Scholarly article on topic 'Conceptualising the prevention of adverse obstetric outcomes among immigrants using the ‘three delays’ framework in a high-income context'

Conceptualising the prevention of adverse obstetric outcomes among immigrants using the ‘three delays’ framework in a high-income context Academic research paper on "Sociology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Social Science & Medicine
OECD Field of science
Keywords
{"Socio-cultural factors" / "Maternal care encounter" / "Immigrant women" / "Health communication" / "Obstetric intervention" / Ethnicity / "Migrant-friendly hospitals" / "Maternity guidelines"}

Abstract of research paper on Sociology, author of scientific article — Pauline Binder, Sara Johnsdotter, Birgitta Essén

Abstract Women from high-mortality settings in sub-Saharan Africa can remain at risk for adverse maternal outcomes even after migrating to low-mortality settings. To conceptualise underlying socio-cultural factors, we assume a ‘maternal migration effect’ as pre-migration influences on pregnant women’s post-migration care-seeking and consistent utilisation of available care. We apply the ‘three delays’ framework, developed for low-income African contexts, to a high-income western scenario, and aim to identify delay-causing influences on the pathway to optimal facility treatment. We also compare factors influencing the expectations of women and maternal health providers during care encounters. In 2005–2006, we interviewed 54 immigrant African women and 62 maternal providers in greater London, United Kingdom. Participants were recruited by snowball and purposive sampling. We used a hermeneutic, naturalistic study design to create a qualitative proxy for medical anthropology. Data were triangulated to the framework and to the national health system maternity care guidelines. This maintained the original three phases of (1) care-seeking, (2) facility accessibility, and (3) receipt of optimal care, but modified the framework for a migration context. Delays to reciprocal care encounters in Phase 3 result from Phase 1 factors of ‘broken trust, which can be mutually held between women and providers. An additional factor is women’s ‘negative responses to future care’, which include rationalisations made during non-emergency situations about future late-booking, low-adherence or refusal of treatment. The greatest potential for delay was found during the care encounter, suggesting that perceived Phase 1 factors have stronger influence on Phase 3 than in the original framework. Phase 2 ‘language discordance’ can lead to a ‘reliance on interpreter service’, which can cause delays in Phase 3, when ‘reciprocal incongruent language ability’ is worsened by suboptimal interpreter systems. ‘Non-reciprocating care conceptualisations’, ‘limited system-level care guidelines’, and ‘low staff levels’ can additionally delay timely care in Phase 3.

Academic research paper on topic "Conceptualising the prevention of adverse obstetric outcomes among immigrants using the ‘three delays’ framework in a high-income context"

SOCIAL SCIENCE

MEDICINi;

Contents lists available at SciVerse ScienceDirect

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

Conceptualising the prevention of adverse obstetric outcomes among immigrants using the 'three delays' framework in a high-income context

Pauline Binder a' *, Sara Johnsdotter b, Birgitta Essen a

aDepartment of Women's and Children's Health (IMCH), Uppsala University Hospital, 75185 Uppsala, Sweden b Faculty of Health and Society, Malmö University Hospital, Malmö, Sweden

ARTICLE INFO

Article history:

Available online 21 August 2012 Keywords:

Socio-cultural factors Maternal care encounter Immigrant women Health communication Obstetric intervention Ethnicity

Migrant-friendly hospitals Maternity guidelines

ABSTRACT

Women from high-mortality settings in sub-Saharan Africa can remain at risk for adverse maternal outcomes even after migrating to low-mortality settings. To conceptualise underlying socio-cultural factors, we assume a 'maternal migration effect' as pre-migration influences on pregnant women's post-migration care-seeking and consistent utilisation of available care. We apply the 'three delays' framework, developed for low-income African contexts, to a high-income western scenario, and aim to identify delay-causing influences on the pathway to optimal facility treatment. We also compare factors influencing the expectations of women and maternal health providers during care encounters. In 2005 —2006, we interviewed 54 immigrant African women and 62 maternal providers in greater London, United Kingdom. Participants were recruited by snowball and purposive sampling. We used a herme-neutic, naturalistic study design to create a qualitative proxy for medical anthropology. Data were triangulated to the framework and to the national health system maternity care guidelines. This maintained the original three phases of (1) care-seeking, (2) facility accessibility, and (3) receipt of optimal care, but modified the framework for a migration context. Delays to reciprocal care encounters in Phase 3 result from Phase 1 factors of 'broken trust, which can be mutually held between women and providers. An additional factor is women's 'negative responses to future care', which include rationalisations made during non-emergency situations about future late-booking, low-adherence or refusal of treatment. The greatest potential for delay was found during the care encounter, suggesting that perceived Phase 1 factors have stronger influence on Phase 3 than in the original framework. Phase 2 'language discordance' can lead to a 'reliance on interpreter service', which can cause delays in Phase 3, when 'reciprocal incongruent language ability' is worsened by suboptimal interpreter systems. 'Non-reciprocating care conceptualisations', 'limited system-level care guidelines', and 'low staff levels' can additionally delay timely care in Phase 3.

© 2012 Elsevier Ltd. All rights reserved.

Introduction

National enquiries into the occurrence of adverse maternal outcomes have been initiated in many countries to reduce global maternal mortality and morbidity (Wildman, Bouvier-Colle, & MOMS Group, 2004; van Dillen, Mesman, Zwart, Bloemenkamp, & van Roosmalen, 2010). Ultimately, the goal is to improve the health and wellbeing of childbearing women. These management instruments provide a mechanism to measure quality at maternal healthcare facilities, especially for emergency obstetric care (Freedman et al., 2007). While maternal death is generally rare across

* Corresponding author. Tel.: +46 18 6115988. E-mail address: pauline.binder@kbh.uu.se (P. Binder).

0277-9536/$ - see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.socscimed.2012.08.010

Europe, some immigrant women show elevated risk for perinatal death (Essén, Hanson, Ostergren, Lindquist, & Gudmundsson, 2000), severe maternal morbidities (Zwart et al., 2011), and dying from direct childbirth-related causes (Luque Fernández, Cavanillas, & de Mateo, 2009; Philibert, Dene ux-Tharaux, & Bouvier-Colle, 2008; Schuitemaker et al., 1998; Schutte et al., 2010).

The Confidential Enquiry into Maternal and Child Health in the United Kingdom (UK) is designed to identify medical outcomes in relation to socioeconomic influences and is used to evaluate clinical service provision, quality of care, and clinical governance (CMACE, 2011; Lewis, 2007). Many of the maternal deaths represented in these enquiries are women who had migrated from low-income African countries and had experienced difficulties in seeking appropriate care and utilising available maternal health services on a continuous basis. A link was concluded between adverse

Pathway toward facility care (common to both income settings)

Original 'three delays' model identifying socio-cultural factors in low-income setting

Care context:

Rural African setting Homebirth is norm Mortality resulting from non-medical factors such as lack of timely care, which result from delays at Phases I, II, and III

Maternal migration effect?

