Communication barriers in maternity care of allophone migrants: Experiences of women, healthcare professionals, and intercultural interpreters

Abstract Aim To describe communication barriers faced by allophone migrant women in maternity care provision from the perspectives of migrant women, healthcare professionals, and intercultural interpreters. Background Perinatal health inequality of migrant women hinges on barriers to services, with a major barrier being language. Their care is often also perceived as demanding due to conflicting values or complex situations. Potentially divergent perceptions of users and providers may hinder efficient communication. Design Qualitative explorative study. Methods A convenience sample of 36 participants was recruited in the German speaking region of Switzerland. The sample consisted of four Albanian and six Tigrinya speaking women, 22 healthcare professionals and four intercultural interpreters (March–June 2016) who participated in three focus group discussions and seven semi‐structured interviews. Audio recordings of the discussions and interviews were transcribed and thematically analysed. Results The analysis revealed three main themes: the challenge of understanding each other's world, communication breakdowns and imposed health services. Without interpretation communication was reduced to a bare minimum and thus insufficient to adequately inform women about treatment and address their expectations and needs. Conclusion A primary step in dismantling barriers is guaranteed intercultural interpreting services. Additionally, healthcare professionals need to continuously develop and reflect on their transcultural communication. Institutions must enable professionals to respond flexibly to allophone women's needs and to offer care options that are safe and in accordance to their cultural values. Impact Our results give the foundation of tenable care of allophonic women and emphasize the importance of linguistic understanding in care quality.


| INTRODUC TI ON
There is incontrovertible evidence that migrant women suffer from more maternal and general health problems than nativeborn women (Bundesamt für Gesundheit, 2012; Keygnaert et al., 2014;Merten & Gari, 2013). The poorer health of migrants has numerous causes. Notably, refugees may have pre-existing health problems resulting from inadequate healthcare services in their home countries or suffer from privation or traumas experienced during their flight. Furthermore, poorer perinatal outcome is also associated with a low socioeconomic status (de Graaf, Steegers, & Bonsel, 2013;Vos, Posthumus, Bonsel, Steegers, & Denktas, 2014).
Factors such as poor living conditions, unemployment, and poverty adversely influence migrant maternal health. The health inequality of pregnant migrants and their newborns does however seem to be primarily related to difficultly accessing perinatal care services in the host country (Bundesamt für Gesundheit, 2013;Cignacco et al., 2018;Keygnaert et al., 2016;Merten & Gari, 2013). It has also been shown that there is a direct link between language barriers and adverse birth outcomes (Sentell, Chang, Ahn, & Miyamura, 2016).

| Background
Access to health care is considered the ability to identify healthcare needs, seek and reach healthcare services, to obtain and use healthcare, and to have those health needs addressed. Healthcare access results from complex interaction of healthcare systems, organizations and providers with the characteristics of individuals and their social and physical environments (Levesque, Harris, & Russell, 2013).
In Switzerland, maternity care is generally given during pregnancy by obstetricians, then during birth and early postpartum by interdisciplinary teams in hospitals, followed by postpartum at home by midwives and finally, for children up to 1 year by communitybased maternity counsellors and paediatricians. A range of other maternity care services and providers complement this standard care, such as antenatal classes, childbirth, and postnatal care in a birth centre or at home, breastfeeding counselling (Navarra & Liewald, 2017). Most of these costs are covered by health insurance. Though the standard of care can be considered high, maternity care services are highly fragmented which may affect migrant women's ability to identify and reach them, thus hindering continuous maternity care provision (König & Pehlke-Milde, 2010).
About 30% of the children born in Switzerland are foreign and 40% have a foreign born mother (Federal Statistical Office [Bundesmat für Statistik BFS], 2018a). Of all documented foreign women who have lived in Switzerland for more than a year, 11.2% do not speak an official language, this number rises to a 25% amongst first-generation non-European nationals (BFS, 2018b). In addition, migrants such as asylum seekers and undocumented foreigners rarely speak a Swiss language. Therefore, healthcare professionals serve a considerable number of allophone migrant women and thus are affected by linguistic barriers which impede effective and equitable health care. We define allophone migrants as migrants who do not speak any of the official languages of their host country.
Research on migrant women's experiences of maternity illustrates that allophone women face many barriers accessing these services. These barriers include lack of awareness of their existence, insufficient support accessing maternity services and discordant expectations between the women and service providers (Higginbottom, Bell, Arsenault, & Pillay, 2012). Moreover, caring for migrant women is perceived as demanding because of their often complex socioeconomic or asylum-related problems. Care providers often observe this group's lack of health literacy and knowledge of the maternity care system. Language barriers and conflicting values further complicate their treatment and care (Boerleider, Francke, Mannien, Wiegers, & Deville, 2013;Kurth, 2013;Pelaez, Hendricks, Merry, & Gagnon, 2017).
Intercultural interpreting effectively ensures mutual understanding and has been shown to positively affect health (Gehrig, Caldéron, Guggisberg, & Gardiol, 2012). In addition to standard translation, intercultural interpretation requires a sound knowledge of the migrant group's socio-cultural frameworks, circumstances and the local structures. Intercultural interpreters are able to move in both the "foreign" and "local" frames of reference and relate them to each other (Interpret, 2017). However, in Switzerland intercultural interpreters' frequency of deployment varies in hospitals since the cost of interpretation services is not regulated. In outpatient treatment it is even more difficult to afford interpreting services.
Healthcare delivery fundamentally depends on effective and efficient communication (Vermeir et al., 2015). The evidence reviewed shows that both, users and providers experience barriers that may professionals to respond flexibly to allophone women's needs and to offer care options that are safe and in accordance to their cultural values.

