Background: Communication with health care providers represents an essential part of access to health care for the over 230 million cross-border migrants around the world. In this article, we explore the complexity of health communication from the perspective of cross-border migrants seeking antenatal care in Cape Town, South Africa in order to highlight the importance of high quality medical interpretation. Methods: As part of a broader study of migrant maternal and infant nutrition, we conducted a secondary data analysis of semi-structured in-depth interviews (N = 23) with Congolese (n = 7), Somali (n = 8) and Zimbabwean (n = 8) women living in Cape Town, as well as nine focus group discussions (including men: n = 3 and women: n = 6) were conducted with migrant Somalis, Congolese, and Zimbabweans (N = 48). We first used content analysis to gather all data related to language and communication. We then analysed this data thematically. Results: Zimbabwean participants described how the inability to speak the local South African language (IsiXhosa) gave rise to labelling and stereotyping by healthcare staff. Congolese and Somali participants described medical procedures, including tubal ligation, which were performed without consent. Partners often tried to play the role of interpreter, which resulted in loss of income and non-professional medical interpretation. Participants’ highlighted fears over unwanted procedures or being unable to access care. Challenges of communication without a common language (and without professional medical interpretation), rather than outright denial of care by healthcare professionals, mediated these encounters. Conclusion: Although there are several factors impeding cross-border migrants’ access to health care, effective communication is a prerequisite for quality care. Free-to-patient professional medical interpretation would not only benefit migrant populations but would benefit the broader community where language and health literacy are barriers to accessing health care. Novel approaches to language access may include technology-enabled professional interpretation.
Fieldwork for this qualitative study was conducted between February and October 2013, and the findings presented in this paper are based on language issues raised during in-depth interviews and focus group discussions. The results presented in this paper are derived from a broader study of cross-border migrants experiences of maternal and infant nutrition in Cape Town, South Africa, where language and communication emerged as an unintended theme [11, 12]. We hold to the constructionist tradition [13] that meaning and analyses are constructed rather than fixed, and this analysis represents a contribution to our understanding of patient experiences of multilingual medical encounters in a specific context of cross-border migration to a LMIC. The study was conducted across greater Cape Town. Cape Town is in the Western Cape province of South Africa where 28% of residents are not born in the province and up to 9% are born in another country [4]. The participants in this sample predominantly resided in rooms with their partner and children, sharing amenities with other families, either in a larger house or apartment, or in illegally subdivided warehouses. Most interviews took place in migrant homes, whereas focus groups took place primarily in more communal settings, including community centres, shops, and a women’s shelter. While one focus group took place with residents of an informal settlement with shack housing, participants more commonly resided in inner city settings across many parts of Cape Town. Although the participants in this sample largely occupied these relatively crowded inner city spaces, it should be noted that cross-border migrants in Cape Town have a wide range of educational and socio-economic backgrounds, and have a broad range of professional roles in South Africa. The experiences of participants in this sample were thus grounded in a specific socio-economic context, in particular, of life in the inner city, and to a lesser extent, in informal settlements. Study participants were selected to participate either in in-depth interviews or in a focus group. Participants had typically resided in South Africa for less than 10 years. For the in-depth interviews, 23 Somali (n = 8), Congolese (n = 8), and Zimbabwean (n = 7) women were purposively selected to include different migrant groups who could provide diverse insights. The interview inclusion criteria was: women over the age of 18 who were pregnant or had given birth in the last 2 years, and self-identified as Somali, Congolese (from the Democratic Republic of Congo, DRC), or Zimbabwean. The majority of in-depth interview participants were married. Nine separate focus group discussions (N = 48) where held with adult Somali, Congolese, and Zimbabwean men (N = 3; n =21) and women (N = 6; n = 27), segregated by country of origin and gender. In the parent study, in-depth interviews were used to attain depth of knowledge about individual perceptions of maternal and infant nutrition, while focus groups provided perspectives on collective meanings and understandings—the collective narrative [14]—related to maternal and infant nutrition in Cape Town. Both in-depth interviews and focus groups consisted of a convenience sample using snowball sampling, an effective technique among hard-to-reach populations [15], and used both non-governmental organization (NGO) contacts and individual introductions to begin meetings with eligible cross-border migrant participants. These NGOs worked across different parts of the city, with various