Interrogating the effectiveness of service engagement for underserved populations in complex health and social care systems: towards an equitable engagement strategy

Abstract

There are increased sector-wide efforts within health and social care systems to engage those with lived experience in service design, delivery, and monitoring – aiming to secure more equitable health outcomes. However, critical knowledge gaps persist around how national whole-system engagement strategies can account for the challenges experienced by populations that encounter exclusion within complex multi-layered systems. This includes a failure to delineate shared challenges across groups, and to develop transferable cross-group frameworks to assist sector-wide change. There is, therefore, a danger that those groups already least heard will be collectively left behind. With a view to informing a more inclusive engagement strategy in Ireland, this national study aims to investigate multi-level (policy and strategic, operational, on-the-ground services, individual) shared challenges impacting engagement for five populations who have been identified as underserved groups in a complex health and social care system, including: (1) those who misuse drugs and alcohol, (2) those who are experiencing homelessness, (3) those experiencing mental health, (4) migrants and those of minority ethnicies, and (5) Irish Travellers. Adopting a mixed-methods approach which draws on an evidence-informed multistakeholder perspective, this study employs data from: focus groups and life-course interviews with lived-experience populations (n=136), five focus groups (n=39) and a national on-line survey (n=320) with population-specific services providers; and national-level stakeholder interviews (n=9). Two cross-group participatory consultative forums with lived-experience and provider participants (n=28) were used to co-produce priority action areas based on study findings. This article presents findings on shared challenges in engaging these groups around leadership and commitment, implementation and action, population capacities, trust, and representation, stigma, and discrimination. Derived from these challenges, six development areas are presented to advance an inclusive equitable engagement approach in Ireland. These comprise: 1) balancing top-down prioritisation, and bottom-up direction; 2) sustaining multi-level, multi-form implementation; 3) measuring effectiveness and action; 4) embedding inclusive equitable engagement; 5) trust as a prerequisite, and outcome; and 6) an equalising, agency empowering agenda.

Background

In recent years there have been increased efforts within health and social care systems to engage those with lived experience in the design, implementation, and monitoring of care services [11, 42] – a process referred to in this article as ‘service involvement’ (SI). ‘SI’ is defined in this study as an overarching umbrella term that incorporates a spectrum (or ladder) of activities that vary in terms of where power and control between service providers, and those who use services lie in the decision-making process. Whether termed Public and Patient Engagement, Public and Patient Involvement or service-user involvement, these strategies aim to embed the voices of target populations to improve services and secure better health outcomes [53]. Despite the rapid growth of system-wide strategies in research scholarship, gaps in knowledge persist around how such strategies can account for the circumstances and barriers of groups who encounter unequal access to care, and who are less likely to be engaged with in a meaningful way within systems [59, 72]. In Ireland these underserved populations can include those who use drugs or alcohol,those experiencing homelessness,those using mental health services; those from migrant, minority ethnic and Roma backgrounds; and the Irish Traveller community [22, 61]. With nations, like Ireland, beginning to adopt whole system approaches to embedding involvement, there is a potential danger that a one-system level approach to engagement will not consider the barriers to SI faced by underserved groups. As such, there is a risk that populations like these will be collectively left behind without due consideration within such approaches [15, 62].

A burgeoning international literature documents factors that detract from effective SI for underserved groups. This includes research on services where there is a longer track record of engagement, such as within drugs and alcohol [14, 50] and mental health supports [45, 47]. It also includes services where involvement is still considered to be emerging, such as in homelessness supports and in the area of migrant and minority ethnicity health [40, 69]. Studies point to barriers related to communication, system familiarity, group-level exclusion and precarity, and entrenched organisational culture and resourcing issues [1, 7, 33, 49, 62].

However, much of the empirical literature concentrates on single populations, with less work delineating shared challenges [67, 68]. Therefore, while population-specific frameworks for supporting some groups are becoming evident [6, 21], more transferable approaches that have relevance for different underserved populations are markedly absent. While it is of course important to identify challenges which may be unique to specific populations, there is also a pragmatic need to understand any shared challenges, in a ‘one system’ approach.

Further, within existing research there is often a failure to represent the views of key actors at different levels of health and social care systems (e.g. national stakeholders; service providers; individual users of services) within the development of these approaches [39]. There is also a critical lack of research on how to support the identification of agreed challenges and solutions amongst these multilevel actors, and amongst multiple sets of groups [71]. Consequently, comprehensive system-sensitive frameworks for underserved populations, that support the advancement of inclusive SI strategies, have been slow to emerge.

These deficits in approach come to the fore with the increasing complexity of contemporary health and social care systems [3, 39]. The rich potential for SI to enhance outcomes has certainly been recognised within such structures [39], but meaningfully incorporating underserved groups’ voices can be challenging given the features that can characterise such systems [48]. This includes: the complex integration of health and social care in some instances, or their complete separation in others [63],the balance and degree of mix in provision (public-private-voluntary) [9], the range of actors and their mandates [34], multi-level system governance, and power imbalances [78],and dis-integration across levels and geographical regions [41]. Underserved populations can, under such circumstances, feel disempowered within complex systems and disillusioned with efforts to implement change [9]. There has not been a sufficient examination of how to bolster and advance SI for such groups within these complex structures.

This article seeks to address these research gaps. With a view to informing a more inclusive strategy, this study aims to investigate multi-level (policy and strategic, operational, on-the-ground services, individual) shared challenges impacting SI in a complex health and social care service system for underserved populations. This aim is couched within the acceptance of a diversity of experiences across and within different underserved populations but is driven by a pragmatic need to understand shared challenges in a system which often develops and implements broader one-size-fits-all strategies. The study utilises an evidence-informed multistakeholder research process that harnesses lived experiences (LE) to agree on priority action areas, and policy and practice messages for the advancement of SI for these groups within the Irish health and social care system.