Impact:
Our results give the foundation of tenable care of allophonic women and emphasize the importance of linguistic understanding in care quality.

K E Y W O R D S
access to health care, communication barriers, health services needs and demand, intercultural interpreter, maternal health services, midwifery, migrants, nursing, professionalpatient relations hinder it. Juxtaposing their experiences and including the view from interpreters as key informants, could reveal divergent perceptions and allow formulating appropriate culturally acceptable measures.

| Aim/s
This study aimed to describe the communication barriers faced by allophone migrant women when accessing maternity care service.
We explored the perspectives of maternity service users, healthcare professionals and intercultural interpreters in the German speaking part of Switzerland. The findings served as a basis for recommendations to improve quality and access to maternity care services.

| Design
This exploratory, qualitative study was designed to reveal processes, interpretive frames, and patterns of communication barriers from multiple perspectives using focus group discussions (FGD; Krueger & Casey, 2009) and problem-centred interviews (Witzel, 2000). It was part of a broader investigation into allophone migrant women's access to maternity care (Origlia Ikhilor et al., 2017); detailed information on that study's methodology is described there. In the study for this article, we used the consolidated criteria framework for reporting qualitative studies (COREQ; Tong, Craig, & Sainsbury, 2007).
The COREQ checklist is available as supplementary online material.

| Participants
Convenience sampling was used to recruit a total of 36 individuals: users, healthcare professionals, and intercultural interpreters (Tables 1 and 2). The group of users was composed of six Eritrean, one Albanian and three Kosovan mothers. They had between one and four children each, with the youngest child's average age being 18 months. They had all been treated in the German-speaking part of Switzerland during pregnancy, birth, and postpartum. Half of the women understood simple sentences in German, the other half only few words or no German at all. Women with serious complications during their most recent pregnancy, childbirth or postpartum were not eligible to take part in our study.
The 22 healthcare professionals, all female, were midwives, nurses, obstetricians, paediatricians and maternity counsellors. Half of them had a formal qualification for transcultural competences.
Most were regularly involved in care or treatment of allophone mothers and families, who, in most cases, speak Tigrinya or Albanian.
The intercultural interpreters were two Eritrean, an Albanian and a Swiss female. They all had at least 1 year of experience translating on-site and over the phone for pregnant women or young mothers.

| Data collection
All interviews were conducted by the two main researchers (PO, GH). The Eritrean's users' data were collected in an FGD led by a Tigrinya-speaking moderator (FA). This FGD was observed by researchers assisted by a translator, all staying in the background. This allowed the researchers to follow the discussion and if necessary, to query or support the moderator. Because of scheduling difficulties, the Albanian-speaking women attended three interviews, two single and one double, conducted by a researcher and an Albanian-speaking translator. Audio recordings, which were transcribed directly into German, were made of all the interviews and field notes were made immediately after the interviews by the researchers. The healthcare professionals were interviewed in two FGDs and the intercultural interpreters had a single interview each, which were recorded and transcribed verbatim. The interviews lasted between 40 and 60 min and FGDs between 1 hr and 45 min and 2 hr.