populations. Including both in-depth and focus group discussions, there were 26 Congolese participants (16 women, 10 men), 21 Zimbabwean participants (16 women, 5 men) and 24 Somali participants (18 women, 6 men). More women than men were sampled due to the original study design which focused on maternal and infant nutrition. This may mean that more language issues were described, as women were less likely to have paid employment or to have had other opportunities to become conversant in English. Nevertheless, maternal care is also one of the key points of care for migrants, and thus language in the specific context of maternal care is an important issue for migrant health [16, 17]. JH-A conducted all the interviews and focus groups. Two focus group discussions, four Somali in-depth interviews and one Congolese in-depth interview also included a trained interpreter to assist JH-A. In-depth interviews lasted between 1 and 1.5 h, and were primarily conducted in participant homes. Focus group discussions took place in community centres and lasted between 1.5 and 2 h. The objective of the broader study was to garner meanings and understandings and dominant discourses related to maternal and infant nutrition in a migrant context. An interview guide framed the semi-structured in-depth interviews [12]. When participants discussed health care it was frequently in relation to accessing healthcare during pregnancy, and how participants felt this related to feeding decisions. Questions in focus groups broadly related to comparing experiences of food provision and nutrition advice for pregnant mothers and new babies living in South Africa versus practices in participants’ countries of origin. In this paper, the focus is on participants’ unprompted discussions that related to language and healthcare provision in Cape Town public hospitals, as well as primary health care clinics. Our use of interpretation in this study has been briefly described in a previous publication [12]. Nevertheless, it is important to discuss in more detail issues of meaning and language related to the our research process, to make the multiple languages used in the study more visible and consider the implications related to conducting a study in multiple languages [18]. Interpretation was not clean-cut; in Cape Town there are multiple South African languages spoken by residents. Amongst Congolese migrants, participants often used multiple languages in a single conversation. As such, making meaning, and settling on specific choices of words in transcripts, involved much negotiation [19]. This negotiation began with the process of simultaneous interpretation, and weighing how to manage interviews alongside interpreters, who were situated as a type of bridge between the participant and the interviewer. In in-depth interviews, we found that trying to make interpretation overly formal circumscribed free conversation, while allowing the interpreter to be a third party in the conversation, made the participant more comfortable. Larkin and colleagues [20] termed this process “mutual reciprocity”. Zimbabwean participants were comfortable using English in the interview setting, though they may speak English, Shona, and/or Ndebele amongst family and friends. Somali participants typically preferred to speak Somali, whereas Congolese participants preferred a mixture of languages (French and either Swahili or Lingala) or had become fluent in English through work or studies in Cape Town. Independent professional medical interpreters then checked the quality of interpretation (comparing audio to English transcripts), discussed the role of the interpreter, and discrepancies and alternative meanings with the moderator/interviewer (JHA). Transcripts were then revised and discussed with the independent professional interpreter to enhance the quality of the final “verbatim” transcript. We recognize that the process of speaking on behalf of any participant group, particularly when engaging in research, presents potential research limitations. In this paper we focuses specifically on categories related to language and communication. These categories were identified during overall data analysis of the broader research question. The broader data analysis was guided by principles of thematic analysis [21, 22]. This analysis began during fieldwork in the form of a research diary, notes, and reflections. After immersion in both focus group and interview transcripts through reading and rereading, an inductive initial analysis was used to generate a codebook and code all transcripts [23]. After developing and defining an initial set of codes, we began to code transcripts and test validity of the codes and the extent to which they seemed to convey the meanings and understandings presented by participants. Once no new categories of codes appeared, the codebook was considered complete and coding of all transcripts- both in-depth interviews and focus groups- were coded using the complete codebook. All transcripts were uploaded to the computer software Hyperresearch (Researchware Inc., 2009, Massachusetts, U.S.A.), to assist with coding, sorting, and data management. Based on our analysis, language and communication in health care settings represented a key theme through which participants understood their experience of maternal and infant nutrition in Cape Town. As such, we performed additional content analysis [24] to focus specifically on instances involving language barriers. This represented a secondary data analysis design, and was conducted by the same researchers who designed, implemented, analysed, and reported on the primary data analysis [25, 26].
N/A