Background

Profound inequalities exist in the access to, experience and outcomes of mental health support for many marginalised or minority communities in the UK [1, 2]. While a number of these characteristics are legally protected by the Equality Act 2010 (race, gender and sexual orientation) [3], and despite National Health Service (NHS) commissioners being statutorily bound by the Health and Social Care Act 2012 [4] to reduce health inequalities, inequities persist, particularly within mental health care where individuals may face an additional level of stigmatisation or discrimination [5]. Accumulating evidence has also established unsatisfactory experiences and outcomes of mental health care in individuals affected by social determinants of poorer health [6, 7] including, but not limited to, socioeconomic deprivation, homelessness and transitional housing and asylum seeker or refugee status [2, 8, 9]. Furthermore, individuals with more than one of these features or protected characteristics are likely to be at further disadvantage, in line with the theory of intersectionality [10].

To reduce the disadvantage associated with these inequalities, meaningful and effective strategies need to be developed. Furthermore, tackling inequalities in mental health care can significantly benefit the wider economy. Consistent evidence suggests that improving access to, and experience of, mental health care can reduce the economic burden of illness and bring long-term cost savings associated with enhanced employability, fewer lost work hours and reduced utilisation of costly health services [11]. To truly bring about effective change and improve outcomes, innovative practice is required at all layers of care to promote cultural, structural and attitudinal shifts.

The NHS has committed to prioritising the reduction of health inequalities in both the Five Year Forward View for Mental Health [12] and, more recently, the NHS Long Term Plan [13]. Contemporary models of mental health care in high-income countries are primarily based on concepts of mental health and well-being derived from Western culture, and as such, do not necessarily consider cultural and social diversity [14]. Therefore, one way to advance mental health equality may be to commission, or improve uptake of, evidence-based interventions specifically targeted at communities that face inequalities. There is a growing literature base of ‘disparity interventions’ which aim to adapt existing interventions in such communities or develop new interventions tailored to the community of interest. To bring about change, it is important to understand what interventions have been attempted, their effects and overall costs, as well as barriers and facilitators to uptake and success.

This review is part of a larger piece of work—Advancing Mental Health Equality (AMHE) [15], which was commissioned by NHS England as part of the Mental Health Care Pathways programme. The review aims to map the existing literature on disparity interventions in order to inform the delivery of more effective and culturally appropriate care and, ultimately, advance mental health equality.

Methods

Study design: systematic map

Systematic mapping reviews aim to draw together existing studies in a specific topic area and develop an understanding of the available data as well as any potential gaps [16,17,18,19,20]. There is no authoritative methodological guidance on how to conduct a systematic mapping review, such as exists for instance for a systematic review [20]. Researchers have gravitated towards a systematic process of study identification, screening and data extraction [20, 21], so that it is clear how the maps have been created. However, research questions are commonly broader in mapping reviews than in systematic reviews [22], study quality is not appraised or graded, and data are presented in a tabular or visual format rather than analysed or fully synthesised [16, 17, 20, 21].

Systematic mapping was chosen as the method for this work since we aimed to understand the studies and data available with a view to establishing further research priorities. The maps are used to consolidate studies in the broad research area of interventions to address mental health inequalities experienced by marginalised or minority communities.

Research questions and objectives

The research questions were as follows:

  1. 1.

    What studies are there on interventions to address or reduce mental health inequalities?

The objective of this research question was to identify the existing interventions which seek to address mental health inequalities.

  1. 2.

    What are the data from economic evaluations for interventions to address or reduce inequalities in mental health care?

The objective here was to identify studies that assess the economic impact of interventions to address inequalities in mental health care.

  1. 3.

    What are the barriers and facilitators to interventions to address or reduce mental health inequalities?

The objective of the third question was to identify factors which act as barriers and those that can facilitate interventions which seek to address mental health inequalities.

