Are district nurses well placed to provide equitable end-of-life care to individuals who are homeless?

Difficulty in predicting disease trajectory

A common theme running through many of the pieces of literature offering some explanation as to why the homeless population does not access appropriate palliative care services is the unpredictability of their disease trajectory. Thomas (2012) performed a comparative analysis of mortality rates between homeless individuals and adults from the general population in the UK. This paper sets the context of society in the UK at present, referring to the economic downturn, welfare cuts and budget cuts to the NHS.

Thomas (2012) analysed four data sets between 2001 and 2009 and substantiated the difficulty in predicting disease trajectory in individuals who are homeless, with a high proportion of deaths being due to drug and alcohol abuse, which are often sudden. A third of all deaths among the homeless population are substance-misuse related. This is higher than the proportion in the general population, even when younger-aged subsets are considered. In addition, Thomas (2012) reported that the risk of suicide in individuals who are homeless is three-and-a-half times the national average. Although this paper does have some limitations, previous studies around this subject have used only coroners' data, which would provide a less in-depth investigation (Keyes and Kennedy, 1992; Grenier, 1996; Jones and Pleace, 2010).

In a qualitative study to discover ways to improve support for the homeless population with advanced liver disease, Davis et al (2011) performed a case note audit and data analysis over 1 year, between 2009 and 2010, of resident's deaths at St Mungo's, a charity that provides accommodation for the homeless in London. They identified 27 eligible case notes out of a possible 56, corresponding to people who had died while in their care during the study period. A mixed-methods approach was used, which increased the validity of the study, wherein volunteers from the home were invited to participate in focus group meetings. Findings around drug and alcohol abuse indicated that many service users were in denial about the extent of their substance misuse, and that in 50% of the cases, service users would ‘binge’ at the EOL. Davis et al (2011) also established that the signs of end-stage liver disease varied greatly among individuals, making it difficult to predict when they were approaching the EOL. A palliative care coordinator role was pioneered by St Mungo's, and hostel staff were empowered to carry on caring for dying service users, rather than having to use secondary care. There is the possibility that district nurses could imitate this role and provide an in-reach service to homeless shelters and accommodation nationally to achieve this. Moreover, the study identified that many hostel staff only had a tacit knowledge of palliative and EOL care and recommended that they needed training to recognise the key physical and behavioural changes that occur at the EOL in people with end-stage liver disease. The role of supporter and educator could be adopted by district nurses. By working collaboratively, both district nurses and hostel staff could learn about each other's roles and provide a better service. Thus, it appears that district nurses could be well placed to provide equitable EOL care for the homeless population.

In support of this view, Hudson et al (2016) and Shulman et al (2018) also noted the difficulties in predicting disease trajectories among patients in this marginalised group dying of liver disease and hepatitis C infection. Hudson et al (2017) used a qualitative approach, with interviews and focus groups, and produced a thematic analysis of the data. Although this was the largest qualitative study to date at the time, it had some limitations. Unfortunately, for ethical reasons, intoxicated individuals were unable to consent to participating in the interviews, and therefore, this vital demographic in the homeless population was under-represented. Although the study only covered three London boroughs, the researchers used national steering groups to triangulate the results, and the groups' findings resonated with Hudson et al's (2017). Their study also asserts the importance of enhanced multidisciplinary team (MDT) support and suggests a joined learning summit to bring together expertise from a range of professional groups to improve outcomes. District nurses could spearhead this summit and help improve the outcomes for service users. Additionally, Hudson et al (2017) indicated that there should be shared responsibility between district nurses, GPs and other members of the MDT in supporting the EOL choices of individuals who are homeless, and they suggest that the MDT should use the Public Health England (2016)Homelessness: applying all our health framework to assist it in achieving this. Within this framework, it is suggested that HCPs not only have to join up care with the usual members of the MDT and look at the issue from a health perspective, but they also have to involve local authority housing, hostel workers, outreach workers and social care workers. Additionally, they should use volunteers or employ people who have experienced homelessness to be able to achieve equitable care for individuals who are homeless at the EOL. District nurses could act as the lynchpin to co-ordinate this.