What influences a woman from a low-income scenario when she seeks or utilizes care in a high-income scenario?

Applied 'three delays' model to identify socio-cultural factors in high-income setting

Care context:

1. Urban western setting

2. Facility birth is norm

3. Mortality or morbidity resulting from which non-medical factors? Lack of timely care? What is influence

of each phase on potential for delay?

Maternal care a matter of course?

What influences maternal providers in a high-income scenario when they provide care to women from a low-income scenario?

Fig. 1. Pathway towards facility care for obstetric emergencies in the low-income, rural and high-income, urban contexts.

outcomes, maternal vulnerability and social exclusion, which supported a partial explanation due to socioeconomic factors. However, limited attention was paid to the socio-cultural factors likely to underlie women's difficulties with care-seeking and consistent utilisation.

Following migration from a high-mortality, low-income country to a low-mortality, high-income country, immigrant women of reproductive age are likely to remain closely influenced by adverse maternal health experiences brought with them from their setting of origin. This assumption is supported by our prior research (Essen, Johnsdotter, et al., 2000; Essen et al., 2002), which we now coin as the maternal migration effect. A definition of migration that supports this assumption is relocation to a setting in stark contrast to the one of origin, involving exposure to unfamiliar social, cultural and economic conditions, where spontaneous assimilation is assumed to be less profound among those migrating as adults (Lindström & Munuz-Franco, 2006; Silverstein, 2005).

The maternal migration effect might be exemplified by women's responses to obstetric interventions, such as caesarean section (CS). Sub-Saharan African women's aversion to CS is due to delivery- or facility-related fears and their increased risk of developing negative attitudes towards future pregnancy after having had a caesarean experience (Collin, Marshall, & Filippi, 2006; Mrisho et al., 2009; Sunday-Adeoye & Kalu, 2011). Studies conducted in western countries among women from comparable settings suggest that, following migration, similar perceptions occur. For example, women from Africa's Horn have described anxiety, fear of dying, or substantial reservations towards CS in Sweden (Essen, Johnsdotter, et al., 2000; Lundberg & Gerezgiher, 2008), the US (Brown, Carroll, Fogarty, & Holt, 2010; Herrel et al., 2004), Norway (Johansen, 2006), and the UK (Essen, Binder, & Johnsdotter, 2011).

Consequently, in both pre- and post-migration contexts, negative maternal responses to CS may explain the low rate of utilisation or the outright refusal of the procedure, despite the indication of severe obstetric complications (Dumont, de Bernis, Bouvier-Colle, Breart, & for the MOMA Study Group, 2001; Essen et al., 2011). Additionally, in low-income settings in Africa, where non-facility childbirth is the traditional norm, utilisation of a maternal care facility depends on prior use for both antenatal care and delivery,

although facility availability is influential (Gabrysch & Campbell, 2009). How women from sub-Saharan settings respond to universal facility care after migrating to another income level is thus an interesting question.

This study uses a holistic approach to explore women's responses towards facility-based maternal care. We consider that the success of system-level initiatives meant to either prevent or avert the occurrence of adverse maternal outcomes should simultaneously prioritise evidentiary insight into the circumstances facing both the pregnant women and the healthcare facility (Maine & Rosenfield, 2001; Thaddeus & Maine, 1994). Our aims are to explore the influence of pre-migration socio-cultural factors on post-migration maternal care-seeking and to elaborate barrier-causing delays to women's utilisation of facility care. Furthermore, we want to understand the reasons behind barriers, if any exist, between immigrant women and maternal care providers during the care encounter. The purpose for identifying these is to inform future audits of maternal or perinatal death with a migration perspective, so that need-based guidelines can be further contextualised for both groups.

One objective is to conceptualise maternal access in relation to women's recalled actions, perceptions and attitudes. These, we anticipate, form the basis of women's maternal care strategies during pregnancy and childbirth. If the maternal migration effect is to demonstrate its influence on women, we assume that it will be exhibited through their care strategies. The additional objectives are to explore maternal care provider experiences within their high-income context, and to compare these to women's experiences during the shared clinical encounter. We thus further assume that the healthcare system standards for medical training and guidelines are personified by maternal care providers, and expressed through their competence as a matter of course (Healthcare Commission, 2005).

The three delays of obstetric care-seeking

To model women's pre-migration socio-cultural experiences, we apply the 'three delays' framework (Thaddeus & Maine, 1994). This framework was initially developed in response to the low intrinsic value of women's wellbeing represented in contemporary

maternal and child health initiatives. Its authors aimed to better understand the challenges faced by maternal mortality intervention initiatives in low-income, high-mortality African settings. The social determinants identified in the framework, including both socioeconomic and socio-cultural factors, were supported by a literature review that encompassed a wider base of evidence than isolated medical factors. The three phases are posited by the authors as viable across all income contexts. However, to our knowledge, applicability to the high-income setting remains hypothetical.

The original three delays framework assumes a lack of timely and adequate care as the foundation of maternal death, but it does not generalise the problem by blaming women for delays in seeking care. It focused on three points (i.e., phases) of potential delay for the timeframe between a woman's first suspicion of an obstetric problem and its outcome. The chronological order is emphasised: the decision to seek care (Phase 1), where delays mainly result from either perceived or actual barriers that create disincentives to act; the infrastructure involved in reaching a medical facility (Phase 2), where delays can result from the actual barriers of cost, and transportation in the form of adequate ambulance and road systems; and finally, the receipt of appropriate and adequate treatment (Phase 3), where delays result from actual barriers at the care facility, such as lack of skilled birth attendants, technological equipment and medical supplies.

Avoiding delays relies on overcoming both perceived and actual barriers. Disincentives in Phase 1 might result from having to negotiate with a partner involved in decision-making, or from a woman's low social status. These may influence her ability to judge the severity of a complication in relation to whether an appropriate care facility is accessible. Perceived barriers from negative expectations rely on a woman's prior experience or those of others close to her. Actual barriers in the African setting are obvious for each phase: 1) the local economic environment can hinder a woman's ability to act; 2) long geographic distances and poor infrastructure make it improbable to reach a health facility; and 3) resources for optimal care may be limited or non-existent.

Fig. 1 illustrates the original three delays framework and our proposed interest in it. The need to gain timely access to facility-based care is represented by the central arrow as implicit in the first and second phases. Overcoming an adverse complication therefore presumes a facility-based solution, which appears to be the component referred by Thaddeus and Maine (1994) as the most applicable across income settings. In the original, low-income scenario, access to a healthcare facility means overcoming delays to timely care for an obstetric emergency, within a rural setting where homebirth is the norm. We conceptualise overcoming a complication in an urban scenario, where facility-based care and childbirth are universal norms, and where the type of facility sought depends upon the type of maternal care required (emergency or non-emergency).

