Preventing drug-related deaths in Scotland: perceptions and experiences of engagement in a “shared care” model of service delivery Open Access

Currently, there is great concern regarding the level of drug-related deaths in Scotland and the need to provide effective responses to problem drug use. Drug-related deaths are higher in Scotland than in the rest of the UK, and drug-related harms and deaths disproportionally affect individuals and families living in the most deprived areas. In Scotland, after a small decrease in 2021/2022, drug-related deaths rose again in 2023 – 1,172. In the most deprived areas, the rate was 53.7 per 100,000 people compared to 3.5 per 100,000 in the least deprived areas (National Statistics, Scotland, 2024). The final report from the Scottish Drug Deaths Taskforce (2022) emphasised that “more co-ordinated, cross-sectoral and holistic approaches are needed across treatment services” (p9) and that “It should be the responsibility of services to join up support, not the individual to develop and navigate their own care plan” (p13). Primary care settings are recognised as key environments for the delivery of services to people who use drugs (PWUDs) and as a core element of a shared care approach. The model of care includes four service “Tiers” that respond to different levels of assessed need. GP care is Tier 2, intended for people who are more stable and “in recovery”. Secondary care is Tier 3, for people who have more complex health and social care needs and riskier behaviour and is normally delivered in specialised alcohol and/or drug treatment services. Tier 4 is for those with very high levels of need and/or vulnerability, with care delivered in hospital inpatient or residential substance use units or wards.

The shared care model in Scotland – which can vary in practice across geographical locations – consists of general practitioners (GPs) who work alongside other staff employed in primary care settings, such as practice nurses, community mental health/addiction nurses and “link workers” (who assist with arranging services and support for patients). The shared care system includes other health and social care services such as community pharmacists, health visitors, social work services and third sector substance use agencies. The primary care team provides “shared care” for PWUDs under the “National Enhanced Service (NES) contract for drug misuse” in conjunction with NHS secondary care substance use services. This entails provision of an enhanced level of care such as medication-assisted treatment, regular patient review, drug testing, blood borne virus testing and immunisation, wound care, harm reduction and recovery interventions and joint working with other local services involved in the care and support of PWUDs.

In the UK, there is a long history of efforts to implement and improve shared care approaches to working with PWUDs (Gerada and Tighe, 1999). Protocols for shared care, issued by the Department of Health in 1995, were not widely adopted, but over the ensuing years, different shared care models emerged – for example, in response to the HIV crisis (Greenwood, 1996), to promote better management of people who inject drugs (Gruer et al., 1997), to incorporate a primary care liaison worker (Dey et al., 2002) and to develop a “community care” approach by including pharmacists and drug agency staff (Keene et al., 2004). As these examples illustrate, the concept of “shared care” was implemented in different ways from one context to another and has shifted over the years. However, whichever shared care model is adopted, the engagement of service users with the system of care and collaboration between service providers are key elements of effective service delivery (Pomey et al., 2015; Bombard et al., 2018).

The aim of this paper is to explore patient and service providers’ perceptions and experiences of engagement in shared care delivery for PWUDs, by drawing on data from interviews with Scottish primary and secondary health-care practitioners, social care providers, policymakers/commissioners, PWUDs and affected family members. The research adopted a systems approach as a framework for the study design and analysis (Crawford, 2020; Rutter et al., 2017). This recognises the interrelationship between factors influencing drug use and drug-related deaths at the individual, organisational (service delivery) and structural levels. It includes the views and experiences of multiple “actors” involved in the topic of interest. While this paper focuses on the interaction between patients, affected family and practitioners at service delivery level, we also draw on the data to contextualise service delivery within the wider care system and the structural influences which facilitate and constrain engagement in service use and shared care delivery.

The term “engagement” was not pre-defined or questioned directly in the topic guide; it emerged as a recurrent theme from the accounts given by participants when they spoke about their interaction and relationships with each other and with services, often without using the term “engagement” at all. When practitioners used the term “engagement”, they generally meant patient willingness to access services, to take up referrals to other relevant services and to remain in treatment.

