“Imagine if I'm not here, what they're going to do?”—Health‐care access and culturally and linguistically diverse women in prison

Abstract Background Women in prison have complex medical needs and poorer health status than the general population. Culturally and linguistically diverse (CALD) women in prison, particularly those with limited English proficiency (LEP), have distinct needs and risk additional isolation, discrimination and marginalization when they are in prison. Objective We sought to examine how cultural and linguistic diversity, particularly LEP, affects the health‐care experiences of women in prison. Design, Setting and Participants We conducted focus groups and semi‐structured qualitative interviews with CALD women and frontline nursing staff in the three female Correctional Centres in New South Wales, Australia. Results Participants comprised 30 women in prison and nine nurses. Both women and staff reported communication difficulties as a significant and additional barrier to accessing and receiving health care. For some women with LEP, barriers to care were perceived as discrimination. Fellow prisoners were often utilized as support persons and informal interpreters (“peer interpreters”) in place of formally trained interpreters. While peer interpreters were perceived as useful, potential challenges to their use were vulnerability to coercion, loss of confidentiality, untrained health advice and errors of interpretation. Conclusion The persistent use of peer interpreters in prison is complicated by the lack of clearly defined roles, which can include informal peer support roles and lay health advice. These are highly complex roles for which they are unlikely to be adequately trained or supported, despite perceived benefits to their use. Improved understanding and facilitation of health‐related communication could enhance equity of access for CALD women in prison.


| BACKG ROU N D
In prison, women can experience a profound loss of autonomy with respect to their ability to manage their health. [1][2][3] Women in prison have complex medical needs and poorer health status than the general population. 4 However, prison can also provide new opportunities for access to health care. 1,5 In 2014, 20% of the 683 females in full-time custody in New South Wales, Australia, spoke a language other than English at home and almost a quarter (23.6%) were born outside Australia, predominantly Vietnam. 4,6 The number of women in prison continues to rise. 7 Culturally and linguistically diverse (CALD) women in prison have been described as the "silent" or "forgotten" few, [8][9][10] referring both to the limited research relating to them 11 and to their additional isolation, discrimination and marginalization within the prison system. [8][9][10][11][12][13] Reports on Australian women's prisons describe barriers to communication with staff and other prisoners for women with limited English proficiency (LEP), including a lack of access to interpreters, to information about prison processes and legislative rights, and to programmes and educational opportunities in their own language, and reduced access to religious practices and ministers. 14 The use of professionally trained interpreters in prisons in Australia and overseas is seen as being suboptimal. 12,15 Using professionally trained interpreters in health care improves the quality of clinical care more than using ad hoc, or no, interpreters 16 and significantly reduces the likelihood of errors. 17 NSW health policy, which applies to the prison health service, mandates that professionally accredited interpreters must be engaged for health-care communication for all patients who are considered by the health practitioner to not be fluent in English. Additionally, all patients should be informed of their rights to an interpreter, except in medical emergencies or where there is a bilingual health practitioner communicating directly with the patient. 18 Informal or untrained interpreter use is widespread in health settings despite being problematic. 19 While patients commonly report a preference for formal interpreters due to the perception of higher quality interpretation, key reasons for use of informal interpreters (often family members) include personal trust and rapport and advocacy. 20 The unique prison context and added vulnerabilities of patients in prison further complicate the issue of interpreter use, especially as informal interpreters are likely to be fellow prisoners, but there is limited research in this setting. Two qualitative studies of interpreter use in Spanish prisons have reported professional challenges of interpreting in the prison environment 21 and the ethical difficulties associated with use of fellow prisoners as interpreters. 15 The use of interpreters in the context of prison health care was not examined, nor did the research focus on the experiences of the prisoners themselves.
In this research, we aimed to explore the complexities of communication and interpreter use for CALD women prisoners accessing prison health care, with a view to improving service access for CALD women in prison. Our research questions were as follows: 1. What is the impact of cultural and language difference on accessing and receiving health care in prison?