The research focuses on five groups identified within the Irish health and social care system’s social inclusion agenda as being at particular risk of exclusion [43, 44], and these are: those who use drug or alcohol services (DA),those experiencing homelessness (HO); those who use mental health services (MH); those from migrant, minority ethnic and Roma backgrounds (MMER); and Irish Travellers (TR). From the outset of this article, it is necessary to recognise that these are not mutually exclusive groups, but they are categorised as these separate populations groups within the Irish health and social care system, albeit acknowledged as having significant overlap in composition. Furthermore, while the focus of this article is most definitely on shared challenges across these groups, to provide a layer of granularity, any notable findings which are specific to the groups individually are highlighted.

Service Involvement and the five populations in Ireland

Ireland represents a particularly interesting case for this analysis. First, the Health Service Executive (HSE), which manages Ireland’s health and social care system, has implemented actions to embed SI in services across its nine Community Healthcare Organisation (CHO) regional areas [35]. This commitment is demonstrated most recently by the HSE’s publication of the ‘Better Together: Health Services Patient Engagement Roadmap’ [43], and by an investment in capacity building to support engagement (e.g. National Patient Forum; CHO-level patient engagement officer roles). While policies solely dedicated to SI for any of the five groups do not exist, there has been a long-standing recognition of the importance of active engagement within Irish general strategies targeting some of the populations. This is particularly evident in relation to partnership approaches pursued within mental health services (e.g. Sharing the Vision’ policy: [26]), but also in Traveller Health (e.g. National Traveller Health Action Plan (2022-2027): [25]). Second, Ireland possesses a health and social care system that is challenging in its complexity. Key features include: (1) a national system but with regional CHO delivery, (2) the integration of health and social care within this system, and (3) a two-tier model that involves a complex interaction of public, voluntary and private elements in the production, delivery, and financing of healthcare [46]. Third, and critically, there has not been a comprehensive effort to examine SI for the five groups in Ireland, with neither LE perspectives on key challenges nor the views of national stakeholders and service-providers being sufficiently captured.

A summary of each of the five populations is now presented to contextualize the groups’ demographic and care need profile in Ireland.

For those who use drugs or alcohol, it is estimated that 15 per cent of the Irish population suffer from Alcohol Use Disorder, and seven per cent report illegal drug use [58]. While certain socioeconomic factors are associated with drug use (poverty,unemployment) in Ireland, there is significant variation in the age and educational background of impacted groups. Different needs are also linked with different sub-groups, such as those who use night-life drugs, those who are older and use opioid drugs, and those in recovery.

Regarding homelessness, in 2024 there were 13,866 people without a home in Ireland, representing the highest figure ever recorded. This included those in emergency accommodation, longer-term supported accommodation, and rough sleeping. Thirty per cent of those experiencing homelessness are aged 45 years and over, and 43 per cent are from migrant backgrounds [27].

In terms of those using mental health services, Ireland is considered to have a higher-than-average rate of mental health issues, with 13-21 per cent of people estimated to experience these challenges. Being a woman, being middle-aged or older, being separated, divorced, or widowed, or having lower education levels can increase risks of mental health problems in Ireland [29].

In relation to migrant, minority ethic and Roma groups, approximately 14 per cent of the population in Ireland are foreign nationals. The largest proportion of migrants come from Poland (15%) and the UK (13%). Although possessing similar education and employment levels as Irish citizens, migrant groups experience a lower median income and higher poverty rates [54]. With respect to ethnicity, Ireland comprises; two per cent Indian, Pakistani or Bangladeshi; 1.5 per cent Black; 1.4 per cent Asian or Asian-Irish, and 0.3 per cent Roma.

Finally, Irish Travellers, who constitute 0.65 per cent of the Irish population, are an indigenous ethnic group with historic nomadic traditions. On all indicators, Travellers report much worse health than the wider population [4]. Sixty-seven per cent of Travellers (compared to 83% of the wider population) report ‘good’ or ‘very good’ health, and 59 per cent and 63 per cent of Traveller men and women respectively reported bad mental health. Suicide rates amongst Travellers are estimated to be six times higher than in the general population (ibid.).

Methods

Study design

The study design incorporated three components. First, the research accounted for the multiple levels within the Irish health and social care system that are implicated in SI [55], that is, those who make policies or strategies at the top, those in the middle who operationalise these strategies, those at the bottom who provide services on the ground, and those who are in receipt of these services. The study recruited participants at the national and strategic level, the level of service coordination and provision, and critically the individual level through members of the five groups. Gathering the multiple perspectives that impact, and are impacted by, SI allowed for an examination of different understandings operating within the care system. Second, the study drew on a voice-led multistakeholder approach to guide this research [30, 79]. The approach helps foster shared understandings of complex challenges and the sort of solutions required to address them. In addition to guiding the overall research ethos, this approach informed two parts of its design: (1) a Project Advisory Board (PAB), which (along with academic experts) included LE perspectives in all study stages, and (2) the selection of data-collection methods that emphasise participants’ experiences and support the co-production of recommendations. Third, the research employed a mixed-method sequential design, incorporating semi-structured interviews, focus groups, a large scale online survey and translation forums. Allowing for synergy between in-depth, detailed insights and more representative knowledge, mixed-method designs have gained prominence in the study of SI in health and social care settings [23, 24]. The study comprised of five interconnected components, completed from March 2022 to July 2023.

Strategic context

The first component addressed the strategic policy, practice, and civil society context of SI, and included nine in-depth interviews with national-level social inclusion, and health and social care stakeholders. Participants comprised of seven women and two men, and included those working across the HSE, government policy departments, inter-agency taskforces, and civil society organisations. The interview guide was organised around categories that included: (1) enablers and barriers to SI; (2) the role of SI in services for the groups; and (3) the role of SI in planning and policy developments.

