Article Text

Download PDFPDF

Compassionate community networks: supporting home dying
  1. Julian Abel1,
  2. Jon Bowra2,
  3. Tony Walter3 and
  4. Glennys Howarth4
  1. 1Weston Hospicecare, Weston-super-Mare, Somerset, UK
  2. 2Living Well, Dying Well, Sussex, UK
  3. 3Centre for Death & Society, University of Bath, Bath, UK
  4. 4Centre for Health and Social Care Innovation, Plymouth University, Plymouth, UK
  1. Correspondence to Julian Abel, Weston Hospicecare, 28 Thornbury Road, Uphill, Weston-super-Mare, Somerset BS23 4YQ and Weston Area Health Trust, Grange Rd, Weston-super-Mare, BS23 4TQ; julian.abel{at}


How may communities be mobilised to help someone dying at home? This conceptual article outlines the thinking behind an innovative compassionate community project being developed at Weston-super-Mare, UK. In this project, a health professional mentors the dying person and their carer to identify and match: (a) the tasks that need to be done and (b) the members of their social network who might help with these tasks. Network members may subsequently join a local volunteer force to assist others who are network poor. Performing practical tasks may be more acceptable to some family, friends and neighbours than having to engage in a conversation about dying, and provides a familiarity with dying that is often lacking in modern societies, so in this model, behavioural change precedes attitudinal change. The scheme rejects a service delivery model of care in favour of a community development model, but differs from community development schemes in which the mentor is a volunteer rather than a health professional, and also from those approaches that strive to build community capacity before any one individual dying person is helped. The pros and cons of each approach are discussed. There is a need for evaluation of this and similar schemes, and for basic research into naturally occurring resource mobilisation at the end of life.

View Full Text

Statistics from


‘It takes a village to raise a child’, says the African proverb. It also takes a village to care for someone as they are dying. The trouble is, few people in rich countries live in villages these days, and for the many in poor countries who do, family and community networks are increasingly compromised.1

In many rich societies, more people want to die at home than actually succeed in doing so,2 3 while in poor countries, many have no option but to die at home.1 Given the economic constraints on the development of expensive professional care, and given the frequency of carer burnout in rich countries often leading to emergency hospital admission in the final days,4 a service delivery model of support for carers has to be supplemented by a community development model. Specifically, one prerequisite to enable people to die at home, and to die well at home, is that they and their carer(s) mobilise their social networks to assist them in the tasks of care. Tasks done by network members may seem small in themselves—walking the dog, doing the shopping, fetching a prescription, sitting with the dying person while the carer takes a break—but together can enable the main carer to continue, especially if the sick person needs someone in the house at all times. Where networks are insufficient, volunteers may be able to assist. This article explores how, in the context of rich Western countries, local resources may be mobilised, focusing on one innovative British project and placing it in the context of related projects. The project is at an early stage and will be evaluated in due course; this article is therefore conceptual, intended to provoke discussion.


Mobilising networks, simple though it sounds, actually presents supportive and palliative care with considerable conceptual and practical challenges.

First, the welcome development of palliative care over the last 50 years in many countries has gone hand in hand with communities' loss of ownership of dying.5 There is now an expectation, particularly in the developed world, that when someone is dying, care will be given primarily by professional carers. Formal ‘care plans’ itemise the care to be performed by paid professionals, and rarely include the carer, let alone any plan for assistance from the carer's social network.6 7 When carers are included, they are usually seen as part of the professionals' team, rather than professionals being seen as a relatively small part of the carer's team. Likewise, the main carer often seeks as much professional support as possible to try and maintain looking after the person in their own home. Even the literature on how to support family carers is suffused with a service delivery model, concerned with how formal services can support informal carers to the exclusion of any consideration of how informal community networks can support carers.4 There is no research literature on if and how families galvanise informal support, and a recent literature review of psychosocial interventions for family carers of palliative care patients makes no mention of support from informal networks and did not list this as a need identified by carers.8 It is entirely possible that some main carers feel it is their duty to do all the care.

