The review question, methodology, and final manuscript check were co-produced through the “Community Solutions for Health Equity (CSfHE) project”, based in a socioeconomically disadvantaged area in England. Non-academic co-researchers led the identification of the research question and the design of the review. The data analysis involved academic co-researchers; however, non-academic co-researchers participated in a consultative capacity and were invited to review the identified themes, as well as their structure and interpretation to generate the final themes. Non-academic co-researchers shaped the interpretation of themes by reviewing the written interpretation of all themes and subthemes, providing comments and suggestions on their accuracy, relevance, and resonance with community and practice-based understandings of service access. These comments were incorporated into the revised theme interpretations, after which co-researchers reviewed the amended version again to confirm that the final interpretation was accurate and relevant. There were no disagreements in the interpretation of themes. Comments were made on a shared document, which enabled contributors to review the material asynchronously and suggest changes directly to the manuscript. Although academic co-researchers retained responsibility for coordinating the manuscript, the use of a shared document supported transparent decision-making by making comments, responses, and revisions visible to all contributors. The collaboration followed INVOLVE’s principles [69] and included co-researchers from the VCFSE sectors and academia, and public representatives, over three sessions. Nineteen initial review questions were refined using the PICO framework [70], prioritised, and discussed with the co-production group, resulting in nine refined questions. The highest-priority question, chosen based on the group’s regional experience, informed this review: “What are the perceived barriers and facilitators of accessing statutory and non-statutory services, in disadvantaged communities?”

The review protocol was pre-registered via PROSPERO (Reference No. CRD42024546994). Also, the review used the thematic synthesis approach, to integrate and analyse findings from included studies [71]. Reporting followed ENTREQ guidelines for qualitative synthesis [72] and PRISMA guidelines [73]. This review was informed by a health equity perspective, with a focus on how overlapping forms of disadvantage may shape experiences of service access. The review sought to interpret service users’ accounts in relation to the structural, relational, cultural, and emotional conditions affecting engagement.

The search strategy was iteratively developed. KS created initial terms, with input from a university librarian. Specific and truncated phrases were derived from common literature expressions. The research team, then, refined and finalised the search strategy.

Four main concepts guided the search:

service use OR access OR barrier OR facilitator

Relevant synonyms were used across all databases (see Supplementary Material 1 for details). MeSH terms were excluded due to relevance concerns and indexing delays. The databases searched were, PubMed (including Medline), Scopus, APA PsycInfo, and Web of Science Core Collection. Limits included publication date, English language, and publication type. Additional records were identified via handsearching strategies, including reference checking and supplementary searches in Google Scholar, Semantic Scholar, and Connected papers, with all retrieved records screened against the same eligibility criteria. The search concluded on 18 June 2024.

Articles before 2003 were excluded due to the NHS Reform and Healthcare Professions Act 2002, which likely affected service accessibility in disadvantaged areas [74] (Table 1). Only qualitative studies exploring service user experiences in England were included, given differences within health systems in the United Kingdom (UK) [75, 76]. Non-English publications, grey literature, opinion pieces, and secondary interpretations were excluded. Primary care studies were included, even when linked to secondary or tertiary care (e.g., maternity services) [77, 78]. Included studies had to focus on disadvantaged populations (e.g., due to socioeconomic status, race/ethnicity, sexual or gender identity, disability) [27]. The review only included direct, primary-source accounts of service users, to maintain focus on lived experiences of target groups [79].

Two authors conducted a single-blind screening. KS screened all articles, RG screened 20%. Inter-rater reliability (Cohen’s Kappa) was κ = 0.813 (SE = 0.044), indicating almost perfect agreement [80]. KS screened all full texts, RG screened 20%. Inter-rater reliability was κ = 0.726 (SE = 0.064), indicating substantial agreement. All disagreements were resolved through discussions between the two reviewers, and no third reviewer was required given the level of agreement observed.