On-the-ground service use/delivery

Twenty focus groups were conducted to explore on-the-ground, multi-stakeholder views of SI in health and social care services for the five populations. For each population, one focus group with related service providers, and three focus groups with those who use services were completed. This meant five discussions were conducted with service providers overall (n=39, on average eight people per group), and 15 with those with LE across the five populations (n=101, on average seven people per group). Of the LE discussions, five were online, and ten were in person, and of these fifteen discussions, eight comprised participants based in Dublin, while seven comprised participants based across the west, north west, midlands and south east of Ireland. For LE discussions, and in consultation with the PAB -some of whom had lived experiences in the same areas as LE participants in this study- different sub-groups were targeted in recruitment to maximise the diversity of population samples. This was also done in response to diverse and specialised needs, and potential challenges in both accessing services, and gathering or giving views on services (see Table 1). Participants were recruited through: snowball sampling; representative associations, networks and local community groups; and with the assistance of specialised HSE services, or community-voluntary providers. Providers were recruited through open and targeted calls distributed through the national networks of the HSE National Offices supporting these populations. For provider and LE focus groups, the discussion guide included questions on what participants felt were challenges for effective SI. All five provider discussions were conducted online. Participatory Learning and Action (PLA) techniques were employed in each focus group. PLA [18] offers practical approaches where power asymmetries may exist. These techniques were designed to bring individual and collective voice into discussions to equalize power between groups and/or to encourage individuals to express their views [19, 75].

Table 1 Sub-populations involved in LE focus group discussions

Life-course engagement

Thirty-five life-course interviews were conducted: 18 online, 15 in-person, and two by telephone. Interviews explored in-depth views on involvement, and how these views were influenced by life-course experiences of care, and individuals’ positions as members of these populations. A secondary purpose was to capture the perspectives of those requiring specific sets of services or those at greater risk of exclusion (see Table 2). As with the focus groups, the interview sample was determined by discussions with the PAB. Participants were recruited in the same manner as used for the focus groups. The final sample, consisting of 14 women and 21 men ranging in age from 23-76 years (mean age: 44 years), comprised the following breakdown across groups: DA = 7; HO= 9; MH = 6; MMER = 8; and TR = 5. The interview involved three parts: (1) an open participant narration to elicit narratives of care service engagement; (2) a visual life-path exercise [36, 38, 70] ato probe on life-long helth biographies; (3) a semi-structured guide to explore topics such as life-course health experiences; views and goals of SI; experiences, barriers, and enablers of involvement; and preferences for involvement. Eighteen online, fifteen in-person, and two telephone interviews were conducted.

Table 2 LE Interviews participant sample

Data analysis: components 1-3

All interviews and focus groups were audio-recorded, transcribed, and analysed using an inductive, semantic thematic approach [13]. All research strands were analysed separately but followed the same approach. A provisional coding framework was developed from an initial reading of transcripts in each strand by multiple team members, and then agreed amongst the team. Transcripts were analysed in an iterative process that sought to merge key elements of the text into descriptive codes, until major inductive themes were identified in each transcript. These themes were used to refine the provisional framework into main themes and sub-themes. NVivo 12 qualitative software was used to organise the data.

Involvement audit

An on-line survey was developed in SurveyMonkey and distributed to service provider organisations working with, or as a part of, the HSE to deliver care services to the five populations. The survey focused on the nature, quality and quantity of SI activity levels and experiences within respondents’ organisations or units, and the sort of barriers and enablers encountered in SI across Ireland. Five survey links were created, with each targeting providers linked to each group. Survey links were distributed: (1) to complete or partial accessible provider lists held by collaborating HSE Offices; (2) through the HSE Offices’ network of regional teams; and (3) through group-specific organisations and associations. After all submitted surveys were reviewed, 320 were deemed to have been completed, with Table 3 presenting the breakdown by group-specific providers. The survey comprised of a 51-item questionnaire, encompassing five modules: (1) organisational details; (2) respondents’ personal knowledge, experience, and views of SI; (3) SI in the respondents’ organisations/units; and (4) population specific factors related to involvement. Specific questions were informed by the findings of earlier research phases, and by the structure of similar surveys within the literature [74, 76] and included Likert scales, multiple choice items and comment boxes. The survey was released for four weeks and took approximately 20 minutes to complete. Survey data were managed and analysed using SPSS 27. Frequency and cross-tabulation statistics were completed to analyse the data.

Table 3 Number of surveys completed by group survey link

Triangulation and translation

The fifth and final component comprised of two strands. First, the findings of each preceding phase were synthesised, supporting the overall interpretation of the research. This involved the integration of qualitative and quantitative analyses at the level of findings’ themes, helping support the identification of overarching conclusions within the empirical data. Second, two multi-stakeholder Consultative Forums were conducted. The Forums helped to validate the findings, and to agree on priority areas and co-produce policy messages to advance SI for the five groups presented in this paper. Each Forum involved a sample of 14 participants (28 people across both Forums) drawn from the focus groups. Across the two Forums, there were 12 providers and 16 LE participants in all. Participants were selected based on representing different sub-groups from different geographic regions. At each Forum, findings were presented in three 15-minute presentations. After each presentation, participants discussed the findings in small mixed groups (4-5 people) with other providers and LE participants for 35-40 minutes. Observations and suggestions were fed into a plenary discussion and used to distil priority areas. Again, to help ensure representation of voice amongst participants, PLA techniques were employed.

Results

Please note, results are based on perspectives from diverse informants. Informants included ‘high-level’ stakeholders (i.e. national-level policy, practice and civil society actors in social inclusion, health care or social care), service providers, and those with lived experience. To help contextualise challenges for effective SI, a brief outline of current SI activities across the five populations will be first presented.