We may contrast the service delivery approach of palliative care with practice at the other end of life, where it is recognised that to care for a new baby, ‘Parents should be able to meet other parents and get support from one another to establish social support networks… By being involved in…networking opportunities, parents develop relationships in their communities, which connect them to the services they need.’9 And social workers, working with clients at all stages of life, have for 30 years ‘recognised the importance of their clients’ sources of informal social support and (made) these resources a focal point in case planning'.10 What seems obvious at the beginning of life and in social work is ignored by almost all those writing about end of life care.

Second, Western palliative care sees the dying person holistically, but as an individual. The patient is perceived as a physical, emotional, social and spiritual being, dimensions all addressed by the palliative care team. ‘Social’ typically refers to a nuclear family, so palliative care professionals relate to a patient-in-family. What we are advocating is a different view: of the ill person as located within a number of social networks, of which healthcare services are but one; the focus is not the individual as a whole person, but as an individual-in-community.5 This relational concept of the person is familiar outside the West, but entails a conceptual shift for Western palliative care.11

Third, there is no research evidence on how—in the absence of any professional encouragement to do so—families mobilise their social networks at the end of life. We do not know from research how formal services can support naturally occurring support networks, because we do not know how those naturally occurring networks operate or fail to operate. Although there is literature on reciprocity in care at other stages in life,12 we know very little about if and how reciprocity norms may change at the end of life.13 We do know that there is considerable debate as to the strength of community networks in general (not just at the end of life). Putnam's celebrated book Bowling Alone claimed local social networks in the USA to be at an all-time low, but his analysis prompted considerable debate and challenge.14 15 Likewise, the current UK government's claims of a ‘broken Britain’ and the need for a ‘big society’ have proved contentious.

There are nevertheless reasons to believe that modern societies have developed in ways that may not always make it easy for family and friends to be routinely involved in end of life care. Thankfully, most people now do not die until old age, but this means that spouses and siblings are themselves older, possibly disabled or frail. Social isolation among older people is common. Their children have full-time jobs, as have their spouses; they may live a long way from their parents and have children of their own. The same is true for extended family members and friends. Lives are very busy and it may be hard for family members to make the commitment to provide regular care. Neighbours too may be busy, or isolated. The result can be care being performed entirely by main carers and professionals, and being experienced as a ‘burden’. However, we also predict wide variations, not only between individuals but also between communities; ethnicity, religion, social class, gender, personality, family and neighbourhood dynamics could all be influential. Some communities may be very well connected, and this may not correlate with affluence or other conventional indices of power.

Compassionate communities

Over the past decade or so, the concept of health promoting palliative care/compassionate cities has attempted to address such issues.5 16 Care for terminally ill people should be returned to communities, within which medical, nursing and other specialist services have an important place. This reverses the present situation in which palliative care is highly professionalised, with the community marginalised. This approach draws on the concept of social capital, which refers to social connections and related norms and trust. Although contested,15 the concept highlights that ‘relationships (social) have obvious value (capital)’.17 So far, the most developed version is in Kerala, India. There, neighbourhood networks of 10–15 volunteers identify the chronically ill people in their area and organise appropriate interventions, liaising with medical services and covering a population of more than 12 million people.18

In this article, we explore two approaches to a compassionate community. The first is to develop compassionate communities so that people can access local networks to resource their desire to die at home. The second is the other way round—to enable people dying at home to access local networks, and thus to develop compassionate communities. This is a chicken and egg question, for both are necessary to the other, but we argue that, although complementary, these are two distinctive routes to the same goal. Both have their place, and it is helpful to set out the distinctive features of each.