| Ethical considerations
Research on migrants is delicate because they are particularly vulnerable. Therefore, participants were fully informed in their mother tongue about the study before signing their consent to participate.
Focus group members were informed of the need to preserve the confidentiality of group discussions. The ethical considerations were set out in detail in the study protocol (Origlia Ikhilor et al., 2017) and the ethics committee responsible issued a formal memorandum that no special Research Ethics Committee approval was necessary.

| Data analysis
Two researchers (PO, GH) coded the interviews and created themes.

| Rigour
Rigour was assured by investigating this topic from three different perspectives. After every interview, participants were given a TA B L E 1 Age of participants, age of youngest infant, and residence duration of users (N = 36)

| FINDING S
Analysis revealed that the participants from all three groups commented on three overarching dimensions. Firstly, the dimension "circumstances and healthcare system" describes how communication is affected by the women's personal circumstances and the organization of the healthcare system. Secondly, "relations and interactions" encompasses communication issues between all participants. Finally, "adaptation of health services" which describes to what extent, given the quality of communication, the needs of the women corresponded with the health care given. From each group's perspective we categorized their responses into three themes (see the middle three columns of Table 3). From there we took the three themes for each dimension and produced "synthesized themes". This enabled us to see what the main issues are and how the three group's perceptions coincided or diverged.

| Challenge of understanding each other's world
The interviews suggested that all three groups genuinely tried to understand reality from the others' perspective, despite the challenge to understand each other's different worlds and the unfamiliar system. Whereas intercultural interpreters sympathized with users because they knew the country of origin and were able to fall back on their own experience in an unfamiliar country, healthcare TA B L E 2 Sample characteristics (N = 36) Characteristics n

Healthcare professionals 22
Sex

3-10 years 8
More than 10 years 14 Frequency of care for allophone migrants One to several times per week 14 One to several times per month 6 One to several times per year 2 Users 10 Nationality Eritrean 6 Kosovar 3 Albanian 1

Mastering German language
Not at all 0

Comprehends single words 1
Comprehends and talks single words 4 Comprehends and talks simple sentences Professional experience 1-2 years 2 3-10 years 2 More than 10 years 0

Frequency of interpreting assignments in maternity area
One to several times per week 1 One to several times per month 2 One to several times per year 1 TA B L E 2 (Continued) professionals sometimes lacked understanding of the women's social environment. The women on the other hand, lacked knowledge of the reproductive healthcare system.
For the allophone women, it was difficult to "navigate through the complex (healthcare) system". Given Switzerland's highly differentiated system it was hard for them to know who to turn to, who was responsible for what and how to assess the usefulness of spe- Although interpreters and healthcare professionals recognized the disorientation of some migrant women interpreters found it easier to "comprehend the circumstances" of allophone migrant women. They explained that existential, rather than health, problems could overshadow a consultation. These situations were frustrating for the interpreters as they were unable to give a solution in the scope of their task: Asylum seekers put great hope in every meeting. The person they speak to is very, very important to them.
But sometimes that person disappoints. And then it's solely that person's fault -mine, the translator's, or the healthcare professional's, because they did not help. (Albanian interpreter)

| Communication breakdown
Both healthcare professionals and allophone migrant women tried to understand each other but were unable to, not through lack of effort but it was just the completely different languages spoken by the parties. This led to communication being reduced to a bare minimum and thus, insufficient to inform women about treatment and adequately address their expectations and needs.
Some women reported very good care experiences despite the language barriers and that professionals were strongly committed to them and considered their situation. Others reported a "lack of rapport", for example, that the relationship between them and the healthcare providers was deficient. Any perceived dominance of the professional over the lay person in their relationship was intensified by the language barriers. Women described this as a relationship gap between themselves and the healthcare professionals: The women felt offended if procedures and interventions were inadequately or not explained. Some women described situations where they assumed they were being discriminated against: The problem was that we were informed that the  The women felt a "lack of agency" and that they were at the mercy of the providers. In many situations, they did not understand the reasons of the medical intervention, so they could not give informed consent. This made them to feel helpless and concerned. Long-lasting bad feelings sometimes lead them to renunciate further care:

