Addressing social inequity through improving relational care: A social–ecological model based on the experiences of migrant women and midwives in South Wales

Abstract Background Migrant and ethnic inequalities in maternal and perinatal mortality persist across high‐income countries. Addressing social adversity and inequities across the childbirth trajectory cannot be left to chance and the good intentions of practitioners. Robust, evidence‐based tools designed to address inequity by enhancing both the quality of provision and the experience of care are needed. Methods An inductive modelling approach was used to develop a new evidence‐based conceptual model of woman–midwife relationships, drawing on data from an ethnographic study of relationships between migrant Pakistani women and midwives, conducted between 2013 and 2016 in South Wales, UK. Key analytic themes from early data were translated into social–ecological concepts, and a model was developed to represent how these key themes interacted to influence the woman–midwife relationship. Results Three key concepts influencing the woman–midwife relationship were developed from the three major themes of the underpinning research: (1) Healthcare System; (2) Culture and Religion; and (3) Family Relationships. Two additional weaving concepts appeared to act as a link between these three key concepts: (1) Authoritative Knowledge and (2) Communication of Information. Social and political factors were also considered as contextual factors within the model. A visual representation of this model was developed and presented. Conclusions The model presented in this paper, along with future work to further test and refine it in other contexts, has the potential to impact on inequalities by facilitating future discussion on cultural issues, encouraging collaborative learning and knowledge production and providing a framework for future global midwifery practice, education and research. Patient or Public Contribution At the outset of the underpinning research, a project involvement group was created to contribute to study design and conduct. This group consisted of the three authors, an Advocacy Officer at Race Equality First and an NHS Consultant Midwife. This group met regularly throughout the research process, and members were involved in discussions regarding ethical/cultural/social issues, recruitment methods, the creation of participant information materials, interpretation of data and the dissemination strategy. Ideas for the underpinning research were also discussed with members of the Pakistani community during community events and at meetings with staff from minority ethnic and migrant support charities (BAWSO, Race Equality First, The Mentor Ring). Local midwives contributed to study design through conversations during informal observations of antenatal appointments for asylum seekers and refugees.


| BACKGROUND
In high-income countries, migrant and ethnic inequalities in pregnancy outcomes persist. In the United States, for example, there is an increasing mortality gap between non-Hispanic Black and all other women, 1,2 where recent data show that Black women are at over three times higher risk of dying from pregnancy complications than White women. 1,2 In the United Kingdom, the most recent Confidential Enquiry into Maternal Deaths and Morbidity 3 reports that women from Black ethnic minority backgrounds are more than four times more likely to die as a result of complications in their pregnancy compared to White women. Similarly, women of mixed ethnicity are nearly twice as likely and women from Asian backgrounds have a threefold risk of dying as a result of complications.
Ethnic inequalities also exist in perinatal outcomes, where UK mortality rates remain exceptionally high for babies of Black and Black British ethnicity: Stillbirth rates are over twice those for babies of White ethnicity and neonatal mortality rates are 45% higher. 4 For babies of Asian and Asian British ethnicities, stillbirth and neonatal mortality rates are both around 60% higher than for babies of White ethnicity. 4 Migrant women and babies are also at increased risk of mortality and morbidity; nearly a quarter of maternal deaths between 2015 and 2017 were women born outside the United Kingdom, 3 and UK mortality reports call for 'continued focus on action to address these disparities' 3 (p. 5), while highlighting the role of midwifery in addressing inequities during pregnancy, birth and the early postnatal period.
One critical factor in midwifery care is the quality of the woman-midwife relationship, with extensive literature suggesting that these relationships significantly impact on women's experiences of care as well as pregnancy outcomes. [5][6][7][8][9][10] Indeed, the 2011 UK of service users and care providers, could be used in education and practice settings to address inequity by increasing mutual understanding, thus enhancing both the quality of care provision and the experience of care.
This gap in understanding and the clear need to create evidencebased tools designed for practical application influenced the authors to develop an original model of woman-midwife relationships, informed by our ethnographic study of relationships between migrant Pakistani women and UK midwives in South Wales 22 (described in full in an earlier issue of this journal). The remainder of this paper describes the development of this new social-ecological model of woman-midwife relationships, which conceptualizes not only interpersonal but also social and ecological factors that may serve as barriers or facilitators to a positive woman-midwife partnership approach to maternity care. Through the visual representation of the model, we aim to present the relationship between a woman and her midwife in a dyadic and holistic way that reflects the experiences of those receiving and those providing the care.
By doing so, our intention is to encourage better informed and culturally aware/sensitive ways of providing services to ethnic minority women in the United Kingdom to address social inequities during pregnancy, birth and postnatally.

