Moving beyond the ‘language problem’: developing an understanding of the intersections of health, language and immigration status in interpreter-mediated health encounters

ABSTRACT Health systems internationally are dealing with greater diversity in patient populations. However the focus on ‘the language problem’ has meant little attention is paid to diversity within and between migrant populations; and how interpreted consultations are influenced by intersecting migratory, ethnicity and sociodemographic variables. Our analysis of the experiences of patients, health care providers and interpreters in Scotland evidences the need to move beyond language, addressing multiple hidden inequalities in health care access and provision that operate in both clinic and, especially, home-based settings. We call for a practice-evidenced research agenda promoting cultural communication across health care and home settings, acknowledging immigration status as a social determinant of health.


Introduction
People move. On 1 January 2016, the number of people living in the 28 Member States of the European Union (EU) who were citizens of non-member countries was 20.7 million, 4.1% of the total EU population. A further 16.0 million EU citizens were living in a different Member State to the one they were a citizen of (Eurostat, 2017). People move for work, education, family reunification andof Communication lies at the core of health care. One implication of increased diversity from population movement is the plethora of languages that now exist within the EU, with an estimate of 160-180 different languages now in daily use in the EU (Vertovec, 2007). This creates challenges for the development of health care policies and for the delivery of equitable, effective healthcare services that are responsive to the needs of all patients, both in terms of language and cultural understanding (Phillimore, 2011;MacFarlane et al., 2012;van den Muijsenbergh et al., 2014). The inability to communicate within health care consultations has recognised consequences. Patients visit their health care provider less (Derose, Escarce, & Lurie, 2007), consultations are longer, there is poorer mutual understanding leading to increased non-compliance with medication and treatment advice (van Wieringen, Harmsen, & Bruijnzeels, 2002) and fewer referrals to other services, especially psychological services (Bischoff, Perneger, Bovier, Loutan, & Stalder, 2003). Other issues include the increased risk of misunderstanding and misdiagnosis (Flores, 2005;Flores, Abreu, Barone, Bachur, & Lin, 2013;Karliner, Jacobs, Chen, & Mutha, 2007), extensive clinical testing required to compensate for poor communication (Priebe et al., 2011) and prolonged administrative procedures.
However, intercultural communication and understanding goes beyond just language. Competency in intercultural communication extends to an ability to ensure shared understandings by people of different social identities (Byram, Gribkova, & Starkey, 2002). In this sense, intercultural communicative competence requires certain attitudes, knowledge and skills in addition to linguistic competence and demands a curiosity and openness as well as readiness to see other cultures without being judgemental. Wider social identities are an unavoidable part of our communicative interactions, and necessarily extend to language, culture, ethnicity, age, gender and, we argue, immigration status. Health, and our understandings of health, are shaped and informed by one's cultural background. As the Lancet Commission on culture and health stated, 'Intercultural health communication is not only about language translation, but also situated beliefs and practices about the causation, local views on what constitutes effective provision of health care, and attitudes about agency and advocacy' (Napier et al., 2014(Napier et al., :1614. Migrants' understanding of their rights and entitlements to health care, and the response of health care professionals to them, can also influence communication and understanding within consultations. This intersection of culture and migrant status with other structural and societal factors can therefore create and compound difficulties in the health care encounter, including a lack of trust in the role of the health care practitioner (O'Donnell, Higgins, Chauhan, & Mullen, 2008).
Widening out the lens to include multiple intersecting variables is valuable because to date, understandings of the ways in which health systems and policies can intersect with different personal characteristics and social inequalities to mitigate or burden individuals, has been largely missing. This has particular resonance for migration research in general (see for example Anthias, 2012;Bastia, 2014;Bürkner, 2012) and for migrant health issues in particular, and provides a way to move beyond a narrow focus on the 'language problem'. Health care access and use is influenced both by intersecting demographic and by personal characteristics (Crenshaw, 1989). These include age, gender, country of origin, religion, health literacy, socioeconomic status, migration statusand wider system and social determinants, for example entitlement based on migration status, structural racism and organisation of the health system being accessed (Castañeda et al., 2015;Derose et al., 2007;Ingleby, 2012;O'Donnell et al., 2013;Phillimore, 2011Phillimore, , 2015Rechel et al., 2013;WHO Regional Office for Europe, 2010). To date, however, few studies have focussed on untangling the intersecting quality of these characteristics and determinants in this way (the exceptions being for example, Derose, Bahney, Lurie, & Escarce, 2009;Gazard, Frissa, Nellums, Hotopf, & Hatch, 2015;Hankivsky & Cormier, 2009;Hankivsky, Grace, Hunting, & Ferlatte, 2012;Mens-Verhulst & Radtke, 2006;Viruell-Fuentes et al., 2012). This is significant because as we shall see in the following review of literature, a 'language problem' approachfor which interpreting is advocated as the solutionglosses over the effects of multiple variables on the experience of effective intercultural communication in health care settings. Flores has found that use of untrained ad hoc interpreters led to poorer quality care, more errors in interpretation and less satisfied patients (Flores, 2005;Flores et al., 2013). Conversely, use of professional interpreters resulted in a decrease in communication errors, increased patient understanding, reduced inequity in service use, and improved clinical outcomes and patient satisfaction (Karliner et al., 2007). However, these studies took a fairly reductionist stance in their approach to the literature. Moreover and importantly, many of the studies identified in our review were located in hospital or family doctor settings, and there was little attempt to consider either multiple perspectives simultaneously (namely comparing the views of patient, clinician and interpreter) nor was there any critique of the tendency to homogenise the 'migrant' population. There was also no acknowledgement of the challenges of interpreting 'verbatim' versus approaches designed to promote wider intercultural communication and understanding. This is well known to readers of this journal, but our point is to highlight here how pervasive the belief in the language problem and interpreting solution is, and how little of the intercultural communication literature has been received into the policy and decision making contexts of health communication literature.
