How do women with social risk factors experience United Kingdom maternity care? A realist synthesis

Abstract Background Echoing international trends, the most recent United Kingdom reports of infant and maternal mortality found that pregnancies to women with social risk factors are over 50% more likely to end in stillbirth or neonatal death and carry an increased risk of premature birth and maternal death. The aim of this realist synthesis was to uncover the mechanisms that affect women's experiences of maternity care. Methods Using realist methodology, 22 papers exploring how women with a wide range of social risk factors experience maternity care in the United Kingdom were included. The data extraction process identified contexts (C), mechanisms (M), and outcomes (0). Results Three themes, Resources, Relationships, and Candidacy, overarched eight CMO configurations. Access to services, appropriate education, interpreters, practical support, and continuity of care were particularly relevant for women who are unfamiliar with the United Kingdom system and those living chaotic lives. For women with experience of trauma, or those who lack a sense of control, a trusting relationship with a health care professional was key to regaining trust. Many women who have social care involvement during their pregnancy perceive health care services as a system of surveillance rather than support, impacting on their engagement. This, as well as experiences of paternalistic care and discrimination, could be mitigated through the ability to develop trusting relationships. Conclusions The findings provide underlying theory and practical guidance on how to develop safe services that aim to reduce inequalities in women's experiences and birth outcomes.


| INTRODUCTION
Women living in areas with the highest levels of poverty in the United Kingdom are 50% more likely to experience a stillbirth or neonatal death. 1,2 These women experience increased rates of premature birth, low birthweight, cesarean, and maternal death. [3][4][5][6] As socioeconomic status decreases, women are more likely to report that they were not treated respectfully, that they were not spoken to in a way they could understand during their maternity care, and that their concerns are not listened to. 5,8 Health inequalities between socioeconomic groups are well documented 7,8 and have been a key priority in many international and United Kingdom initiatives, including the World Health Organization's (WHO) "Global strategy for women's and children's health" 9 and the "Better Births" National Maternity Review. 10 Lower socioeconomic status is often accompanied by other complex social factors associated with adverse outcomes 5,[11][12][13][14][15] (Table 1). It is hypothesized that a lack of antenatal care and engagement with maternity services is directly linked to poor maternal and neonatal outcomes; therefore, policies are often focused on improving access to care. 9,[16][17][18] A secondary analysis of the United Kingdom's National Maternity Survey 5 showed that the most deprived women in the United Kingdom were 60% less likely to have received any antenatal care when compared to the least deprived women. Reviews of maternal and neonatal deaths [2][3][4]14 have found that women with social risk factors present real challenges for maternity services, with communication lapses between hospitals and the community health care setting.
Marmots' review of the social determinants of health encourages the development of partnerships, with those affected by social inequities working with their health practitioners. 11 Central to this approach is the development of a system that empowers women to have a real say in decisions that affect their lives, and that recognizes their fundamental human rights. 18,19 These values are echoed in the National Institute of Clinical Excellence (NICE) guidelines for women with complex social factors, 20 which called for a reorganization of maternity services to improve antenatal care for this population and identified gaps in evidence with respect to effective service provision. Continuity of caregiver is a key government priority in an attempt to improve poor outcomes for women, with priority to be given to black and minority ethnic women alongside those living in the most deprived areas. 16,17 This is currently a far cry from the reality of a fragmented United Kingdom maternity system. A large, national United Kingdom survey 21 reported 65% of women did not have a named midwife during pregnancy, and subgroup analysis of disadvantaged groups found inequalities in access to care, information, and interactions. 21 Compared to women receiving standard care, a recently updated Cochrane review 24 found that women who received continuity of care from a known midwife experienced significantly fewer preterm births, fetal losses, neonatal deaths, and clinical interventions and greater satisfaction. The review does not report on whether outcomes differed for socially disadvantaged women but recommended that future research should explore this population and the mechanisms underpinning the improved outcomes. Positive outcomes, including less clinical intervention, shorter hospital stays, fewer neonatal unit admissions, and increased liaison with multidisciplinary services for women with social factors, have been associated with continuity of care models in the United Kingdom. 25,26 There remains a paucity of evidence and professional agreement with respect to what models of care are effective in meeting specific population needs, and why some are more effective than others. Group antenatal care has also been identified as a possible way of reducing health inequalities for socially disadvantaged women, but the evidence to date is limited. 27,28 It is not known Conclusions: The findings provide underlying theory and practical guidance on how to develop safe services that aim to reduce inequalities in women's experiences and birth outcomes.