Starting with the community

This approach applies well-established principles of community development to end of life care, and there are elements of this in the UK's Dying Matters Coalition.19 This can include providing education in different settings about how we culturally and historically have reacted to death, raising awareness in schools and workplaces about end of life issues, as well as attempting to make discussions of death normal rather than taboo or medicalised. Older people are encouraged to consider the decisions they will need to take as they face the end of life. Such initiatives take steps towards deprofessionalising care, returning end of life care to the community and building up social capital that can then be mobilised when citizens come to the end of their life. Specific examples are the St Christopher's Hospice Schools Project20 and awareness raising among pensioner groups in the London borough of Lewisham. This approach has clear advantages. Its building of community capacity is not dominated by medicalisation, and therefore has the potential to normalise and demedicalise dying. Non-health service groups take the lead in raising death awareness and capacity building in the community. This approach has disadvantages too. It may be embraced by existing voluntary groups or by new groups that not all dying people will know about, so some may be left without community help. Even those who do know what help is available have to opt for it. So there is no guarantee that every person towards the end of their life will be able to access the available help. Community groups will also have to liaise with formal health services, and there is a danger that to the uncoordinated professionals already crowding the dying person's house are added a number of volunteers who may or may not be coordinated among themselves, let alone with the professionals.

Starting with the dying person's existing networks

An alternative approach, which can be developed even in the absence of a compassionate community, is for every dying person to be offered a mentor to help them identify their existing networks and other sources of support. The mentor role has a long history in empowering disadvantaged youth and in mental health, and has recently been developed in the English county of Devon, which now pays mentors to care for isolated and/or depressed older people.21 In palliative care, the mentor role is central to a project in the greater Sydney area of Australia, where mentors help families galvanise their support networks; a recent evaluation highlights some very positive benefits of this scheme.22 There are plans to replicate this by a hospice organisation in Ireland (the Milford Care Centre, Limerick) and by a new voluntary group in England (Living Well, Dying Well).

In these palliative care interventions, the mentor is a volunteer. Referral to the mentor by the existing healthcare team could be automatic, as in Limerick, or may be hit-and-miss depending on whether the person or the healthcare team knows about the mentor scheme, as in Living Well, Dying Well. Volunteers need to be recruited before any of these palliative care schemes can get going, so to some extent they require the pre-existence if not of a compassionate community, at least of some compassionate volunteers within the community.

The scheme we will now describe is being developed through the hospice at Weston-super-Mare, England, and adopts a slightly different strategy, with the mentor role being taken by a healthcare professional such as a community nurse. When it is clear that the person is not going to get better, on average 3 months before death, a healthcare professional who is close to the patient's family discusses end of life care with them. If they wish a home death, the mentor then asks the carer to identify the different tasks that will be necessary for the person to be looked after at home. These tasks are the routine jobs such as cooking, cleaning and shopping as well as personal care. The next step is to identify the social network of family, friends and neighbours who surround the household. A mapping exercise can then take place in which they can take tasks on a weekly basis that can help to support both the main carer and the patient. Professional care can then be slotted in around the gaps, depending on the need. A coordinator of informal care is appointed; this could be the main carer, another family member or someone else in the network. The mentor role is thus short-term, kick-starting the process.

If started early enough, care can be increased as the patient becomes less well. This means that caring, which is more like a marathon than a sprint, is likely to last longer and be less susceptible to carer burnout and the need for emergency hospital admission. In addition, support and care from the network can continue into bereavement for as long as needed. In the Sydney project, this was demonstrated to mitigate the social isolation and loneliness that can so easily follow death.22

A subsequent phase in the Weston scheme is to develop volunteer networks of people who have helped a friend or neighbour in their journey of dying. Once an existing network has looked after someone, some of the people within the network may be prepared to do something similar for another person in their neighbourhood who may be more isolated, that is, network poor. There are a number of potential advantages to this approach compared with more traditional uses of volunteers. People who have given care may be comfortable with specific roles. These may be varied, such as a teenager walking a pet, someone taking a carer shopping, someone who sits with a patient or provides personal care to a greater or lesser degree. Carers and patients may find it easier to accept an unfamiliar face within the house if that person has already cared for somebody who has been through a terminal illness. Within a neighbourhood, there may also be indirect connections between family and volunteers. It may also be easier to link networks once an initial contact has been made. Some families have extensive and very strong networks while others have very limited support structures around them. Linking an extensive network to someone who is network poor may help to significantly increase care in the community and make the difference between caring for someone in their own community and institutional care. Quite where these volunteer networks are located organisationally has yet to be determined, but one possibility is to link them to general practitioner practices.