Hsieh has tried to unpack the triadic nature of the interpreted consultation. In an elegant ethnography, she observed hospital-based interpreted consultations conducted in Mandarin Chinese (in which Hsieh was fluent); she then interviewed a wide range of professional interpreters (Hsieh, 2008). This study found that interpreters could assume four potential roles, often within one consultation. The predominant role was that of 'conduit', in which they strove for invisibility and to reinforce the professional-patient dyad. The second role was that of advocate, although not all interpreters were comfortable with this role. Other roles were that of cultural mediator or broker and, finally, that of a trained professional equal to the clinician as opposed to someone who was 'merely' bi-lingual and providing 'only' an interpreting service. Greenhalgh, Voisey, and Robb (2008) conducted work in UK general practice, interviewing a range of health care professionals, interpreters and patients. They found that general practitioners were central to the routinisation of interpreting in the practice; interpreters, even paid professional interpreters, were relatively powerless to influence practice routine and ways of working. Migrant patients have their own issues, with many wanting to have a professional interpreter rather than a family member or friend interpret for them, as they have concerns about confidentiality and appropriateness of the interpreting when ad hoc interpreters were used (Barron, Holterman, Shipster, Batson, & Alam, 2010;MacFarlane et al., 2009).
We have, however, identified a number of gaps here. First, there is an almost entire focus on spoken language translation, with little acknowledgement of other forms of communication. Emotional expressions and meanings are often not interchangeable between languages (Flores et al., 2013) and non-verbal communication from the patient to the practitioner can be difficult to read, especially if the presence of the interpreter draws the focus of the health care provider away from the patient (Tribe, 2009). Secondly, there is little acknowledgement of diversity within migrant populations or the interplay of immigration status, interpreting and intercultural communication. One Canadian study that did consider interpretation in relation to the migrant journey, suggested that different migrant groups have different language capacities, particularly in relation to discussing health (McKeary & Newbold, 2010). A third gap is that most studies we reviewed were located in hospital settings or in family physician clinics, such as general practices in the UK. We were not able to identify any research set in community or home-based settings where, arguably, the dynamic between patient and health care provider may be different with a wider set of variables shaping the social identities of interlocutors, and therefore affecting competencies in practice.
Our study aimed to explore the different experiences of practitioners, interpreters and migrants in interpreted consultations across a range of community and home-based encounters, with a longterm aim of using these narratives to develop training and support materials for primary care and interpreting service providers. In the process of conducting that work, we became interested in the extent to which a more nuanced understanding of intersecting variables of difference might help us to understand the challenges faced in interpreted consultations. This is our focus in this paper, where we use an intersectional perspective to inform our exploration of how the interconnecting variables of immigration status and migratory journey interconnect with language, to influence health care provision and help-seeking behaviours and to shape experiences of interpreted encounters.

Background to the study and methodology
We conducted this study in the city of Glasgow, which is host to one of the largest populations of refugees and asylum seekers in the UK under the dispersal policy as well as having a history of hosting large communities of migrants, particularly since the 1960s due to rapidly increasing intra-European mobility of migrants. Politically, the Scottish Government has adopted a pro-immigration position in the face of an ageing population and declining birth rates, making Scotland a distinct research context in relation to immigration within the wider UK. This was most recently demonstrated in December 2017 with the launch of a consultation on electoral reform, which would mean extending the right to vote to everyone legally resident in Scotland, which would include refugees, and again in January 2018 with the launch of the Scottish Government's second New Scots (http://www.gov.scot/Resource/0053/00530097.pdf) integration strategy.