To develop the research questions, we worked with stakeholders and methodologists [16], scoped the literature and informally reviewed the evidence base to identify relevant studies [23]. The purpose was to determine suitable research questions and use these to develop the approach to study identification, as well as pilot inclusion criteria and data extraction for the mapping review [23, 24]. We convened a group of stakeholders to inform the development of this review as part of the AMHE resource [15] developed at the National Collaborating Centre for Mental Health (NCCMH). Stakeholders included people with lived experience of mental health problems, informal carers and people with one or more of the characteristics outlined in Table 1. We also consulted experts in the field of mental health and equalities research; mental health care professionals, including psychiatrists, mental health nurses, approved mental health professionals (AMHPs); and staff working in equality lead roles in the NHS. They contributed to the development of the research questions through a series of focus groups and workshops [15]. They were also integral to the work on the categorisation of barriers. Stakeholders were informed of the progress of the review and had an opportunity to advise further in subsequent meetings and via email.

Definitions

Mental health inequalities and inequities

‘Mental health inequalities’ are defined in this work as differences between population groups in their mental health status and outcomes, including the following:

  • the prevention of mental ill health;

  • access to and experience of mental health care; and

  • outcomes associated with mental ill health.

‘Mental health inequities’ are avoidable inequalities between population groups. They arise from social and material inequalities within society, such as discrimination, stigma and distribution of wealth and resources [25]. This study considers interventions addressing mental health inequities; however, because there is a lack of clear differentiation between definitions of inequity and inequality within the literature, the term ‘mental health inequality’ is used.

The population, intervention, outcomes and study designs

Population

In this work, the targeted population is defined as people who meet at least one criterion from each of the two types of criteria pertaining to: (a) disorder/problem type AND (b) having one or more specific characteristics.

  1. A.

    Population: Disorder/problem type

People who have a diagnosis, or who are at risk, of any of the following mental health conditions/disorders or problems:

  • anxiety

  • bipolar disorder

  • antisocial behaviour and conduct disorders (in children and young people)

  • depression

  • eating disorders

  • mental health problems in the pregnancy and postnatal period

  • personality disorders

  • psychosis and schizophrenia

  • self-harm.

This list is derived from the National institute for Health and Care Excellence’s (NICE) categorisation of guidance and pathways for the above specified conditions [26]. We have cross-referenced the list with the International Classification of Diseases’ (ICD-10) [27] classification of mental health disorders for conditions that fall under the categories of:

  • F20-29: Schizophrenia, schizotypal and delusional disorders

  • F30-39: Mood (affective) disorders

  • F40-48: Neurotic, stress-related and somatoform disorders

  • F60-69: Disorders of personality and behaviour in adult persons

  • F91: Conduct disorders (in: Behavioural and emotional disorders with onset usually occurring in childhood and adolescence)

  1. B.

    Population: Having specific characteristics

    People who have one or more of the characteristics outlined in equality impact assessments used by NICE in the development of guidelines. These characteristics served as a basis; we broke them down to identify the specific areas that we wished to focus on (see Table 1).

Table 1 Population characteristics (protected or other) considered for inclusion in this review
Full size table

Intervention

In this work, an ‘intervention’ refers to any type of purposeful act, programme, system or deliverable that has been put in place with the intention of addressing or reducing mental health inequality or that is targeted at a specific group at risk of experiencing mental health inequality. These can include treatment interventions, targeted adaptations of existing treatments, policies, intentional organisational or structural changes.

Outcomes

This relates to the research questions, such that for research question 1, the outcomes relate to the effectiveness of interventions measured, for example, using relevant clinician- or patient-rated scales (such as symptom severity scales or quality of life measures) or access rates; for research question 2, the outcome is economic in nature, such as a cost-benefit; and for research question 3, the outcome is any actual or perceived barrier or facilitator to intervention uptake and/or success for which themes were extracted.

Study design

By research question:

  • Research question 1: any primary study evaluating effectiveness

  • Research question 2: any economic evaluation

  • Research question 3: any primary study evaluating barriers and/or facilitators to intervention uptake.

Systematic reviews were included but they are reported separately to studies identified above and in their own table (see Additional file 3). Editorials, commentaries and letters were not included.

