befriending, mentoring, Community Navigators, social group schemes. People who use befriending or Community Navigator services reported that they were less lonely and socially isolated following the intervention. The outcomes from mentoring services are less clear; one study reported improvements in mental and physical health, another that no difference was found. Where longitudinal studies recorded survival rates, older people who were part of a social group intervention had a greater chance of survival than those who had not received such a service. Users report high satisfaction with services, beneting from such interventions by increasing their social interaction and community involvement, taking up or going back to hobbies and participating in wider community activities. Users argued for exibility and adaptation of services. One-to-one services could be more exible, while enjoyment of group activities would be greater if these could be tailored to users preferences. When planning services to reduce social isolation or loneliness, strong partnership arrangements need to be in place between organisations to ensure developed services can be sustained. We need to invest in proven projects. Community Navigator interventions have been shown to be effective in identifying those individuals who are socially isolated. Befriending services can be effective in reducing depression and cost-effective. Research needs to be carried out on interventions that include different genders, populations and localities. There is an urgent need for more longitudinal, randomised controlled trials that incorporate standardised quality-of-life and cost measures.

October 2011

Review date: October 2014

RESEARCH BRIEFING 39

Introduction
This is one in a series of research briengs about preventive care and support for adults. Prevention is broadly dened to include a wide range of services that:

promote independence prevent or delay the deterioration of wellbeing

resulting from ageing, illness or disability

brieng focuses on services aimed at reducing the effects of both loneliness and social isolation. Although the terms might have slightly different meanings, the experience of both is generally negative and the resulting impacts are undesirable at the individual, community and societal levels.

What is the issue?

There are a number of population groups vulnerable to social isolation and loneliness, (e.g. young care-leavers, refugees and those with mental health problems). Nevertheless, older people (as individuals as well as carers) have specic vulnerabilities owing to loss of friends and family, loss of mobility or loss of income.3 In consequence, there has been a policy concentration on this group.47 The statistics on population ageing in the UK (and in many developed countries) are well known. Those aged 60 and above currently account for approximately 20 per cent of the population and this proportion is expected to rise to 24 per cent by 2030.8 In the next 20 years, the population of those aged over 80 will treble and those over 90 will double.9 In exploring prevalence, it is estimated that across the present population aged 65 and over, between 5 and 16 per cent report loneliness,10 while 12 per cent feel socially isolated.9 In looking at the experiences of a nationally representative sample, Victor et al11 found that 2 per cent of individuals reported that they were always lonely, 5 per cent that they were often lonely and 31 per cent rated themselves as sometimes lonely. Such gures are likely to expand with increasing family dispersal and growing numbers of older people and the older-old those aged 80 and over.12

delay the need for more costly and intensive services.

Preventive services represent a continuum of support ranging from the most intensive, tertiary services such as intermediate care or reablement, down to secondary or early intervention, and nally, primary prevention aimed at promoting wellbeing. Primary prevention is generally designed for people with few social care needs or symptoms of illness. The focus therefore is on maintaining independence and good health and promoting wellbeing.1 The range of these wellbeing interventions includes activities to reduce social isolation, practical help with tasks like shopping or gardening, universal healthy living advice, intergenerational activities and transport, and other ways of helping people get out and about. Just as the range of wellbeing services is extensive, so too is the available literature examining how well they work. For this research brieng, the focus has been narrowed to the effectiveness and cost-effectiveness of services aimed at preventing social isolation and loneliness. Our review question was: To what extent does investment in services that prevent social isolation improve peoples wellbeing and reduce the need for ongoing care and support? While social isolation and loneliness are often used interchangeably, one paper2 examined the distinct meanings that people attach to each concept. Loneliness was reported as being a subjective, negative feeling associated with loss (e.g. loss of a partner or children relocating), while social isolation was described as imposed isolation from normal social networks caused by loss of mobility or deteriorating health. This 2

Why is it important?
Perhaps not surprisingly, social isolation and loneliness impact on quality of life and wellbeing,1315 with demonstrable negative health effects.12 Being lonely has a signicant and lasting effect on blood pressure, with lonely individuals having higher blood pressure than their less lonely peers. Such an effect has been

Preventing loneliness and social isolation: interventions and outcomes

found to be independent of age, gender, race, cardiovascular risk factors (including smoking), medications, health conditions and the effects of depressive symptoms.16 Loneliness is also associated with depression (either as a cause or a consequence) and higher rates of mortality.9,15,17,18 A recent meta-analysis found that people with stronger social relationships had a 50 per cent increased likelihood of survival than those with weaker social relationships. In understanding such a gure, this would mean that by the time half of a hypothetical sample of 100 people had died, there would be ve more people alive with stronger social relationships.19 As the authors argue, the inuence of social relationships on the risk of death are comparable with well-established risk factors for mortality such as smoking and alcohol consumption and exceed the inuence of physical activity and obesity.19 Such negative impact on individuals health leads to higher health and social care service use, while lonely and socially isolated individuals are more likely to have early admission to residential or nursing care.15,18,20 The benets to individuals and the wider community of reducing loneliness or social isolation are therefore self-evident. For the individual, mitigating loneliness will improve quality of life.3,14,15,20 Similarly, such changes may impact on subsequent health and social care service use, limiting dependence on more costly intensive services and contributing to the healthy ageing agenda8 by compressing morbidity.21 Supporting social engagement also provides benets to the wider community. Reducing social isolation enables a possible harnessing of potential contribution to the community through, for example volunteering2224 and caring responsibilities. Given such individual wellbeing, health status, nancial and wider community imperatives, there has been a national and international policy consensus4,7,27,28 that support must be provided to ameliorate social isolation and to reach those living with or on the brink of loneliness.3 There is less clarity as to the most effective type of intervention or the sector responsible for delivery (e.g. statutory or third sector). As will be

discussed, the available interventions and their evidence base have been developing incrementally.