K E Y W O R D S
experiences of care, maternity services, socioeconomic status and ethnicity T A B L E 1 Social factors associated with increased risk divided into two groups [2][3][4][5]13,15,16,19,22,25 Women who find services hard to access RAYMENT-JONES ET Al. whether tailored models of care improve outcomes related to social deprivation, for example, the duration of breastfeeding, parent-infant bonding, and childhood obesity. It is also not known how acceptable these models of care are for women with complex social factors, and whether they are seen as supportive, stigmatizing, or potentially isolating. A systematic review 22 found that the effectiveness of specific antenatal care programs to reduce infant mortality for socioeconomically disadvantaged women has not been rigorously evaluated. A further synthesis of women's views and literature focusing on the initiation of antenatal care by black and minority ethnic groups in the United Kingdom 23 identified a range of barriers experienced by women including unfamiliarity with the system, inadequate interpretation services, a lack of cultural sensitivity, and impersonal care. The review suggested several interventions to overcome these barriers such as continuity of care, improved resources, and education but concluded that existing examples of specialist models of care should be fully evaluated and tested before they could be implemented into the wider NHS.
An analysis of the evidence presented in the Lancet series on midwifery 29 recognized the importance of women's experience in improving clinical outcomes and indicated future research investment should address "right care-tailored to individuals, weighs benefits and harms, is person-centred, works across the whole continuum of care, advances equity, and is informed by evidence, including cost-effectiveness." Therefore, this synthesis focused on how women with social risk factors experience maternity care in the United Kingdom, in order to advance theoretical understanding of the conditions required to increase the positive impact of care for this population. The aim was to uncover mechanisms that affect women's experiences of maternity care and develop program theories to be tested in a subsequent realist evaluation. Social complex*, social Factor*, vulnerab*, socioeconomic, socioeconomic status, SES, depriv*, poverty, poor, disadvantag*, level of education, low education, low prestige, social class, disparit*, inequalit*, inequit*, discriminat*, impoverish*, low income, social* exclu*, social isolat*, homeless*, refuge*, immigra*, asylum*, non-native language, language barrier*, minority ethnic*, ethnic*, black and minority ethnic, BME, sexual* abuse*, abuse*, domestic abuse*, domestic violence, intimate partner violence, IPV, physical abuse*, emotional abuse*, victim of abuse, sex worker*, adolescent*, young mother*, teenage*, single mother*, traveller*, travelling community, roma*, mental health, perinatal mental health, safeguard*, social care, social service*, child protection, substance abuse, drug abuse, addict*, alcohol*, alcohol abuse, disabil*, physical disabil*, learning disabil*, emotional disabil*, Female genital mutilation, FGM, Female circum*, HIV Positive status, HIV Methodology Included-qualitative literature or the qualitative data within mixed-methods research Excluded-any literature published before 2010 to reflect the response to recommendations of the NICE 20 maternity service guideline for women with complex social factors Experien*, encounter, perception, view*, feel*, felt, remember*, recollect*, access*, engage*, communicat*, trust*, comfort*, uncomfort* people respond to interventions using contexts, mechanisms, and outcome configurations, 30 for example, how women in a particular context respond to an aspect of their maternity care (the mechanism), and what is the outcome of this response. This was thought to be the most appropriate methodology for the review question posed as it not only recognizes the complexity of social risk factors and maternity services, but also allows the structured development of program theories to break these complex phenomena down into more manageable hypotheses to test what works in improving women's experiences of maternity care. This synthesis was undertaken through regular collaboration with a patient panel consisting of recent maternity service users with social risk factors, and a panel of international experts in health inequalities and maternity care. Both panels advised on the review aims, search criteria, data extraction process, analysis, and identified gaps in the literature.