Search strategy

Study identification (literature search) was undertaken by a qualified information specialist. The following bibliographic databases were systematically searched:

  • MEDLINE and MEDLINE In-Process via Ovid

  • Health Management Information Consortium (HMIC) via Ovid

  • Applied Social Sciences Index Abstracts (ASSIA) via ProQuest

  • Social Policy & Practice via Ovid

  • Sociological Abstracts via ProQuest

  • Social Services Abstracts via ProQuest

  • PsycINFO via Ovid.

The full search strategy can be found in Additional file 1 and takes the following form: (terms for mental health) and (terms for inequalities and reduce) and (terms for economics, meta-analysis, systematic reviews, observational studies, randomised controlled trials (RCTs) and barriers and facilitators). The searches were not limited by language and spanned the period from January 2008 to December 2018, a timeframe the authors considered to be within resource limits [28] and agreed with select stakeholders. The search strategies were reviewed by the research team using the PRESS checklist [29]. Resources did not permit the inclusion of the Embase database.

Study selection

Studies were double-screened according to the pre-determined inclusion criteria. Title/abstract screening was undertaken using the desktop Rayyan application [30] and resulted in 97.1% agreement. Disagreements were resolved by discussion with the wider review team.

Inclusion criteria

For all research questions, to be included in the review the studies had to:

  • Focus on a population with:

    1. a.

      Mental health disorders, conditions or problems that meet the definition for population in this review, and

    2. b.

      Focus on a population group with protected or other characteristics identified as at risk of experiencing mental health inequalities (see Table 1), and

  • Address an intervention, as defined by this review, focused on addressing or reducing mental health inequalities, and

  • Meet the following criteria for one or more of the research questions:

  1. a.

    Research question 1: be a primary research study (any study design) or systematic review reporting effectiveness findings for an intervention or interventions

  2. b.

    Research question 2: report the findings of an economic evaluation; include sufficient detail regarding methods and results; the study’s data and results to be extractable (full economic evaluations that compare two or more relevant options and considered both costs and consequences; costing analyses that compared only costs between two or more interventions; and non-comparative studies were all included)

  3. c.

    Research question 3: be a primary research study (any study design) or systematic review identifying and categorising, describing or explaining, potential barriers or facilitators to intervention uptake or success.

Data extraction

Where possible, data were extracted from title/abstract, which is consistent with methods of the other mapping reviews [20]. Where study abstracts were insufficient in providing the data required for extraction, lacked clarity or there was any doubt, full texts were retrieved and the relevant data extracted. For primary studies, we extracted study aims, study design, population (sample), population characteristic(s) associated with inequality, intervention details, intervention types and strategies (as applicable), comparator (as applicable) and outcomes. We also extracted currency for primary studies answering research question 2 and outcomes and/or themes for those answering research question 3. For systematic reviews we extracted study aims, included studies, population characteristic(s) associated with inequality, intervention details, intervention types and strategies, comparator(s) (as applicable) and outcomes. For systematic reviews to answer research question 3, we extracted themes pertaining to barriers and facilitators. Data extracted by one researcher were always double-checked by another. The data extraction tables are set out in Additional files 2 & 3.

Study quality (risk of bias)

Study quality was not appraised.

Results

The overarching findings from the study identification and screening processes are reported to PRISMA reporting guidance; a PRISMA flow chart [31] is included (Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram of research studies search

Full size image

The inclusion criteria were met by 128 studies; 115 were primary studies [32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146] and 13 were systematic reviews [147,148,149,150,151,152,153,154,155,156,157,158,159]. Relevant data were extracted and tabulated to map the studies (additional file 2); separate tables summarised relevant systematic reviews (see Table 2 and additional file 3). Maps were used to consolidate primary studies that addressed the broad research area of mental health inequalities. Using Adobe Illustrator [160], we also developed a visual representation of the mapping review in the form of a mapping diagram (Fig. 2).