Building on the work by Hudson et al (2017), Shulman et al (2018) undertook qualitative research to explore the gaps in EOL care services for individuals who are homeless. Using the same interviews, Shulman et al (2018) found that it was notoriously difficult to predict prognosis due to service users continuing to misuse substances, even when the person was dying of a condition with a more predictable disease trajectory, such as cancer. They added to Hudson et al's (2017) study by uncovering that hospices were reluctant to accept a person who is homeless due to them potentially occupying a bed for extended periods of time due to having no onward place to go. They also found that advance care planning (ACP) was not occurring as frequently as it should be. In order to better predict disease trajectory, Shulman et al (2018), too, highlighted the need for better joined-up services. Thus, district nurses could be well placed to promote in-reach services in hostels, collaboration and joint training.

A Swedish qualitative explorative single-case study by Hakanson et al (2015) further highlights the need for the development of tailored, person-centred palliative care approaches for those marginalised groups that do not benefit from existing palliative care policies. The study was set in a support home specifically designed for eight homeless persons. Its aim was to provide people who are homeless with healthcare, including palliative care, and social support to enhance wellbeing. This support home is unique in Sweden because of its tolerance of ongoing substance abuse by its residents.

Hakanson et al (2015) suggested that due to the unpredictable disease trajectory, it is difficult to assess residents in terms of EOL and to begin the ACP process. The Leadership Alliance for the Care of Dying People (2014) policy paper One chance to get it right suggested that communication is key in achieving this. The DHSC (2015) EOL strategy added that the patient's family should be involved in this process, but Hakanson et al (2015) found that this may not be simple in the case of homeless individuals, since homeless people are more likely to have witnessed the traumatic deaths of friends, which causes high levels of anxiety in those who are dying. Additionally, they often lack relatives who can act as advocates for them. Hakanson et al (2015) also found that ACP may not be easily achievable for homeless people, due to the high level of mental health issues and drug abuse within this population. Therefore, staff were reluctant to start the ACP process at all, for fear of causing more harm to the patient than benefit. This is another example of why a ‘one-size-fits-all’ policy for EOL does not reach the most vulnerable and marginalised groups of society and a more flexible approach is needed.

According to DHSC (2014) guidance EOL: achieving quality in hostels and for homeless people, hostel workers are the most appropriate key workers for homeless residents who are dying. Webb (2015) used a qualitative approach with semi-structured interviews to help understand the views of hostel staff on EOL care among the homeless and to ascertain whether the staff agreed with the DHSC (2014) guidance. This was a multi-site study, so the results were triangulated, and the use of a diverse sample of participants enhanced the transferability of the findings to the possible views of all hostel workers. Webb (2015) is relevant to the present literature review, as it is a good indicator of whether government policy and guidance is practical and achievable in practice. Hunter and Killoran (2004) have suggested that many government policies are merely paper exercises, and the findings of the Webb (2015) study substantiate the possibility of this particular DHSC guidance being an example of the policy–practice gap, as none of the participating hostel staff were even aware of the existing guidance.

Webb (2015) also observed that hostel staff found it difficult to predict disease trajectory in their client base because of the inability to access their clients' medical notes. Participants discussed that homeless people would not give out personal and highly sensitive medical information to them until a rapport had been developed. Although the DHSC (2014) guidance encourages hostel staff to engage with HCPs and social care agencies, hostel staff do not even have access to the same records as these various agencies. This also made the clients very frustrated, as they often had to repeat painful stories to different agencies.

It was also mentioned in Webb's study (2015) that even if hostel staff had access to relevant and accurate medical information, they had no training in recognising physical deterioration and would not know how to deal with it when it occurred. A nurse-led training initiative undertaken by a palliative care specialist nurse who provided training to hostel staff in basic palliative care principles was reported by Cole (2017), and it could form a solution to this problem. The course used a bottom-up approach and asked for the views of staff and clients on how to improve EOL care. In addition to the training offered, the hostel has a fortnightly drop-in clinic attended by a GP and a nurse to coordinate care. Although in this example, the training was led by a palliative care specialist nurse, the possibility of district nurses providing a similar ‘in-reach’ service could be explored. Further research into the generalisability and repeatability of this approach should be undertaken with district nurses playing the roles of coordinator, to determine whether they could provide training on EOL care and offer support to hostel staff.