What sorts of interventions are used in reducing social isolation or loneliness?

We have classied the wide variety of interventions to address social isolation or loneliness as one-to-one interventions, group services and wider community engagement.3,13,14

One-to-one interventions
These include: befriending,13,14,17,23,24,29 mentoring8,9 and gatekeeping (Community Navigator or Waynder initiatives).14,29,30 Befriending has been dened as an intervention that introduces the client to one or more individuals, whose main aim is to provide the client with additional social support through the development of an afrming, emotion-focused relationship over time.17 The process of the intervention differs between individual programmes, but usually involves volunteers or paid workers visiting an individual in their own home (or place of care) on a regular, usually non-time limited basis. Other models have evolved to include telephone and group befriending.13,14 31 The type of assistance that each befriender provides can also differ, but always includes companionship and may involve provision of transport and the completion of small errands such as picking up medications or shopping. Befrienders work with an extremely wide range of populations: those living with health problems (e.g. individuals with dementia and their carers, those with ongoing mental health problems); those who are going through a transitional life phase (e.g. young people leaving care); and those who want the opportunity to access and enjoy social activities within the community, but who need some support to do so (e.g. those with learning difculties, older people with mobility problems). Many of the befriending schemes have emerged from the community level to ll the social and emotional gap that may not be met by 3

RESEARCH BRIEFING 39

existing statutory health and social service provision31 and are run through community or voluntary organisations, although funding can be provided from the statutory sector. Mentoring concentrates on achieving agreed individual goals: Mentoring is dened as a relationship between the volunteer and the individual, based on meeting agreed objectives set at the outset and where a social relationship, if achieved, is incidental.32 Mentors will work with the client (often) on a short-term basis, and thus one key goal is to provide clients with the necessary skills and abilities to ensure that they are able to continue and sustain any achieved change following withdrawal of the service.8,9 As with befriending schemes, mentors work within the umbrella of community or voluntary organisations and across populations, including the most vulnerable (e.g. young offenders, refugees, victims of domestic violence). Waynders or Community Navigators are usually volunteers who provide hard-to-reach or vulnerable people with emotional, practical and social support, acting as an interface between the community and public services and helping individuals to nd appropriate interventions. The structure and processes of this type of service vary across localities and are dependent on population need. For example, those Community Navigators working with frail older individuals may carry out a series of home-based face-to-face visits to discuss concerns and plan, alongside the older person, what service or community provision may be benecial. For less frail populations a telephone conversation may be more appropriate, followed by written information that the individual can access and take forward if they so choose.27,33

groups13,14,30 that cover a number of areas (e.g. bereavement, friendship, creative and social activities, health promotion). Their structure and way of working depend on the needs of the population to whom the intervention is addressed. For example, a group focused on social activities can be open to all14 while another wishing to build self-efcacy and independence for older socially isolated women would be restricted to the original group members to ensure an appropriate sense of sharing and safety.3436 Such groups can be highly structured to achieve specic aims30 or more organic, developing activities dependent on the interests of the group members.33 Facilitation of groups can be peer-led or carried out by specialist staff within health and social care.18,30 Social group schemes also include those focused on rehabilitation and health promotion. For example, one programme, Lifestyle Matters, involved individuals aged 60 and over living in the community and attending an eight-month course (two hours per week) at which they explored a number of healthy living areas including health and ageing, health through physical and mental activity and endings and new beginnings.30

Wider community engagement

Wider community engagement includes programmes that support individuals to increase their participation in existing activities (e.g. sport, use of libraries and museums)3 as well as to use and join outreach programmes37 and volunteer schemes.2224,28 One example of an outreach programme is the professionally conducted choir run by the Levine School of Music in Washington, DC.37 Older people (mean age 79) were recruited to the choir and attended weekly singing rehearsals for 30 weeks as well as putting on public performances of their work. Volunteer schemes are extremely broad, involving the structured engagement of befriending or mentoring or, for example, community organised Time Banks that use hours of time rather than currency and where the type of support volunteers undertake depends on their own skills as well as the needs of the wider community.29

Group services
Supportive interventions that fall within group services include day centre-type services (such as lunch clubs), and social group schemes which aim to help people widen their social circles.3 The number and extent of services is thus broad. Those interventions within social group schemes incorporate self-help and self-support 4

Preventing loneliness and social isolation: interventions and outcomes

What does the research show?