Table 2 Systematic reviews for research question 1
Full size table
Fig. 2
figure 2

Mapping diagram of primary studies. *Studies examining these characteristics were only included if they also looked at other characteristics. **Specific populations are those who may have a specific set of characteristics and experiences placing them at risk of experiencing mental health inequalities and therefore may already be defined by multiple characteristics (e.g. refugees). Circle size area is asscociated with the number of primary studies that consider a population characteristic. The lines between the circles indicate where characteristics were considered together, while the thickness of the lines indicates the frequency of the association. 49 studies examined 2 population characteristics; the most frequent association was between age and socioeconomic factors. 24 studies examined 3 characteristics, of which age, race and socioeconomic factors were the most frequent associations. 8 studies examined 4 characteristics. 3 studies examined 5 characteristics. 3 studies examined 6 characteristics

Full size image

We identified 94 studies that addressed research question 1, 6 that addressed research question 2 and 36 that addressed research question 3; some studies were relevant to more than one question. An existing taxonomy of disparities interventions [161] was used and modified by way of expansion (Table 3), to categorise interventions by type (access, early intervention, intervention, prevention) and strategy (Table 4). Target populations in primary studies were categorised by a range of characteristics and characteristic sub-types (Table 5). It should be noted that characteristic sub-types may or may not be mutually exclusive; therefore, the ‘Number of studies by characteristic’ column in Table 5 does not offer a summative count. The mapping diagram (Fig. 2) presents the findings of the mapping review with reference to the number of primary studies that consider different population characteristics and the frequency with which multiple characteristics are considered together across studies.

Table 3 Modifications made to expand the taxonomy of disparities interventions used in this study
Full size table
Table 4 Summary of intervention strategies for tackling inequalities identified during research question 1
Full size table
Table 5 Populations identified and number of included studies by characteristic across all research questions
Full size table

For research question 1, we identified 94 studies of interventions that aimed to address mental health inequalities: 85 primary studies and 9 systematic reviews. A total of 74 unique interventions were identified across primary studies.

We categorised primary study target populations according to the following characteristics: socioeconomic factors (n = 65), age (n = 46), race/ethnicity (n = 29), location (n = 17), pregnancy and maternity (n = 14), sex (n = 10), ‘specific intersectional groups’ (homeless people, youth offenders and refugees; n = 6), ‘other’ (n = 6), disability (n = 3), sexual orientation and gender identity (lesbian, gay, bisexual, transgender, queer and others (LGBTQ+); n = 2) and religion (n = 2) (see Table 5). The mapping diagram (Fig. 2) presents the number of primary studies that consider different population characteristics using circle area, e.g. the circle for socioeconomic factors is largest, while the circle for religion is smallest. The lines connecting the circles indicate characteristics considered together, with thickness of the connecting lines showing frequency of that association. Forty-nine studies examined 2 population characteristics; the most frequent association was between age and socioeconomic factors. Twenty-four studies examined 3 characteristics, of which age, race and socioeconomic factors were the most frequent associations.

Countries in which primary studies were conducted were also recorded (Table 6), with the majority conducted in the USA (n = 34) and the UK (n = 17).

Table 6 Countries in which primary studies included in the review were conducted, by research question
Full size table

Intervention strategies used most frequently in primary studies were providing psychological support (n = 45), delivering education and training (n = 29), engaging the community (n = 26) and other—culturally adapted interventions (n = 26) (Table 4). The most frequently reported intervention strategy in systematic reviews was other—culturally adapted interventions (n = 5) followed by providing psychological support (n = 4). As with the primary studies, most of the reviews focused on targeting populations based on socioeconomic factors (n = 6). The systematic reviews we identified, including information on intervention types, strategies and target populations, are summarised in Table 2.

For research question 2, only 6 economic evaluations were included (Table 7); these were cost-effectiveness studies of which 3 were also identified in research question 1 [100, 119, 152]. Five studies examined population variables related to socioeconomic factors: 3 studies were of children [100, 108, 152], 1 looked at pregnant women [79] and another at adults eligible for Medicaid in the USA [119]. The last study examined the cost-effectiveness of a health-check intervention for people with learning disabilities [121]. All three cost-effectiveness studies of children with low socioeconomic factors examined the Incredible Years parenting programme. All cost-effectiveness studies were conducted in either the USA or Europe.