Methods

The high-income setting and motivation for recruitment

This study design is contextualised for immigrant women who received maternal care from the UK National Health Service (NHS), London, UK, before 2005. The NHS offered socialised access to newly centralised maternity services with well-equipped facilities at the time of this study. Service provision was also becoming increasingly reliant on evidence-based healthcare, intended to standardise care providers' level of expertise (Healthcare Commission, 2005). During the year of data collection, this setting had the highest proportion of births to first-generation

immigrant women in Europe (21%; Sobotka, 2008). The Confidential Enquiry into Maternal and Child Health for 2003-2005 identified a maternal death rate of 13.95 per 100 000 maternities (inclusive of direct and indirect deaths), with the highest risk among ethnic 'Black African' immigrant women (Lewis, 2007). This rate was six times the number of maternal deaths for the overall population, and the majority of these occurred among women from Somalia and Eritrea.

Recruitment and data collection

The study was conducted in greater London between 2005 and 2006 and was approved by the Riverside Research Ethics Committee of the NHS, London (reference 06/Q1401/15). Participants were recruited for individual or focus group interviews by either the snowball sampling or purposive sampling technique (Bernard, 2006; Binder, Borné, Johnsdotter, & Essén, 2012).

Recruitment resulted in a sample of 54 immigrant women from sub-Saharan regions in Africa and 62 NHS maternal care providers. The women represented low-income settings (Somalia, Ghana, Nigeria, Senegal, and Eritrea), according to the World Bank GDP Index for 2005 (World Bank, 2005). Ethnicity was self-defined according to country of birth. Age range for women was 18-48 years, and time spent in the UK was 1 and 20 years. Range of parity was 1-10 children. Inclusion criteria for women was "currently pregnant" or "having had at least one child within the NHS system", and residence within the study area at the time of data collection. Maternal care providers were doctors or midwives at five hospitals within the study area. Inclusion was based on years of experience (>5) in caring for women of British and non-British ethnic backgrounds. Care providers represented multiple ethnic profiles (4 Somali, 34 other sub-Saharan or Caribbean, 21 white British, and 3 Asian), as self-defined by country of birth.

Each tape-recorded interview took 30-90 min Box 1 contains the initial semi-structured, open-ended questions that were asked by an obstetrician (BE), and sometimes accompanied by an anthropologist (SJ). Language-specific interpreters were used as necessary.

Analysis

Data collection, analysis and interpretation of findings supported a constructivist hermeneutic, naturalistic study context and

Open-ended interview questions. Women:

How have you experienced maternal healthcare within the British system?

Which procedures have you experienced?

If you needed help understanding your care provider, how did you manage?

If you have given birth in your home country, what medical procedures were available to you?

How did you manage if you had a problem during the pregnancy or birth?

Regarding your needs for care, how do you know what to do when you are pregnant? Care providers:

What are your experiences with providing maternal care to immigrant and non-immigrant women?

How and in what ways do immigrant women respond to the care that you provide?

What is your experience with the use of medical interpreters?

What do you think about immigrant women's personal care-seeking behaviours?

How do you define challenges to care and how do you manage if you are faced with them?

In what ways do you use maternal care guidelines in your practice if you encounter a problem?

emergent design (Lincoln & Guba, 1985) as a proxy for medical anthropology. All analyses were conducted by PB. Interviews were transcribed in English. The unit of analysis was comprised of answers to the above open-ended questions along with the follow up probing. During the primary analysis, conceptual intuitions were identified from the text data by multiple readings and constant comparison of informant responses across the interviews. This process of developing intuitions forms the basis of our etic insight into the informants' emic perceptions, beliefs, and attitudes, as used in anthropology and elsewhere (Essen et al., 2011; Harris, 2001). The secondary analysis consisted of a triangulation of the findings: from the primary analysis, the intuitions from the women were applied to the original three delays framework, and the intuitions from the providers were applied to relevant aspects of the NHS context. That latter information was drawn from Lewis (2007), CMACE (2011), and the guidelines for 'Maternity Matters' (Department of Health, 2007). This triangulation allowed for the generation of explanatory models, as guided by Kleinman (1980), about the reasons for delay experienced during care-seeking and the maternal care encounter.

Findings

The path to facility care: deciding to seek care

If both women and providers articulated care experiences as satisfying, then trust was intuited as implicit and most likely to facilitate optimal care-seeking. This occurred for the majority of the women and providers. However, "outliers" were evident: factors of broken trust were interpreted as mistrust, distrust, and feigned trust. Box 2 provides a sample of text used to interpret broken trust.

Mistrust reinforced negative expectations of care quality or the ability to provide quality care. These included women's lack of confidence in certain hospital policies, procedures, or providers, as well as care providers' negative responses, e.g., frustrations, over unmet expectations. Mistrust manifested early in the pregnancy, during non-emergency care-seeking. As early as the first ultrasound scan, an active choice was made by some women not to adhere to a provider's recommendations, e.g., for amniocentesis, or not to use a certain clinical facility, even if located close to home. Either these women attended antenatal care at a cross town clinic or they travelled great distances for a second opinion from a provider with 'an acceptable reputation' (as far away as Germany, Italy, Luxembourg, or Sweden).

When negative expectations combined with misconceptions about care procedures, then distrust became manifest as low-adherence, delayed care-seeking, late-booking, or outright refusal of preventive interventions. As an extension of mistrust, attitudes of distrust became evident primarily during emergency care-seeking among those women who articulated decisions to refuse care outright, e.g., for CS or induction of labour. Such attitudes were also tied to suspicions about whether certain interventions were actually necessary or about whether the care provider or NHS intended purposeful malpractice, i.e., being told the pregnancy was abnormal when it resulted in a healthy baby.

Women's adherence to advice or recommended protocols was not perceived by providers as a guarantee of sincere trust or autonomous care-seeking. The concept of feigned trust is implied: "In our culture you don't really ask 'why?' Also, if you go to a hospital in Ghana, you don't ask questions as a patient because the doctor always knows best... it is just to go ahead and do whatever, to accept whatever ... No questions asked." (Other sub-Saharan provider, doctor 1) Similarly, "Particularly in Ghana, for example, women have been very used to the idea of doctors." (Other sub-Saharan provider, doctor 2) One consultant having a different ethnic profile than

Ghanaian stated: "And some African ladies sometimes seem to be so disinterested. I don't think they are disinterested. Some Ghanaian women, they just come to antenatal clinic and you can't even hear them talking . they are whispering." (White British doctor 1)

The path to facility care: identifying and reaching the medical facility

All women had appropriate knowledge about the NHS emergency call number, expressed as "Easily dialling the number '999'". (Somali woman 1, seven years in UK, four children) Only one woman described minor delays due to transport concerns: "My husband and I were waiting to see if the pain got worse. We did not want to cause trouble in the system." (Somali woman 24, two years in UK, three children) No women described being turned away due to lack of available space, and one provider specifically stated, "We don't turn people away, ever ... It is very open access." (White British doctor 8) Factors of communication, such as limited ability to articulate either care needs or care advice, was interpreted as the main barrier to access. One woman prioritised her language needs, depending on which facility she was visiting, "When I visit the GP, I do not need an interpreter. Here in hospital, yes." (Senegalese woman 1, one year in UK, one child) Overall, Somali women had the most difficulties with language and articulated concerns about becoming a problem to the provider, "Everyone will be fed up with you if you can't understand what they are saying — if you can't talk to them ... If you can't speak you will be ignored." (Somali woman 17,16 years in the UK, two children)