The outcomes of interventions to prevent social isolation
The wide variety of interventions and their different outcome measures make it difcult to be certain what works for whom. The only clear nding is that there is, as yet, no conclusive empirical evidence that computer and/or internet usage impacts on loneliness, physical or psychological outcomes.14,30,38 Some evaluations, either singly reported or incorporated within systematic reviews, have argued that such interventions are effective in reducing loneliness.30,39 However, small samples and inadequate matching of comparison or control groups have led to unreliable outcomes. For example, in one study that provided computer and follow-up internet training to 12 older people the authors stated: We cannot conclude that the reduction in loneliness observed among the participants could be attributed to the intervention.39 In contrast, there is some evidence that group interventions (e.g. closed self-help groups) are more effective than one-to-one support (e.g. telephone support services).13,14 Nevertheless, when individual studies are explored, there are differential outcomes: some group activities have no impact while there are specic one-to-one interventions that are seemingly effective. In bringing together the available evidence for this section of the brieng, 12 papers are included that used validated outcome measures and assessed the effectiveness of seven interventions, while four papers brought together ndings within systematic reviews. The nal paper29 carried out decision modelling to assess cost-effectiveness. In exploring the impact of interventions (or intervention types), changes across three outcomes are reported: loneliness, health and wellbeing (including mental health), and health service use.

intervention. For one-to-one interventions, evidence was available that people who used Waynder or Community Navigator services became less lonely and socially isolated following such contact.14,30 Similarly, an evaluation of a US paid befriending intervention reported that appropriate companionship had been provided, mitigating loneliness.23 Somewhat less denitive ndings were seen within evaluations of group services or interventions. Two systematic reviews identied closed self-help or support groups as effective in reducing loneliness and social isolation.13,14 The single studies provided helpful wider descriptions of the structures and processes of such groups, although differential outcomes were reported. A 12-week closed group that aimed to develop self-efcacy in terms of social integration, and focused each week on different topics relating to friendship, found no change in loneliness. Those individuals who used the intervention were seemingly still as lonely after the course as they were before.34,35 Nevertheless, a further closed model that included social group activities (art and inspiring activities, group exercise and discussion and therapeutic writing and group therapy) reported that 95 per cent of the participants (mean age 80) felt that their feelings of loneliness had been alleviated during the intervention.20 Within the Washington choir (a group activity focused toward wider community engagement), it was found that although there was a slight decrease in loneliness at follow-up, this was not statistically signicant and there was little difference between the intervention and comparison groups.37

Health and wellbeing

Within this area, changes in a number of outcomes were explored: depressive symptoms, physical health, health-related quality of life and mortality. Those one-to-one interventions reporting a reduction in depressive symptomology included a paid befriending initiative based in the US.34 A systematic review on the effectiveness of befriending also supported such outcomes, nding that befriending had a modest but signicant effect on depressive symptoms in the short and long 5

Reduction in loneliness
Achieving a reduction in individual loneliness was reported across very different types of

RESEARCH BRIEFING 39

term when compared with usual care or no treatment.17 Individuals involved in befriending interventions reported that they felt less depressed following the intervention. The nding of a 0.27 standardised mean difference (95 per cent CI, 0.48 to 0.06) did not meet the National Institute for Health and Clinical Excellence (NICE) depression guidelines. It is necessary to demonstrate a standardised mean difference of 0.5 or above if the technology is to be adopted. Nevertheless, as the authors argue, these effect sizes of befriending in the short and longer term are not substantively different to those associated with conventional treatments in primary care such as collaborative care and counselling.17 Two mentoring initiatives found divergent outcomes. A non-randomised observational study reported that improvements in individual depressive symptomology were maintained at 12 months follow-up.9 Nevertheless, a second study (a case controlled trial) that explored the same community mentoring intervention working with socially isolated people for up to 12 weeks to restore older peoples self-condence, self-esteem and social identity found there were no robust improvements in depressive symptoms, physical health, social activities, social support or morbidity.8 This same trial reported that the intervention group demonstrated poorer outcomes, reporting signicantly less improvement in health status (as measured through the EQ-5D) than the control group. A number of group initiatives improved health and wellbeing. Members of the Washington choir reported improved physical health and a reduction in falls in contrast to the comparison group.37 A signicant improvement in subjective health was also reported by those older people taking part in the social group activities art and inspiring activities, group exercise and discussion and therapeutic writing and group therapy.15 This latter study also explored differences in survival (or mortality). At two years, survival was 97 per cent in the intervention group and 90 per cent in the control group a statistically signicant between-group difference.15 6

Health service use

Of the papers selected for inclusion in this review, only two group-based interventions explored before and after service use.15,37 Cohen et al37 reported that while self-reported visits to the primary care practitioner (GP) rose in both intervention and comparison groups, those in the intervention group reported fewer visits (mean of 6.73 per person compared with 10.84). Pitkala et al15 measured hospital bed days, physician visits and outpatient appointments. Across all services, the intervention group had signicantly less usage.

Peoples views on the interventions

The concentration on quantitative outcome measures meant that there was necessarily limited reporting on users experiences. Those interventions that included appropriate qualitative methods (interviews, diaries or observation) and analyses were a group activity model,15,20 a mentoring outreach programme,9 a befriending initiative23 and a closed group educative programme.3436 In general, it would seem that users reported high satisfaction with the services. They felt they had beneted from such interventions and (perhaps more importantly) recognised that they had changed specic areas of their lifestyle.36 For example, users of the short-term mentoring outreach service reported that they had increased their social interaction and community involvement, taking up or going back to hobbies or wider community activities. They also said that their self-esteem had improved and that they felt physically and mentally better. They had increased their physical activity, were sleeping better and had reduced their medication.9 For those physically frail or housebound users, one-to-one befriending or mentoring interventions able to visit on a regular basis were particularly welcome: It has meant everything to me. It has helped me so much. With my mind, I mean, its taken so much loneliness away and worrying.23 Peer support and cohesion (where discussed) were central to a positive experience of the interventions.15,20,3436 Few users gave constructive criticism as to their experience of the intervention. Such silence is