Gaps in existing systems

Through the process of analysing the literature, it is apparent that many of the failings in equitable EOL for homeless people are due to gaps in the inflexible and unsuitable systems available. These gaps not only cover a lack of appropriate places for homeless people to receive EOL care but also poor education and training for all staff working with homeless people and the inflexibility of rigid institutional and organisational norms.

In an attempt to identify gaps in services in providing equitable EOL care for homeless people, Davis-Berman (2017) conducted a qualitative needs analysis by interviewing key informants from all of the services available to this vulnerable group in a medium-sized city in the American Midwest. Although this is an American study and there are some differences due to the differing welfare systems between the US and UK, its findings are consistent with those of other research, and it adds to what is known about equitable EOL care for homeless people. Davis-Berman (2017) identified that service users felt stigmatised by health and social care workers and also felt that they were often treated with disrespect. Evidence arose to suggest that homeless people were being treated differently in hospitals than members of the general public were, especially if they attended without an advocate. This is typical of a lack of knowledge and understanding that has manifested in stigma. The study also reported that staff were reluctant to enter certain neighbourhoods or homeless shelters to provide care. This represents a system failure, which could be overcome by putting the structure in place to ensure that staff feel safe to provide care wherever and to whomever needs it. Following a risk assessment, rather than relying on hearsay, district nurses in the UK could overcome any potential risk by putting the appropriate safeguards in place. They could visit in pairs, carry a lone worker device or inform colleagues of their whereabouts and the anticipated length of visit.

The view that stigma is a barrier to equitable EOL care is supported by Shulman et al (2018), who identified that HCPs do not understand the complex needs of homeless people, who have often experienced trauma, have underlying mental health issues and are abusing alcohol and drugs. Their study points out that the inflexibility and inexperience of staff add to this stigma. In the same vein, Webb (2015) discovered that HCPs treated homeless people differently due to their limited understanding of the complex health needs of this population, and that five of seven of the homeless people interviewed in their study outlined that they would be referred from one service to another with no one taking responsibility for the provision of care. This is further reinforced by Hakanson et al (2015), who reported that residents felt they were treated with disrespect by HCPs and social workers. In addition, Davis et al (2011) found evidence of GP receptionists becoming gatekeepers, and reported that homeless clients would feel intimidated in the waiting room and, as a result, would refuse to see their GP despite being registered. This is an education issue, which could be overcome by training staff on factors that affect homeless people's daily lives. As a result, staff might understand why it is sometimes difficult for homeless people to attend appointments and why their behaviour is not the same as that of the general population and may be viewed as disruptive. District nurses could arrange this training for their own staff and involve people who have experienced homelessness to assist.

A similar qualitative study was carried out by Krakowsky et al (2012), who conducted semi-structured interviews to explore how palliative care services in Toronto, Canada, could better serve the homeless population. They found that using a collaborative training approach to address stigma and the lack of knowledge around palliative care was beneficial. This study identified that homeless people are reluctant to access services available to them at the EOL because of negative experiences suffered previously. Krakowsky et al (2012) also observed that hostel staff are not equipped with knowledge about the complex palliation needs of this group of individuals, who often lack family or someone whom they can trust, which increases their need for an advocate. The role of the district nurse in this case could be to advocate for the homeless. In support, Webb (2015) pointed out that hostel staff have a very limited understanding of what palliative care is and their role in providing it, with one participant stating that they thought that suicide prevention was part of EOL care. Webb (2015) also argued that there was no joined-up working between HCPs and hostel staff, which is something that would certainly need to be tackled to overcome this barrier to equitable care. District nurses providing in-reach care to hostels could encourage this and probably smoothen the lines of communication between care providers and recipients, thereby improving the latters' EOL outcomes. In addition, Hakanson et al (2015) found that more fluid, less structured opportunities to build a rapport with residents to allow them to share their stories helped the staff to understand the residents better, and this in turn helped to facilitate good ACP and support. In order to achieve this, the visiting district nurse would have to ensure an atmosphere that would make residents feel comfortable in talking to them. One aspect that needs to be taken into consideration here, however, is that it may be difficult to be close to the residents while maintaining professional boundaries, as observed by Hakanson et al (2015).