| METHODS
The development of the inductive model was undertaken in three phases, which are described in the following sections: Participants in interviews and observations included 11 NHS midwives working in the industrialized South Wales region of the United Kingdom; seven first-generation migrant Pakistani women who were between 3 and 6 months pregnant and receiving NHS maternity care in the same region; one migrant Pakistani woman who was the mother of another participant; and one language interpreter (female) who had also migrated to the United Kingdom from Pakistan and experienced the UK maternity system. Length of residency in the United Kingdom ranged from 2 to 15 years, with a mean length of residency of 7 years. The focus was on migrant Pakistani women specifically, as at the time of study design, Pakistan was the second most common country of birth for non-UK-born mothers, 23 and Pakistani women were at significantly increased risk of infant maternal mortality when compared to all other ethnic groups in the United Kingdom. 10 The review of relevant media involved searching for news stories, healthcare policies, social media posts and government legislation relating to immigration, maternity care, ethnic inequalities and healthcare provision. Searches were carried out under the 'news' and 'scholar' advanced options of an internet search engine, using keywords such as 'migrants', 'midwifery', 'inequalities' and 'policy'. Additionally, hardcopies of local papers were skimmed for relevant stories on a daily basis, and government and health board websites were searched for relevant policies every 6 months. Social media posts were also scanned on a frequent basis. Findings from this study highlighted the complexity of relationships between women and midwives, and suggested a number of influential nested and interrelated themes. These included the role of family relationships; participants' relationships with culture and religion; understanding of different healthcare systems; attitudes towards authoritative knowledge; and perceived function of communication of information. 22 A full description of the study methods and key themes can be found in the findings paper, published in a previous volume of this journal. 22 2.2 | Phase 2: Exploring existing models of woman-midwife relationships In the next stage, we began by comparing early data from our underpinning research with existing models of healthcare relationships to explore the ways in which our themes, later named as concepts, aligned and differed. As is common in the healthcare literature, conceptual models are developed from research in an embryonic fashion, where key themes developed from the data are translated into model 'components' or 'concepts'. Conceptual models of midwifery provide different ways of looking at practice and alternative ways of working with women and their families. 24 They can also provide a framework for organizing education and identifying research questions. 24 Early on in this process, we realized that the influence of social and political contextual factors on the woman-midwife relationship was reminiscent of a theory of child development, namely, the 'Ecological Systems Theory' 25 (EST) developed by Bronfenbrenner.
EST proposes that child development should be viewed from a social-ecological perspective, placing the child in the centre of a 'layered' system, with all layers interacting and influencing development. These layered systems range from the child's individual interactions to the social and political context in which the child is raised.
To correctly study human development, Bronfenbrenner argues that one has to see within, beyond and 'across' how these systems interact. The emphasis on systems interaction resonated with our ethnographic findings, as woman-midwife relationships were influenced by a number of social and ecological factors (our themes) that could not be studied in isolation.
We then discovered that the EST theory had previously been applied to models of healthcare relationships. For example, Hummell and Gates 26 built on the systems approach, suggesting that healthcare relationships operate within a series of 'nested dimensions'. The concept of nested dimensions also resonated with our findings, suggesting that to understand the quality of healthcare relationships and their impact on women's experiences and outcomes, it is crucial to acknowledge the complexities of interacting systems that might influence these relationships. A model proposed by Higgs 27 similarly posits that each person exists within a network of multiple relationships, which variously impact on their encounters with healthcare professionals. Although relevant to our emergent research findings, these existing models focus on each person's individual factors such as knowledge, attitudes and beliefs separately, thus neglecting the interactions between, in this case, the woman and the midwife's own social-ecological influences. In short, we felt that existing models were not sufficiently sensitive to the interaction and nested complexities occurring between women and midwives that we were identifying in the data. We therefore began the process of creating a new conceptual framework to better understand our data and capture the relational complexities of partnership working therein.