A few providers found telephone interpretation easy to use. However, the interpretation service was mostly described as complicated and either unavailable or unutilised. "It is very good if you have an interpreter but... we haven't been fully geared towards that." (Other sub-Saharan provider, doctor 3) Another provider remarked: "During labour? In cases where the person speaks a little English, we don't call for interpretation. But when the person doesn't speak at all, that is normally where the interpreters come in." (Other sub-Saharan provider, midwife 15) Still another commented: "Okay, so they come to their first pregnancy with an interpreter . but by their second pregnancy, perhaps they should have learned English. But we will tell them to use an interpreter because it is a good practice not to use a relative." (White British midwife 1)

Sometimes the limited insight of women into NHS procedures was attributed by providers as an arbitrary cultural factor or to the impossibility of the situation. "There are a lot of problems with cultural things because they cannot really understand or appraise what needs to be done or is being done. I think some of them don't give the consent because they don't understand the information." (Other sub-Saharan provider, midwife 18) In reference to ultrasound scanning:

It must be a misunderstanding. The women interpret our explanations as an X or Y abnormality. When they are explained about screening, maybe they feel that they are actually being told that there is an abnormality — rather than that there is a risk of an abnormality. If we have lots of women having scans, who think they are being told that something is wrong and then having normal babies . there must be a problem with communication. (White British doctor 4)

End of the path: receiving adequate and appropriate facility-based maternal care

In contrast to the access difficulties described for African women in the original framework, all the women in our study reached a care facility, although some were late. Additionally, no informants indicated delays resulting from lack of essential medical supplies.

Factors of trust leading to mutual misconceptions about care or type of care.

Immigrant women's articulations that diverge from a trust-based experience

Level of broken trust

Care providers' articulations that diverge from a trust-based experience

I thought maybe ... they would take me to the operating Mistrust

theatre. I've also heard that epidural can easily go to the spine ... I thought maybe I can get a doctor who isn't really experienced and he can do some stuff that I can regret. (Somali woman 8, over 20 years in UK, three children) When I had fainted they ... gave me medicine. That medicine helped me, but after all that I went abroad, to Germany, to see another specialist. (Somali woman 3, 10 years in the UK, 10 children) I am the kind of person who doesn't want to take any injection, any epidural or gas or anything. I preferred to be in control of my body because of what I had heard from other women. Some of them have back problems . some of them, when they had the gas, they didn't know what they were doing. I've also heard that epidural can easily go to the spine. You know all this sort of stuff that women talk about. (Somali woman 16, 20 years in UK, three children) If I had family here that would have been different. ... Because this person would have been in this situation and they would know exactly what to expect or what to do. ... As for maybe doctors and midwives, they are talking of things that they have probably heard or read somewhere, you know, as in learning wise. Between them ... I would probably trust the person who had been in the situation before, to be honest. (Nigerian woman 2, five years in UK, one child) The antenatal doctors who told me that my baby was abnormal and Distrust

the doctors who actually delivered the baby were different. But . the doctor who delivered said ... "He is not abnormal". ... I'm afraid that these earlier doctors during the scans were telling a lie. For the birth, I thought that when they gave me anaesthesia and I became unconscious, maybe they can now change the baby ... I suspected they would hand me a handicapped baby. . Half the Somalis have the same problem. (Somali woman 21, 10 years in the UK, 10 children).

In Somalia ... you've got doctors saying, "I'm a doctor" but they are not really. You get so desperate that you want to trust anyone. Here, in London, there is no trust either because when you are giving birth, they do not respect you enough, the majority of them. They are not very

polite and not very respectful. They just want to do their job, get in there and get out. They do not care too much for person in needs, and they don't care too much for comforting. (Somali woman 38, one year in UK, five children). I told them that I still wanted to deliver naturally [vaginally]. In Somalia, they do that even if it is a breech ... (Somali woman 9, eight years in UK, six children)

I did not listen to the doctors or midwife ... I refused and just kept pushing and pushing. (Somali woman 5, five years in UK, three children)

[Somali clients] may also have a completely different attitude and approach to illness or antenatal care. So, what that means in practice is that the way we are set-up is such that there is a small section of the population that do not necessarily access the facilities as they are currently configured. I come from a developing country myself and quite often people's first response to be ill is not to go to the hospital until it is so obviously abnormal or they are so unwell that they can't go around on daily business.... People from this part of the world do not recognise illness for what it is. And when they do, they don't do it early enough and even when they recognise it at the end of the day they are not necessarily going to seek . you know, health facilities. (Other sub-Saharan provider, doctor 4)

[Western] doctors do not appreciate how much information the Somali patient needs; or the other side, which is that Somalis distrust authority from the beginning. (Somali doctor 2) Somalis have never had a health system they can rely on, and they still have that mentality. Many times, women are simply not satisfied with answers they get from a doctor. Especially when they are used to delivering vaginally, the question becomes 'Why are these people operating so much?' It is not as simple as trying to unravel and figure out why; there are many reasons. (Somali doctor 1) ... the Somali women, they come with such different social circumstances, and especially about caesarean section, there is distance because we assume they understand their fears in the same way that we do, in actual terms. (White British midwife 8) [Her refusal of emergency caesarean section] was very difficult. I get really cross about how selfish Somali behaviour is. She hadn't any idea what effect she had on us. (Asian doctor 1) ... and maybe some Somali women do go late there, you know ... if the woman came late, late in labour to the hospital and they don't have her background. . Some women go from one hospital to another hospital... they are booked for example at [one hospital] and they go and have their baby somewhere else. I know about women who do it on purpose to go late because they are afraid of having caesarean section . so they say let the labour progress itself and they go late. .It is very difficult. (Other sub-Saharan provider, doctor 14) I guess a lot of women might not even have an ultrasound, if they don't book or they book late. They won't have the same degree of care, will they? I have never really thought about it... for them. Possibly in their own country they wouldn't have the scan. And that's probably why they don't book as early as we would . we would book about eight to twelve weeks but they would not come until about twenty weeks. So, they would miss out on some care and the treatment wouldn't be equitable. They would miss out on certain treatment, but I am not sure if they value it the same. (White British midwife 6)

Instead, several other factors were identified: ability to manage reciprocal trust (influence from Phase 1) and congruent communication (influence from Phase 2) during the care encounter, as well as consistent availability of clinical care guidelines, congruent conceptualisations of care, and adequately staffed care facilities. Box 3 provides the most salient examples from the data.