Preventing loneliness and social isolation: interventions and outcomes

not necessarily surprising: few older people feel able to risk negative comment when they are reliant on any service. Nevertheless, some spoke about the rigidity of their intervention, arguing for more exible provision. For example, within the befriending programme, one user suggested that it would be more helpful if the befriender could sometimes change their usual visiting time and day.23 Adaptation was similarly important to users within the short-term mentoring intervention, with their enjoyment of activities mediated by the extent to which the mentors could tailor these to the users preference, abilities and level of condence. Similarly, for those individuals with more severe health problems or disabilities there was a request for greater mentor support a need for the mentor to be available longer than the 12-week limit or to visit more often within the existing timeframe. Users also reported the importance of a skilled mentor. If mentors were unable to encourage users in the right way, users felt disempowered and less condent, feeling blamed for their lack of progress.9 A nal barrier to full use of interventions was that of transport. Users reported that lack of available transport limited those activities that could be attended or any meetings with each other outside the intervention.9 To overcome these difculties, some pilot programmes provided transport to the venue by minibus.15,20 Unfortunately, it is likely that cost would prohibit such arrangements if an intervention was rolled out across a wider locality.

documents, longer-term funding,33 absent or minimal criminal record checks3). Nevertheless, commentators cited a number of components central to ensuring the effectiveness of any intervention. In exploring the planning stage of any service, there was a need to be aware of and use existing community resources and to build community capacity.14 Older people should be involved in any planning as well as implementation and evaluation,13,14 and be enabled to choose and (re)structure the content of any ongoing programme.9,20 There was a need for high-quality selection, training and ongoing support of facilitators, coordinators and volunteers.1315,30 There were some indications in the literature that volunteers belonging to the same generation, sharing common culture and background, were likely to be more effective in building relationships with a service recipient.13 However, in a further study, volunteers being of the same age was not seen as a central requirement.24 There was some support for strong external management or facilitation of any group intervention. Older people should be allowed to self-select to groups, and there is a clear need for facilitators to assess individuals appropriately and thus place them with others having similar interests.15,18,20

Implementing effective interventions

Perhaps the rst task is to identify those populations within localities that are at risk of, or suffer from, social isolation or loneliness. New procedures through the joint strategic needs assessment and health and wellbeing boards42 could respond by allowing such identication, supported by appropriate and rigorous consultation and involvement of older people: If schemes to target loneliness in older people are to be effective, they must involve older people at every stage, including planning, development, delivery and assessment. Often the vital step of asking what people want is missed out when designing services.3 Prior to designing any programme, there also needs to be a clear understanding across organisations and individuals about what is being prevented whether those interventions being considered are to ameliorate social isolation or mitigate loneliness.11 7

Organisational implications
Effective interventions (e.g. befriending,17,23,24 Waynders,14,29 and creative group sessions9,15,20) can and do work in day-to-day services. Perhaps the most important factor and one rarely discussed in the empirical papers is the need for health and social care statutory services to successfully work alongside the voluntary sector.7, 22,40,41 Volunteers (supported through a voluntary agency) delivered ve (of the seven) interventions. Yet there was no discussion as to the need for appropriate partnership arrangements or those effective structures or processes that could ensure available services and volunteers (e.g. appropriate tender

RESEARCH BRIEFING 39

Certain projects need to be funded and implemented. Waynder or Community Navigator interventions have been effective in identifying those individuals who are truly socially isolated or lonely14,29 and in ensuring signposting to appropriate services. Similarly, there is good evidence that befriending services are effective in reducing depression and cost-effective when compared with usual care.17,23,24,29 Creative groups tailored for differing interests and needs lead to reductions in loneliness and re-engagement with the wider community,13,14,20 and demonstrate that the deteriorating health effects of loneliness may be reversed by an intervention which socially activates lonely, elderly individuals.15 Nevertheless, as discussed above, good practice needs to be embedded within such programmes in relation to:

the selection and training of volunteers consistency and rigour in assessment processes
(including the incorporation of levels of loneliness in any medical assessment10)

It was estimated that for befriending schemes, a typical service would cost around 80 per older person within the rst year and the reduced need for treatment and support would provide about 35 in savings. The authors also argued that such savings would be likely to continue in future years. When factoring in the quality of life improvements as a result of the reduction in depression17,33 it was argued that the monetary value would be around 300 per person per year, well exceeding the costs of the intervention. The economic benets from Community Navigators would seem to be greater.29 Knapp et al estimated that the cost per person would be a little under 300. To this they added the costs of a visit to a Citizens Advice Bureau or Job Centre Plus, bringing the total cost to 480 per person per year. Nevertheless, they estimated that the economic benets (e.g. move into employment, fewer services used) would amount to approximately 900 in the rst year. Costs were also provided by Pitkala et al15 in their follow-up study of those individuals involved in the closed activity groups.20 The total cost of health service use (hospital bed days, physician visits and outpatient appointments) was 1,522 per person per year in the intervention group, compared with 2,465 in the control group. This statistically signicant difference between the groups of 943 was greater by 62 than the costs of the intervention 881 per person.

ongoing support and encouragement for

participants to continue attendance

programme exibility allowing the targeting

and tailoring of interventions.