A case study approach (Yin, 2014), using qualitative data collection methods, facilitated a detailed description of the contexts in which shared care is delivered and offered insights into factors affecting the quality of shared care for patients who use drugs. The case study sites were two GP practices located in large urban areas of extreme deprivation. Both practices were signed up to their Health Board’s National Enhanced Service contract for problem drug use. The practices differed significantly in staffing complement, team composition and patient list size, but each had a GP partner with a longstanding interest in caring for PWUDs. Table 1 provides a profile of the two study sites.

Semi-structured qualitative interviews (of approximately 60 min each) were conducted with 34 participants, eight GPs, four other primary care staff, two pharmacists, three NHS substance use staff, three third sector drug agency staff, four policy makers/commissioners, six patients who use drugs and receive medication-assisted treatment and four affected family members. Interviews were conducted between April 2021 and March 2022 (xx). Additional forms of data collection, including focus groups and expert workshops, were used to “sense check” findings emerging from the interviews, but we do not draw upon these data in this paper.

Professional participants were recruited via email by purposive sampling in each case study area. Recruitment packs including participant information sheets were made available to a lead GP contact on each site, who invited patients and family members who met the inclusion criteria. Those who expressed interest and gave permission to their GP to share their details were contacted by the researcher. Patient and family member participants received a £20 supermarket voucher.

As data collection took place during the COVID-19 pandemic, interview methods were informed by the NHS Scotland/Scottish Government guidance for medical settings. All patient interviews and some family member interviews were conducted in-person at their GP practice at times when it was possible to meet on site. All professional interviews and the remaining family member interviews took place remotely, using either telephone or video conferencing software.

The topic guide for professional interviews included questions on aims when working with people at risk of drug-related death; the model and efficacy of local systems of care; influence of relevant policies and guidelines; and the impact of the COVID-19 pandemic on service provision. Patients and family members were asked about their and/or their relative’s health and social care needs; their experiences of accessing care and support from different services in their local area and any gaps in this system; and the impact of the COVID-19 pandemic on their access to services. Interviews were audio recorded and transcribed verbatim by a professional transcription service. Transcripts were checked, pseudonymised and uploaded to NVivo12 for coding and exploration. PWUD pseudonyms were assigned by researchers.

In the first instance, the analysis was conducted along thematic principles using the well-established framework developed over time by Braun and Clarke (2006, 2022, 2023). The first analysis was conducted on the whole data set; one member of the research team (KC) performed the initial deductive and inductive coding using a coding frame developed by team members (AO’G, AW, KC, BT) to reflect the study aims as well as allowing for emergent themes. Engagement was one of the emergent themes. A subset of the research team (BT, AW) undertook secondary coding for this paper, using an inductive approach in coding the data to examine the concept of “engagement” and the meaning bestowed by different people, as it emerged in participant accounts. Recurring issues, concepts and themes related to engagement and prevention of drug-related harms and deaths were identified, summarised, compared and refined. Practical aspects of engagement, such as access to services, information sharing and communication, were highlighted as well as relational aspects of engagement between patients and practitioners and between different professionals working in different service contexts.

Ethical approval was obtained from the NHS Scotland Research Ethics Committee (IRAS ID: 287852). All participants provided informed consent.

The research findings are presented under six themes. The first theme sets the scene by sketching out the complexity of patient needs and the complexity of the treatments systems which need to be navigated by patients and professionals alike. We then consider how engagement is influenced by: patient characteristics, attitudes and experiences; organisational and systems level challenges; and professional differences and tensions. The fifth theme reports how practitioners had devised strategies to address engagement challenges and reduce the risk of suicide and drug-related deaths. Finally, we highlight participants’ comments regarding the importance of macro-level social/economic factors and the impact of wider social structural issues on the shared care system and the potential to provide an effective response to drug-related deaths. The views of patients, family members and practitioners are reflected across these themes.