| Imposed health services
The physician explained the operation [caesarean section] as well as possible. I just said yes, yes, I couldn't say anything else. For example, I couldn't ask, "Can I wait another week, can I wait until my due term?".
(Eritrean woman) Healthcare professionals found themselves caught between "trying and resigning" that is, having to decide whether to try their hardest to communicate or come to terms with the situation. Often, professionals resigned themselves to frustrating communication and comforted themselves with the fact that the mother and baby's health was fine, and that Swiss health care is almost certainly better than that in the woman's country of origin: We Non-involvement in the decision processes triggered a range of negative sentiments in women, from mild concern to existential anxiety, especially in emergencies. Emergencies were a big strain for all involved, including interpreters: We had to bring a woman in for an emergency op-  Both users and providers struggled to understand each other's circumstances. Their frustrations, however, were centred on different issues. Allophone women were often disorientated in the Swiss healthcare system, which is highly complex and typified by a variety of different providers. They sometimes expected help not only for health issues but also for many other concerns including unsolved existential problems even though healthcare professionals did not have the authority to directly respond to social or legal issues. Depending on their residence status, migrants access healthcare services in different ways: permanent foreign residents must fend for themselves while refugees are supported-to a lesser or greater extent-by local authorities (Origlia Ikhilor et al., 2017). A study on how Swiss asylum centres organize healthcare provision showed that sexual and reproductive healthcare needs of women asylum seekers are not sufficiently addressed and there is a lack of translation and gender sensitive approach of women (Cignacco et al., 2018). There is an urgent need to better target health care at allophone women's needs and women need guidance through the Swiss maternity healthcare system to be able to make informed decisions.

| D ISCUSS I ON
The providers were challenged when trying to grasp the social system of migrant women. Although for the women, the family represents an important moral and practical crutch, healthcare professionals did not always succeed in integrating relatives appropriately into the care process and often perceived them as a hindrance.
Other studies have described relatives sometimes feeling unwanted, highly worried, and often lacking adequate information too (Bihr & Kaya, 2014). The patterns of interaction of Swiss healthcare professionals with individualistic socialization and migrants with sociocentric socialization often fundamentally differ (Domenig, 2007). That said, in most cases the social network of migrant women is often restricted compared with how it was in their homeland. Self-selected reference persons should be addressed actively by healthcare professionals. They need to be informed comprehensively and their fears and concerns also taken into account (Bihr & Kaya, 2014). The efficient involvement of the family is a resource and a protective factor. Moreover, intercultural interpreters can support healthcare professionals in assessing the migrant women's family system and identifying available intra-family resources.
Breakdowns in communication were frustrating for all involved.
The allophone women reported some discriminatory behaviour from healthcare professionals. Although providers and interpreters made no direct statements about this, the professionals did express expectations about the integration efforts of women and families. It is therefore conceivable that expectations were said unconsciously and led to users' negative feelings. Discrimination and professionals' unintended, unfavourable attitudes naturally discourage women from engaging in consultations (Origlia, Jevitt, Sayn-Wittgenstein, & Cignacco, 2017). Transcultural competence development can help to sensitize professionals to their behaviour and unintentional discrimination. Essentially, transcultural competence in healthcare is the ability to interact effectively with migrants across cultural boundaries (Domenig, 2007). To do this, professionals must first learn to perceive their own lifeworld through self-reflection. Together with background knowledge for example, about migration processes, they can better classify and understand migrants' circumstances and environment. Finally, narrative empathy allows professionals an appreciative, respectful attitude towards migrants. By placing narration at the centre of treatment, carers can reflect on their own prejudices and discriminatory behaviour (Domenig, 2007). In our study, almost half of the interviewed professionals had completed training in transcultural competence. In practice, healthcare professionals generally feel less proficient in intercultural tasks than medical skills (Hudelson, Perron, & Perneger, 2011).
According to the interviews the most common strategy to overcome language barriers was communication with gestures. Communication with gestures seemed to some extent to be approved by the healthcare professionals, although they were aware of its limitations. Moreover, if mother and baby could be healthily discharged, the result of their treatment was considered sufficient.