Prior to extraction, procedures were piloted. KS and PJ extracted qualitative primary source data from service users in relation to barriers and facilitators of service access into an Excel database. RG checked 20% of entries for accuracy and completeness. Data from each included article were reviewed and extracted under the relevant headings in the extraction form.

The Critical Appraisal Skills Programme (CASP) Checklist for Qualitative Research [81] was used to appraise the methodological quality and rigor of the studies. KS conducted the appraisal, with 20% checked by RG. A Kappa value of 0.750 (SE = 0.043) indicated substantial agreement. All disagreements were resolved through discussions. No articles were excluded based on quality. The quality assessments are presented in Supplementary Table 2.

This review followed a thematic synthesis approach [71]. NVivo 12 was used to manage and code extracted qualitative data. The process involved three key steps: [1] line-by-line coding of service user data to capture initial concepts, [2] development of descriptive themes by grouping similar codes, and [3] generation of analytical themes that offered broader interpretations grounded in the data. KS led the analysis, with BM and BK reviewing the coding and interpretations. While data extraction was deductive (focusing on barriers and facilitators), theme development was inductive. Interpretation was informed by a health equity perspective, with attention to intersectionality and to how overlapping forms of disadvantage and exclusion shaped experiences of service access across studies. Reflexive practices were maintained throughout to mitigate bias and ensure analytical rigour.

Studies were grouped into pre-COVID (n = 34) and during and post-COVID (n = 8) based on the year of data collection. This division was used because the pandemic substantially altered the organisation and delivery of health and social care services, including the disruption to face-to-face provision and increased socioeconomic pressures affecting service engagement. There was no distinction made between during and post-COVID, because these studies did not specify whether their data collection took place during or post-COVID. Supplementary Table 3 was used to descriptively compare the representation of themes and subthemes across these periods, which was intended to identify broad patterns in subtheme representation across studies.

After removing duplicates, 1,796 unique records were screened. Based on title and abstract exclusions, 321 full texts were assessed. A total of 42 studies met the inclusion criteria and were included in the review (Figure 1).

Supplementary Table 4 presents the study characteristics, including sample characteristics, study setting, service focus, and data analysis approaches. The 42 included studies were published between 2005 and 2024 and were conducted across diverse locations in England, including the North of England (n = 8), Yorkshire (n = 7), the Northeast (n = 3), Northwest (n = 9), Midlands (n = 7), London (n = 7), Southeast (n = 1), and Southwest (n = 5), although some studies focused on specific cities and localities and a small number did not report a clearly defined location (n = 6). Sample sizes ranged from 8 to 90 participants, and the included populations represented overlapping forms of disadvantage, including ethnic minority and migrant communities (n = 18), people experiencing homelessness (n = 2), disabled people (n = 6), young people (n = 2), older adults (n = 3), people with long-term conditions (n = 16), people using substances (n = 9), asylum seekers and refugees (n = 2), and women with complex social and health-related vulnerabilities (n = 11). Participant characteristics frequently overlapped across studies, which reflected the intersecting social, economic, and health-related vulnerabilities represented in the review. The studies examined a broad range of statutory, voluntary-sector, and community-based services, including maternity care (n = 6), primary care (n = 2), mental health (n = 9), sexual health (n = 2), smoking cessation (n = 3), social prescribing (n = 2), welfare advice (n = 1), advocacy (n = 1), screening (n = 3), substance use (n = 2), migrant support (n = 7), homelessness support (n = 2), pharmacy services (n = 1), youth services (n = 2), and health promotion (n = 3). All studies used qualitative analytic approaches (n = 42), most commonly thematic analysis (n = 31) and framework-based analysis (n = 15). Other reported approaches included grounded theory-informed analysis (n = 2), constant comparative methods (n = 5), focused ethnography (n = 1), abductive thematic categorisation (n = 1), participatory evaluation (n = 1), and theory-informed approaches such as the Theoretical Domains Framework (n = 2), COM-B model (n = 1), Silences Framework (n = 1), and relational autonomy (n = 1).