2.
How do women in prison and health-care providers manage these impacts?

| ME THODS
Using an inductive qualitative approach, we conducted focus groups and individual interviews with CALD women and individual interviews with prison health nurses. This study was undertaken in connection with a larger project into health-care transitions of women leaving prison. 22

| Setting
Interviews were conducted in three women's prisons in New South Wales, Australia. They ranged from low-to high-security settings, and one was the remand prison for the state. Health care in these correctional centres is delivered under a Board-governed network under the state health department. It is predominantly a nurse-led model of care, which requires patients to be triaged by nursing staff prior to seeing medical practitioners. 22

| Sampling
Sampling of the women in prison was purposive for variation in age, cultural background, length of custody, health conditions and health-care utilization to increase data richness. 23 We defined CALD women as those women who were born overseas (in a country that did not have English as its primary language) or those who were born in Australia and spoke a language other than English at home. We excluded those born in Australia who spoke English at home to focus on the impact of cultural and linguistic differences. 23 Potential participants were identified by nursing and custodial staff and by review of a list of current inmates. They were then invited by nursing or custodial staff to meet with the researchers. Peer recruitment also occurred by asking a nurse-nominated woman to invite other women or friends they felt would be interested in participating. Nurses were purposively sampled to include a variety of patient care roles.

| Data collection
KW conducted the focus groups and most of the interviews, with PA conducting 2 interviews and 1 individual interview jointly with KW.
As PA was a general practitioner at one of the research sites, she did not conduct any staff interviews or interviews with women to whom she had provided health care.
Interview questions covered how cultural and linguistic diversity affects prison life and health-care delivery, peer and formal interpreter use in prison and the women's experiences of health care in prison. We defined a peer interpreter as a fellow prisoner from the same cultural background who spoke more English than those they were assisting, but who was untrained in interpreting. A formal interpreter was defined as an interpreter from outside the prison with professional qualifications and training in medical interpreting, used via phone or face-to-face.
We obtained written informed consent after explaining via the participants' chosen interpreter, offering the women participant information and consent forms in English or translated in the woman's language. Formal interpreters were offered to all women. To ensure the credibility of the data when informal interpreters were used, we explicitly explored the interpreter's perspectives on the interview topics during the focus groups to identify any points of divergence from non-interpreted interviews. 24 In the focus groups, at times, two or three people acted as informal interpreters for the other women, giving multiple points of interpretation. We also used communication techniques to foster mutual understanding by avoiding multicomponent questions, beginning with open-ended questions and progressing to more specific queries ("the funnel approach") 25 and checking the interviewer's understanding of the response. 26 Interviews and focus groups were audiotaped and spoken English transcribed verbatim. Two women also provided written data during the focus groups, which they explained was a summary of issues they had experienced with the health service. Focus groups containing background discussion in Vietnamese (the only language spoken other than English in the focus groups) were professionally translated and transcribed into English. These transcripts were then discussed with a bilingual general practitioner who acted as a cultural advisor. Both methods of translation aimed to provide cultural context to the data and to reduce the risk of distortion of the results through translation. 24,25,27,28

| Ethical issues
Transcripts were deidentified and stored securely. Any urgent or serious clinical issues that had not previously been addressed arising from the interviews were passed on to health staff, with the participant's permission, for follow-up through formal routes. Interviews were conducted in private health clinic rooms, in prison visitation rooms or in privacy in the prison cottage dwellings. Guards and health staff were present nearby but could not hear or see the interview. Given the constraints of prison access, these were considered the most neutral space available. 25,27 Interviews with the nurses took place in private rooms in the health clinic. A $10 AUD payment was made into each woman participant's in-prison account in keeping with usual research practice in NSW prisons.