Cost and cost-effectiveness

No evaluation or research study included an analysis of cost-effectiveness: Research into cost-effectiveness is especially sparse, with little economic research even into programmes with evidence of effectiveness.30 Nevertheless, limited cost data were provided in two papers. Decision-modelling was used by Knapp et al29 to demonstrate the economic impact of befriending interventions and Community Navigators, compared with what might have happened in the absence of any such service. The likely care pathways of individuals were modelled and the costs and outcomes at each stage estimated. Along with the costs of formal service provision, those unpaid resources and opportunity costs provided by family and/or informal carers were included. 8

Gaps in the research evidence

How the research evidence could be improved
The research evidence on loneliness and social isolation has developed incrementally over the last two decades, beginning to build an evidence base of what works for whom. Although it could be argued that there is no longer the dearth of evidence found in earlier systematic reviews,13,14 evaluations within this area are still compromised by weak methodologies.30 As with much research in statutory social care and third-sector provision, future evaluation needs to concentrate on appropriately measuring

Preventing loneliness and social isolation: interventions and outcomes

(rather than merely assessing) quality-of-life outcomes and cost-effectiveness. To ensure that changes in quality of life can be robustly measured, there rst needs to be recognition that cross-sectional research (although often providing good snapshots) will not necessarily allow attribution of effect. If we are to know whether specic programmes are able to change individuals quality of life, or impact on their care pathway, those referred to any initiative need to be asked their views before the start of the intervention as well as following such contact. Qualitative semi-structured or in-depth interviews are invaluable in being able to tease out views and support theoretical and thus policy development within specic practice areas. Nevertheless, such research does not necessarily allow for assessment of the impact of the intervention. A wider use of standardised quality of life measures needs to be embedded within any evaluative practice. This will allow for measurement of change, as well as supporting comparisons across other programmes or interventions that may similarly have used such measurement tools. The measurement tools of EQ-5D (a health-related quality of life tool) and ASCOT (a social care related quality of life tool) have been identied and nationally adopted within the outcomes frameworks as tools to measure quality in health and social care, and in future will allow for a broader comparison across delivery models.43 However, although comparing outcomes from different interventions provides insight about the level and extent of effect, it is now becoming essential to include a comparison or control group (preferably with randomisation) within any rigorous evaluation. We need to be able to separate out what would have happened to the individual if they had not received the service the counter-factual argument. Without control or comparison groups, there can be over-interpretation of the data: single-group pre-post and non-randomised comparison studies yield larger mean effect sizes relative to randomised comparison studies.12 For example, two papers included within this review explored the outcomes of an outreach mentoring service.8,9 The rst9 used an

appropriate measure of quality of life and collected data from the same individuals before the intervention and one year later. The results were positive with statistically signicant improvements in mental health and health utility scores. Such continuing outcomes could indicate to commissioners that such an intervention is worth investment. However, in the second evaluation,8 a prospective controlled trial was carried out. Two groups were recruited, one receiving the intervention and one usual care, with data collected from each group. No signicant between-group differences were found. That is, the group receiving the intervention demonstrated no better physical or mental health outcomes than if they had not been offered the service. As the authors comment, the between-group trial data did not reect improvements in mental health status and in depressive symptoms that were reported in the earlier observational study. These very different ndings when a comparative group is included perhaps emphasise the necessity of carrying out controlled and preferably randomised trials that incorporate multiple methods, rather than simply qualitative and observational research. Measurement of cost-effectiveness is complex and, as has been discussed, is rarely a core part of any evaluation of preventative services. If cost-effectiveness is to be measured appropriately, key tools need to be included to collect data on:

individual service use before and after the

intervention

organisational set-up and implementation

costs44

the level and extent of informal carer

support45

the use of wellbeing measures (e.g. ASCOT,

GHQ-12 or EQ-5D) to derive the social care or health-related quality adjusted life year (QALY) gained by the project or intervention. Such best practice in evaluations would enable a robust and rigorous assessment of whether the piloted or planned implementation is cost-effective as compared to usual care. Nevertheless, it is also recognised that available 9

RESEARCH BRIEFING 39

funding and indeed the expertise of the evaluators may not allow such an approach. Knapp et al29 used a decision-modelling approach to show the economic impact of different initiatives to support community capacity-building. Such an approach permits cost calculation without the necessity of more costly and lengthy primary research, and provides a baseline for commissioners and policy-makers from which to make decisions. However, the emphasis in this paper is (not surprisingly) on documenting the outcomes. This leads to minimal description of the raw data underlying the modelling, or the level of sensitivity analysis undertaken. To allow appropriate assessment of whether the recommended outcomes are robust, greater detail needs to be provided.

voluntary organisations working alongside the community.33 Nevertheless, there is little robust outcome data on those interventions that have included BME communities,46 rural populations3 or the most frail and excluded those in care homes, refugees etc. From reviewing the primary research, there are some concerns about whether those socially isolated or lonely individuals are being appropriately reached by some of the interventions. For example, those receiving a befriending initiative were found to score well on social integration and wellbeing when compared with normative scores for elders in the US.23 For the group choir initiative, the comparison group reported a higher level of loneliness than the intervention group.37 Similarly, for other group interventions, success was dependent on group members being well motivated and wanting to make substantial life changes, with high dropout rates where perhaps motivation was absent.3436 Windle et al30 noted that a disappointing feature of the papers included within their systematic review was the disproportionate number focusing on relatively healthy older people in the community, predominately women. With few exceptions we know little about older people in long-term care facilities, notably those who are frail or over 80. Few interventions were targeted at alleviating poverty and none at older people from ethnic or sexual minorities. Within this review, although there are papers that include the most frail and lonely,9,18 there are few that incorporate all populations the majority of individuals being white and female thus limiting how far successful outcomes can be extrapolated.18, 24,3436,38