The CASP quality appraisal showed that most studies demonstrated strong methodological quality, with 35 studies scoring 8 or above. Nearly all studies clearly stated aims (n = 41), used suitable methodologies (n = 42), and applied appropriate research designs (n = 41), recruitment strategies (n = 40), and data collection methods (n = 41). Most addressed ethical considerations (n = 39) and clearly conveyed their value (n = 41). The review team identified common limitations that included limited reflexivity around researcher-participant relationships (n = 33), unclear findings (n = 13), and inconsistent data analysis rigor (n = 10) (Supplementary Table 2).

Four analytical themes were identified in the synthesis, each comprising of related barriers and facilitators represented through subthemes. The themes are presented in an order that reflects service engagement as a process that moves from factors shaping initial access, to the wider social and cultural contexts influencing engagement, to experiences within services, and finally to the emotional meanings and consequences associated with those encounters. Although analytically distinct, the themes were interconnected which reflected the complexity of service access and engagement.

A descriptive comparison of subtheme representation suggested broad continuity rather than a marked shift between pre-COVID and during/post-COVID studies. All four themes were represented across both periods, which indicated that the core barriers and facilitators shaping access persisted over time. However, the clearest pattern was within the subtheme ‘The role of socioeconomic positioning in service access’, which appeared proportionally more often in during/post-COVID studies than in pre-COVID studies, with greater attention to communication, service responsiveness, and the practical demands of digitally mediated access. Community and locational enablers, cultural and linguistic influences, and discrimination, mistrust, and exclusion also remained prominent in the during/post-COVID studies. By contrast, some subthemes, particularly Engagement Motivated by Holistic Care and Lived Experience, were less visible in the during/post-COVID subset. Given the smaller number of during/post-COVID studies, these differences ought to be interpreted cautiously as these were represented by shifts in emphasis.

This co-produced qualitative review aimed to synthesise qualitative evidence on the perceived barriers and facilitators of accessing statutory and non-statutory health and social care services among disadvantaged communities in England. The thematic analysis found that disadvantaged communities experienced service access as difficult to initiate and sustain because services were fragmented, hard to understand, culturally and linguistically inaccessible, and poorly aligned with the material constraints of everyday life. Engagement was facilitated when support was local, trusted, relationally consistent, culturally responsive, and able to address practical and emotional needs alongside formal care. The overall findings are supported by international evidence and WHO guidance on community-centred and integrated care that shows access improves when services respond to the social and economic conditions shaping inequality and when local provision is organised in ways that reduce fragmentation [9–11]. The review also explored health equity [123] and intersectionality [124] perspectives, which placed engagement within wider structures of poverty, discrimination, exclusion, and unequal resource distribution [16–18, 26]. Consequently, access depended on whether services remained practically usable across successive encounters within the structural, relational, and emotional pressures of everyday life. The sections below, follow this argument by examining four cross-cutting mechanisms (Supplementary Table 5), navigation, trust, emotional and psychological safety, and service alignment, that connected the analytical themes across the findings. The cross-cutting mechanisms move beyond a descriptive account of barriers and facilitators and are used to demonstrate the relationships between the themes, and to show how disadvantage accumulated across the access pathway.

Accessing and navigating services depended on confidence, transport, language proficiency, and familiarity with institutional procedures. However, these resources were often unavailable among disadvantaged groups, especially where structural and intersecting forms of disadvantage already shaped everyday life. A Norwegian study involving Syrian refugees similarly confirmed that service users may formally have entitlement to care, but still face difficulties because navigating the system requires understanding how it works and dealing with unfamiliar institutions [125]. These demands in the review, shaped whether services were practically reachable. Navigation therefore emerged as an unevenly distributed burden, which highlighted how access depended on both service permeability [22, 24] and the practical and communicative capacities users needed to seek and reach care [19, 126]. Routine administrative requirements therefore became exclusionary, requiring repeated effort from users with the least capacity to absorb those costs. This interpretation is consistent with research from the USA [127] and Argentina [128] on treatment burden in healthcare, which shows that learning and compliance are often shifted onto users. Contrasting evidence highlighted that remote service can work better in rural settings with low-income adults, which may reflect access barriers rather than rejection of remote care itself [129].