| Data analysis
Thematic data analysis was undertaken, 29 facilitated by the use of NVivo software (version 9, QSR). Transcripts were initially opencoded by KW and then refined. During this stage, PA and WH independently analysed a selection of transcripts and emergent ideas and concepts were iteratively discussed and tested by returning to the data to develop preliminary themes. Data from women and nurse participant groups were initially analysed separately and then compared across and between groups to provide different perspectives and arrive at the final themes. Memo writing was used extensively throughout the process to provide an audit trail. These themes and subthemes are presented below with illustrative quotes.

| The impact of cultural and language difference on prison life
All participants reported that women from CALD backgrounds experienced substantial barriers to health care and difficulties in everyday prison life, particularly isolation, difficulty in adjusting to prison life and loss of autonomy. These barriers were greatly increased when women had trouble communicating in English and even more profound if there were no other women who spoke the same language at their prison. Some women perceived racism and discriminatory treatment from staff to be the main cause of disrupted health care, such as cancelled appointments and delayed investigations.
Social networks of women from the same language group increased the availability of peers for interpreting, knowledge sharing and support. Informal networks had developed so that women sought medications such as creams or paracetamol from other women, rather than go to the health service themselves. Receiving treatment and medications from health staff could be seen as a victory for the group.

| Health Communication and the use of interpreters
Both women and nurses reported that communication difficulties were the most significant barrier to health care for CALD women in prison.

| Deciding on interpreter use
Some nurses felt that formal interpreter services were underutilized. These staff frequently used interpreters and valued them for their ability to improve ease of communication and offer confidentiality for the woman, while also improving the therapeutic relationship because they regarded arranging one as an act of respect. A few nurses did not perceive any significant communication problems and did not use interpreters because they felt they largely dealt with simple health issues for which limited English sufficed.
I think unless there is a formal interview, then you don't worry too much about the interpreter, because you will get across-they'll get across whether it's a cold, or it's a sore ear, or sore throat. I don't have a problem language wise.
(Nurse 1) Most women felt that a formal interpreter should be offered for all significant health interactions, unless the complaint was minor, such as a common cold. (laughter)

VS1: Yeah. Now let's say that person is competent in ah… this industry and someone else is competent in, specialises in another industry in interpreting. It's not as if you can interpret in whatever situation you want. Then you don't interpret correctly, they don't understand those meanings, in specialised area. And specialised in ah… ah… the breasts let's say and you send an interpreter who speak in another matter then he won't know what the special term for breast is.
VS2: Wait let her say a few words back to her or else she will think that we are saying something bad about her.

| Using formal interpreters
Some participants, both women and nurses, expressed ambivalence about formal interpreters. There was particular concern that formal interpreters were not always accurate, including in their medical interpretation. Face-to-face interpreters were thought by some to be too expensive and time-consuming to organize. Unpredictable prison schedules and transfers meant that some nurses and women preferred to expedite the consultation using the resources at hand, rather than wait for a formal interpreter (including telephone interpreters) and risk not having a consultation at all.

| Peer Interpreters
Peer interpreters were reported to be preferentially used by prison and health staff. Some women and staff felt that, at times, peer interpreters could be better placed as communication brokers, as they were more likely to know the woman and the prison system and to use language that was adjusted for the woman's level of understanding.
One time this lady-she still got me there but the doc-

| Perceived challenges with peer interpreters
Some nurses and women of LEP were concerned with the lack of confidentiality of peer interpretation. Additionally, prison dynamics could mean women were vulnerable to private information being used against them, and peer interpretation could be affected by conflicting agendas. she's our people and she wouldn't even help us.
(Woman participant 1-peer interpreter) Peer interpreters could also be put in the position of assisting, or choosing not to assist, other women when they did not think their claims were reasonable.
Some of these problems that she's got is actually visible, you can see it, and then there's just some really outright silly ones where I think that's just a bit selfish on her behalf to be asking clinic staff about certain stuff like that.
(Woman participant 14-peer interpreter) They described unease at being caught in conflicts between women, health staff or prison staff, or if they could be seen as complicit if the interpretation involved informing on other women or required interpretation of very personal or bad news.
Nobody wanted to interpret for this lady 'cause apparently she was a trouble-maker…Even though I knew she was causing problems I felt bad because nobody wanted to do it for her, so I did it…and it was very private and I didn't want to know that stuff.