This paper draws on a research project that explored the development of ethical approaches to healthcare interpreting. The project itself emerged from discussions between members of the project team through their involvement in GRAMNet (a Glasgow University-based international research network) 1 and health care providers. These providers had identified challenges in the delivery of equitable health care to non-English speaking patients in a range of clinical settings including home settings, combined with a frustration with existing training materials that fell short of recognising the reality and messiness of interpreting practice. Exploring issues of intercultural communication requires a range of perspectives and knowledge (van den Muijsenbergh et al., 2014) and this was reflected in the interdisciplinary project team, with expertise in primary care, psychology, nursing and midwifery, intercultural communication and language, education, sociology and social anthropology, working alongside a non-academic third sector partner. This interdisciplinarity allowed us to explore more thoroughly issues of intercultural communication. Our sampling strategy targeted three main groups: interpreters, health care providers and service users. We contacted health care professionals working with interpreters in the community and home settings, in particular community midwives and health visitors. In Scotland, such staff are usually linked to family practices, so we identified four practices in neighbourhoods with high percentages of dispersed asylum seekers and refugees (Scottish Refugee Council, 2010). The project researcher (TP) contacted the relevant line managers of healthcare staff to seek permission to approach them. Having already identified that research into interpreted healthcare tends to focus on family practices, we also recruited sexual health specialists via direct contact with a community-based health care service what had expressed an interest in improving its delivery of interpreted health care. Interpreter participants were contacted through a 'snowball' sampling technique (Bryman, 2008) via the professional networks of the study team. Migrant service users were recruited with the support of the third sector project partner (BEMIS 2 ) who identified potential participants from their service users. Finally, we contacted two African community organisations in Glasgow and invited members to participate. Service users were offered travel expense reimbursement and light refreshments.
We collected our data through a combination of individual in-depth interviews and group discussions with 19 research participants. These interviews focussed on the current practices, coping strategies and knowledges of working in multilingual contexts, as appropriate to each group (see Table 1). The interview schedule was informed by the wider literature review and organised thematically around: training (applicable to professionals and interpreters); understanding of interpreting and translation; role of interlocutors in the healthcare encounter; expectations in the health care setting of interlocutors; service provision levels; experiences of/concerns about equality of access to services (specifically for migrant service users); understanding and intercultural communication competence (ICC) and managing understanding in encounters of difference; constraints, opportunities and challenges of multilingual health care encounters. Each interview sought to understand the dynamics of the interaction that occurs in translation in health care, and to shed light on the specific professional, personal and emotional issues from each perspective. All interviews, which lasted around one hour, were transcribed verbatim. Four required translation into English: two interviews were conducted in Urdu with and translated by the representative from BEMIS and two interviews were conducted and translated from French by TP who has extensive experience of conducting qualitative research in French. A constructivist grounded theory approach was used to inform data analysis, with constant comparison across interviews (Charmaz, 2006). Broad themes (e.g. access, understanding, confidence and stigma) were initially identified from the interview transcripts. Each theme was then further then broken down and coded into sub-themes (e.g. knowing where to go, immigration status) by TP and IJ. Codes and themes were re-assessed and interpretations discussed and agreed with the wider project team. The University of Glasgow College of Social Sciences Ethics Committee granted ethical approval.
A key output of this research project was the production of five short educational films on interpreting in health care settings. These films were developed to support health care practitioners, interpreters working and migrant service users. The resource is freely available in DVD format, online (Vimeo and YouTube) and in a range of digital formats.

Findings and discussion
Our analysis identified a complex interplay of multi-dimensional markers of difference, in addition to language differences, relating to migratory, ethnicity and socio-demographic variables that shape subjective experiences of health care and highlight the value of intercultural communication competency in practice. We present these findings here.

Migratory trajectory and 'knowledge'
Overall, basic levels of treatment were being met for the service users interviewed. They had come to the UK through two different migratory trajectories: seeking asylum and as migrant workers. Each had successfully managed to register with a family doctor and had learned how to use primary health care services relatively quickly, reflecting the Scottish position of full access to the National Health Service for all, regardless of immigration status. Our research suggests however that levels of information about, and familiarity with, services and rights and entitlements are closely tied with migratory trajectory. The Polish migrant worker, who also worked informally as an interpreter, described how many Polish migrants she knew developed their knowledge base prior to leaving Poland: They [Polish migrants] will usually check it online. There are quite a lot of websites with information for the Polish Diaspora, and there are separate sections for education, health, other services, so they will read about it […]. And as well through friends so they will tell them 'today that you need to do A, B, C, D … ' (INT5) Some of the healthcare professionals interviewed, including one who worked in the community and one in a specialist setting, supported this view of different levels of information and familiarity depending on the migration trajectory.