Table 7 Cost-effectiveness studies identified for research question 2
Full size table

For research question 3, we identified barriers and facilitators to interventions aimed at addressing mental health inequalities, including the populations with whom certain interventions are used (Table 8).

Table 8 Types of barriers, populations at risk and facilitators as identified in the literature
Full size table

Using the input from the lived experience members of the stakeholder group, we categorised the types of barriers reported in the literature (36 studies) into 8 groups: (1) limited treatment options and service limitations, (2) perceived or real discrimination, (3) access to care, (4) financial constraints, (5) communication issues, (6) awareness of available services, (7) trust in services or ‘the system’, (8) appropriateness of available services. Of the 36 included studies, 34 reported information on barriers in their findings, while only 20 reported on facilitators (Table 8).

Discussion

Research question 1: What studies are there on interventions to address or reduce mental health inequalities?

The majority (80%) of primary studies focused on targeting populations based on socioeconomic factors (n = 65), age (children and young people as well as older adults; n = 46) and race/ethnicity (ethnic minorities and indigenous/aboriginal populations; n = 29), indicating that these populations are most frequently targeted in interventions to address mental health inequalities in the published literature. However, it should also be noted that the majority of the included studies were conducted in the USA or the UK, limiting the ability to generalise these findings to other countries. We identified very few primary studies targeting populations on the basis of religious affiliation (n = 2), sexual or gender identity and sexual orientation (LGBTQ+; n = 2) and disability (n = 3), and none of the systematic reviews targeted these populations. These findings warrant further investigation as we are unable to conclude, through use of a mapping review, the reasons for these observations. It is likely that these findings might be indicative of the current state of the literature on availability of interventions for these populations. Further and more focused research on interventions designed for these specific populations is needed.

Identification of intervention strategies used to address inequalities across studies, by frequency, is an interesting finding of this mapping review. However, on its own this finding is of limited use and would be more informative when analysed in conjunction with other research on the effectiveness of these strategies with different populations, to better understand what works for different at-risk groups.

Research question 2: What are the data from economic evaluations for interventions to address or reduce inequalities in mental health care?

Only 6 economic evaluation studies were identified, limiting our ability to adequately address this research question and form a representative picture of the state of the cost-effectiveness literature regarding interventions aimed at addressing mental health inequalities. Half of the included studies were also analysed in research question 1, where the authors had performed a cost-effectiveness analysis as part of the study.

Research question 3: What are the barriers and facilitators to interventions to address or reduce mental health inequalities?

The identification of 8 types of barriers in the literature suggests that addressing inequalities in mental health care may be hindered by several factors, other than those that tend to be most commonly discussed, such as access to care [147, 162,163,164,165] and service integration [166]. The identification and categorisation of barriers and associated facilitators to interventions aimed at tackling inequalities is, therefore, a useful outcome of this mapping review. An understanding of barriers and facilitators can inform future intervention design, clinical practice, service organisation and methods of care delivery. We were able to identify where interventions are likely to encounter challenges in meeting their aims, while also summarising solutions and potentially helpful guidance around what could work. We were also able to identify the population groups that appear at risk of experiencing barriers. Future research should look more deeply at the effectiveness of facilitating factors in improving equality for people from at-risk population groups.

Strengths and limitations

This review demonstrates a number of important strengths. First, the protocol was informed by expert opinion with sustained input from expert stakeholders, including those with lived experience of mental health problems and related inequalities. This approach ensured that the direction of the research was both person-centred and reflected the priorities of the target populations. The input from clinicians also ensured the research held practical applicability and could be translated to clinical practice. In addition, the search strategy was broad, incorporating a range of pervasive cultural, social and intersectional inequalities. Similarly, the articles covered a wide geographical range across several continents and were not excluded on the basis of language. This approach is particularly pertinent for health equity research and ensures that geographical inequalities are also represented within the literature base.