For PWUDs living in the study areas of deprivation, participants noted the additional complications of comorbid health conditions related to injecting and polydrug use, such as asthma, COPD, blood borne viruses, liver disease, hypertension and cancers along with untreated trauma and mental health, stigma and social isolation and their frequent contact with housing, social services, child protection and criminal justice systems. The physical problems associated with long-term drug use, the resultant premature ageing and the link with drug-related death was commented on:

Another GP linked premature ageing and drug use to the deprived circumstances of the communities the patients lived in, noting that poor health and low life expectation was the norm within those communities:

Participants were clear that responding to such multiple and complex needs requires an integrated “shared care” approach between primary care teams including: generalists, secondary care services (“specialists”), social care organisations and others such as the prison, welfare benefit and housing support services. Shared care could provide a holistic approach to the multiple health and social care needs of people who use drugs, as illustrated by a GP who commented:

However, systems of care were described as complex, difficult to navigate and often unfit for purpose:

There was agreement that practitioners experienced considerable challenges in delivering shared care and that patients often found the system inaccessible and unsuited to their needs. Describing the shared care system as a “pathway that flows from the community to a primary care setting into a specialist setting where there’s access to a variety of different services which hopefully the person can engage with”, policymaker/commissioner B voiced the common concern that it was “quite difficult for the patient to feel they want to engage […] sometimes I think patients are made to feel they have to engage with things that they perhaps feel as being impediments rather than advantageous”. GPs, in particular, commented on the problems patients experience in moving between primary and secondary level services, “people don’t pass between the tiers” (GP M) and “it’s chaotic drug users where it doesn’t work” (GP B).

Overall, comments from participants revealed the problematisation and individualisation of “engagement” at both service user and practitioner level. On the one hand, engagement was presented as dependent on the service user, with comments about “lack of engagement” frequently attributed to patient attitudes, behaviour or negative experiences of services. At the same time, successful shared care delivery was seen to rely heavily on individual practitioners “folk who are very engaged – enablers – an enthusiast” (Policymaker/Commissioner B). Indeed, considerable determination appeared to be required to make shared care work at practitioner level, as participants reported difficulties with collaboration across services and between professionals with different occupational backgrounds, professional ethics and guidelines and different roles, which could result in lack of co-ordinated holistic care.

Thus, reasons affecting engagement encompassed individual-level factors, organisational and systems-level factors and tensions between different professional groups and different practice approaches. As one third sector worker told us, “Sometimes barriers are the client’s perception, the client’s behaviour, the client’s chaotic lifestyle and sometimes it’s the setup of the services, you know” (Third sector worker B). Although not a focus of the interviews, participant comments revealed awareness of how wider socio-economic and political factors were also important drivers of drug-related deaths and constrained what could be achieved within shared care services.

Thus, all participants in this study saw service engagement as vital to the reduction of drug-related risks and harms and the prevention of drug-related deaths. While patients and family members usually spoke about their interaction and relationships with individual practitioners and how that influenced engagement with services, service providers frequently referred to engagement within the shared care system and the issues that arose from the need to collaborate across professional groups and across services.

Accounts from all participant groups detailed problems patients had in engaging with services because of their personal circumstances, their current physical and mental health status, past experiences of services or attitudes towards accessing services. Comments from family members and PWUDs had a strong focus on how relational experiences influenced trust and willingness to engage. Both good and bad experiences were reported:

Accounts also highlighted how intrusive questioning and stigmatising environments acted as barriers to service use:

Practitioners also recognised relational factors and commented, too, on the practical barriers to engagement, such as limited resources to access services, problems in keeping in contact with PWUDs who may have no stable address and the need to be aware of the individual’s current ability to engage (e.g. being in crisis):