Communication styles also added to this burden, in the current review. Unclear information, poorly explained pathways, and professionalised and medicalised language left users to interpret systems that were already difficult to approach. The negative effects of these challenges were previously confirmed whereby communication problems were found to be part of the burden itself that are not separate from administration [127]. Such issues arose from institutional fluency that could not be assumed among disadvantaged groups. This aligns with health equity research showing that engagement depends on unequal access to the material and social resources needed to use care [18, 26]. Community-linked and peer-based approaches reduced these difficulties by making services easier to find and enter through familiar settings and social relationships, consistent with evidence from the USA [47] and Zimbabwe [48] showing that community-based and outreach-oriented models can improve uptake and timeliness of care in underserved populations. Although previous literature indicated that service user navigation improved screening and reduced time to treatment initiation [130], our findings extend this evidence by indicating that navigation itself operated as a stratified demand, with the greatest burden falling on those least resourced to manage complexity and repeated effort.

Trust shaped whether services were entered and re-entered, with engagement tied to whether services were experienced as dependable. This suggests that trust was a key condition of access and played a more foundational role than in service models that place more emphasis on post-entry care [131]. Trust appeared to be already active before contact, especially where earlier encounters involved dismissal, stereotyping, cultural invalidation, and poor follow-up, thus suggesting that willingness to re-enter services was being shaped before access even occurred [22, 132]. This supports existing literature by showing that prior experiences of care can shape future candidacy before a new service encounter begins [22]. On the other hand, mistrust developed cumulatively through repeated encounters with fragmented pathways and experiences of dismissal and stereotyping. These patterns were most visible among racially minoritised participants. Such findings are line with previous evidence that showed marginalised patients’ experiences with restrictive healthcare systems and clinicians whose biased behaviour and lack of interest undermined meaningful care conversations [133]. Also, long waits and crisis-led thresholds left participants uncertain that help would follow through, further aligning with evidence that integrated care can improve user experience [50] and continuity [52]. However, trust was strengthened through everyday encounters in which service users experienced empathy and informal contact from providers. These experiences shaped whether support felt dependable and worth returning to, in line with evidence from Canada showing that culturally competent, co-designed services can build trust and improve uptake [42]. Previous evidence also highlighted that peer support and volunteer schemes can facilitate access by offering advocacy [53]. In the current review community and voluntary services were often experienced more positively, because their less hierarchical and more locally embedded forms of care helped service users feel that their needs and circumstances were understood, while also supporting repeated contact.

Emotional and psychological safety shaped whether support felt possible to disclose during service engagement. Fear, shame, anticipated judgement, and emotional overwhelm appeared to restrict engagement across a range of settings, particularly where services were experienced as insensitive to vulnerability. In these circumstances, emotional safety set the terms on which support could be approached and sustained, as research in similarly marginalised contexts has likewise shown [134]. Such user responses were closely tied to prior interactions and service environments rather than to individual reluctance, with formal health settings often experienced as more dismissive and judgemental than informal and peer-based environments. However, evidence from marginalised populations shows that formal healthcare settings can also become safer and more supportive when providers use destigmatising practices [135] and receive training that increases empathy and reduces prejudice toward underserved patients [136]. The findings also demonstrated service users experiencing shame, which was linked to homelessness, substance use, sexual activity, and other stigmatised experiences, while concerns about gossip and reputational harm were particularly salient in youth and school-based settings. Shame and fear therefore appeared as socially patterned responses, shaped by what services communicated about who could speak safely and on what terms. Previous evidence supports these observations by indicating that shame acts as a barrier to speaking out and disclosing problems to professionals because of expectations of judgment [137].