We should also acknowledge the caveats. First, while a systematic approach was taken, it is not comparable with a full systematic review as only select databases were chosen; we did not search Embase and we were unable to search grey literature. To address research question 3, the search approach taken used a pragmatic cluster of search terms to focus the study identification on the barriers and facilitators to advancing mental health equality. The aim was to identify relevant studies within a manageable volume of studies to screen. It is possible that some studies have been overlooked by this approach and future work may look at the use of different terms. The timeframe for our literature search was narrowed to 2008–2018. We decided on a decade span between the authors and with input from stakeholders, as it was considered to be within resource limits, following guidance in the Cochrane Handbook [28]. As a result of this approach, it is possible that a relevant body of literature has been missed. In particular, considering the lack of literature pertaining to research question 2, it would be beneficial to conduct a full and focused systematic review in the future, to both provide a more encompassing review of the literature and gain a greater insight into the economic effectiveness of these interventions. Furthermore, as this is a mapping review, we did not assess the quality of each study, nor did we determine the effectiveness of the interventions. As such, the findings report the breadth of the data available, but it would be inappropriate to make any firm recommendations. Data were extracted from title/abstract where possible, with extraction from full text only performed when abstracts either lacked the data required or were unclear. While this method of extraction is both efficient and consistent with other mapping reviews [20], it does carry a risk of potentially relevant information being missed.

A further limitation arises from the problematic nature of categorising communities in the literature. The LGBTQ+ movement represents a diverse range of communities, each with unique needs and social perceptions; for example, the transgender community may face entirely different social issues and experiences of care compared with the lesbian community, and so on. Within the literature, however, these distinct groups were often amalgamated under one overarching label and, as such, it was difficult to separate the findings for lesbian, gay, bisexual, transgender and queer communities. Similarly, the Black, Asian and minority ethnic (BAME) label amalgamates a number of distinct cultures and ethnicities with heterogeneous experiences. Much of the literature did not make any further distinctions, once again limiting our ability to draw distinct conclusions for each component group. This approach limits the practical applicability of academic research concerning these communities and highlights a further pervasive inequality to which they are subjected. Ultimately, this requires a more nuanced, accurate reporting within future academic research involving these communities to ensure care can be tailored to their unique needs.

The stipulation applied to our inclusion criteria regarding population characteristics for age, sex and pregnancy and maternity is potentially problematic. This is because, to be considered eligible for inclusion in the review, studies of these populations needed to include people who have another characteristic or need that puts them at risk of experiencing inequalities. Our reasoning for this decision was to avoid including studies of participants from these groups who did not have an identifiable inequality. For example, studies of children and young people as such may not focus on addressing inequalities but might have been included inappropriately had it not been for the caveat applied here. Still, even though necessary, it is possible that the caveat meant that some potentially relevant studies were excluded.

Based on these limitations, we recommend a full systematic review for each of the characteristic sub-types to gain greater insights into the effectiveness of interventions to tackle mental health inequalities and inform national delivery of care for different population groups.

Conclusions

The mapping review indicated that the majority of mental health inequality interventions identified in this study focus on addressing socioeconomic factors, race disparities and age-related issues (most of which pertain to children and/or young people). The majority of interventions tend to use providing psychological support and delivering education and training as strategies. The review also identified population groups who may be at risk of experiencing barriers to interventions aimed at addressing inequalities. This knowledge is useful for commissioners and service providers seeking to understand what can be done to support the advancement of mental health equality for different populations. The information gained from the mapping review should be used to inform the direction of further research that could influence local commissioning and service provision.

The mapping review was useful in assessing the spread of literature across sub-topics and identifying the highly researched areas (which include interventions aimed at minority races; addressing socioeconomic factors; and age-related inequality issues) versus the prominent gaps (including interventions aimed at marginalised religious groups; the differing and unique needs of groups within the LGBTQ+ community; and people with disabilities). This map supports the identification of these potential gaps in existing research and assists in setting out future research priorities.