| Literature search
This realist-informed, systematic synthesis of qualitative primary studies focused on the maternity care experiences of women with social risk factors using Pawson's 30 5 stages of a realist synthesis. Two independent researchers reviewed 1830 papers by title and abstract according to the search strategy and inclusion criteria ( Table 2). Fifty-two full-text papers were reviewed and 22 papers included ( Figure 1) (See Table  S1 for an overview of included studies). Included studies were quality-appraised using a validated checklist 53 and generally assessed as high quality (Table 3). Although it was important to report on the quality of the studies, they were not weighted RAYMENT-JONES ET Al. T A B L E 3 Quality assessment of included papers in synthesis of how women with social risk factors experience United Kingdom maternity care 54

Lephard and Haith-Cooper
according to quality during the analysis as the purpose of this synthesis was to collate program theories and CMO configurations ready to test in a subsequent realist evaluation.

| Data extraction
A data extraction tool was devised and completed for each paper to identify explanatory contexts (C), mechanisms (M), and outcomes (0), and to develop program theories arising from these configurations. Program theories were constructed using "if….., then…" sentences. For example, "migrants who arrived in the country late in their pregnancy or had re-located or been re-dispersed from elsewhere in the UK (C), were unable to register with a GP in sufficient time to access maternity services before birth (O)" was converted into the following program theory: "If women who arrive in the country late in their pregnancy or have been re-located or re-dispersed from elsewhere in the UK are able to book maternity care directly with a midwife, then barriers to early access will be overcome and those who have difficulty registering with a GP will not be excluded." This process ensured transparency in converting findings into tangible, testable hypotheses or "program theory." A total of 354 program theories were constructed from the findings of the 22 included studies. This collected the voices of 936 women with various social risk factors. Program theories were organized using data analysis software 53 to uncover themes and develop middle-range theories as recommended by Forster e al 55 to increase transparency in decision making. This process enabled similar theories to be condensed, the extraction of theories specific to certain social risk factors, and the identification of conflicting theories. These conflicting theories give insight into what works in different contexts and for different populations. 56 Once all papers had been classified according to the social risk factors included and the model of maternity care received and similar program theories condensed, 85 program theories remained. These final theories were grouped into the most commonly occurring themes and further refined into eight CMO configurations.
Middle-range theories help conceptualize complex reality so that empirical testing of the more specific program theories becomes possible and generalizable. [57][58][59] This conceptualization aided the development of the final CMO headings and has enabled a theoretically informed approach to the design of the subsequent realist evaluation, with the theories incorporated into the interview guides.

| RESULTS
The full findings of this synthesis are detailed in 85 program theories (45 general theories and 40 that are specific to different social risk factors) and referenced to relevant included studies to demonstrate transparency (see Table S2). For the purpose of presenting a concise overview, the program theories were refined into eight overarching CMO configurations under three thematic headings (Table 4): System Resources, Relationships, and Candidacy. The CMO configurations are not ordered in relation to importance as all are thought to be important in impacting outcomes depending on the specific contexts identified. Quotes from women are included to add meaning and illustrate findings in the included studies.
The Resource theme included (a) access to maternity services and (b) appropriate antenatal education, (c) interpreter services, (d) practical support, and (e) continuity of care, these were particularly relevant for women who are unfamiliar with the National Health Service (NHS) system and those living chaotic lives. For women with experience of trauma, abuse, and discrimination, or those who lack a sense of control, (e) the ability to build a relationship with a health care professional was key to regaining trust in the system and control over what happens to them and their baby. The "Candidacy" theme recognized that women with social risk factors are more likely to experience paternalistic care and highlighted the impact of (f) health care professionals' assumptions based on race, class, ability, age, and other sources of oppression. This might be overcome by placing services in local communities where health care professionals are immersed in local cultures and recognize the strengths and assets held by women and their communities. Lastly, many women with social risk factors perceive health care services as a system of (g) surveillance rather than support, impacting on engagement and meaningful support. This could be mitigated through the ability to develop trusting relationships, health care professionals' knowledge of safeguarding and reporting mechanisms, and processes put in place to ensure women's safety.