| Phase 3: Creating a new social-ecological model of woman-midwife relationships
The process of developing a new model was undertaken in parallel with data collection and analysis. For example, key analytic themes identified in the data were translated into social-ecological concepts ( Table 1). The lead author produced a number of prototype models to visually represent how the socio-ecological concepts appeared to interact (i.e., how the social-ecological influences on the woman interacted with the social-ecological influences on the midwife). Each iteration of the prototype model was discussed with the coauthors for refinement and agreement and, in turn, informed further analysis of transcripts, observations and field notes. The model continued to be refined until all data had been analysed and there was consensus amongst the authors. The process of creating the model can, therefore, be described as both inductive and iterative. The following sections further unpack the model by explaining how the various elements of the model dynamically interact, with direct reference to our previously published research findings. 22

| The three key concepts
Three key concepts were developed from the three key themes and findings of the underpinning research: (1) the Healthcare System; (2) Culture and Religion; and (3) Family Relationships. From our original findings, it was clear that participants placed different 'weighting' on the importance of these concepts in terms of their influence on the woman-midwife relationship. 22 For example, the 'Healthcare System' was the most commonly discussed concept by midwives when interviewed about their relationships with women, 22 where the biggest source of tension in their relationships with migrant Pakistani women was perceived to result from women's late arrival, or nonattendance, at NHS antenatal appointments.
In contrast, the 'Healthcare System' was discussed less frequently by migrant Pakistani women. 22 Instead, women mostly focused on the impact of 'Family Relationships' and 'Culture and Religion' on their relationships with midwives. 22 It was therefore important to demonstrate the different weightings of these concepts GOODWIN ET AL.

| Weaving concepts
The two 'weaving concepts' were developed from the 'weaving themes' identified from the original study data. 22  'Whenever we have a baby, we follow our elders. Our grandparents, our mother-in-law, our mothers-we follow them. We don't try to follow what the midwife wants to say to us-what the midwife is saying for safety. We don't bother-frankly speaking we don't bother…We follow our grandparents, our mother-in-law and our mothers'.
Women's relationships with mothers/mothers-in-law 'They're inclined to talk for them as well… You're not quite knowing what the lady herself is thinking. I think mothers-in-law can be quite, the dominant relative. So they're inclined to, the mother-in-law, if she comes, to sort of dominate the consultation'.

The involvement of mothers-in-law Family relationships (midwife)
'Pregnant women shouldn't fast. And I always find that if they are fasting then I'm kind of lecturing them "no-you shouldn't be fasting" and that kind of thing, and they do get a bit funny about it. Because they want to do it, and I'm saying no you shouldn't do it. And that can cause a bit of-you can see that they're not happy that I'm saying no you shouldn't'.

Cultural practices Culture and religion (midwife)
'The traditions at home are completely different and how many visits they get-some are quite surprised at how many they get and some are surprised that they're not getting more. So yeah-it is different…You know-the whole maternity system at home'.
Understanding different healthcare systems Healthcare system (midwife) 'I would listen to the midwife. Because she's obviously the person who's more experienced in that. But then it's tradition… and you kind of respect tradition as well. I don't know-it's a bit difficult. How would you balance it?'

Authoritative knowledge Authoritative knowledge
'What I dislike is people who come in and they've got a list of demands. "You need to write me a letter for housing. You need to do this-you need to do that" That's all they want! Care isn't always a priority for them…they'll only come when they want something'.