There was acknowledgement from practitioners that they, too, found the shared care system difficult; collaboration across services was frequently inadequate and failed to match patient needs. Some of the difficulties were because of the location and unappealing aspect of services which deterred patients from taking up referrals:

Even where services were provided close by, some patients were reluctant to attend, but according to one GP:

Access to mental health care was reported as particularly difficult in one case study area because Community Mental Health Services had very low capacity and operated a high threshold for access, including some services choosing to exclude people with problem drug use from treatment. This was seen as an especially important risk factor for drug-related deaths:

Third sector worker B described how people with complex needs struggle to engage with services:

This view was endorsed by another who spoke about the need for closer working with a mental health nurse or finding an easier, more timely way to access mental health services:

Systems-level barriers to engagement are perhaps exemplified by problematic transitions between services. For example, one GP describes the challenges arising when PWUDs are released from prison at short notice:

Other problems included poor communication between services, lack of shared information about the patient and insistence that the patient demonstrate “commitment” to treatment before being accepted into the service. Examples of these organisational and structural problems, given by a link worker and a pharmacist, illustrate the difficulties.

The link worker’s account:

The pharmacist’s account:

Pharmacists are often the service with the most frequent patient contact, as PWUDs come to collect their prescriptions weekly or even daily. They noted how exclusion from “full integration” in the system of care meant that often they struggled to report worrying changes in patient presentations back to GPs; for example, they had to rely on the same GP phone line as patients use and, therefore, had to navigate lengthy queue systems.

Commenting on the shared care model, one GP voiced a common concern that an increase in part-time working, and the move to team working, especially where teams work in shifts, means that there is not a “go-to” person responsible for the patient. This can result in a “collusion of anonymity” whereby “if everyone’s responsible, no one’s responsible and everyone’s anonymous, and if something really bad happens to a drug user there might be eight or nine people who might be responsible for that, so maybe no one’s responsible” (GP C). As a result, the shared care model becomes less effective in addressing patient needs and reducing the risk of drug-related death. The problem is compounded by relational aspects as third sector worker A explained:

Practitioners felt that these organisational and systems-level issues were compounded by lack of adequate resources. According to third sector drug worker A, the service was “woefully underfunded […] in a climate of decreasing resources”. Another person commented that:

Deficit in workforce capacity and capability to provide the high-quality care required to meet the complex needs of the population affected by drug use was frequently mentioned and participants stressed the need for service delivery to be supported by a skilled and appropriately resourced multidisciplinary workforce, both within primary care and also out with primary care.

Issues also arose concerning collaboration between practitioners with different occupational backgrounds. Practitioners spoke about having different goals, different interpretations of the patient’s situation and different views on appropriate action regarding patient care. For example, a mental health/addiction nurse reflected on relationships with social workers and how this affected patients:

Working with different models of care and feeling that there was a status difference between the different professional groups was also raised as an issue:

For GPs, the reality of responding to the needs of their patients frequently went against current policy trends, especially the focus on a “recovery approach”; as one GP mentioned:

Most GPs felt similarly, that for the patients they managed, a “harm reduction” approach was more appropriate than the “recovery approach” practised more widely by secondary substance misuse services and some third sector services. One GP, for instance, discussed the involvement of a secondary substance misuse service as “disruptive”:

Such tensions and differences in treatment approaches between practitioners in the shared care system risked undermining patient trust in services and patient engagement. For instance, one third sector worker commented on differences in prescribing practice between services and how it affected patient engagement:

Both practitioners and patients noted the strategies they used to encourage or secure the kind of service engagement they deemed necessary and appropriate.