| DISCUSSION
This synthesis systematically identified qualitative literature that focused on the experiences of maternity care in the United Kingdom for women with social risk factors and used realist methodology to uncover the contexts and mechanisms that led to positive or negative experiences. These contexts and mechanisms were coded and developed into CMO configurations, providing a set of program theories to test and compare women's experiences in future research and evaluation of services. The findings contribute to knowledge by providing detailed insight into how different social risk factors affect women's ability and willingness to access and engage with services. The realist methodology takes the findings of the 22 included papers deeper by unearthing potential mechanisms that may improve or worsen experiences.  social risk factors. The included studies covered a range of social risk factors that were often multiple and overlapping. Black and minority ethnicity, and asylum seeker/refugee status were the risk factors most commonly focused on, and although the vast majority of the studies found that the participants were socially deprived, only four of the 22 papers used social deprivation in their inclusion criteria. By focusing on single social risk factors when designing research or services, the complexity of social deprivation and oppression may be overlooked and deficits within the system disregarded. For example, the growing body of literature on the "healthy migrant" phenomenon shows that many first-generation immigrants often have better physical and mental health than the indigenous populations of many developed countries. 60,61 This suggests that it is not that a person is not native to a country that puts them at risk of health inequalities, but it is growing up in a place where that person might be perceived as different that has a greater bearing. This synthesis found that for black and minority ethnic women, asylum seekers, and refugees, it was the language barrier and unfamiliarity with the United Kingdom system that had the biggest impact on how they accessed, engaged, and experienced their maternity care. This leads us to the concept of intersectionality. Although intersectionality was not explicitly discussed in the included studies, it became a clear factor in how women experienced maternity care. Oppressive institutions of racism, sexism, ableism, classism, etc, are interconnected, impact on health inequalities, 62 and cannot be separated when trying to understand why some women experience maternity care differently to others. One example of this is found in Bradbury-Jones' study 36 where the women felt that not only they were perceived as less able to make decisions because of their disability, but also this was compounded by health care professionals' judgments about the domestic abuse they had experienced. Five of the eight CMO configurations related to system resources: access, interpreter services, education, practical support, and continuity of care. This closely reflects the findings of Hollowell et al's 23 review of black and minority ethnic women's experiences of maternity care. A frequent finding in both papers was the importance of communitybased care, allowing women and midwives to integrate with the local community, and ease access to services for women who lack resources or are not able to travel far to hospital appointments.
The importance of relationships was so apparent in the program theories that it became a key middle-range theory. There is a wealth of literature on the benefits of continuity of care on women's outcomes. [23][24][25][26] This synthesis found that for women whose trust has previously been broken, either through interactions with professionals, or previous trauma and abuse, the development of a trusting relationship with a health care professional results in increased confidence, for women with immigration issues, including processes of payment as a non-United Kingdom resident, and ability to signpost women to confidential advice 2. HCP's ability to explain the reasoning behind reporting safeguarding concerns, the process of assessment, and discussion of what "meaningful support" means to the woman 3. Women's involvement in the process of reporting safeguarding concerns in an open manner that encourages them to identify their needs 4. Processes are in place that protect the woman from being put at risk of harm, for example, women whose abusers or traffickers may control or observe access to services are given the opportunity to self-disclose in safe environment and disclosures are followed up safely and sensitively