In terms of how vulnerability was conceptualised, emotional strain also intensified during periods of heightened care and dependency, such as pregnancy and caregiving, which supports wider research showing that vulnerability is shaped by individual need and by services’ response to acute dependence [134]. In the current review, where that sensitivity was absent, service contact became harder to initiate and maintain even when provision was available. By contrast, opportunities to speak openly without fear of dismissal, within care was experienced as ethical and non-judgemental, and helped create the psychological safety needed for continued engagement. This suggests that access was strengthened through care that reduced emotional threat and made disclosure feel safer, which is in keeping with international evidence that culturally responsive and person-centred forms of care can improve access and resilience and that relational models can support health through emotionally supportive engagement [13, 53].

Continued engagement was more likely where services aligned with cultural expectations and socioeconomic realities. This suggests that access depended on service availability and on the fit between service organisation and users’ ability to seek, reach, and engage with care [19, 126]. Across the findings, support became easier to use when it was delivered through familiar settings, communicated in culturally and linguistically intelligible ways, and organised around the practical constraints of childcare, work, and financial strain. However, where that fit was absent, availability did not translate into use, because services assumed economic flexibility and stability [19, 138], in line with evidence from the USA that demonstrates ordinary service arrangements can become unusable when they are designed around economic assumptions that do not mirror the lives of many users [139].

On the other hand, misalignment led to disengagement even where services were formally available. The need to prioritise basic needs made support harder to reach, subsequently reflecting wider evidence that service users’ financial situations and fragmented service geographies can constrain access for disadvantaged communities [27, 37, 39]. Digital-first systems and service designs that assumed stable internet and phone access also made engagement harder to initiate and sustain, supported by evidence that digital tools can reproduce inequalities when differences in physical access and digital access are not addressed [57, 59]. In this sense, some services appeared misaligned practically and socially, and where services failed to recognise culturally shaped understandings of help-seeking, support became harder to approach and easier to withdraw from. UK-based evidence similarly demonstrated that cultural mismatches can reduce engagement through bias and limited cultural competence [30, 31], while evidence from South Korea [41] and Canada [42] showed that culturally competent, co-designed services can improve uptake among disadvantaged communities.

The comparison suggests that the pandemic did not fundamentally alter the overall structure of access inequities, but it appeared to have exposed and, in some instances, might have intensified pre-existing disadvantages. The during/post-COVID studies placed greater emphasis on socioeconomic positioning, particularly the practical constraints associated with service responsiveness, remote contact, and digital access. This suggests that digital exclusion may have amplified inequalities for disadvantaged groups, especially where access depended on stable internet, phone credit, digital literacy, and confidence in dealing with altered systems. Also, the continued prominence of community and locational enablers indicates that familiar, trusted, and locally embedded forms of support remained important routes into care. Thus, the post-pandemic learning point is to ensure that services are flexible, multimodal, culturally and linguistically responsive, and organised in ways that reduce the unequal labour of accessing support.

The findings extend Levesque et al.’s framework [19] by highlighting how access inequities accumulated across the care pathway in disadvantaged contexts. Early difficulties in perceiving, seeking, and reaching care often constrained later possibilities for obtaining and engaging with support [19]. Subsequently, service access operated as a cumulative process in which limited institutional familiarity and material insecurity progressively narrowed opportunities for action. The implication is a move away from treating barriers as discrete interruptions in access and towards understanding how systems can organise disadvantage over time. In relation to Candidacy Theory [22], access was shaped by the practical and social demands involved in becoming intelligible to services. Service permeability and judgements of need were points at which disadvantage could improve or deteriorate. The findings add specificity by showing that access depended on whether service users could make claims in forms that institutions recognised as credible, possessed the social and communicative resources expected by services, and felt able to re-enter after earlier encounters. Trust, shame, and emotional safety therefore formed part of the production of eligibility itself. The two frameworks are connected by the question of fit, which operated as a selection mechanism embedded in ordinary service design. Formal provision was converted into unequal access when routine service expectations presupposed temporal flexibility, institutional confidence, communicative fluency, and material stability more typical of advantaged social positions. This helps explain how service-user fit and institutional permeability can turn routine arrangements into mechanisms of exclusion.