In the case of patients and their families, this not only included action to secure the help they needed but also it meant trying to control relationships and interaction with service providers to manage the level and nature of their engagement. One PWUD, for example, asked his lawyer to put him on a drug treatment and testing order rather than community service as he felt that would get the kind of assistance he needed. He commented:

Another person described how she approached social services for help to settle her daughter with her grandmother while she dealt with her drug use. They were unwilling to help because, they said:

Attempts to control the interaction with service providers, for instance by limiting details of personal information and experiences, were mentioned by patients and family and could lead practitioners to feel that the patient’s main goal was to secure a script. As one GP said:

Failure to control the situation was acknowledged as a reason for refusal to engage further with the service or the practitioner. Family members spoke of their frustration with PWUDs who told practitioners what they thought they wanted to hear or told lies to avoid suggestions for treatment. For example, one mother told us, “we spoke to a psychiatrist and that up there as well after one of the incidents one night; we sat and spoke, but I think John knew as well what he could say when they were, like, assessing him, they would ask him things, I think he knew exactly what to say so that they wouldn’t section him” (Angela, Family member).

Practitioners were aware of the organisational, structural and relational problems and told us how they attempted to navigate the system both for themselves and their patients. They gave examples of strategies to engage patients and recounted how they tried to overcome some of the barriers to patients’ access to services and their reluctance to attend secondary care services. One key strategy was careful management of prescriptions:

Other responses included providing personal support to introduce a patient to a different service, accompanying them on the first visit and acting as “advocate” because “Getting out the door and getting in a new door is the big issue for people” (Link worker B). The importance of reducing stigma prompted attempts to “normalize” help seeking:

Using opportunistic encounters to provide health care and maintaining the relationship by “rolling with resistance”, keeping the patient “visible” and waiting till the patient is ready to provide information were also seen as ways of sustaining engagement over time. Offering social prescribing as well as health related referrals and talking to patients in a way they can relate to were useful communication and engagement strategies. Talking about contact with the community psychiatric nurse, a family member mentioned:

This study was located in two very deprived areas and the impact of deprivation at individual and community level was recognised by participants as an important factor in drug use and drug-related deaths. A practice pharmacist highlighted the interaction between physical and mental health, social inequalities and drug use:

One of the mental health nurses made the important point that the effects of deprivation were cumulative and that effective responses were needed long before people were at risk of drug-related death:

PWUDs and their family members commented on the risks within their local communities. These included being exposed to violence as well as being surrounded by people encouraging the use of drugs:

There was consensus among practitioners that providing care for people with complex problems required a “shared care” approach to offer an appropriate range of physical health, mental health and social care responses. However, the study results illustrate that engagement in service receipt or service delivery is situated within a complex network of factors interacting at individual, organisational and systems levels to facilitate or hinder engagement. Patients, family members and practitioners were all aware of the difficulties in securing and maintaining engagement of patients in the services. These findings are not unique to the care of PWUDs. As suggested in a review by Pinto et al. (2019, p. 9), relational (as well as practical) barriers to engagement “exist in the intersection of client and provider, client and organization, and client and systems”. Our results indicate that the relational elements of engagement between different sectors of the treatment system and different professional groups are equally as important as patient-related interactions. Studies of partnership approaches to health service provision (or “joint working”) have reported similar findings, in that navigating complex systems of shared care are difficult for all concerned, and inter-professional differences in policies, organisational aims and working practices make engagement across the services and professions problematic (Hunter and Perkins, 2014).

Caring for PWUDs in primary care is perceived as challenging. Lack of clinical expertise, training and inadequate resources and support systems (Dey et al., 2002; Matheson et al., 2003), the need for health-care professionals to re-conceptualise their role in caring for these patients (Ford, 2011) and difficulty in integrating a response to the social determinants of health within shared care models (Sullivan et al., 2017) have been enduring problems. Research on engagement has largely responsibilised and individualised “problems” of engagement to different actors in the field, including patients, practitioners and services. In doing so, engagement “problems” and “solutions” are largely framed as a (dys)function of the shared care system and the individuals within the care system. Although this way of thinking about and defining the problem emerged in our study, participants’ engagement narratives also revealed political messages about the “problems” of a flawed care system, for example, by highlighting deficits in resources, unequal power relations and structural inequalities.