Commenting on the suitability of hostel staff to deliver palliative care, Hudson et al (2017) noted that hostel staff are recovery-focused and are therefore less likely to accept palliation, especially when the cause of death is reversible, for example, with substance misuse. Additionally, Davis et al (2011) discovered that hostel staff experienced a great emotional burden and found themselves supporting bereaved family members. This again reflects a gap in the system that could be overcome by ensuring that hostel staff know which services are available to signpost grieving relatives towards, such as pastoral and bereavement care, something that district nurses could help facilitate.

McNeil et al (2012) undertook a large qualitative study to identify barriers to EOL care for homeless people. They conducted semi-structured interviews with 54 health and social care professionals in six Canadian cities; the scale of this study undoubtedly increases the reliability and generalisability of its findings. McNeil et al (2012) found that facilities that prohibit patients from using drugs and alcohol and have restrictive codes of conduct exclude access to palliative care for the homeless. They also uncovered evidence of systematic exclusion due to mental illness, as a result of which patients were being denied services as they were misidentified as disruptive. On the basis of their findings, they recommended harm reduction strategies as beneficial. This means a move away from the medical model to acknowledge that the homeless may be misusing substances, and that the service provided to them needs to be flexible, such that it should allow substance misuse in a controlled way to achieve the main goal of providing equitable EOL care. In support of this view, Hakanson et al (2015) explained that treatments can successfully be provided simultaneously with the use of illicit drugs if staff have the appropriate knowledge and skills. Further, McNeil et al (2012) indicated that EOL care specialists lacked experience in mental health issues and substance abuse and that collaborative training as well as harm reduction training would decrease discrimination and improve patient outcomes. They suggested that hostel staff and other community agencies involved with homeless populations be linked to EOL care services, as the former have more experience and are more trusted by homeless people and can therefore act as mediators between homeless patients and HCPs.

These problems are also evident in an Australian case study conducted by MacWilliams et al (2014) involving a typical homeless EOL patient with the pseudonym ‘Ray’. This case study identified the challenges faced by Ray and the services and HCPs he came into contact with. While inconsistencies in care could potentially affect anyone, MacWilliams et al (2014) looked at the specific challenges faced by homeless individuals that make it harder for them to access equitable EOL care than for the general population. They concluded that homeless people often do not engage with recommended treatment pathways, and that referrals are made by HCPs with no follow-up, as a result of which homeless people often slip through the net. MacWilliams et al (2014) also discovered that due to concerns around the safe storage, compliance and safety of opioid medication, patients would often not receive adequate pain relief. Similarly, Davis-Berman (2017) observed that clinics were not dispensing pain relief due to fear of abuse. Safe storage and administration of opioid medication was also cited in Shulman et al's (2018) study, which also found that the problem was further complicated by EOL care being delivered in hostels that have no clinically trained staff. Webb et al (2018) suggested transdermal patches for pain relief as a solution to this problem. Controlled drugs such as fentanyl or buprenorphine patches can be transported and administered by district nurses for use if the drug has been prescribed as patient specific (Nursing and Midwifery Council, 2010), although district nurses should always follow their own trust's policy on medicines management before considering this.

The participants in Shulman et al's (2018) study voiced concerns about whether hostels are an appropriate place to receive EOL care, citing barriers such as lack of support for staff, lack of staff confidence and safeguarding concerns. They highlighted that hostels are often noisy and chaotic and are not an ideal place to die. In support of this view, Davis et al (2011) added that the impact on other hostel residents also needs to be considered. To tackle these issues, MacWilliams et al (2014) recommended a specialist coordinator role, which is intended to provide a link between HCPs, hostel staff, and, potentially, social care services. It should also aim to offer support to homeless individuals and those supporting them. This role could be adopted by district nurses if the appropriate collaborative training and hostel in-reach service are established.