We find that people in the asylum system, are much less knowledgable about services and rights especially rights to an interpreter compared to say EU nationals, who I think seem to be just better informed about how the NHS works. And if people are irregular migrants, there are clear issues about knowledge of services. (SHS1) The way we see it, and I'm talking about everyday interactions, levels of understanding about the NHS really do vary, and for asylum seeking women they know very little compared to other migrants, like your Poles for instance. (CMW2) Although this might be interpreted as EU migrants generally having better access to information and understanding of systems; the non-EU migrant worker (from Cameroon) presented a similar experience of using her wider European transnational networks in the UK and France to build a picture of the different services available in the UK. However, the experience of claiming asylum is characterised by much greater uncertainties, not solely related to health but also housing, welfare support and education (O'Donnell et al., 2008;Phillimore, 2011). We found that asylum seeker and refugee journeys into health care were characterised by a lack of knowledge about how it worked, how to access services or their rights to an interpreter.
When you arrive, you don't even know where you are … you don't know where to go or for what. You have no idea, then even to ask for an interpreter … you've never used an interpreter, you don't know … you are on your own. (SU4) When we came here, we did not know about Scotland or England, we didn't read the history or the culture. These are just the circumstances. If they [health care providers] assume you know everything it justs adds to our problems. So we need to watch [what happens in reception areas] and listen and find out for ourselves. (SU1) The main obstacle? It used to be language, but now I think it's information about services. But then I have been here for so long, I feel confident that I can find a way to information if I need it. I learn from my friends who use services too. (SU3) Asylum seeker participants described 'feeling their way' through the system to become better informed. They did this through informal conversations with other asylum seekers and asylum advocates they met in various sites across the city. Many would be opportunistic momentsconversations in immigration reporting centres, at bus stops, post offices and local shopsbut also through attending local community drop-in centres, from volunteers, and from advocacy groups like the Scottish Refugee Council. For some, it was seeing an interpreter working in a health centre that made them realise their entitlement.
Our data suggest a qualitative difference between the migrant workers and asylum seekers in their ability to be informed about NHS services, and one that is not surprising; the former able to use online resources to build a picture of the NHS prior to arrival, the latter uncertain as to their destination and so unable to prepare. This has implications for service providers in communicating information about services, rights and entitlements. One of the Health Visitors (HV2) belonged to a team that had employed a Slovakian-speaking support worker to work directly with the local Slovak and Romanian Roma populations to improve service knowledge and uptake and to be able to communicate more effectively across multiple social identities (Byram et al., 2002). These different responses suggest that access to information, knowledge and service familiarity are closely tied up with levels of individual capital and migratory trajectory of the migrant service user, and reveal how the immigration gateway may have a direct effect on health outcomes.
All of the participants in this study recognised that there is a period of adjustment to life in a new environment and to the way services are delivered and accessed. They each spoke about the socioeconomic, emotional and practical stressors of managing in a new country. Most often these related to the bureaucratic stress linked to immigration status, in relation to either asylum claims, or entitlements to benefits (Weishaar, 2008). This was complicated by immigration status-related stigma, combined with a lack of a wider support network to facilitate building knowledge of health care systems (Feldman, 2006). For some, this translated as a reliance on service providers (health care practitioners and interpreters)particularly in the 'community' settingto provide additional support in terms of signposting, advice and information. Interpreters in particular explained that it was commonplace to be asked by a patient for help and support with letters from the UK benefits agency, housing or UK Home Office. This highlighted the 'navigator' role which they were expected to assume, thus marking a shift beyond the 'conduit' role previously identified (Hsieh, 2006). Health care providers tended to deny that they too expected this of interpreters, whilst acknowledging the benefit of interpreters' bringing their experiential learning to bear in appointments, especially where they themselves lacked knowledge of wider supports such as community groups, or local faith groups. I think to be honest, with interpreters with asylum seekers, you did find that they were all fairly clued up, and they did have a really broad knowledge base because they were doing a lot of work with asylum seekers … you know, to signpost them […] it did help I suppose if they were telling me something I didn't know, and there's always things you don't know, so signposting does help. (HV2) There was a mixed response from interpreters on their desire to perform any 'navigator' role. Two felt uncomfortable about this, that it pushed their professional boundaries, while the three others interviewed could see it as part of their remit as a community interpreter. When faced with the various role expectations of others, they discussed the adoption of specific 'performance' strategies to manage these conflicts and act out their interpreter role: The interplay between migratory trajectory and knowledge also extended to healthcare providers. It became clear during interviews that a lack of background knowledge affected health care practice and practitioners' abilities to make appropriate health assessments, particularly because a medical record was usually unavailable. I