What the evidence does not tell us

When measuring the impact of projects on specic outcomes (reduction in loneliness, improvements in health and wellbeing and changes in health service use), one of the remaining questions is whether these can be implemented in the English social and health care system. In short, are these projects transferable? Those Waynder initiatives that successfully reduced loneliness were UK-based,14 while in contrast evidence on reduction of loneliness by befriender interventions was drawn from the US.23,24 The latter provided direct payments to befrienders, a structure unlikely to be replicated in England. Nevertheless, the process itself the provision of companionship and small errands to vulnerable individuals is already in place in the UK31,32 and providing positive outcomes.17,31 Similarly, the nding that closed social groups diminish feelings of loneliness is not necessarily negated simply because the research took place in Finland.15,20 These activities are replicable and it is argued that those projects that produce the most effective outcomes are designed by, or tailored to the needs of, the older people themselves and the locality. The greatest strength of successful upstream or lower-level provision is that the interventions do not demonstrate (or demand) complicated structures or processes. By far the majority are provided by volunteers and delivered through 10

Implications from the research

There is good evidence that one-to-one interventions such as befriending and Community Navigators reduce loneliness and improve health and wellbeing.14,17,20,23,24 Users report high satisfaction with the services and there are some indications that involving users in the planning, implementation and

Preventing loneliness and social isolation: interventions and outcomes

evaluation of the programmes improves outcomes. Nevertheless, interventions also need to permit exibility of delivery and necessary adaptation to the needs of the population. Where we have the evidence, both types of intervention appear to be cost-effective when compared with usual care.29 For social group interventions15,20 and wider community initiatives37 there is similarly good evidence that appropriately facilitated cultural and health-related interventions reverse the deteriorating effects of social isolation and loneliness.15 Some interventions may well be promoting and delivering promising practice despite the fact that there is little robust evidence of their effectiveness although changes in outcomes may well be happening. For example, the mentoring intervention that supported and empowered older people8,9 reported that where differences were found between the intervention and control groups, the intervention group had poorer outcomes. In part, such a lack of evidence is due to the insufcient size of the groups, the methods selected (purely qualitative, rather than standardised measures, cross-sectional rather than longitudinal approaches) and a lack of randomisation to enable an understanding of the impact of the service. Similarly, there have

been few studies of population sub-groups that might enable us to understand for whom such interventions may be most appropriate. It is possible that this review understates the benets that can be derived from small services providing emotional support to those who are socially isolated or lonely. The necessity for such preventative projects is supported by the coalition government: When people develop care and support needs, our rst priority should be to restore the individuals independence and autonomy. With the solid basis provided in the Spending Review for social care, there is no reason for councils to restrict support to those with the most intensive needs. This not only serves local people poorly, it is a false economy.42 Nevertheless, as the extent and depth of the real reduction in social care spend begins to bite, it may be very easy for councils to refocus their provision away from such services to concentrate on secondary and tertiary prevention strategies in order to avoid more immediate admissions and readmissions. While our ndings are mixed, they also demonstrate that the contribution of wellbeing services to health improvement is worthy of attention by both social care and health resource commissioners.

11

RESEARCH BRIEFING 39

Useful links
The Campaign to End Loneliness A campaign which draws on research and inspiration from across the UK to offer ideas to both individuals and those working with older people. www.campaigntoendloneliness.org.uk/ AgeUK, Social Inclusion and Loneliness research hub Provides links to academic research units, charities, and funders which focus on social inclusion and loneliness topics in ageing. www.ageuk.org.uk/professional-resourceshome/knowledge-hub-evidence-statistics/ research-community/social-inclusion-andloneliness-research/ Department of Health, National Evaluation of Partnerships for Older People Projects (POPP) The nal evaluation of the POPP programme, which was funded by the Department of Health to develop services aimed at promoting older peoples health, wellbeing and independence. www.dh.gov.uk/en/Publicationsandstatistics/ Publications/PublicationsPolicyAndGuidance/ DH_111240 Support Line Offers condential support including for people who feel socially isolated. www.supportline.org.uk/ Do-it For information relating to volunteering. www.do-it.org.uk/

Related SCIE publications

Social Care TV: Prevention: promoting wellbeing www.scie.org.uk/socialcaretv/ Social Care TV: Prevention: Reablement www.scie.org.uk/socialcaretv/ Social Care TV: The mental health wellbeing of elders in black and minority ethnic communities videos www.scie.org.uk/socialcaretv/ SCIE Research brieng 22: Obstacles to using and providing rural social care www.scie.org.uk/publications/briengs/ brieng22/index.asp SCIE Research brieng 35: Black and minority ethnic people with dementia and their access to support and services www.scie.org.uk/publications/briengs/ brieng35/ SCIE Research brieng 36: Reablement: a cost-effective route to better outcomes www.scie.org.uk/publications/briengs/ brieng36/ SCIE Report 38: Supporting black and minority ethnic older people's mental wellbeing: accounts of social care practice www.scie.org.uk/publications/reports/ report38.asp SCIE Report 41: Prevention in adult safeguarding www.scie.org.uk/publications/reports/report 41/index.asp