Level and quality of service involvement

Findings suggest that overall, there is substantial SI happening with the groups, but this activity is not always of high quality, varies across populations, and does not sufficiently engage groups. Within the provider survey, 93 per cent of respondent organisations conducted SI of some form, with minimal differences in rates of activity across the groups. For some populations, such as those who use MH services, and the TR community, it was clear that there were long-standing examples of partnership approaches. For others, such as MMER communities, and those experiencing homelessness, SI was generally growing but embryonic. However, limitations were found in existing SI activities for all groups despite the high levels of activity. First, there was a concentration on less intensive SI: comment boxes or questionnaires were the most frequently reported means of SI (51%), followed by focus groups (41%) and membership of committees and boards (37%). Second, all study informants indicated there were was still a tendency towards more superficial engagement, with activities more likely to take place at the levels of ‘engaging’ (individuals express views and influence some decisions), ‘consulting’ (individuals invited to provide views in surveys/meetings) or ‘informing’ (telling individuals about a service), rather than ‘co-production’ and ‘co-design’ levels – the exception being MH and TR services, where co-production was found to be the most common level of engagement. While the potential of SI was recognised by national-level stakeholders, there was frustration around a lack of progress given investments made. Correspondingly, many LE participants had rarely or never been asked for their views, and if they had, their participation tended to be less involved (e.g. once-off, surveys).

Challenges

Five challenges were identified as impacting SI effectiveness for the five groups. Findings on each challenge are drawn from all study informants and data-collection strands: national stakeholder interviews; provider focus groups and survey; and lived-experience focus groups and interviews.

Leadership and commitment

Across all informant groups, but especially national-level stakeholders and providers, leadership and commitment were highlighted as key influences on SI for the populations. Although positive examples were discussed, participants were concerned about gaps and deficiencies. At the national strategic level, some stakeholders noted that systemic change was required. For many, this was about HSE and government leadership showing a commitment to SI, and leading the implementation of a well-considered strategic direction, as this stakeholder interviewee explains:

We need leaders in the health service who see the value of service user involvement…That it’s not just a concept, that it is something that is modelled and something that is practiced. And when…our colleagues who see something working for example, at national level, then they're more likely to say well look we can try replicate that… and do it locally…(IN.SH.05¹)

Similarly, providers felt there was little meaningful high-level leadership, with some noting that the SI responsibility appeared to fall to smaller, on-the-ground organisations. Almost half of respondents (47%) reported that SI for the groups was not given adequate attention within national policy – with MMER (66%) and TR (75%) provider organisations highlighting this most. Forty-six per cent also disagreed that SI was sufficiently prioritised for the marginalised groups they serve within the HSE – an issue for those delivering MH services (61%) more than any others. While LE participants did not address specific deficiencies in policy, several individuals said that they didn’t feel considered by those higher up. This is illustrated by the following quote from a younger LE participant who had used child and adolescent mental health services:

Yeah, I just feel adolescents are sort of looked down upon sometimes, not all the time. It's sort of like we’re not stupid, we’ve a lot to offer, we have a lot to say. Our opinions are valid as well. (FG.PWLE.MH.CA)

Stakeholders and providers felt strongly that buy-in at the regional (CHO) and local organisational level was critical to give SI visibility; to help create an ethos of SI; and to ensure SI is embedded within an organisation’s daily and strategic activities. Nevertheless, almost a quarter of respondents (23%) felt that the level of internal leadership was insufficient. This was more of an issue for those in MH (36%) and MMER services (41%). Focus group provider participants were more likely to draw attention to these issues, with many speaking about the absence of organisational support, resulting in tokenistic engagement. This attitude can be seen in the following quote from a service-provider in the drug and alcohol supports sector:

I suppose I’d like to see stronger leadership and funding to support this. And not service user involvement or engagement being just kind of the added thing to the agenda… if somebody else has time to do it. But it’s always kind of the soft option or the, you know, it’s the thing that you do when you’ve ticked every other box…(FG.SP.DA)

Several LE participants opined that some of the professional occupations that they engaged with neither valued their views, nor SI processes per se, like one man who had used mental health services for severe and enduring mental health challenges:

You’d have to challenge…this particular doctor, on his arrogance, because he is an arrogant man, and his theory is that ‘If I don’t think of it, it’s not going to be used.’…It has to be his idea… I don’t know whether he would see us as being fit to talk to him…(FG.PWLE.MH.SE)

Such experiences echoed the national stakeholders’ and providers’ accounts, where SI was described as being perceived by professions as a threat to traditional power dynamics:

So, there’s a kind of a tension there always, I guess, with professionals in terms of who knows more and who has the authority to say what should happen… And we saw it in terms of peer support, hugely… you know, some of the professions were hugely threatened at having peer support workers. (IN.SH.06)

Provider participants argued that SI should be built into service plans as it is too often relegated, restricted, or ring-fenced into a small number of activities. Providers also noted that engagement was often only assigned to a small team or an individual, and its existence often precarious and rarely sustainable, within organisational work-programmes.

Implementation and action

Implementation and action emerged as a dominant theme, incorporating a multifaceted set of issues -related to resources and training, measurement and accountability, and follow-through and impact- that detracted from effective involvement. All informants raised lack of SI implementation as a challenge, with many feeling that positive change was either not happening or not happening fast enough. Some described an endemic implementation deficit where plans were often not executed, damaging confidence in SI processes, as is illustrated by the following quote from a stakeholder interviewee:

…I think that that the lack of implementation, or I suppose slow implementation in some of those key areas…becomes a key challenge. Because if you have a number of consultative processes, which are exhaustive and that have key recommendations for implementation, but don’t happen. How do you ensure you know I suppose confidence in the system? …(IN.SH.08)

As this LE interviewee who was from a migrant background states, SI outcomes which are not implemented, do not generate meaningful change:

It’s useful if they do something about it or with this information. If it’s just data and...in a file, a paper file in a cabinet, that’s no use at all. (IN.PWLE.MM.07)

Three factors were thought to drive implementation challenges. First, in terms of resources and training, 69 per cent of provider respondents agreed that there was a lack of SI resources, with again a higher proportion of those providing MH (77%) and MMER (94%) services agreeing. Informants noted the importance of programme budgets for implementation, and they had specific ideas around how such budgets should be allocated. Stakeholders and providers were strongly focused on the need for dedicated staff to avoid overburdened workforces:

And the doing of good service user engagement takes time away from [service provision]. So, the challenge will be what doesn’t get done...If your resources are finite, you're competing with the waiting list or the productivity. (IN.SH.09)