So far, this project is in its early stages. The specialist palliative care community nurses are becoming more familiar with talking about community networks and are developing expertise in how to approach the subject effectively. Its use is becoming more routine. Obstacles and benefits have already been observed, but indications so far are that compassionate community networks will have a place in end of life care at Weston. The second phase of the project, which involves the development of already established networks for people in a neighbourhood who are network poor, has not been started. This phase requires specific administration and expertise and funding is being sought.

The eventual benefits are not only for dying people and their carers, they are also for those who gain from helping out by becoming more aware of end-of-life issues. Even more widely, the building up of social capital benefits a wide range of people throughout the community, including some who may have no contact with the dying person. Both affluence and poverty, along with communications technologies and mass transport, detach many from their local community; nevertheless, it is when vulnerable bodies need to be looked after at home that local networks of support remain indispensable. The three instances where this need is time-limited rather than potentially open-ended (as in long-term care of the chronically sick or disabled) are looking after a new baby, getting neighbours to care for a pet when the owner goes on holiday, and dying: each requires hands-on local assistance, which in turn both needs and generates local social relationships. Childrearing and dog ownership are implicitly recognised as the two motors most likely to re-generate neighbourliness where affluence undermines it; we suggest that dying could be a third.13 Evidence from the Sydney project confirms this, showing that both quality and quantity of networks remained at an increased level a year after the death.22

Advantages of compassionate community networks

In the Weston scheme, network identification becomes a routine part of advance care planning. Task identification and network identification are therefore not something that families opt into; they have to address them if they wish for a home death. Once hospice staff see network identification as an integral part of care planning, networks can be mobilised even in fragmented communities. The first port of call is a person's existing networks; only when the gaps in what these networks can provide are identified are volunteers and professionals considered. Because the scheme is initiated by a healthcare professional, there is the possibility of coordination between network members, volunteers and professionals.

The scheme employs a radical model of volunteer recruitment and training. In the role of a friend, neighbour or extended family member, people train themselves—practically, emotionally and socially—on the job, before they formally apply to join a volunteer group to assist people they do not know. If they subsequently become formal volunteers, they will require some further training and security checks, but these simply top up the pre-existing self-training. Volunteer training is likely to be inexpensive, and volunteer retention is likely to be high.

Disadvantages of compassionate community networks

There are also, however, potential pitfalls with the mentor role being taken by a health professional rather than a volunteer. In the UK, healthcare professionals typically inform cancer patients when the disease becomes incurable, so an advance care conversation may be initiated. The trajectories of dying from heart disease, stroke or multiple organ failure in old age are like cancer was 50 years ago, that is, rarely identified to the patient by health professionals. Until this situation changes, it is difficult to see how health professionals can take the mentor role with many non-cancer conditions.

Even with cancer, as discussed earlier, the mentor role requires a radical shift in how doctors or nurses see themselves and are seen by families—from helping patients to enabling citizens, from being powerful to being empowering, from service provider to community developer, from expert to mentor. Even if the scheme works well within the pilot hospice in which it is driven by enthusiastic leadership, if rolled out more widely it may succumb to traditional service delivery assumptions.

Assumptions on the part of carers may also militate against network mobilisation. Carers may want to attempt to do all the care themselves. Some wives feel it their duty, and theirs alone, to care for their husband. Some will find it more burdensome to coordinate a network of care than to do the various tasks themselves. Others may be concerned about privacy, not wishing neighbours to be involved in personal care or to enter any part of the house other than the kitchen or living room. Carers may be willing to ask a neighbour to help for a few weeks, but what if a 2-month terminal cancer diagnosis extends to 2 years? What about other diseases where the trajectory is much less predictable than with cancer? Issues of reciprocity become crucial here.

It is also quite possible that families that respond positively to an invitation to mobilise their networks would have mobilised them anyway.