Discussion

Our main findings are illustrated in Fig. 2. Point A contextualises the Thaddeus and Maine approach. It is a given constant in their low-income scenario that the potential for delayed decision-making begins in Phase 1, when a pregnant woman in a rural environment, who is anticipating a homebirth, perceives disincentives to seeking facility-based care for an obstetric problem. Delays contributing to her decision-making are also significantly influenced by her perceptions about actual barriers of infrastructure in Phase 2 and of care quality in Phase 3 (identified by blue arrows).

Fig. 2, point B, represents factors likely to influence the receipt of emergency and non-emergency care in our urban, high-income setting, where facility-based care is the norm. The given constant is the potential for delays beginning in Phase 3. When a care-seeking pregnant woman enters the scenario during the care encounter, her actual receipt of optimal treatment is critically influenced by her perceptions about decision-making in Phase 1 and infrastructure in Phase 2 (indicated by solid red arrows). Actual accessibility delays (e.g., those related to infrastructure and transportation) at Phase 2 are greatly minimised in this context (indicated by dashed red arrow).

Governance of clinical service provision: a meeting of care strategies

Because the NHS offered socialised access to well-equipped facilities at the time of this study (Healthcare Commission, 2005), we have assumed an emphasis on care provider expertise and their expectations during the care encounter. The health system

Factors in Phase 3 leading to delayed reciprocity in the care encounter.

Phase 3 factors most likely to Evidence to support constructions of the Phase 3 factors create delay-causing barriers

Reciprocal trust One consultant called another consultant. The other consultant called another consultant. ... They were going to inject me

on my stomach so they could check if my baby has a problem. When they told me, I said 'No, you cannot do that, because what happens if my baby has that problem? Would my baby still be treated when it's still in my womb?' Then, they all said 'No'. I replied that I would rather wait until I give birth to my baby instead of worrying myself too much. (Somali woman 13, seven years in UK, one child)

And even if it's her third or fourth baby ... if we say: "You should be induced" she just vanishes the whole time and the midwives go and find her and she says she is coming the next day. But we can't force them in. ... we know that our duties end with making sure that they fully understand. ... You keep talking and keep talking and ... we see them discharging themselves from hospital with preeclampsia. (White British doctor 9) Congruent communication We had a baby who died quite soon after birth and an investigation into that found out more information from the family and

the coroner. ... These discussions brought up things about her past that we have not been told as professionals. That obviously affected the way that she felt and acted and things. But we only know about those things if she chooses to communicate them. (White British midwife 16)

We could have easily prevented this three hour delay . the woman arrived late and then she was not able to understand what we were talking about, and by the time we called for an interpreter it was too late. (Asian doctor 2)

Interpreters were on hand, but my children speak very well English so they used to translate for me. (Somali woman 39, nine years in UK, two children)

Consistency of care guidelines Oh yes, the guidelines on consent are based on the experiences that women refuse treatment, but no guidelines [exist] about

how to manage her refusal. . It is very difficult for doctors to stand by while the baby dies but you can't do anything about it if the woman is competent. (White British doctor 11).

I am quite lucky in the training that I have had by virtue of where I have worked ... you know I have just sort of picked up a lot... by having to look [clinically] after these patients and finding out the dos and don'ts. But I didn't have any sort of formal training or any 'How to' description, no. (White British doctor 8) Congruent care I warn everyone that they will tell you your baby has no head when you have that scan. And, later, if it takes too long, they will

conceptualisations just operate. No questions or anything. (Somali woman 22, 14 years in UK, three children)

The health care is different than in our country. . But to have the baby, that's not . I mean, you have to go to the places where there is a lot of people and you have a baby. You don't have a kitchen where you can boil the water . you have to stay three, four people in one room. (Eritrean woman 1, one year in UK, one child)

Even in my own country . we are exposed to doctors from other cultures and of different sex. I don't see this as any problem. As far as I am concerned, if you know your job and you are very professional about it, that's what matters. (Nigerian woman 1, eight years in UK, three children).

I am female, so it is not a problem. I have been asked to cover for male colleagues, but I do try to be clear with patients that their

health is a greater priority. . Making that point is not always successful, but I feel it my duty to point it out. (White British doctor 7)

What's more is that, these women haven't conformed because they haven't attended their antenatal appointments and now they are ill and we have to look after them. . But if they would have done what we said then everything would be okay. I mean,

my role is that I actually look after a number of immigrant women who actually don't always conform to how we think they should. (White British midwife 5)

I think that people sometimes come from countries where there is no such thing as a GP so they have no idea of such a service existing. . Actually, if it ends up in tragedy because they refuse to take the care offered . we struggle against their God. (White British doctor 10)

Adequately staffed facilities Quite a lot of times we don't have enough midwives so that women can choose. ... And in the general labour ward we are

under so much pressure anyway that women's choices are just kind of pushed into the background, really. (White British midwife 2).

The actual care that you gave in the past was midwifery, now you are becoming a social worker, their auntie, their uncle, their mother, you know, everything, because their support network has broken down as a whole. (White British midwife 4)

standards for medical training and care guidelines are therefore embodied as provider strategies for service provision as a matter of course. A contemporary problem of limited availability of maternal care providers was identified in the literature as resulting from centralisation of maternity services, and some authors suggest that this aspect has severely restricted women's choice and free access to different levels of care (Baker, Choi, Henshaw, & Tree, 2005). A number of our informant providers expressed their concerns about heavier workloads or about having to respond to incompatible care protocols, e.g., guidelines that exist for obtaining consent from a woman who refuses treatment while missing are those about what to do when she refuses. The latter aspect has been reported in the literature (Essen et al., 2011). Incompatible care conceptualisations, which can result in unmet expectations for service provision and might be reflective of women's poor health literacy, have also been explored, e.g., Vangen, Johansen, Sundby, Traeen, and Stray-Pedersen (2004), Carroll et al. (2007), and Ng and Newbold (2011).