Adequate training investment was also flagged as critical, to assuage staff resistance, and to build inclusive relationships between providers and groups. Nonetheless, less than half of provider respondents stated that their organisation facilitated training. Finally, remuneration for LE participants was discussed as something which, when deficient or absent, could make them feel unappreciated, especially when substantive contributions of time and money are involved. The importance of this is illustrated by the following quote from a LE participant in who had used drug and alcohol services:

Probably money…. [would help]…I do a free service on a Thursday night on Zoom for [treatment centre]. I do a speaker talk and it’s no charity. I enjoy doing it and get a great benefit from it but there’s time you’re going down there and you’re on your own diesel and you’re driving down as well and you’re just thinking, ‘Fuck’s sake’, you know, selfishly enough. I shouldn’t be thinking like that but I’m a human…(IN.PWLE.DA.02)

Second, in terms of measurement and accountability, many informants stated that successful implementation must be driven by evidence. Stakeholders spoke about the need to demonstrate how SI generated improvements in outcomes:

... so engagement feels like the right thing to [do], it makes us all feel warm for doing it. But really, in terms of delivery of services, what has it done? … we do have to measure it as we go…and we have to know that it’s making a difference. (IN.SH.06)

Some argued that key performance indicators should be established to foster a culture of accountability and assessment. Otherwise, SI would not be truly prioritized as a sector wide agenda. Thirty-two per cent of provider respondents reported that SI evaluations were not conducted.

Third, the importance of SI activities generating follow-through impact was highlighted by all. However, many stakeholders and provider participants stated that SI was often beset by a lack of continuation and completion, with services rarely empowered or resourced to make changes based on the perspectives of PWLE. Some stakeholders suggested that this may be more prevalent in the HSE due to its scale and the difficulty in applying changes consistently. LE participants believed SI frequently had no impact with some pointing to long-standing consultation, with no obvious change. This attitude is made clear in the following exchange between two Travellers in a focus group discussion:

P1: … we have done this kind of [thing] over and over and over…

P3: … but there's been research…that has shown the same barriers, the same causes, same triggers… What I’m getting at is, is there a need to swing this […] completely around on the services and say, ‘What are ye doing and what are ye not doing?’…like we’re the most over-researched group of people and I’m sick of it. (FG.PWLE.TR.DU)

Population capacities

All study informants recognised that the capacities of the five groups, and their life-course circumstances, needed specific consideration in SI. With respect to psychological readiness, some LE participants spoke about their on-going efforts to regain life stability, and about not feeling ready to participate – whether because they were unwell or were concentrating on their own recovery. The following quote is from a person who had experiencing sleeping rough in homelessness:

If you asked me, I wouldn’t get involved anyway, because…at the moment now, I wouldn’t be ready to talk to other people. I wouldn’t be able to give advice to help people or help me change anything. Because I’m just changing myself. I’m only getting a grip on my own life at the moment… (FG.PWLE.HO.RS)

LE participants also spoke about how accumulated exclusions and traumas can dissuade people from participating. Fifty-five per cent of provider respondents agreed that precarious conditions and exclusion amongst these groups could disrupt involvement, with those providing MMER (94%) and TR (88%) services more likely to agree. Furthermore, 73 per cent of providers, and many LE participants described how complex health issues can affect the abilities of some members of the populations to contribute, particularly in group settings. This is illustrated well by the following quote from an older man with a history of drug use and mental health challenges:

….there’s an awful lot of psychiatric problems… goes with the drugs. You know, you wake up full of… you know the anxiety!…. Well it’s a horrible thing for an addict, you know?... That your whole insides is like a bunch of butterflies, and you’re so nervous. And you’re not afraid of anybody in particular, but yet there’s an inside fear, you know, that needs to be sorted out before you can start to think [about talking to anyone] one-on-one. (FG.PWLE.DA.OO)

With respect to self-esteem, LE interviewees discussed how they or their peers may place little value on their own opinions and/or believe they are not ‘qualified’ to take part. Many of these participants described how they would feel especially nervous in engaging with those in positions of power, such as clinicians or other professionals. This low level of self-esteem was often attributed to their situations, for example being displaced, being in recovery, experiencing homelessness, or experiencing depression. To overcome these barriers, LE participants argued that it may also be appropriate to encourage family and/or keyworkers to support participants during SI processes. Moreover, it was suggested that boosting a participant’s confidence and lessening feelings of intimidation was essential, as can be seen in the following quote from a focus group participant who provides services to Travellers:

So, there is a need really to, I suppose, support that work more and to provide I suppose better… an understanding of why their voice is of equal if not more important at those meetings and for them to feel confident and have the confidence to participate fully in those settings I think is really, really important. (FG.SP.TR)

With respect to understandings of SI, LE participants highlighted that a lack of familiarity with SI and its key processes may mean that some individuals would be reluctant or under-prepared to take part in engagement. To counteract this, it was asserted that SI training, or at the very least a briefing on SI processes, should be provided to participants to allow them to understand their roles and responsibilities. However, some providers -like the one quoted below- also felt that these forms of supports can be too resource intensive, under current resource constraints:

... we would have people in recovery on our board, but we don’t have a specifically designated service user...because...there needs to be a lot of kinds of proprietary work done, there needs to be a lot of mentoring done, there needs to be a lot of development done with the service-user to...bring them to that space.... (FG.SP.DA)

Trust

Informants noted that trust was a key challenge upon which the success of SI for the five populations often hinged, with 53 per cent of provider respondents feeling that these groups were less likely to have confidence in SI processes.