12

Preventing loneliness and social isolation: interventions and outcomes

References
1. Department of Health (DH) (2008) Making a strategic shift towards prevention and early intervention: key messages for decision makers, London: DH, www.dhcarenetworks.org.uk/_library/ Resources/Prevention/CSIP_Product/ MSS_-_Key_Messages.pdf, accessed 24 January 2011. Cattan, M. et al. (2003) 'Alleviating social isolation and loneliness among older people', International Journal of Mental Health Promotion, vol 5, no 3, pp 2030. Age UK Oxfordshire (2011) Safeguarding the convoy: a call to action from the campaign to end loneliness, Oxfordshire: Age UK. Department for Work and Pensions (DWP) (2005) Opportunity age: meeting the challenges of ageing in the 21st century, London: DWP. Department of Health (DH) (2001) National service framework for older people, London: DH. Ofce of the Deputy Prime Minister (ODPM) (2005) Excluded older people: social exclusion unit interim report, London: ODPM. Ofce of the Deputy Prime Minister (ODPM) (2006) A sure start to later life: ending inequalities for older people, social exclusion unit nal report, London: ODPM. Dickens, A.P. et al. (2011) 'An evaluation of the effectiveness of a community mentoring service for socially isolated older people: a controlled trial', BMC Public Health, vol 11 p218 Greaves, C.J. and Farbus, L. (2006) 'Effects of creative and social activity on the health and well-being of socially isolated older people: outcomes from a mulit-method observational study', The Journal of the Royal Society for the Promotion of Health, vol 126, no 3, pp 133142.

11. Victor, C.R. et al. (2005) 'The prevalence of, and risk factors for, loneliness in later life: a survey of older people in Great Britain', Ageing and Society, vol 25, no 3, pp 357375. 12. Masi, C.M. et al. (2011) 'A meta-analysis of interventions to reduce loneliness', Personality and Social Pyschology Review, vol 15, no 3, pp 219266. 13. Cattan, M. et al. (2005) 'Preventing social isolation and loneliness among older people: a systematic review of health promotion interventions', Ageing and Society, vol 25, no 1, pp 4167. 14. Findlay, R.A. (2003) 'Interventions to reduce social isolation amongst older people: where is the evidence?' Ageing and Society, vol 23, no 5, pp 647658. 15. Pitkala, K.H. et al. (2009) 'Effects of pyschosocial group rehabiliation on health, use of health care services, and mortality of older persons suffering from loneliness: a randomised, controlled trial', Journal of Gerontolgy: Medical Sciences, vol 64A, no 7, pp 792800. 16. Hawkley, L.C., Thisted, C.M. and Cacioppo, J.T. (2010) 'Loneliness predicts increased blood pressure: 5-year cross-lagged analyses in middle-aged and older adults', Psychology and Aging, vol 25, no 1, pp 132141. 17. Mead, N. et al. (2010) 'Effects of befriending on depressive symptoms and distress: systematic review and meta-analysis', British Journal of Psychiatry, vol 196, no 2, pp 96100. 18. Ollonqvist, K. et al. (2008) 'Alleviating loneliness among frail older people: ndings from a randomised controlled trial', International Journal of Mental Health Promotion, vol 10, no 2, pp 2634. 19. Holt-Lunstead, J., Smith, T. and Layton, J.B. (2010) 'Social relationships and mortality risk: a meta-analytic review', PLoS Medicine, vol 7, no 7, doi:10.1371/journal.pmed. 1000316 20. Savikko, N., et al. (2010) 'Psychosocial group rehabilitation for lonely older people: 13

2.

3.

4.

5.

6.

7.

8.

9.

10. O'Luanaigh, C. and Lawlor, B.A. (2008) 'Loneliness and the health of older people', International Journal of Geriatric Psychiatry, no 23, pp 12131221.

RESEARCH BRIEFING 39

favourable processes and mediating factors of the intervention leading to alleviated loneliness', International Journal of Older People Nursing, vol 5, no 1, pp 1624. 21. Jagger C. et al. (2011) 'Capability and dependency in the Newcastle 85+ cohort study: projections of future care needs', BMC Geriatrics, vol 11, no 21, doi:10.1186/ 1471-2318-11-21 22. Bowers, H., et al. (2006) Making a difference through volunteering: the impact of volunteers who support and care for people at home, London: CSV. 23. Butler, S.S. (2006) 'Evaluating the Senior Companion Program: a mixed-method approach', Journal of Gerontological Social Work, vol 47 (1-2), pp 4570. 24. Rabiner, D.J., et al. (2003) 'The impact of the Senior Companion Program on quality of life outcomes for frail older adults and their families', Home Health Care Services Quarterly, vol 22, no 4, pp 126. 25. Charlesworth, G. et al. (2008) 'Does befriending by trained lay workers improve psychological well-being and quality of life for carers of people with dementia, and at what cost? A randomised controlled trial', Health Technology Assessment, vol 12, no 4, pp 178. 26. Torp, S. et al. (2008) 'A pilot study of how information and communication technology may contribute to health promotion among elderly spousal carers in Norway', Health and Social Care in the Community, vol 16, no 1, pp 7585. 27. Age Concern and Help the Aged (2009) Prevention in practice: service models methods and impact, London: Age Concern/Help the Aged. 28. World Health Organisation (WHO) (2002) Active ageing: a policy framework, Geneva: WHO. 29. Knapp, M. et al. (2010) Building community capacity: making an economic case, PSSRU Discusssion Paper 2772, London: PSSRU. 14