Compassionate community networks are only part of the home care jigsaw. Training informal carers in practical nursing, 24-hour availability of expert medical and nursing advice, and compassionate care leave from work are some other all-too-often missing pieces.3 In countries without any economic support for carers, the main task that many carers need help with is earning a living. In African countries ravaged by AIDS, the carer may have to work in her fields while caring for successive family members who are terminally ill; the task she really needs help with is a lot more than walking the dog. Yet, the remaining fit people in the village and in her family, mostly children and older people, are already stretched.1

The two approaches compared

The approach that starts with the community assumes a particular model of change, namely that talking about death changes people's attitudes, which then allows capacity building within communities which can in turn generate the networks that may help specific individuals as they near the end of life. It is premised on the assumption that death is a taboo topic in modern societies,19 even though this assumption has been critiqued for at least 30 years.23 24 The network approach, by contrast, does not require communities first to initiate conversations about death and dying. Rather, it invites people to perform tasks. ‘You know your neighbour Fred? Well, he hasn’t much longer and Freda wants to look after him at home, and needs the grass cut once a week. Can you help?' Ariès argued that the modern problem with death is one not so much of unmentionability as of unfamiliarity: with people dying in old age in hospitals and institutions, most people are not familiar with what dying is like.25 What compassionate community networks do is enable people to become familiar with dying. The 14-year-old who is pleased to help out by walking her dying neighbour's dog gets to see the person and the carer every day, and so learns about dying.

In compassionate community networks, the model of change is: get people to help with specific tasks, they will feel good about this, their knowledge about dying and their attitudes about death will change, which will then build capacity within the local community. In other words, the first approach believes you must change people's attitudes before their behaviour will change; the second approach changes people's behaviour, believing that attitudes will then fall into line with the new behaviour. Research in health psychology has moved away from models that emphasised the need for education to change attitudes in order to change behaviour; current models of behavioural change indicate that attitudes, behaviours and habits can all influence each other.26 So death education and the compassionate community can be a product of, as well as a means to, caring for a neighbour.

Both approaches are part of the care jigsaw. There are several ways to enable more people to die at home, and what works in one community may not work in another. In the UK, some established ethnic minority communities, traditionally problematised for their low take-up of palliative care services, may have much to teach other communities about the resourcefulness to be found within community. Some more recently arrived ethnic minority members, for example those seeking asylum, may have exceptionally poor local networks, with their family spread around the world, and may need high inputs of volunteer and professional assistance.

Given that the concept of compassionate community networks is innovative, and given that there are several ways of empowering people to die at home, it is important that diverse projects are developed and funded, and that they are evaluated, to see what works and why. Outcomes—most immediately an increase in home deaths but also the building of social capital in the community—need to be documented and evaluated. But processes also need to be researched, for example through action research, to see how doctors, nurses, families and neighbours encourage, re-negotiate or block the mobilisation of local resources, and why; if they have good reasons to block the process, these reasons need to be taken seriously if pilot projects are to be rolled out more widely. Also necessary is basic research into how people naturally mobilise networks at the end of life, without any encouragement from health professionals.

Conclusion: natural evolution?

Compassionate community networks could be argued to be a natural evolution for palliative care, at least in Britain. If dying people are to be perceived holistically, that must mean relationally. If people want to die at home, they must mobilise their networks, and if their networks are not strong enough then volunteers need to be mobilised. British hospices have led the world in family centred palliative care, in promoting home dying and in developing volunteering, so promoting compassionate community networks is a natural next step.

And yet, compassionate communities are a most unnatural evolution for palliative care. Kellehear identifies the reasons.5 First, dying people will need to be seen not as patients, but as citizens. Second, palliative care's vision of the whole individual has driven the expansion of a palliative care ‘empire’, involving ever more specialist disciplines within the multi-disciplinary team. Promoting compassionate community networks requires building not occupational capacity, but community capacity. Third, although palliative care may look beyond the patient to the family, it rarely looks beyond the family to the community or sees the whole person as an individual-in-community.

Enabling more people to die in their own homes therefore entails palliative care abandoning some cherished assumptions and reinventing itself. How completely that is possible, we will see. But the gains—not only for patients and carers, but also for their families, friends and neighbours, and for communities as a whole—could be considerable.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.