Some of the women booked late for antenatal care. All women who expressed concerns about a sudden change in their pregnancy

also sought emergency treatment, and while most described attending to their concerns immediately, some described waiting. Once at the facility, however, some of these avoided or refused treatment irrespective of their timely care-seeking. We interpret this choice to be the result of emergency-like responses (e.g. heightened negative reactions) to non-emergency care. This included heightened reactions towards routine ultrasound scans and screening procedures, and also appeared to strongly influence the women's rationalisations about later care choices. Such perceptions were articulated mostly among the Somali women in our study, but this finding has negative implications for all women who present late with symptoms for premature rupture of membranes, preeclampsia, haemorrhage or sepsis. The original framework authors discussed the influence of women's fears towards labour interventions as they relate to care-seeking, but not specifically towards non-emergency antenatal care. With facility-driven outcomes as the foundation of our study, this finding expands the original framework to include women's 'negative responses to future care-seeking' — initiated early during routine

Factors influencing care-seeking and utilisation of facility-based care and obstetric

outcome

Phases of Delay

Phase I: Deciding to Seek Care

A. Low-income, rural context

Original 'three delays' framework

Perceived socioeconomic/cultural factors

Recognition of illness severity ■ Women's economic and education status Consideration of distance, transportation, cost

Previous satisfaction with service provider

Phases of Delay

Phase I: Deciding to Seek Care

B. High-income, urban context

Migration 'three delays' framework

Perceived socioeconomic/cultural factors

Presence of broken trust: mistrust, distrust, feigned trust

Negative responses to future care seeking (rationalisations for late-booking, low adherence / refusal of treatment)

Phase II: Identifying and

Reaching Medical Facility

Actual accessibility of facilities factors

J Distribution of health facilities Distance and travel time to reach facility Transportation Cost versus ability to pay

Phase II: Identifying and

Reaching Medical Facility

Actual accessibility of facilities factors

Mutual language discordance Reliance on interpreter service Limited relative barriers to facility distribution Limited barriers from cost, distance, time

Phase III: Receiving Adequate and Appropriate Treatment

Actual quality of care factors

Poorly staffed or unskilled facilities Poorly equipped facilities for blood, equipment or pharmaceuticals Inadequate management of diagnosis and action

Phase III: Receiving Adequate and Appropriate Treatment

Actual quality of care factors

Reciprocal broken trust during care encounter Reciprocal incongruent language ability Inconsistent clinical care guidelines Misconceptions about reciprocal care Poorly staffed facilities Lack of clinical guidelines

Fig. 2. Comparison of factors influencing women's care-seeking and utilisation in two income contexts. Arrows indicate influence of factors on delays in a specific phase. Broken arrow indicates reduced influence.

antenatal care — as something that may prevent a woman from taking action in the future.

Gabrysch and Campbell (2009: pp. 5—7) identified factors underlying care-seeking of skilled, i.e., facility, delivery services for emergency (complicated) and preventive care (uncomplicated) childbirth in low- and middle-income settings. Among the multiple factors identified in their literature review, those relevant to our findings include 'perceived quality of care', 'ANC use', and 'previous facility delivery'. It is likely that our informants were susceptible to decision-making guided by perceptions that relied upon "culture and fear of procedures". However, for our high-income setting, we were not able to glean that "familiarity with services" increased their use. It is potentially interesting that this factor is apparently influential in low-income settings, when women can choose to adopt facility care over traditional homebirth while still in their setting of origin, but appears to be less influential after a woman has migrated to a high-income setting. We suggest that this maternal migration effect creates a scenario where women rationalise choices based on prior non-existent or limited experience with facility-based care, and that, when they are situated in a migration context they are more likely to exhibit slow acculturation. Implicit within this suggestion is Gabrysch and Campbell's factor, 'ethnicity, religion, traditional beliefs', which provides rationale for skilled service use based on "certain cultural backgrounds, beliefs, norms and values [that] may decrease care-seeking". Our findings, however, are inconsistent with their factor 'complications', i.e. previous exposure to complications leads to women's increased awareness of risks and "should increase use of skilled attendance". Among our informants, purposeful use of facility care was evident, but the extent to which women are aware of certain medical risks requires further study.

Governance of care quality: optimal reciprocity

Non-reciprocating trust during the care encounter in Phase 3 existed because, in Phase 1, broken trust was mutually held by both the women and the providers. During emergency encounters,

women's distrust was the strongest influential factor creating actual treatment delays, which resulted in provider frustration over the unmet treatment expectations. In non-emergency encounters, all forms of broken trust were influential, but especially distrust for its potential to create future delays during emergency treatment. Mutual broken trust therefore has potential to delay the conveyance of preventive maternity advice, the timely detection of prior nutritional deficiencies or co-morbid conditions, and timely referral.

Trust is identified in the literature as universal when it is reciprocated by the care provider and the pregnant woman. Similarly, reciprocity occurs when the provision of trustworthy care is synonymous with women's care-seeking strategies (which satisfy a woman's own obstetric goals) and when she adheres to the provider's recommendations (Tanassi, 2004). Lack of trust among immigrant Somali women in relation to acute and non-acute labour interventions has been described for other migration contexts in the US and UK (Carroll et al., 2007; Herrel et al., 2004; Straus, McEwen, & Hussein, 2009). Low levels of mutually-held trust can also magnify divergent conceptualisations about emergency labour care, especially as they relate to emergency CS, and can heighten vulnerability for women if they refuse treatment and for providers if they are not properly supported by health system care guidelines (Essen et al., 2011).

We have assumed at the outset of this study that, in a migration context, women's access to a health facility is situated in her having access to a care provider's competence. Most women placed high importance on provider competence, which clashed with some providers' presumed reduction of women's preferences for gender-congruent care. We have elaborated this finding elsewhere (Binder et al., 2012). In our modified framework, access to a competent care encounter is reliant upon mutually-proficient language capacity in Phase 2. If the encounter in Phase 3 is dependent on a suboptimal or suboptimally utilised interpreter service, then the likelihood for treatment delays for maternal and foetal complications is especially high. A poorly operating interpretation service obviously denies a reciprocal articulation of care needs and demands. The providers are likely to express frustration (as in Kravitz, Helms, Azari,

Antonius, & Melnikow, 2000) when incongruent language creates life-threatening time delays during emergency treatment. Similarly, women who are reserved about using unskilled language assistance are still likely to prefer a familiar interpreter. The latter is supported by the literature (Davies & Bath, 2001; Herrel et al., 2004).

Delayed provision of service in response to unarticulated needs during a care encounter can support the maternal migration effect if women are inexperienced with this communication skill. In the US, problems of communication and literacy acted as barriers to healthcare access for a specific clan of Somali refugees (Carroll et al., 2007), and when combined with mutual unmet expectations and care conceptualisations, care quality was perceived as low (Pavlish, Noor, & Brandt, 2010). More generally among refugee and non-refugee African groups, communication barriers exhibited during emergency treatment have been shown to increase the risk for maternal and perinatal mortality (Essén et al., 2002; Gagnon, Zimbeck, Zeitlin, & ROAM Collaboration, 2009; Jonkers, Richters, Zwart, Ory, & van Roosmalen, 2011).

Several studies have shown the need for appropriate interpreting services when caring for non-western immigrants (Harper Bullman & McCourt, 2002; Carroll et al., 2007). The use of informal or personally-known interpreters is now strongly discouraged for UK maternal care (CMACE, 2011). In fact, skilled interpreter use is currently ranked as one of the top ten recommendations for optimal service provision in response to the increased number of maternal deaths linked to suboptimal levels of provider-client communication. However, the current care guidelines for medical interpretation services have yet to satisfactorily address how to overcome such limitations as provider reservations about placing blind trust in an interpreter's level of skill, even if they are medically trained, or about women's strong distrust towards using a stranger for this purpose (Binder et al., 2012; Davies & Bath, 2001).