This was in the first instance related to system structures. Service-provider and LE participants highlighted that there can be a lack of trust in the state, and in HSE services specifically – with those who use services sometimes slow to believe their participation is anything more than tokenism. Several national-level stakeholders and providers described how marginalized populations can perceive themselves to be powerless. This was particularly evident amongst MMER and TR groups, with some migrant participants being influenced by their experiences with their home-country institutions. This is explained well by the following quote from a Roma participant:

It’s very important to mention that when you have marginalised Roma community come in from Romania, Slovakia, and they come in from very bad conditions….It’s a totally new system for them. And it’s about building the trust… It’s the most important thing for them to understand that they can trust the HSE. (FG.PLWE.MM.RO)

Trust issues were also related to mistrust of individuals and in the interpersonal relationships within these systems. Several LE participants questioned the values and motivations of some professionals working in care services, such as the person quoted below who had experience of using drug and alcohol support services:

They’ll turn around and they’ll say they’ll listen to us. They hear us, but they’re not fucking listening. You know? They’re not doing anything… they’re in it for the money. Or they’re in it for a promotion to get more money, so they are. (FG.PWLE.DA.OO)

However, it was also clear that several LE participants felt the issues were broader, where they described having to continually open-up and ask for help because of upsetting situations, intensifying their sense of vulnerability. These experiences meant participants felt they were less likely to talk openly about their opinions or emotions in an SI situation. Such instances were exacerbated further by high staff turnover, as the following quote from a person experiencing homelessness illustrates:

She was my keyworker... But I really, really trusted her, you know? I had big trust in her. And then when she left, they were trying to assign all these different keyworkers… When you open up to someone then you have to open up to another person, then you have to open up to another person. Then when you move out of that place, you’ve to open up to five more people. (FG.PWLE.HO.SA)

In addition, trust issues were related to concerns about confidentiality and information misuse. Although some participants acknowledged that sensitive information was often provided in care settings, this was usually under privacy protections and/or helped to access needed services. In contrast, purported benefits of SI did not justify such disclosures. A greater concern was how being openly critical of a specific service might impede access to that service, with some members of the populations feeling their trust had been betrayed in the past, like the person experiencing homelessness who is quoted here:

They asked me to do an interview with [organisation name], right? I was like, what’s it for? Oh, a newsletter. Then I found out it’s for the accounts or something?… The lady printed out everything I said and I was high that day as well…. and I was like, fuck. The manager came and gave it to me and she was like: ‘Do you want to sign it?’ I was like, I thought this was anonymous... I said, 'What is this for? Why are people lying to me?’ (FG.PWLE.HO.SA)

Finally, trust issues were related to fear and feelings of a lack of safety. Service-provider participants and LE interviewees highlighted how frightening it can be for people who have experienced threats to personal safety to speak-up. For some, these fears were based on experiences of harm, with the following quote, from a woman who had experienced sexual exploitation, illustrating this dynamic:

I’m struggling mentally really. My self-confidence is gone, I don’t have it at all. I don’t trust men. Even if a man looks at me, I don’t think that’s a single girl he may like, I’m just thinking everything what’s happening to me is written all over me or that’s what he wants to do. (IN.PWLE.MM.06)

Representation, stigma and discrimination

Challenges related to representation, stigma and discrimination emerged strongly within the research. For the five populations, past service experiences were used to highlight concerns that such challenges may feature within SI itself, or as cautions of broader system features that detracted from SI capacity to affect positive change.

Representation was a key concern for some groups, national stakeholders, and providers. This first related to the fundamental need to represent the populations due to long-standing marginalisation, wherein those included tended to be characteristically ‘mainstream’, as one LE Traveller participant highlighted:

…you’ll get the white, settled person answering [surveys]. That’s who normally answers them…you’ll get the same kind of people having the time to do that and the ability to do that. There’s challenges…If that form was sent out to my mother-in-law or any members of the family that couldn’t read and write and they’re not really understanding what it is, they wouldn’t do it. That’s just reality. (FG.PWLE.TR.DO)

A second more dominant concern related to the diversity of population sub-groups being overlooked. While 51 per cent of provider respondents reported difficulties in recruiting a diverse participant sample, stakeholders noted how problematic it is to only include ‘the usual suspects.’ Whether this reflected a lack of resources or a lack of understanding, its impact can be a dilution of difference. For others, such under-representation reflected a more pointed neglect of particular groups, or complex intersectional experiences.

All populations spoke about the prevalence of stigmatising processes that arose from negative beliefs based on circumstances or prescribed identities. Ethnic-based stereotyping was most prominent for the Traveller and Roma communities. However, other forms of stigmatisation were an issue for all groups, with people describing being spoken down to, and being seeing as lesser. One male interviewee experiencing homelessness described how a stigmatised attitude about his drug-use history resulted in damaging assumptions about a current health condition:

I got shingles one time living up in the park, I didn’t sleep for five nights. Took three seizures on [street name]. It comes up on that screen – heroin addict, methadone addict. ‘Ex’ doesn’t matter…they say: ‘Is there any pain anywhere else?’ I go, ‘I’m in awful pain here’, and they turn around and go, ‘Ah yeah’… they’re thinking: ‘You’re looking for fuckin’ opiates.’ (IN.PWLE.HO.05)

The normative nature of stigmatised attitudes could also be internalised. This was described by some -including the LE participant quoted below, who had used mental health services- as having manifest as a reluctance to seek support or accept a diagnosis or treatment:

I gradually stopped…[taking my medication]… I started skipping days …but I got really bad then and… my attitude towards anti-depressants was an unhealthy one and [my psychiatrist] was saying to me, ‘If you had a heart condition…you’d be on medication and you’d take that medication without question…’ (IN.PWLE.MH.04)

Experiences of being stigmatised or stereotyped meant participants were less likely to want to engage in a meaningful way in SI processes.