sometimes do try to find out about where exactly the patient is in the immigration process, because it can affect how they present and what they present with. My understanding of the asylum process is poor and we could probably do with some training on that … for example if someone has been trafficked … But we don't usually have that information. And probably trying to elicit that at the start of an appointment, then it might be seen to be you are going to drop somebody in it with the authorities. (SHS2) Again here, this openness to intercultural communication and understanding beyond language match offers opportunities for better practice and demonstrates how a move away from a language problem focus can contribute to health care provision underpinned by social justice values. Where training on the asylum and immigration system had been available, it was considered helpful: A few of us went on a maternity action asylum seeking training day for midwives. So we do have an understanding of the asylum process [… ..]  The cultural differences of patients, depending on their migratory route and previous experience and knowledge of health care systemsin addition to language capabilitypotentially impacted upon the level and quality of health care the health practitioners were able to offer: You don't have access to any level of past medical history in different countries … so you try to get a good history. Obviously, their level of understanding can be quite poor too, and if they have had access in their own country then it is very different to here. (CMW1)

Intersecting differences and hidden inequalities
Despite health care providers' efforts to provide an equal service to all patients, the asylum seeker and refugee service users felt stigmatised due to their immigration status; this often began in the clinical reception area, when health centre receptionists would question them on their immigration status: When I arrive at reception, the receptionist always asks me if I am an asylum seeker. In front of everyone! If I need another appointment, they [health care receptionist] book it at the end and ask me if I need another interpreter and if I am an asylum seeker. Why do they need to know? (SU4) The current political rhetoric in the UK often frames migrants as undeserving 'health tourists' (Home Office, 2014), fuelling wider public debates about certain types of migrants being a 'drain on resources' (Hanefeld, 2017). Following Castañeda et al. (2015), we also found that deservingness was inflected by immigration status, as well as language and accent. Service users in our study reported that the emphasis on their immigration status was an unanticipated focus in the clinical appointment, which left them feeling that it overshadowed their physical complaint. 3 This intersection of culture, ethnicity and migrant status compounded difficulties, resulting in a lack of trust in clinicians and health support staff and a feeling of not being taken seriously.
I have this back problem […] I told the doctor about it and he referred me to physio. She said you are not responding and after three sessions I am going to discharge you. I said why? I want to do the exercises. When we fled from Pakistan, I couldn't bring my record. First they were not ready to believe I had this problem, they thought I was traumatised and that's why I am feeling pain in the back and neck. She thought there isn't real pain […] just it was related to my asylum. Even at one point she said you need a psychologist not a physio. And when I told the doctor and he agreed [with the physio]. I said, see, this is an insult. (SU1) Even when there are well-intentioned efforts to provide equitable health care services, the dynamic interplay of overlapping variables including country of origin, ethnicity, language, regional and local identities, and migration experience all added to the complexity of the interactions. For example, there was some evidence of potentially stigmatising narratives amongst practitioners in relation to the credibility of individual narrative's relating to immigration status: I'm just laughing because sometimes you'll ask what their status is and they'll kid on they don't know … or they just [shrugs her shoulders]. (CMW3) The interpreters interviewed also reported incidents during consultations they attended whereby negative comments about the patient would be made either directly to them, or to the patient about their rights and entitlements to services. These were bound up in wider socio-demographic factors relating to ethnicity, education and health literacy, class and migratory trajectory. This Indian doctor came in, he wasn't even the consultant … he did ward rounds. And I introduced myself as the interpreter, and he blanked me out, talking to the patient -'you understand me?'and the patient said 'uhhuh'and he said 'so you don't even need an interpreter'. He asked him 'how long have you been in this country?' and the man said '15 years' and he said 'so you have been here 15 years, and you haven't learned English, that's terrible, you should know how to speak English and not need an interpreter'. He said this and I was standing right there. The patient felt really bad. And he said well 'I came here when I was very young. I was always working in takeaways in the back kitchens, I never got the chance to learn English and I'm always working, and I always wanted to learn, but now I am too old to learn and I've got responsibilities, so I need to keep on working' … and I'm translating all this back to the doctor. Obviously he doesn't have enough English to say all this so he's saying it in Punjabi, and the doctor is saying in English 'no, no, no … there is no excuse, if you live in a country you have to learn the language' … (INT4) The above clinician's approach illustrates the intersectional nature of social identities, how this comes to influence presumptions about the 'other' and subsequent social interactions (Anthias, 2012;Byram et al., 2002). In this example the patient's immigration status became socially classed and genderedhis socio-demographic background influenced the type of work he was able to find as a male Pakistani migrant worker. These data are revealing, showing that in frontline health care work with migrants, multiple social structures and processes interweave to produce specific social positions and identities (Anthias, 2012).