By our choice of quoted material, we illustrate that women who have lived for many years in the UK continue to uphold perspectives about maternal care that may conflict with the NHS preventive care routines. These women remained hesitant about using spinal analgesia and about trusting the information and advice from providers. They were also willing to go outside the UK for a second opinion. However, potential susceptibility for social vulnerability, slow acculturation or dissociation from the NHS requires further exploration. Our findings suggest a need for identifying socio-cultural risk factors, in addition to socioeconomic influences, especially because incongruent care conceptualisations appear to exist and some negative attitudes are in such stark contrast to the clinical guidance of care providers.

Potential limitations

Our use of a diverse sample population coupled with a study design that approximates medical anthropology avoids the limitations which usually confound qualitative analyses that miss 'how' and 'why' differences (Daly et al., 2007). The design is purposefully triangulated to the setting context and original three delays framework to lay the foundation for highlighting a migration perspective in clinical audits. The hermeneutic process is designed for topical saturation. Our assumptions were also identified at the beginning of the study and addressed within the findings and discussion for internal consistency. We are not able to transfer findings beyond the UK to high-income medical systems that are inconsistent with socialised care. However, the factors identified here provide a platform for coherence to similar settings. To avoid potential bias created by snowball sampling, we ensured that information relayed about informant experiences was understood from the individual's point of view, whereas second-hand or

descriptive information about others' experiences was kept distinct (Lambert & McKevitt, 2002). Similarly, to avoid the risk of lost or misinterpreted meaning by the use of an interpreter, informants were asked to elaborate first-hand responses during follow up, which posed questions on the same topic from different angles.

Conclusions

Medical anthropologists are currently missing from the panel of experts reviewing the UK Confidential Enquiries into Maternal and Child Health (CMACE, 2011). Our findings suggest that the addition of this expertise could complement the efforts of the NHS to understand why adverse obstetric outcomes occur in this high-income setting. The maternal migration effect further provides the groundwork for conceptualising delay-causing socio-cultural influences that can have broad implications for global maternal health initiatives (see e.g. Freedman et al., 2007). The identification of delay-causing factors allowed us to generate two explanatory models (Kleinman, 1980) for our migration three delays framework. These may provide reasons for why it is still 'too far to walk' for optimal facility-based care for immigrant women who have migrated from low-income settings:

1) Mutual broken trust between women and maternal care providers may result in delays at the facility level, expressed as women's choice for late-booking, non-adherence, or inappropriate decision-making, and as provider frustration resulting from the inability to impart optimal treatment.

2) Unmet expectations among providers are worsened by a lack of consistent clinical guidelines, insufficient staff, and poor interpreter services, which are among factors of infrastructure management that can hinder medical providers from delivering optimal maternal healthcare.

Acknowledgements

The authors would like to thank Sir Sabaratnam Arulkumaran, professor at St. Georges Hospital, and Mr. Onsy Louca, clinical director at Northwick Park Hospital, for facilitating the research process. The Swedish Council for Working Life and Social Research (FAS 2005—2006) and the Faculty of Medicine at Uppsala University, Sweden, are acknowledged for support.

References

Baker, S., Choi, P., Henshaw, C., & Tree, J. (2005). 'I felt as though I'd been in jail': women's experiences of maternity care during labour, delivery and the immediate postpartum. Feminist Psychology, 15(3), 315—342. Bernard, H. (2006). Research methods in anthropology: Qualitative and quantitative

approaches (4th ed.). Oxford: Altamera Press. Binder, P., Borné, Y., Johnsdotter, S., & Essén, B. (2012). Shared language is essential: communication in a multiethnic obstetric care setting. Journal of Health Communication: International Perspectives. http://dx.doi.org/10.1080/ 10810730.2012.665421, Retrieved 29.07.12. Brown, E., Carroll, J., Fogarty, C., & Holt, C. (2010). "They get a C-section.. .they gonna die": Somali women's fears of obstetrical interventions in the United States. Journal of Transcultural Nursing, 21(3), 220—227. Carroll, J., Epstein, R., Fiscella, K., Gipson, T., Volpe, E., & Jean-Pierre, P. (2007). Caring for Somali women: implications for clinician—patient communication. Patient Education and Counseling, 66(3), 337—345. CMACE. (2011). Centre for Maternal and Child Enquiries: saving mothers' lives: reviewing maternal deaths to make motherhood safer: 2006—2008. The eighth report on confidential enquiries into maternal deaths in the United Kingdom. bjog: An International Journal of Obstetrics & Gynaecology, 118(Suppl. 1), 1 —203. Collin, S. M., Marshall, T., & Filippi, V. (2006). Caesarean section and subsequent fertility in sub-Saharan Africa. BJOG: An International Journal of Obstetrics & Gynaecology, 113(3), 276—283. Daly, J., Willis, K., Small, R., Green, J., Welch, N., Kealy, M., et al. (2007). A hierarchy of evidence for assessing qualitative health research. Journal of Clinical Epidemiology, 60(1), 43—49.

Davies, M. M., & Bath, P. A. (2001). The maternity information concerns of Somali women in the United Kingdom. Journal of Advanced Nursing, 36(2), 237—245.

Department of Health. (2007). Maternity matters: Choice, access and continuity of care in a safe service. London: DH Publications.

van Dillen, J., Mesman, J. A., Zwart, J. J., Bloemenkamp, K. W., & van Roosmalen, J. (2010). Introducing maternal morbidity audit in the Netherlands. BJOG: An International Journal of Obstetrics & Gynaecology, 117(4), 416—421.

Dumont, A., de Bernis, L., Bouvier-Colle, M. H., & Breart, G., , for the MOMA Study Group. (2001). Caesarean section rate for maternal indication in sub-Saharan Africa: a systematic review. Lancet, 358(9290), 1328—1333.

Essen, B., Binder, P., & Johnsdotter, S. (2011). An anthropological analysis of the perspectives of Somali women in the west and their obstetric care providers on caesarean birth. Journal of Psychosomatic Obstetrics & Gynecology, 32(1), 10—18.

Essen, B., Bodker, B., Sjoberg, N. O., Langhoff-Roos, J., Greisen, G., Gudmundsson, S., et al. (2002). Are some perinatal deaths in immigrant groups linked to suboptimal perinatal care services? BJOG: An International Journal of Obstetrics & Gynaecology, 109(6), 677—682.

Essen, B., Hanson, B. S., Ostergren, P. O., Lindquist, P. G., & Gudmundsson, S. (2000). Increased perinatal mortality among sub-Saharan immigrants in a citypopulation in Sweden. Acta Obstetricia et Gynecologica Scandinavica, 79(9), 737—743.

Essen, B., Johnsdotter, S., Hovelius, B., Gudmundsson, S., Sjoberg, N. O., Friedman, J., et al. (2000). Qualitative study of pregnancy and childbirth experiences in Somalian women resident in Sweden. BJOG: An International Journal of Obstetrics & Gynaecology, 107(12), 1507—1512.