| D ISCUSS I ON
Women entering prison may suffer shock, fear and disempowerment. 2 In this study, both the CALD women in prison and their treating nurses found communication across language and culture challenging and that this could potentiate stressful prison experiences and further disempower the women. Interpreters were often not offered in health interactions, and this could be a significant barrier to health-care access. Failure to offer a formal interpreter for a person with LEP needing health care can constitute a breach of human rights 12 as well as being clearly outside health service protocols and the evidence base supporting the use of formal interpreters. [16][17][18] Nevertheless, participants in our study perceived there were both pros and cons of using formal and peer interpreters in the prison health setting (Figure 1). The value, we believe, in recognizing and exploring the issue of informal interpretation in prison is to clarify the risks and benefits and understand the perpetuating factors behind the practice.
Informal interpreters lack training and professional obligations, such as those required by The Australian Institute of Interpreters and Translators code of ethics and standards of practice. 31 In primary health care outside the prison, the risks of using untrained family members include errors of interpretation due to varying language ability and lack of knowledge of medical terminology, as well as distortion of information due to conflicting roles and agendas of family members as interpreters. 19,20 Yet, they are commonly used for reasons of personal preference (related to trust, support and advocacy), lack of resources or awareness of resources available. 19,20 In the prison context, the risks of using untrained peer interpreters are further compounded. Variations from expected professional standards such as accuracy, impartiality, professional role, confidentiality and respect have been seen to occur among non-professional interpreters in prison settings 15 and can lead to unintended and negative consequences. 15 There are few boundaries between the peer interpreter and patient as fellow inmates within a complex prison hierarchy, 8 where women of LEP are of low status and vulnerable to being "stood over". 13 In the forced or "artificial" prison community, women cannot easily seek alternative health care and the consequences of errors of interpretation, conflicting agendas and transgressions of confidentiality, may therefore be greater. These could include physical or emotional harm from others in the prison, disciplinary action or loss of privileges due to (inadvertent or not) security breaches. 32,33 Additionally, as demonstrated in our research, prisoners may be at risk of being coerced into a peer interpreter role, even if they Our study demonstrated the persistence of peer interpretation was sometimes due to convenience or because formal interpreters were not considered. However, peer interpretation could also be the preferred choice. Some nurses avoided external phone interpreters due to security concerns, despite the Telephone Interpreter Service (TIS) being nationally accredited and approved, with an ongoing quality assurance program. 34 Issues of quality observed by some nurses and patients were similarly found in the Performance Audit Report of TIS, 35 suggesting that while they remain the gold standard of interpretation, are not without challenges of their own.
Training health-care providers in the use of interpreters may improve their understanding of communication difficulties of patients with LEP and their skills in identifying and managing the risks and complexities of such consultations. 21,36 Our study provides evidence on the need to promote the liberal use of formal interpreters among prison health staff and to inform women they have the right to ask for a formal interpreter. A shared and informed decision-making process would respect the preferences of the women using such services, while also acknowledging policy and best practice recommendations.
This would also promote the autonomy and empowerment of the women should they explicitly choose to utilize a peer interpreter.
Other drivers of peer interpretation in prison appeared to lie in the mediation of trust and in the advocacy they afforded to both staff and patients, particularly as "insiders" of both prison culture and their own cultural background. Cultural capital can be a benefit that at times surpasses the quality of the exchange itself. 20 In our study, barriers to access were perceived by some to be due to racism and discrimination. For women with a background of trauma and abuse, negative interactions with health-care providers can have profound emotional impacts and difficulty accessing health care in prison can be interpreted as deliberate blocking of care. 22 The pre-existing relationships and rapport the peer interpreters had with both parties may have decreased some barriers to care. Some peer interpreters regarded their role as affirming and satisfying, increasing their motivation to act in this role.
It is apparent that peer interpreters in prison may take on informal roles that are in keeping with peer support workers. Prison peer support programmes are an emerging approach to bridge health service gaps; the research suggests benefits exist, but evidence is generally limited. [37][38][39] These programmes utilize prisoners who are formally trained and employed in either paid or unpaid roles and include peer support and health education activities. [37][38][39] They have the potential to reduce barriers to health care and empower CALD women through advocacy and support, while promoting cohesion within the prison community 39 and supporting its rehabilitative function. 40 Figure 2 summarizes the formal peer worker role and its potential benefits in the prison setting.
Benefits for the peer support workers in prisons include positive personal growth, satisfaction and improved physical and emotional health; however, benefits are less well defined for recipients of the support. 37,38,41 It should be noted that the trust and power inherent in peer support worker roles in prison may incur additional F I G U R E 1 Perceived benefits and challenges of formal and peer interpreters security risks, such as distribution of contraband by the peer support worker. 37,38 Further research is needed.

