Country.

Of the 39 studies, fifteen came from Australia [16–30], twelve from the United States [31–42], seven from Europe and the United Kingdom [43–49], and five from Canada [50–54].

Target population.

In the literature review, 13 articles did not specify the language group studied [22,31,34,35,37,41,42,46–49,51,54]. For example, Cicolello and Anandarajah studied “members of ethnic and racial minority populations” [31, p870]. The article by Yarnell and colleagues focused on “recently immigrated versus long-standing [Canadian] resident decedents” [54, p1480]. While they found an association between language proficiency and end-of-life care, they did not define specific languages, ethnic groups, or countries of origin. Javanparast and colleagues explored Australian PC policy in the context of priority underserved populations, including those with language discordance, but do not specify which language(s) [22].

The Australian Bureau of Statistics (ABS) has clearly defined Standards for Statistics on Cultural and Language Diversity (CALD) to facilitate data collection and analysis. There are four main concepts used to define CALD: (1) country of birth, (2) language spoken at home, (3) English proficiency and (4) Indigenous status [55]. Critiques are that the term CALD has been overused as a blanket statement and has been misused at times – labeling groups based on race or ethnicity, without considering Aboriginal status, country of birth, or language proficiency [56]. Nine of the 15 articles reviewed from Australia used only the term CALD without further specifying a target population based on ABS criteria [17,19–21,23–25,27,28]. Five Australian articles discussed how their target population met CALD standards per the ABS [16,18,26,29,30]. For example, Leonard and colleagues narrowed their focus on Australians who speak Arabic, Hindi, Mandarin, or Australian Aboriginal languages at home [26]. Bloomer and colleagues categorized their target CALD groups in relation to religious affiliation, not language [16].

Beyond the five Australian studies of specific CALD communities discussed above, 12 articles clearly defined the population of study. Here, seven studies identified the target population based on the regional demographics [39,40,43–45,50,53]. For example, de Voogd and colleagues explored the experiences of the three largest migrant groups in the Netherlands: Turkish, Moroccan, and Surinamese [44]. Eckemoff and colleagues, on the other hand, studied Russian-Americans as the first author had a connection to this community [32].

Four studies synonymized language with country or region of origin [33,36,38,52]. Glaser and colleagues focused on individuals with limited English proficiency born in Latin America but did not specify languages spoken [33]. Similarly, Khosla and colleagues did not specify a language, they identify their target population as “…people with origins from India, Pakistan, Bangladesh, Nepal, Sri Lanka, Bhutan, Maldives…” [36, p181].

Study design.

Most studies reviewed (n = 23) employed qualitative designs (e.g., ethnography [50,52], grounded theory [19,31,34,35,39], phenomenology [29,47], and case study [17,51]) and analyses [21,23–25,27,30,32,33,43,46,48,53]. There were nine quantitative studies, including retrospective cohort [28,37,42,54] and cross-sectional/survey research [18,38,40,41,49]. There were seven mixed methods studies [16,20,22,26,36,44,45].

Study settings.

Studies were conducted in several settings that span community, hospital and other settings. Three studies took place in the community hospice environment [31,37,51], ten occurred in the general community/outpatient environment, or through a homecare agency [26,30,32,33,36,38,40,45,50,52]. Three articles reviewed research conducted in the specialized palliative community care environment [19,20,44]. Weber and colleagues worked with health care providers in a specialized palliative care team that followed clients between community and inpatient settings [49], while De Graaff and colleagues explored experiences of health care providers working in community and in the hospital [43]. Fourteen studies took place in the inpatient/hospital setting [16,18,21,23–25,27,28,39,41,42,47,53,54]. There was one study that took place in a long-term care facility [29]. Six studies defined varied study settings, or did not clearly define the location [17,22,34,35,46,48], including Javanparast and colleagues, who performed reviews of intersectoral government documents then interviewed key government stakeholders in their work environment [22].

Study participants.

The experiences of clinicians, including nurses, physicians and other allied health professionals involved in the provision of palliative and end-of-life care were specifically explored nine studies [19,23,27,36,39,47,49,52,53], while seven others also included the perspectives of patients and families [16,29,31,32,42,43,52]. De Graaff describes this phenomenon as the communication triad, “…triangular communication between the care provider, the patient and a close relative who helped to resolve language problems” [43, p3146]. Zapata and colleagues studied the perspective of both health care providers and patients [42].

There were eight studies that focused solely on the perspective of the patient and their family [18,20,25,28,33,41,45,54]. Three studies focused on the perspective of medical interpreters [21,24,50]. Five studies explored the experiences of administrators and educators involved in end-of-life care [17,34,35,37,48]. For example, Semlali and colleagues aimed to refine training for health professionals in cross-cultural PC and included the perspectives of educators with expertise in ethno-cultural education and training [48]. Another recent study completed a fulsome review of current government policy and legislation that guides PC and enmeshed those findings with the expert opinions of government and policy leaders in their region [22]. The unique perspective of volunteers working in a community hospice was explored in Jovanovic’s study from 2012 [51]. There were five studies that explored the public’s ideas around education in palliative and hospice care in linguistically diverse communities [26,30,38,40,44].

Participant characteristics.

While most studies reported some characteristics of the participants, they rarely reported findings by characteristics, and certainly, none explored the intersection of these characteristics on study results. Overall, age [16,18,20,29,30,32,33,36,38,40,41,44–47,51,52,54], sex [18,20,23–27,30,32,36,39–41,43,44,46,47,49,51,52,54] and occupation [16,19,21–23,27,29,31,32,34–36,39,43,47–50,53] were the most widely reported participant characteristics. Some studies also reported on ethnicity [29,31,36,39,41–44,51] language [18,25,27,28,31,38,40,41,45,49,51], birth country [20,28,30,32,33,38,47,49] or migrant group [26]. Only a few studies reported on the education [32,38,51], and income or socio-economic status [33,54] of participants. One study reported on race [31] and rural/urban status [54]. Broom does not discuss participants in their work [17].