| 471
RAYMENT-JONES ET Al. safety, and empowerment. It also reduced women's perceptions of discrimination, manipulation, and coercion by people in power. Although "relationships" was found to be an occurring theme in this synthesis, the concept of trust was tied in closely to this. Women described the impact of trust in health care professionals and trust in the system as a whole. Literature on the theoretical perspectives of trust describes these two aspects, suggesting that trust in a person can act as a moderator/mediator when there is distrust in a system. 63,64 However, this protective factor is vulnerable to the trusted person not being there. A conflicting program theory identified that for some women, particularly those with social care involvement, it was more important that the whole service is perceived as safe, respectful, understanding, and kind, rather than one trusted professional in a wider toxic environment. The data from women who expressed this were linked to perceptions of surveillance, which may explain why the thought of one known health care professional might be perceived as intimidating, and building a relationship may be viewed as an invasion of privacy. It should be noted that the vast majority of included papers reflected standard maternity care and that those women who had experienced a form of continuity did not report negative perceptions of surveillance and valued the relationship they had with their health care practitioner/support person. Dismantling the belief that accessing health care services equates to relinquishing control may have long-lasting consequences on women's social interactions, help-seeking, and parenting. Conversely, if women with social risk factors, particularly those that contribute to disempowerment, experience paternalistic care through being denied choice and perceive health care professionals as lacking warmth, patronizing, arrogant, and stigmatizing, then they will remain disempowered and feel undervalued, and their low self-confidence will increase. Candidacy, defined as "the ways in which people's eligibility for medical attention and intervention is jointly negotiated between individuals and health services," 65 was the umbrella concept for two CMO configurations: "assumptions" and "surveillance." The concept suggests that a woman's "candidacy" for maternity services is materially, culturally, and organizationally constructed. For example, it is well known that more deprived women access preventative health care services less than more affluent women, 5,66 and have higher use of emergency services. 67 Candidacy is thought to be at play here, with factors such as help-seeking in response to crisis symptoms rather than to prevent poor health, the normalization and acceptance of poor health, and fear of blame from health care professionals apparent across many of the included studies. Again, these factors were found in Hollowell et al's review, 23 with barriers to initial access, lack of interpreter services, discrimination/disrespectful care, and health care professionals' lack of cultural knowledge affecting how women perceived their candidacy for services. The findings of this synthesis extend these findings further by proposing that if the value of accessing maternity services for the purpose of monitoring, prevention, and support is communicated across the communities in which women live, through community-based services and relationship building, then women would not view the purpose of the service as simply the treatment of ill health, and access care earlier in pregnancy.

| Strengths, limitations, and gaps in literature
Overall, the studies included in the synthesis were assessed to be of high-quality and they reported on studies conducted with women with a range of different social risk factors. However, the number of studies reporting women's socioeconomic status was limited. Only two of the studies reported specialist models of care, with the remaining studies reflecting the experiences of standard maternity care. This meant that the development of program theories for what works in improving women's experiences was often drawn from negative experiences and inverted to a positive program theory. To test those theories, a full evaluation of how women experience specialist models of care is required.
A further limitation of the synthesis is the cutoff date of 2010 in the inclusion criteria (see Table 2), potentially restricting the depth of the findings. This criterion aimed to reflect the NICE 20 guidance for women with social complex factors and to compare findings with previous systematic reviews of women's experiences of antenatal care. 22,23 With these limitations in mind, the findings of this synthesis add depth and detail in what works, for whom, in what circumstances, and how, to existing recommendations from the international wider literature. 5,8,9,13,18,22,23 There were some themes that were expected to be reported but were not. These included the recognition of women's personal strengths and assets, and the impact of their community. This may be because the women interviewed felt these were not important, because the research approach did not explore these themes, or because they were not included in final published work. The assumption of deficit-that people are a burden on the state rather than a resource-with respect to low-income people, asylum seekers, refugees, and migrants was sometimes apparent in the reported experiences of women but was not made explicit in the discussion sections of the studies. In addition to this, despite the growing body of evidence into the "healthy migrant effect," the papers included in the synthesis did not explore inequities in health service use, experiences, and outcomes for second-or third-generation descendants. Tudor Hart's 68 "inverse care law"-the principle that those most in need of care are the least likely to receive it-was also evident in the findings of many included studies but not discussed. For example, do health care professionals "do more" for more affluent women? Do women with lower socioeconomic status have lower expectations of maternity services? Further research, using qualitative realist evaluation methodologies with all stakeholders, will help to answer these questions and test the program theories put forward in this synthesis.

| Conclusions
The findings of this synthesis provide both an underlying theory and practical guidance on how to develop safe, person-centered maternity services for women with social risk factors that encourage early access and meaningful engagement and reduce the discrimination and fear this group of women often experience. The synthesis contributes to knowledge by identifying how women with different social risk factors experience care in different ways, resulting in specific program theories tailored to more individualized need. The CMO configurations developed will be tested in a realist-informed evaluation of two specialist models of care (one community based and one hospital based) within areas of significant health inequity in London, United Kingdom. The synthesis also highlights potentially significant gaps in the literature, such as the impact of discrimination on outcomes and experiences, potentially stigmatizing service provision, or the protective factors of community and family support. These knowledge gaps should be explored in future research and considered when planning services for this vulnerable population.