The interpreter in this incident also found it difficult to challenge the clinician, especially as her own position in the interaction was being devaluedby being ignored. The clinician was perceived to be in the more powerful social and professional position in the interaction, and the interpreter felt, like the patient, it was not her place to challenge him. Both here and in the earlier discussion, our data reveal ways in which practitioners' own attitudes can influence their how they do their job. Discriminatory practices often go unchallenged not just because patients do not always have the language to do so, but also because practitioners employ a deserving/undeserving narrative as a means of asserting their knowledge, status and power in how they do their jobs. This reflects the many complex asymmetrical and uneven relations of power as they play out in a triadic interpreted consultation (Bhatia & Wallace, 2007;Derose et al., 2007;Ferguson & Candib, 2002;Priebe et al., 2011;Shackman, 1984). Moreover, the interpreter's own precarious employment status as a sessional worker on a zero hours contract undertaking work on an ad-hoc basis to meet service provider needs meant she did not want to appear troublesome for fear of losing appointments. Again, here, there is a jostle for power with potentially very real and material consequences for interpreters, who see themselves at risk of losing work if they 'speak out'. Consequently, interpreter silence may be understood as a tacit agreement, thus making them complicit in discriminatory work practices.

Intercultural communication in non-clinical 'community' settings
As described in the introduction, most research on health care interpreting to date has been conducted in hospital or primary care clinical settings, such as the family practice. However, the 'home' or 'community' setting represents a significant area of practice for midwives, health visitors and to a lesser degree family practice doctors and mental health practitioners. It is a place where age, gender, socioeconomic background, ethnicity and immigration status intersect in complex ways, presenting different challenges both to the delivery and experience of health care and interpretation services. An important finding in our study was that the non-clinical setting presented quite distinct challenges for health care providers and interpreters dealing with the different demands of superdiverse communities. Although they still operated on an appointment system, the health visitors and midwives felt there was greater opportunity in this more 'relaxed setting' to explain to patients what to expect from their care, especially maternity care for female patients. The home setting also provided a supportive context for empathic professional practice that was more attuned to developing a relationship between the health care provider, interpreter and service user.
Sometimes we think professionalism means distance when it actually means a good relationship. (HV1) Context is particularly relevant when viewed through the lenses of intersectionality, diversity and intercultural communication. Context here relates not only to the physical location of the health care interaction, but also to socio-demographic factors, particularly the classed and gendered locations of participants within that social interaction. Status differentials which may be apparent in the clinical setting shift in the 'home setting', where the health care provider and interpreter occupy a space that is unknown and unfamiliar to them, and where they are effectively the 'guest' of the service user.
The health care providers and interpreters identified the home setting as a space and place where attentiveness to intercultural communication and respect of cultural values and beliefs had to be sharper, in order for health care to be effectively delivered specifically around privacy, intimate examination, managing the body in a community space and dealing with sensitive issues. Getting the job done effectively required creativity, particularly when there was not the same access to the objects of practice to aid intercultural communication, such as privacy screens and private rooms.
Although it might appear a potentially more relaxed setting, the home is also a place where the multiple dimensions of difference interact in distinct ways revealing a nuanced interplay of ethnic and culturally related and socio-demographic variables. Health care providers and interpreters spoke of having to contend with managing boundaries, respecting the home space, working spontaneously with unpredictable seating arrangements, respectfully managing hospitality and negotiating cultural norms around communication. There were often other people present (children, partners, extended family members, friends), potentially resulting in multiple languages being spoken during the appointment, and multidirectional conversations occurring in ways that had not been anticipated.
There's more noise, more people, it can be more of a challenge, but I think sometimes the patient prefers it as they are more relaxed. (INT1) Interpreters and service users described the home visit as one that was potentially more empowering for the patient, giving them more opportunity to ask questions and engage with the health care provider. We suggest this, albeit slight, shift in power relations is an important feature of home setting health care that is often overlooked and underplayed; it potentially provides a foundation for service user action to redress the power imbalances that often reproduce health inequalities in the clinic setting.
I can ask some more questions, sometimes when you are with the doctor, you feel you don't have time, but when it's at home, there is more time, to listen to you, you know. (SU4) The community setting is also a place where the health-related effects of adapting to life in a new setting can come into sharp relief. As described earlier, support was often required on a wider range of issues which impact on health outcomes and behaviours (e.g. housing, welfare, employment, education) and which can affect migrants in particular ways, especially in relation to rights and entitlements (Bolderson, 2011;Weishaar, 2008).