All informants discussed how the five groups can be treated in a discriminatory way, with a view that discriminatory experiences were commonplace. For those with intersecting experiences, or statuses, for example people in homelessness with a history of drug use or addiction (as the participant quoted below was), discrimination was more keenly felt:

[The food was] very, very poor and that’s in [temporary accommodation service]. I got food poisoning six times out of that food. I ended up in hospital six times…No, what they do, instead of putting the fresh dinners out, they’re putting the old dinners to the front and the fresh dinners to the back… because the way they look at it is you’re an ex-addict or you’re this or you’re that. (IN.PWLE.HO.04)

Discrimination experienced by different races or ethnicities was most often encountered as racism. For Roma individuals, it was felt that providers treated them poorly because of their ethnicity, particularly where English language proficiency was low. However, it was Traveller participants and their representatives who were most vocal about poor-quality engagement due to racism, for example all TR service organisations reported ‘fear of discrimination’ as the primary SI barrier in the survey. For individuals who experienced discrimination, the impacts were substantial, as the following quote illustrates:

…the discrimination that we face daily, if it’s going into a shopping centre, into a restaurant, into a pub. If it’s walking past a settled person on the street… The minute a Traveller opens his eyes in the morning and takes a foot into the outside world till the last foot he takes at night back inside his own door, he faces discrimination…you’re being looked down at. You’re being contradicted. You’re nearly put as a liar… There’s so many consequences that go on and go on and go on that we daily face, and it’s a constant daily struggle… (IN.PWLE.TR.05)

Discussion

This article aimed to investigate multi-level shared challenges impacting SI for underserved populations in a complex health and social care system. Drawing together national stakeholder, service provider and LE perspectives in a mixed-method study strengthened the identification and verification of these challenges. However, three limitations are acknowledged. First, it was not possible to capture the full diversity of sub-groups within populations given the study’s scale. Second, although recruitment strategies were targeted, the views of those least engaged may not have been captured sufficiently. Third, the low provider response rate for some population-related services restricts the generalisability of survey findings. Despite these limitations, the study provides significant insights into SI for the groups. The findings are discussed in relation to the international literature, and the Irish health and social care system context. Based on the findings and the outline priorities coproduced in the Consultative Forums, development areas for operationalising a more inclusive and effective strategy for SI are then described.

In order to address challenges related to buy-in and commitment from leadership, research shows that SI will be ineffective in driving sector-wide service quality when it operates in isolation, without strategic direction or a multi-level structure that prioritises, governs, and regulates activities [64]. Where the culture of a system or its organisations fails to embrace SI, then a genuine engagement ethos is unlikely to evolve or to persist beyond time-bound projects. For implementation, existing literature highlights, as did this study, the interdependency between resourcing, monitoring, and evaluation as SI success factors [15]. Studies have shown the difficulties in leveraging resources within, what are in many countries, the marginalised service sub-sectors that cater for excluded populations [16] – with subsequent difficulties in embedding partnership approaches that avoid tokenism [57]. The remaining challenges were similarly intertwined. Like others have shown, the populations’ capacities for SI require specific attention [31, 77]. While training may provide greater SI ‘know-how’, factors related to psychological readiness and self-esteem represent more profound barriers [5]. Pervasive concerns regarding trust amongst informant-groups reflect previous studies where this has been found to be an important factor in SI [80]. However, it is the multifaceted set of failures that stand-out in this analysis. This reinforces the populations’ perceptions that they are neither valued, nor that their views will be acted upon. For many, poor representation, stigmatisation, and discrimination were the principal drivers of distrust, as well as contributing traumas that undermined SI [64]. When these experiences were reproduced by engagement (e.g. tokenistic or degrading consultations) SI emerged as another mechanism that stripped power and chipped-away at group self-worth.

Given this article’s focus on shared challenges across groups, it is necessary to delineate collective barriers to SI that go beyond those linked to general populations. First, challenges appear to be of a greater magnitude than those facing the general population. This is because of entrenched shortfalls in engagement over time which have left structural deficiencies around processes, resources and knowledge for SI with populations which experience marginalisation. It is also due to the homogenisation of inter-group diversity, that is, the lack of recognition of and flexibility for the varied identities and experiences (and intersections therein) among disadvantaged groups. And it is due to biased perceptions of reduced capacity or agency amongst populations which have been disenfranchised [12]. Second, the populations’ circumstances and experiences introduce additional SI considerations. Psychological readiness, trust, and representation, stigma, and discrimination, as well as the exclusion endured by group members, simply do not feature to the same degree for mainstream populations due to the systemic challenges faced by these groups [23]. Third, reflecting many informants’ concerns, workforce training and development remains under-funded and under-supported, resulting in deficiencies in knowledge and understanding around meaningful service involvement for these populations. Again, it is within this wider context of sector marginalisation that the challenges need to be understood.

While this article is about shared challenges across all groups, it is also important to note that there are group-specific differences in how challenges are constituted. For those who use DA services, SI was more challenged by problematic stigmatised associations, the lack of inclusion of sub-group voices or needs, and by substance-related physical and cognitive impacts [80]. For those experiencing homelessness, disconnection from care systems and issues around transience, complex needs, and, again, stigma intensified practical barriers and necessitated greater sensitivity and flexibility [28, 56]. For those using MH services, there were benefits from a longer track record, and even innovation leadership in SI, with greater awareness and engagement evident. Nonetheless, there remained a damaging legacy from a neglect of voice, and an absence of person-centred approaches, in earlier care experiences. For MMER communities, engagement activities were undermined by the homogenisation of sub-groups and differential migration and settlement experiences [51], along with their patterning by age, gender and health status [37]. For the TR community, issues related to building sufficient involvement capacity in such a small population, discrimination impacts on trust, and the palpable fatigue from consultation that produces little change. These group specific circumstances dictate that regardless of shared challenges, there should always be flexibility for group-specific tailoring which supports broader ‘all system’ approaches.

It is also worth drawing out the influence of Ireland’s complex health and social care system within the findings. The hesitance of some clinical and managerial professions to accept LE perspectives within decision-making processes, and to deconstruct hierarchal profession-led governance structures, is indicative of change barriers in complex care system reform [8]. Without addressing power imbalances or supporting an open dialogue on the impact of SI for all system actors, LE voices will remain a minor part of practice [10, 65]. The findings also testify to an implementation deficit within Ireland’s political culture [20], with ensuing challenges for creating complex system change. Of note were the damaging effects of having little or no resourcing to execute large-scale agendas, or to generate ‘follow-through’ in local initiatives. These dynamics are complicated further by the reliance on small-scale voluntary organisations to not only deliver services [32], but to lead on engagement. As many providers are operating under significant constraints [66], questions arise about how realistic engagement ambitions are for the groups. Given that the success of SI in complex systems is largely determined by context characteristics [8, 17], macro-level environments, CHO organisations and other meso-level settings in Ireland must be considered together as enabling locally led implementation. These various effects of Ireland’s care system complexity compromise efforts to address psychological capacities, and representation, stigmatisation, and discrimination in SI. However, it is their exacerbation of mistrust that are perhaps most pronounced – doing little to persuade groups that they are valued, and that their views will be acted upon.