The inverse care law, which describes how people living in deprived areas and most in need of health care are least likely to receive it, persists in the UK National Health Service, despite attempts to address the challenges and support primary care services in deprived areas (Fisher et al., 2022; Blane et al., 2023). In Scotland, Blane et al. (2023, p. 4) report that, since 1999, “there have been numerous policy commitments to strengthening general practice in deprived areas, recognising the key role of GPs – as part of integrated multidisciplinary primary care teams – in reducing or mitigating health inequalities. […] However, there remains a major implementation gap between Scotland’s policy ambitions to address health inequalities and sustainable delivery on the ground”. The mismatch between need and service provision was highlighted in a study comparing clinical encounters in affluent compared to deprived areas. Researchers found that the greater need of patients in deprived areas was not met by longer consultations, higher GP patient centredness and higher perceived GP empathy (Mercer et al., 2018). The effects of the inverse care law are particularly pertinent when considering patients with complex problems, high levels of mental and physical multimorbidity and, in the case of PWUDs, at greater risk of suicide and drug-related deaths.

The impact of structural inequalities on drug-related deaths, although acknowledged as important, has received less research attention, especially with respect to considering how macro-level conditions influence the environments in which drug use and drug-related deaths occur and constrain effective primary care responses. Thus, while engagement is understood as a fundamental component of care systems that can prevent drug-related deaths, it also needs to be reframed so as not to divert attention away from political choices and responsibility for the kind of society in which engagement is made possible or not.

In the Scottish context, the relationship between drug-related deaths, deprivation, health and social inequalities cannot be ignored (University of Strathclyde, 2022; Miall et al., 2022; Scottish Government, 2023; Finch et al., 2023). Although it has been argued that there is insufficient evidence to claim deprivation as the key driver of Scotland’s higher drug-death rate (van Amsterdam et al., 2021), a recent report indicating lower life expectancy in the most deprived areas, argues that, “Drug related deaths are a prominent part of the decline in life expectancy – and the growing inequalities gap” (Miall et al., 2022, p. 6). The report notes that the biggest single contributor to mortality in 15–44 year-olds, is drug deaths (p50) and, as stated earlier in this paper, inequalities in drug deaths are greatest for areas with large and relatively deprived urban centres (p50). In addition, the evidence points to an accumulation of multiple sources of deprivation, likely to be most prevalent in the most deprived areas (Finch et al., 2023). This accords with the perceptions of participants in our study. Participants all “situated” their views and experiences within a wider context, for example, by drawing attention to structural inequalities, historical and cultural factors influencing drug use and drug-related harm, political influences, constraints imposed by commissioning and contracting processes, organisational structures, the location of services and the built environment. These played an important role in what was considered possible to achieve in primary care and what standards of care could reasonably be expected.

Scottish policy recognises the importance of addressing health and social inequalities. However, as Finch et al. (2023, p. 8) emphasise, “there is a persistent and growing implementation gap” which has resulted in a lack of progress towards solving the multiple complex health and social problems, including drug-related deaths, linked to deprivation. As Miall et al. (2022, p. 11) argue, health and social care services have an important role in tackling inequalities, and the findings from our study show how practitioners are attempting to devise strategies to enhance engagement at the individual level. But actions at individual and service delivery level are only part of the picture. As demonstrated by our study findings, prevention of drug-related deaths requires changes both in the environments where drug use and drug-related harms occur and in the shared care system of service delivery. This includes consideration of how to address the “collusion of anonymity”, exacerbated by the increasing move towards team working, referred to by more than one of the practitioners in the study. It also includes consideration of how to facilitate ease of access to the necessary range of services within the patient’s locality and how to reduce tensions arising from different professional and practice approaches to care. To improve engagement and provide adequate, appropriate support for primary care teams to deliver high quality care, there is a need to reframe narratives about “engagement” as a relational practice and a shared endeavour in the prevention of drug deaths.