The Intersectionality [124] perspective demonstrated why the above restrictions are unequally distributed. In the current review, access is characterised by institutional production of exclusion, where disadvantage is generated through service rules, professional judgement, communication norms, and eligibility practices that organise access unevenly. This means that such concepts are better understood as social relations that structure how services are encountered, whose claims are more readily recognised, and which forms of disadvantage are more easily converted into delay or exclusion. The findings also supported Health equity [123] stance on the institutional organisation of unequal access by indicating that inequity was routed in the routine design of services, which shaped whose needs were more easily articulated, were more likely to endure, and where disengagement became more likely. The implication is that access inequality should be understood as a property of service systems as well as a pattern in user outcomes, with disadvantage reproduced where ordinary service expectations are aligned with more advantaged social positions.

The findings suggest that improving access requires commissioning and service models that are explicitly organised around equity, rather than assuming that availability alone will produce uptake. In practice, this means using flexible, outcomes-based commissioning arrangements that support hybrid delivery across outreach, in-person, and digital routes, and that allow providers to adapt how support is offered in response to local barriers to access. Within ICSs, co-production should be embedded as a routine feature of governance and commissioning, with funded and ongoing involvement of service users and VCFSE partners in decision-making, and with clear accountability for demonstrating how lived experience has shaped service design, delivery, and review. Access is also likely to depend on culturally and linguistically responsive provision, protected non-digital routes into care, and relational continuity for people with complex or intersecting needs. In this context, relational continuity should be treated as a measurable quality outcome, assessed through consistency of contact, follow-up, reduced need to repeat personal histories, and service users’ experiences of trust and emotional safety.

Co-produced with contributors from the VCFSE sector, academia, and the public, the review was grounded in the lived realities of marginalised communities. Its focus on qualitative self-reported accounts of service access enabled an in-depth and multi-dimensional analysis of how structural, relational, cultural, and emotional conditions shaped engagement, which strengthened the review’s health equity perspective. Some limitations should also be noted. The review included only English-language publications and studies conducted in England, which reflected its focus on service access within the English health and social care context; however, the volume of relevant literature from England made this a worthwhile and policy-relevant undertaking. On the other hand, this may have excluded potentially relevant insights from similar disadvantaged settings elsewhere. Grey literature was excluded that may have contained additional qualitative evidence relevant to community-based service access.

Future research should build on the gaps in representation within the current synthesis. Greater attention is needed to voices where forms of disadvantage intersect in ways that remain underexplored, including combinations of ethnicity, gender, disability, migration status, sexuality, and poverty. More evidence is also needed from underrepresented service contexts and community-based models, to clarify how peer-led and community-led provision operates in practice. Given the exclusion of grey literature and the restriction to English-language studies conducted in England, future research would benefit from drawing on a wider range of community-based evidence and service settings. Longitudinal and co-produced research would be valuable in showing how access is shaped across repeated encounters rather than at single points of contact.

In conclusion, this review highlights that service access depends not only on whether services exist, but on how they are experienced and aligned with people’s lived realities. For disadvantaged service users, unclear pathways, cultural mismatch, financial strain, and low institutional trust limit their ability to engage with care. Improving access therefore requires service design that responds to the social, material, and emotional realities of people’s lives. This includes working with service users to shape care that is relevant and responsive. Co-production can support this, but only if it addresses structural barriers and supports equal participation. For planners and commissioners, the challenge is to move beyond provision alone and toward services that are approachable, consistent, and meaningfully shaped with the communities they aim to support.