Towards an inclusive, effective strategy for SI

Key development areas can be identified to enable an inclusive, effective SI approach more readily for the five populations in Ireland. These areas are rooted in the findings and are based on the outline priority areas and messages agreed by the LE and provider participants within the study’s Consultative Forums. While specific to the Irish context, the development areas are likely to hold relevance to other jurisdictions with complex systems, as indicated by challenges documented internationally.

Top-down drive, bottom-up direction

A balanced, coordinated development of strategy in relation to SI is required for the groups. This must be driven at national policy and HSE-CHO implementation levels but informed by on-the-ground local needs. While progress has been made, it is disconnected and not at a scale to generate transformational change. Coordinated strategic efforts, in terms of commitment and a resourced national development programme, are required to address leadership, governance, prioritisation, and implementation challenges. Put simply, more money, more training, more capacity building, more regulation, and more measurement are needed. Again, local actions in complex systems requires macro-level enablement [8, 17]. A national development programme must, however, integrate on-the-ground organisational learning, and LE perspectives, within its formulation. This is critical, practically to enhance effectiveness and symbolically to assert inclusive ambitions.

Sustained multi-level, multi-form implementation

SI for the five populations must be expanded, encompassing different system levels and different kinds of engagement. Given the current concentration of SI on individual-level service delivery, a greater embedding of engagement is needed to inform organisational service programme design and organisational direction. Critically, to impact key resourcing decisions, a greater investment in SI structures within national and regional strategic-level decision-making processes is also needed. The findings indicate the need for multiple forms of SI across these levels. While a shift is required from the dominant less-involved techniques (e.g. short surveys) to co-production and other co-form approaches, all methods can serve a purpose depending on circumstances, and preferences. A pre-engagement consultation with a diverse sample of targeted populations is likely to help guide initial choices.

Measuring effectiveness and action

A key success factor underpinning effective SI for the five groups is the degree to which their SI contributions are truly harnessed. The need for action was asserted by all informants, including calls for implementable tasks; measurable impact; the desire for change; or in the case of individuals and groups, the feeling of ‘being listened to’. For people with LE, it was the commitment to ‘follow-through’, to do something, that was often critical. Measurement and accountability, in terms of governance, responsibility and key performance indicators, must be introduced to drive this implementation. There is also a case for stronger integration between SI stakeholders and structures across levels, enabling continuity between practices and recommended follow-through actions.

Embedding inclusive engagement

SI inclusivity must be improved for the populations. Flexible, accessible strategies need to be developed to address sub-group needs, and structural barriers – including tailored supports to circumvent logistical, financial, and psychosocial restrictions. Inclusive communication (multi-strand, and multilingual modes) may help overcome language and literacy difficulties, while cultural inclusion approaches may counter discrimination and support cultural identity recognition. These features reflect components of inclusive single-population engagement models [73]. Fostering understandings of intersectional positions is also necessary, including the multidimensionality of challenges that individuals face, and their utilisation of services that crosscut groups because of intersectional needs. Narrow prescribed categorisations restrict our view of these populations to single-service user identities. Engagement agendas, therefore, must be sensitive to overlapping group composition [60] and shared experiences (housing,deprivation,recovery pathways).

Trust as a prerequisite and outcome

SI for the five populations is intrinsically linked to trust in engagement, and health and social care structures. Trust is critical in fair, reciprocal interactions in complex systems [2] and in recognising and addressing disparities [81]. Building trust must be a constituent feature of the early stages of involvement. However, links between trust and SI also reflect a mutually reinforcing relationship, where SI may contribute to rebuilding trust in systems. Sustained engagement, implementation, accountability, and action will all help establish trusting relationships. But it is transparency that is likely to be key: transparency around the SI purpose, around what it involves, and around the constraints that may impede follow-through actions. Therefore, while desired outcomes may not always be achievable, a genuine honest process can be.

Equalising, agency empowering agenda

SI must drive an equalising approach for the five groups, in terms of care access, opportunities to express voice, and well-being outcomes. It should aim to address power imbalances that can develop within a traditional, professionally driven care system, and the deprivations that stem from how systems treat these groups. Being engaged in service design and delivery should empower individual and group agency, control, and efficacy with respect to the care received. There is a value in bestowing upon SI the ambition to create clearer pathways to social citizenship [52] for such groups – engraining civil, political, and social rights within health and social care expectations. Capacity building investments will assist in empowering agency, but so too would a life-course trauma-informed approach that draws out more considered understandings of how people’s circumstances evolved over time. Central to this is a resource allocation model that prioritises outcomes and that distributes an appropriate share of resources to SI for the five populations.

Conclusions

A key contribution of this article has been to isolate shared challenges with respect to SI across the five populations, and to identify common elements of a more inclusive, effective pathway to engagement in health and social care. Through coordinated efforts and investments, these challenges need to be recognised and addressed, and these elements need to be embraced. Otherwise, there is a danger that SI will lose traction, will be perceived as meaningless, and will become more of a destructive than constructive force for individual and group agency within complex systems. Across all study informants there was a collective will to elevate the position of individuals and groups in diverse and marginalised populations as contributors to, and leaders of, their care services. It is critical that this collective will is appropriately supported, otherwise those most in need will remain outside of the system. In doing so, there are likely to be ethical and social inclusion gains, but also gains in terms of system efficiency and system financing, where services are more likely to be effective and cycles of treatment and support more impactful.