| Limitations
In our study, we aggregated our analysis of women from diverse CALD groups. Although there were strong recurring themes relating to language and cultural difference compared to the general prison population, differences between cultural groups may have emerged with further analysis of larger numbers of participants.
The custodial setting where interviews took place would have limited participants' ability to respond freely, although they were eager to report both negative and positive experiences.
Peer recruitment of the focus groups enabled existing networks of women to participate together. While this meant that existing hierarchies and relationships were reproduced, it also meant that groups were more homogenous in culture and language, potentially reducing the effect of power relationships between cultural and language groups on data collection and increasing comfort and ease of communication among participants. 25 In addition, individual interviews were conducted where women were able to speak without the constraints of group and community dynamics.
All but one woman declined to have a formal interpreter for their interview, despite our original planning to provide this for everyone.
They may have avoided formal interpreters due to lack of trust in prison outsiders and a strong fear of stigma in their external community, from which the formal interpreter may have come, 12 but this was not explicitly explored during the interview given it was being peer-interpreted. This may represent some bias towards women who preferred not to use formal interpretation and thus use peers.
It could also mean that communication was suboptimal in some individual interviews.
Using a peer interpreter in focus groups has likely influenced our findings on the use of peer interpreters, particularly the discussion on the risks and disadvantages of their use. However, there were also benefits to the women choosing to use them.
There was further richness to the data due to her insider understanding of the women's experience in prison, an important consideration with bilingual moderators. 24 The women already had rapport with the peer interpreter and were more likely to respond openly. 24 Box 2 provides an example of candour evident among the women, as well as the peer interpreter's ability to translate this. Through purposive sampling and using individual interviews, we were able to canvass countervailing views on the use of peers and explored the views of women for whom peer interpreters did F I G U R E 2 Benefits of peer support worker role not exist. We also recognized the potential for confirmation bias towards a western point of view associated with translations coming through a peer interpreter who was westernized and potentially more educated, 24 but this would not have been reduced by the use of formal interpreters.

| CON CLUS IONS
It is essential to overcome communication barriers in order to provide quality health care for CALD women in prison. At times, healthcare providers and women in prison prefer peer interpreters despite best practice recommendations to use formal interpreters. The persistence of their use may be due to their attributes as an informal peer support person and the current failings of prisons to meet the communication needs of women of LEP. However, the peer interpreter role is highly complex for which they are likely to be inadequately skilled, trained or supported. Improved understanding and management of the complexities of communication with both formal and peer interpreters could enable better quality of care and equity of access for CALD women in the prison health service setting.

ACK N OWLED G EM ENTS
The authors would like to acknowledge GPET for funding and Justice

Health & Forensic Mental Health Network and Corrections NSW for
Facilitating the Research. We would also like to acknowledge and thank Dr Thi Thao Cam Nguyen and Josephine Burton for their comment. We acknowledge and thank the participants.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
Ethics approval was sought and was given by the ethics committees

CO N FLI C T O F I NTE R E S T
PA is a general practitioner and board member with JH&FMHN.