When you are in the home setting, you deal with all sorts, not just health … Letters from Benefits, schools. I see that as our job too, if we have time. (HV2) It can be argued then that the concrete effects of immigration policy play out in the delivery of health care and are always present. This is not just in the physical space, which is particularly relevant for asylum seekers who are often dispersed to areas of poor quality social housing, but also in the range of topics covered during a health care home visit that extend beyond the strict remit of 'health care'.

Final discussion and conclusions
This paper has set out some ways in which the challenges of intercultural communication and the practical problems of linguistic diversity and migration trajectory implicitly and explicitly shape different aspects of the health care encounters between service users and service providers of culturally diverse backgrounds. The location of such encounters, whether in the community-based clinic or home setting, also influences the interactions between practitioner, interpreter and patient. Our work found that acknowledging the diversity of migrant populations is a helpful vantage point from which to view the newness and novelty of increasing diversity in our setting (Phillimore, 2015), and that the intersectional perspective proved fruitful in exploring the negative politics of hidden multiple inequalities (Viruell-Fuentes et al., 2012;Werbner, 2013).
Our findings reveal the need to widen out axes of inequality in health care encounters to include migration-related variables, such as migration trajectory and status and indeed the conditions of stay. We argue that attending to how these variables intersect specifically with language can provide better insights into how inequalities are reproduced and maintained, as well as indicating ways they may be challenged, by service users, interpreters and health care providers. Moreover, this approach allowed us a better understanding of the challenges and tensions of intercultural communication in interpreted consultations and revealed how certain axes of difference, particularly language and ethnicity-related factors, come to be privileged over others, for example immigration status, rather than building understanding around the more nuanced interrelationship at play. Our findings also confirm that over-simplistic understandings of difference and too narrow a focus on language barriers in the development of health policies and services fail to adequately describe the intercultural complexities involved providing health care and developing wellbeing to increasingly diverse populations.

Strengths and limitations of the study
It must be emphasised that the healthcare providers who chose to participate went to great lengths to work in culturally sensitive ways with their patients under sometimes very difficult conditions; that they were willing to discuss this in detail is a key strength of this work. Another has been our ability to explore the separate perspectives of health care practitioners, interpreters and service users in one study. This has proven illuminating and in our view offers a foundation for further work exploring intercultural communication of the three parties in the same encounter. Finally, we have explored these experiences in the community and home setting, a largely unexplored site of health care provision with quite specific yet subtle shifts in roles and boundaries, and greater direct exposure to multiple variables, which co-condition integration and settlement outcomes. Although such variables might not always be directly related to health issues, they will clearly affect help-seeking behaviour and health outcomes.
This study did have limitations, however. We were not able to explore specific encounters between service providers, users and interpreters, partly due to the well-recognised methodological challenges of conducting such ethnographic work in a second language (Hsieh, 2007(Hsieh, , 2008. However, we believe this contribution marks an important step away from the dominant singular perspective models towards a more integrated account of the interpreted encounter. The study is small in scope, and a larger sample of service providers would have allowed a comparative analysis of practice across health provision more widely. This would also allow for a more nuanced analysis of the privileging of gender for example in health care provision, as our community midwives and health visitors were all female. The work of community midwives and health visitors affects the potential to examine the implications of gender and how this intersects with language, culture and immigration status. This, however, has not been addressed in this paper primarily because this was not a central feature of our study, which had an over-arching aim of developing educational material across healthcare settings and genders. Moreover, our service providers and interpreters reflected on work with men and women. Nonetheless, we think this would be an interesting and valuable area of further inquiry. We faced particular challenges in recruiting the interpreter sample due to interpreter availability; scheduled interviews were often cancelled at last the minute because of work commitments, thus highlighting a real challenge in doing research with this group of professionals. The service user sample was skewed slightly towards the experiences of asylum seekers and refugees, who have a number of non-health related needs that can directly affect health outcomes. This reflects the service links that the study team have; a larger sample of service users would have allowed better comparison and contrast between migration trajectories. Nonetheless, the data from interpreters suggested that immigration status, migratory journey and ethnicity-related factors are equally relevant for non-asylum service users. The effects of this interplay on health inequalities is an area that requires more work. Our language sample was also limited, and broadening this would provide an opportunity to analyse ways in which certain languages come to be mediated by other hidden inequalities as these might relate to assumptions about 'race', ethnicity, education and socio-economic background (Derose et al., 2007). Despite these limitations, the participants and their experiences were sufficiently varied to enable us to develop a conceptual understanding of how health care providers, interpreters and service users navigate intercultural communication.