Perceived purpose of communication/appointments
Communication of information humour, women often appeared to reject these exchanges in favour of more transactional encounters where medical testing and information provision were the focus of the consultation. Midwives reported that transactional exchanges such as requests for letters or signatures for other agencies (i.e., housing support) led to feelings of frustration, and many felt that they had been 'used' by women in such instances.

| Contextual factors
The social and political contextual factors within which the woman-midwife relationship unfolded, depicted as the merging purple into the white background in Figure 1 The model presented in this paper attempts to minimize this risk by emphasizing the individuality of relationships and encouraging actors to reflect on how the diagrammatic representation of the conceptual model should be adapted to fit their own specific relationships, rather than generalizing solutions to all relationships. Indeed, a strength of our model is that it aligns with the idea of cultural 'humility', rather than that of cultural 'competence'. The humility approach encourages healthcare professionals to continually engage in self-reflection and self-critique, rather than attempting to 'know' about the culture of the 'other'. 37 As such, cultural humility addresses many of the critiques of cultural competency models. For example, the cultural humility approach explicitly acknowledges power differences between healthcare providers and service users 20 and advocates for practitioner self-reflection on what bias and assumptions they may bring to the provider-client relationship when working with people from different backgrounds. 20 Awareness of this power imbalance is especially important for migrant service users who may already feel at a disadvantage due to issues with navigating the healthcare system or accessing the appropriate care. 22

| Implications for practice
It is proposed that this model be used as a tool to encourage healthcare staff to reflect on how their beliefs, assumptions and values may influence their relationships with women. The concept and visual representation of differential weighting could be used to facilitate understanding of the dynamic complexities of the woman-midwife relationship and the challenges that authentic partnership working present. By visually highlighting potential differences in healthcare, cultural and social priorities, as well as differences in the expectations and experiences of midwives and the women they care for, potential misalignments in priorities and/or knowledge could be identified and openly addressed, rather than both parties operating on the basis of differing and unstated expectations and norms.
For example, while navigation of the healthcare system (i.e., attending all appointments and arriving on time) was emphasized by midwives as a key factor in their relationships with women, this was not a common theme in conversations with women. By using the model to identify that women may be unaware that they are not navigating the NHS system 'effectively', and that this is causing their midwife to view their relationship negatively, it may be possible to address and align expectations and improve relationships. Improvement of these relationships may, in turn, promote more of a partnership approach to maternity care.
Although the model presented in this paper was developed from a study of migrant women, 22  with women may be negatively impacted, we hope that practitioners may take steps to avoid this happening, therefore improving relationships, increasing understanding and ultimately contributing to a reduction in inequalities.
In addition to self-reflection, we propose using the model's diagrammatic representation to facilitate discussions between women and midwives on these issues. The grid format provides a simple comparison of options allowing service users to work through key questions and answers, comparing their options, highlighting which issues matter most to them and discussing these key questions in more detail with their healthcare provider. Using our model, new option grids could be developed to facilitate discussions between midwives and BME women about traditional cultural practices during pregnancy, providing evidence for the safety of these practices alongside women's beliefs and thoughts regarding the cultural value of these traditions. The popularity of option grids appears to be growing, 39 as they demonstrate respect for service users' views, while also providing healthcare providers with an objective, nonconfrontational way of approaching and recording potential safety concerns. Documented use of such grids can also provide evidence of midwives' information-sharing and so may reduce anxieties around professional accountability. 39

| Implications for education
There is growing emphasis on the critical role played by relational

| Strengths and limitations
This study builds on previous models of healthcare relationships to develop a robust evidence-based tool to support practice. The model was derived from data collected from observations and interviews with Pakistani women with recent migrant status, living in a particular South Wales, 22 and has yet to be tested with other groups. However, the principles enshrined can be applied to other contexts and situations, and as such, this model has the potential for transferability to other settings where similar inequities exist. For example, situations where inequalities, discrimination and/or stereotyping impact on woman-midwife relationships (i.e., social class or urban/rural divides).
Further testing and refinement of the model would be needed to test its utility in diverse settings and to indicate how services could be adapted to enhance a partnership approach to care and decision-making. This model, along with future work to further test and refine it in other contexts, has the potential to address certain aspects of inequity by facilitating future discussion on cultural issues, encouraging collaborative learning and knowledge production and providing a framework for future global midwifery practice, education and research that has equitable partnerships at its heart.