30. Windle, G. et al. (2008) Public health interventions to promote mental well-being in people aged 65 and over: systematic review of effectiveness and cost-effectiveness, Bangor: Institute of Medical and Social Care Research. 31. Mentoring and Befriending Foundation (MBF) (2010) Befriending works: building resilience in local communities, Manchester: MBF. 32. Philip, K. and Spratt, J. (2007) A synthesis of published research on mentoring and befriending, Manchester: MBF. 33. Windle, K., et al. (2009) National evaluation of partnerships for older people projects: nal report, Kent: PSSRU. 34. Kremers, I.P. et al. (2006) 'Improved selfmanagement ability and well-being in older women after a short group intervention', Aging and Mental Health, vol 10, no 5, pp 476484. 35. Martina, C.M.S. and Stevens, N.L. (2006) 'Breaking the cycle of loneliness? Psychological effects of a friendship enrichment program for older women', Aging and Mental Health, vol 10, no 5, pp 467475. 36. Stevens, N. (2001) 'Combating loneliness: a friendship enrichment programme for older women', Ageing and Society, vol 21, no 2, pp 183202. 37. Cohen, G.D. et al. (2006) 'The impact of professionally conducted cultural programs on the physical health, mental health, and social functioning of older adults', The Gerontologist, vol 46, no 6, pp 726734. 38. Slegers, K., Van Boxtel, M.P.J. and Jolles, J. (2008) 'Effects of computer training and internet usage on the well-being and quality of life of older adults a randomized, controlled study', Journals of Gerontology: Series B, Psychological Sciences and Social Sciences, vol 63B, no 3, pp 176184. 39. Fokkema, T. and Knipscheer, K. (2007) 'Escape loneliness by going digital: a quantitative and qualitative evalution of a

Preventing loneliness and social isolation: interventions and outcomes

Dutch experiment in using ECT to overcome loneliness among older adults', Aging and Mental Health, vol 11, no 5, pp 496504. 40. Downs, M. et al. (2008) Health in Mind programme evaluation: nal report to the programme board, Bradford: University of Bradford. 41. Martin, K., Springate, I. and Atkinson, M. (2010) Intergenerational practice: outcomes and effectiveness, London: Local Government Association. 42. Department of Health (DH) (2010) Equity and excellence: liberating the NHS, London: DH. 43. Department of Health (DH) (2011) Transparency in outcomes: a framework for

quality in adult social care a response to the consultation and next steps, London: DH. 44. Curry, N. (2006) Preventive social care: is it cost effective? London: King's Fund. 45. Francis, J. and McDaid, D. (2009) 'SCIE's work on economics and the importance of informal care' in L. Curtis (ed.) Unit costs of health and social care, Canterbury: PSSRU. 46. Manthorpe, J., et al. (2009) '"We are not blaming anyone, but if we don't know about amenities, we cannot seek them out": black and minority older people's views on the quality of local health and personal social services in England, Ageing & Society, vol 29, 2009, pp 93113.

About the development of this product

Scoping and searching
Focused searching began in November 2010 and was completed in January 2011. The scope included peer reviewed papers reporting evaluations of interventions aimed at reducing social isolation and loneliness. Priority was placed on systematic reviews and controlled effectiveness studies. Poor quality studies with no discernible methodology, no outcome measures and no control were excluded from synthesis. Papers published before 2000 were excluded.

Peer review and testing

The authors have research and topic expertise. The brieng was peer reviewed internally and externally.

About SCIE research briengs

SCIE research briengs provide a concise summary of recent research into a particular topic and signpost routes to further information. They are designed to provide research evidence in an accessible format to a varied audience, including health and social care practitioners, students, managers and policy-makers. They have been undertaken using methodology developed by SCIE. The information on which the briengs are based is drawn from relevant electronic databases, journals and texts, and where appropriate, from alternative sources, such as inspection reports and annual reviews as identied by the authors. The briengs do not provide a denitive statement of all evidence on a particular issue. SCIE research brieng methodology was followed throughout (inclusion criteria; material not comprehensively quality assured; evidence synthesised and key messages formulated by author): for full details, see www.scie.org.uk/publications/briengs/methodology.asp SCIE research briengs are designed to be used online, with links to documents and other organisations websites. To access this research brieng in full, and to nd other publications, visit www.scie.org.uk/publications/briengs/

15

RESEARCH BRIEFING 39

Latest SCIE research briengs

20 21 22 23
The implementation of individual budget schemes in adult social care Identication of deafblind dual sensory impairment in older people Obstacles to using and providing rural social care Stress and resilience factors in parents with mental health problems and their children Experiences of children and young people caring for a parent with a mental health problem Childrens and young peoples experiences of domestic violence involving adults in a parenting role Mental health and social work Factors that assist early identication of children in need in integrated or inter-agency settings Assistive technology and older people Black and minority ethnic parents with mental health problems and their children

30

The relationship between dual diagnosis: substance misuse and dealing with mental health issues Co-production: an emerging evidence base for adult social care transformation Access to social care and support for adults with autistic spectrum conditions (ASC) The contribution of social work and social care to the reduction of health inequalities: four case studies workers: outcomes for older people, staff, families and friends

31 32 33

24

34 Communication training for care home

25

35

Black and minority ethnic people with dementia and their access to support and services Reablement: a cost-effective route to better outcomes Mental health service transitions for young people Mental health, employment and the social care workforce Preventing loneliness and social isolation: interventions and outcomes

26 27

36 37 38 39

28 29

A full list of SCIE research briengs can be found at www.scie.org.uk/publications

Social Care Institute for Excellence Fifth oor 2-4 Cockspur Street London SW1Y 5BH

tel: 020 7024 7650 fax: 020 7024 7651 www.scie.org.uk

Registered charity no. 1092778 Company registration no. 4289790

RB3911