Future directions
Reflecting on future research directions, an intersectional approach to intercultural communication holds considerable promise for improved understanding of how the complex interplay of social differences is navigated and performed. In the UK, present immigration legislation proposes restrictions to health care for certain migrant categories, exposing the urgent need for more research into the structural effects of immigration policy on health care access and delivery. The participants in this study demonstrated their ways of coping in complex social interactions, highlighting the value of more practice-evidenced research into the experiences of accessing and providing health care to an increasingly diverse population. Our data revealed how discriminatory practices might reproduce and widen inequalities in access to healthcare for specific migrant populations. This suggests more training measures are needed that move beyond diversity and/or race-awareness training and which employ a more rounded approach to understanding how varied social identities and multi-dimensional markers of difference come to be produced and reproduced in healthcare encounters. In addition, interpeters may also require additional support particularly in challenging encounters when they may wish to report instances of discrimination in healthcare practices without the risk of losing work.
In terms of moving forward, part of the broader programme of work will be to work with key stakeholders (National Health Service, academic institutions, interpreting service providers and third sector organisations) to focus on embedding the project film outputs into existing or developing educational and training programme on rights, entitlements and good practice, within an anti-racism and anti-discrimination framework. Finally, any future research agendas on intercultural communication in health care should begin to address the 'community/home' setting, where arguably there is an even greater need for interculturally sensitive, empathic practice and awareness of the intersections of health, language, immigration status and culture that produce specific social positions.

Notes on contributors
Dr Teresa Piacentini is an experienced researcher, interpreting practitioner and activist and she has spent most of her professional and academic career working and researching mainly with asylum seekers, refugees and migrants in Scotland. She is a steering group member and Co-Convenor of the Glasgow Refugee, Asylum, and Migration Network (GRAMNet (http://www.gla.ac.uk/research/az/gramnet/about/). She is a Lecturer in Sociology and teaches migration studies across the undergraduate curriculum and since 2000 she has been conducting research broadly into the experiences of refugee settlement in Scotland. Her research is focussed on concerns with settlement, identity and belonging; ethical interpreting in health care provision; researching anti-discriminatory practice in health care interpreting; community activism and mobilisation; and refugee camps and patternings of settlement experiences in France and Jordan.
Professor Catherine O'Donnell's research interests lie in the broad field of migration studies covering the various aspects of social, cultural and political life affecting migrants' experiences of 'settlement', integration and belonging. A primary healthcare scientist, her research focus on the organisation and delivery of primary care services, particularly for marginalised populations, and the evaluation and routinisation of primary care policy into practice. She has a particular interest in the application of theory to research and in the integration of mixed methods in communitybased research. A key area of work is migrant health, in particular the role of language and communication in cross-cultural consultations in primary care and understanding the experience of asylum seekers, refugees and migrants when accessing and using primary care services in both the UK and Europe. She is an Honorary Fellow of the Royal College of General Practitioners, a member of the Advisory Board for the European Forum for Primary Care, and the current Chair for the Society for Academic Primary Care.
Professor Alison Phipps holds a UNESCO Chair in Refugee Integration through Languages and the Arts at the University of Glasgow and Professor of Languages and Intercultural Studies. She is Co-Convener of Glasgow Refugee, Asylum and Migration Network (GRAMNET). She was Distinguished Visiting Professor at the Waikato University, Aotearoa New Zealand 2013-2016, Thinker in Residence at the EU Hawke Centre, University of South Australia in 2016, Visiting Professor at Auckland University of Technology and is Principal Investigator for the £2 million AHRC Large Grant 'Researching Multilingually at the Borders of Language, the body, law and the state.' In 2012, she received an OBE for Services to Education and Intercultural and Interreligious Relations in the Queen's Birthday Honours. She is a Fellow of the Royal Society of Edinburgh, the Royal Society of Arts and of the Academy of Social Sciences. She has undertaken academic and artistic work (including producing and directing theatre) in, amongst others, Palestine, Sudan, Aotearoa New Zealand, Australia, Jamaica, Ethiopia, Germany, France, USA, Portugal, Ghana. She has produced and directed theatre and worked as dramaturg and creative liturgist with the World Council of Churches from 2008 to 2011 during the Decade to Overcome Violence.
Dr Ima Jackson is a Senior Lecturer at Glasgow Caledonian University. She is an experienced clinician, researcher and project manager and she has spent most of her career working with marginalised groups: initially pregnant women in the poorest parts of London and Glasgow, and in more recent years with refugees, asylum seekers and other migrants in Scotland. She is a community engaged researcher and works with migrant and established communities of people of colour, supporting them to evidence their experiences towards the Academy and towards policy.
Dr Niamh Stack is a Senior Lecturer in Psychology. She is currently the Director of Teaching for the School of Psychology and as part of this role, she is interested in research on effective pedagogy, curriculum enhancement and graduate skills. In addition, she is the Research Director for the Scottish Network for Able Pupils which is hosted in the University of Glasgow and she researches and publishes in the areas of high ability/giftedness and atypical development.