Freedman, L. P., Graham, W., Brazier, E., Smith, J., Ensor, T., Fauveau, V., et al. (2007). Practical lessons from global safe motherhood initiatives: time for a new focus on implementation. Lancet, 370(9595), 1383—1391.

Gabrysch, S., & Campbell, O. M. (2009). Still too far to walk: literature review of the determinants of delivery service use. BMC Pregnancy Childbirth, 9(34), 5—7.

Gagnon, A. J., Zimbeck, M., Zeitlin, J., & , ROAM Collaboration. (2009). Migration to western industrialised countries and perinatal health: a systematic review. Social Science & Medicine, 69(6), 934—946.

Harper Bullman, K., & McCourt, C. (2002). Somali refugee women's experiences of maternity care in West London: a case study. Critical Public Health, 12(4), 365—380.

Harris, M. (2001). The rise of anthropological theory: A history of theories of culture. New York: Thomas Crowell (Re-issue).

Healthcare Commission. (July 19, 2005). The state of healthcare 2005. United Kingdom.

Herrel, N., Olevitch, L., DuBois, D. K., Terry, P., Thorp, D., Kind, E., et al. (2004). Somali refugee women speak out about their needs for care during pregnancy and delivery. Journal of Midwifery & Women's Health, 49(4), 345—349.

Johansen, R. (2006). Care for infibulated women giving birth in Norway: an anthropological analysis of health workers' management of a medically and culturally unfamiliar issue. Medical Anthropology Quarterly, 20(4), 516—544.

Jonkers, M., Richters, A., Zwart, J., Ory, F., & van Roosmalen, J. (2011). Severe maternal morbidity among immigrant women in the Netherlands: patients' perspectives. Reproductive Health Matters, 19(37), 144—153.

Kleinman, A. (1980). Patients and healers in the context of culture: An exploration of the borderland between anthropology, medicine, and psychiatry. Berkeley: University of California Press.

Kravitz, R., Helms, L., Azari, R., Antonius, D., & Melnikow, J. (2000). Comparing the use of physician time and health care resources among patients speaking English, Spanish, and Russian. Medical Care, 38(7), 728—738.

Lambert, H., & McKevitt, C. (2002). Anthropology in health research: from qualitative methods to multidisciplinarity. BMJ, 325(7357), 210—213.

G. Lewis (Ed.), (2007). Saving mothers' lives: Reviewing maternal deaths to make motherhood safer 2003—2005. London: The Confidential Enquiry into Maternal and Child Health (CEMACH).

Lincoln, Y., & Guba, E. (1985). Naturalistic inquiry. New York: Sage.

Lindström, D., & Muñuz-Franco, E. (2006). Migration and maternal health services utilization in rural Guatemala. Social Science & Medicine, 63(3), 706-721.

Lundberg, P., & Gerezgiher, A. (2008). Experiences from pregnancy and childbirth related to female genital mutilation among Eritrean immigrant women in Sweden. Midwifery, 24(2), 214-225.

Luque Fernández, M. A., Cavanillas, A. B., & de Mateo, S. (2009). Excess of maternal mortality in foreign nationalities in Spain, 1999-2006. European Journal of Obstetrics & Gynecology and Reproductive Biology, 149, 52-56.

Maine, D., & Rosenfield, A. (2001). The AMDD program: history, focus and structure. International Journal of Gynecology & Obstetrics, 74(2), 99-103.

Mrisho, M., Obrist, R., Armstrong Schellenberg, J., Haws, A., Mushi, A., Mshinda, H., et al. (2009). The use of antenatal and postnatal care: perspectives and experiences of women and health care providers in rural southern Tanzania. BMC Pregnancy Childbirth, 9(10), 1-12.

Ng, C., & Newbold, K. (2011). Health care providers' perspectives on the provision of prenatal care to immigrants. Culture, Health & Sexuality, 13(5), 561-574.

Pavlish, C., Noor, S., & Brandt, J. (2010). Somali immigrant women and the American health care system: discordant beliefs, divergent expectations, and silent worries. Social Science & Medicine, 71 (2), 353-361.

Philibert, M., Dene ux-Tharaux, C., & Bouvier-Colle, M. (2008). Can excess maternal mortality among women of foreign nationality be explained by suboptimal obstetric care? BJOG: An International Journal of Obstetrics & Gynaecology, 115 (11), 1411-1418.

Schuitemaker, N., van Roosmalen, J., Dekker, G., van Dongen, P., van Geijn, H., & Bennebroek Gravenhorst, J. (1998). Confidential enquiry into maternal deaths in the Netherlands 1983-1992. European Journal of Obstetrics & Gynecology and Reproductive Biology, 79(1), 57-62.

Schutte, J., Steegers, E., Schuitemaker, N., Santema, J., de Boer, K., Pel, M., et al. (2010). Rise in maternal mortality in the Netherlands. BJOG: An International Journal of Obstetrics & Gynaecology, 117(4), 399-406.

Silverstein, P. (2005). Immigrant racialization and the new savage slot: race, migration, and immigration in the new Europe. Annual Review of Anthropology, 34, 363-384.

Sobotka, T. (2008). The rising importance of migrants for childbearing in Europe. Demographic Research, 19(Special Collection 7, article 9), 225-248, Retrieved from. http://www.demographic-research.org/Volumes/Vol19/9/19-9.pdf.

Straus, L., McEwen, A., & Hussein, F. (2009). Somali women's experience of childbirth in the UK: perspectives from Somali health workers. Midwifery, 25(2), 181-186.

Sunday-Adeoye, I., & Kalu, C. A. (2011). Pregnant Nigerian women's view of cesarean section. Nigerian Journal of Clinic Practice, 14(3), 276-279.

Tanassi, L. (2004). Compliance as strategy: the importance of personalised relations in obstetric practice. Social Science & Medicine, 59(10), 2053-2069.

Thaddeus, S., & Maine, D. (1994). Too far to walk: maternal mortality in context. Social Science & Medicine, 38(8), 1091-1110.

Vangen, S., Johansen, R. E., Sundby, J., Traeen, B., & Stray-Pedersen, B. (2004). Qualitative study of perinatal care experiences among Somali women and local health care professionals in Norway. European Journal of Obstetrics & Gynecology and Reproductive Biology, 112(1), 29-35.

Wildman, K., Bouvier-Colle, M. H., & , MOMS Group. (2004). Maternal mortality as an indicator of obstetric care in Europe. BJOG: An International Journal of Obstetrics & Gynaecology, 111(2), 164-169.

World Bank. (2005). World Bank classification of gross national index. Retrieved from. http://search.worldbank.org.

Zwart, J. J., Jonkers, M. D., Richters, A., Ory, F., Bloemenkamp, K. W., Duvekot, J. J., et al. (2011). Ethnic disparity in severe acute maternal morbidity: a nationwide cohort study in the Netherlands. European Journal of Public Health, 21 (2), 229234.