Consequently
It was unclear to what extent the professionals recognize the true implications inadequate communication. Possibly they tend to resign themselves to the situation because they have no other option.
Even with involvement of intercultural interpreters, communication between healthcare professionals and allophone women could be inefficient, especially if the expectations of a consultation differed. Sometimes too many professionals were included in a consultation or too much information given at once, which overwhelmed the women. Maternity care counselling on health education and promotion can risk hitting the women with a wall of information that does not apply to individual needs and concerns.
It is much more efficient when professionals are allowed to have a solid and continuous relationship with women (Sanders, Hunter, & Warren, 2016).
Emergency situations were particularly stressful for all involved.
Professionals were faced with a dilemma: they needed to give comprehensive information and at the same time they had to take immediate therapeutic measures to safeguard mother and child.
Obstetric emergencies, neonatal complications and poor quality of provider interactions can cause great anxiety and even traumatize women because they cannot assess the extent of the danger to themselves and their child (Simpson & Catling, 2016

| Strengths and Limitations
In several meetings a group of researchers (PO, GH, EK, JP, EC) reviewed topics and determined how well the themes matched the coded data. Carefully reviewing the findings and interviews as a group allowed the researchers to analyse and interpret the data to generate agreed on interpretations and eliminate misleading ones (Braun & Clarke, 2006).
Despite different perspectives being used in this study the sample of service users was relatively small and limited to only two population groups with different socio-cultural characteristics and migration histories. Therefore, conclusions about other allophone migrants are only possible to a limited extent.
The predefined exclusion criteria had to be relaxed during the recruitment, for example, few women were able to have a simple conversation in German. Nevertheless, their data gave insight into communicative challenges about their experiences. Furthermore, during the interviews it became apparent that some mothers had had complications during their last birth but were willing to talk about their experiences anyway. Therefore, they were not excluded retrospectively.
All the Eritrean women were recruited through a single person who was also involved as an intercultural interpreter in the obstetric consultation of some of them. Moreover, this interpreter was present at the FGD and translated for the researchers. The participation of a known person may have caused a loyalty conflict among the women and triggered some socially desired answers.
In the foreign language interviews and FGD some information was not immediately and fully accessible to the researchers. We therefore put significant effort into instructing the interpreters/ moderator so that they had a comprehensive understanding of the method. No back translation of the Tigrinya and Albanian transcripts was. However, when uncertainties arose, the researchers could compare the transcript and the recorded verbal translation. Despite these limitations, there is methodological consensus that useful knowledge can be gained (Kruse, 2012;Lauterbach, 2014).

| CON CLUS ION
There is a significant range of communication barriers in maternity health care provision that lead to adverse short-and long-term results. Measures to dismantle barriers include first and foremost, secured and financed intercultural interpreting services. As these are not always available on site for example, in emergency situations, therefore telephone services must be considered. Only through interpretation can quality treatment be achieved and legally required informed consent obtained. Lack of agency and understanding in emergencies can be traumatic for women and is stressful for healthcare professionals too. Awareness must be raised that communication with gestures is not enough.
Healthcare professionals' transcultural competencies must be continuously developed and integrated into education and practice.
Continuous reflection may also raise awareness of discriminatory behaviour. Training in the use of plain language and avoidance of information overload would lower language barriers and make translation easier. Their advice should focus on women's concerns and continuity of care should be prioritized.
Healthcare settings for allophone mothers must be more flexible and offer care options that are in accordance with the women's cultural values. Obstetric guidelines, however, may not offer the option of a low-intervention, more natural approach. We would therefore recommend that guidelines should be counterchecked to see if they can be adapted.
Furthermore, migrant women need support navigating an unknown healthcare system. This is not just relevant to pregnant women. To encourage integration, women should be given health system information as early as possible and given the opportunity to process this information with trained peers, integration officers or healthcare professionals.
In summary, better communication during maternity care cannot only be achieved through better individual attitude or professional skills but it also needs appropriate conditions at institutional and policy levels. Improvement will not only help achieve better mother and child health but also help engage healthcare professionals in their challenging task.

ACK N OWLED G EM ENTS
We thank the Swiss Federal Office of Public Health and the Swiss Agency for Combating Racism for providing fund for this study.
Special thanks to Martin Benedikt Christ for his work preparing the manuscript and Daniel Foord for his editorial contribution.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R S CO NTR I B UTI O N S
Made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data.