Terminology and definition of care received.

While there are considerable differences between the role and definition of palliative, hospice, and end-of-life care [1], the terms were often used synonymously and interchangeably within the reviewed studies. For example, Leonard and colleagues uses the term End-of-Life Care in their title but provide definitions that align with the WHO’s definition of PC [26]. Palliative Care was the main term used in 16 of the studies examined [17,20,22,24,25,27,28,39,42–44,47–49], but was only clearly defined in ten of those articles [20,22,24,25,39,42,43,47,48,52]; definitions were iterations of the WHO’s definition of PC [3]. Kirby and colleagues concede that PC is contextually iterative, not clearly defined, and often open to interpretation [24]. Ten studies used the term end-of-life care [16,19,23,26,29,32,41,46,50,54], only five defined their interpretation end-of-life care [23,26,29,32,54].

Hospice care was explored in six articles [31,34,35,37,40,51]. While two did not clearly define the term [31,34], the remaining had considerable variation in their definitions [35,37,40,51]. For example, Lindley and colleagues defined hospice as a “… model for quality, compassionate care for people facing a life-limiting illness or injury” [37, p2]. Jovanovic defines hospice as care for “…dying patients who had six months or less to live, and who wished to die at home” [51, p165].

Advanced care planning is an important component of end-of-life care, allowing for an individual and their family to clearly define goals-of-care and priorities including limitations of therapy, location of treatment, and expectations around pain and symptom management [1,12]. Four studies focused on advanced care planning and discussions, and the term was clearly and consistently defined in all four studies [18,21,30,38].

Dharni and colleagues studied cancer patients but made specific reference to advanced disease and end-of-life care [45], similar to Glaser and colleagues [33]. Khosla used the term “seriously ill” to describe the health status of the population of interest, but also referred to end-of-life wishes and decision-making [36].

The impact of language discordance on the quality of palliative care.

Several studies focused on the impact of language discordance on the quality of PC. Non-English speakers often felt neglected and isolated from their care providers because of language discordance, which fostered an environment of mistrust [45]. Individuals who do not speak the same language as their health care provider have a harder time communicating needs, which limits the quality of assessment by the clinician [28,33]. The length of time in America and degree of language comprehension was found to influence care pathways at end-of-life [41].

Communication barriers contributed to unmanaged pain [25], inadequate symptom management [21], and poor quality of care [51]. Glaser and colleagues found that confusion and misunderstanding after medical appointments led individuals to make decisions they may not have made if they had understood what was being communicated [33]. Individuals with language barriers were also found to be significantly less likely to have advanced directives and resuscitation wishes defined in the context of their care [20], and were more likely to receive aggressive care at the end of life (e.g., cardiopulmonary resuscitation, ventilation, and dialysis) and to die in the intensive care unit [54]. Initial findings indicated that persons with a language barrier had better symptom management, but it was surmised that CALD patients were more likely to under-report their symptoms, giving the appearance of adequate symptom management [28].

Culture and care.

Cultural context is difficult to grasp and goes beyond the words used to describe a phenomenon. Semlali et al noted that people tend to revert to their mother tongue when experiencing severe illness [48]. In an interview with palliative and hospice care nurses, a nurse talks about the importance of emotional language at the end of life: “you can’t fully express yourself, your fears, desires, emotions… unless you can do that in your mother’s tongue [23, p273].

Certain cultures have beliefs and practices that may not be fully understood by PC providers. For example, Wardle and Bennett explored the important role that food and feeding played in PC among Punjabi Indians in Australia [29]. Dharni and colleagues described the story of an Islamic patient who could not communicate with their English-speaking medical team the effect of not being able perform their religious ablution had on their health [45].

Family structures are also important. Eckemoff and colleagues identified that there is a common Russian cultural practice of making decisions for sick loved ones, which conflicts with the focus on autonomy in Western societies [32]. A few studies noted how cultural beliefs about death itself played a role in care. For example, some believe that talking about death will cause death, and therefore avoid talking about death altogether [19,52]. Hayes and colleagues quoted a participant discussing advance care planning: “If I’m talking about death, I’m going to actually make it more likely to happen” [21, p4]. Broom and colleagues noted an urgent need for a “… systematic analysis of real-time family negotiations to reveal the complex interplay of culture, language, news telling, clinical judgement and broader negotiations around transitions to the end-of-life” [17, p1045]. Cicolello and Anandarajah highlight the need for a “cultural interpreter” to navigate these complexities [31].

Cultural competence.

Mislabeling or working under preconceived ideas that a certain group will respond in a certain way can have devastating effects and lead to further barriers to health care delivery [57].

Over half of the studies identified the importance of cultural competence in the context of the delivery of PC [16,20,22,27,29,31,32,34,35,39,43–53]. Green and colleagues found that cultural generalizations made by health care providers contributed to negative patient experiences [20]. Cross-cultural care educators discussed the importance of education for health care providers that highlights self-reflection, positioning individual socio-ethno-cultural backgrounds, and identifying preconceptions as ways to mitigate stereotypes that may be perpetuated by health care providers [48].

The experience of palliative care providers.

Several studies focused on the impact of linguistic barriers on PC providers. Eckemoff and colleagues found that coordinating and working with interpreters was time consuming and difficult to coordinate [32]. Milberg and colleagues reported that some nurses described guilt related to the difficulty they experienced in truly empathizing, and being genuine and supportive, to individuals with language barriers [46]. For example: “you have to keep inside things that you would usually say out loud to the patient and their family” [46, p6]. Clinicians have improved job satisfaction when feel they are able to support clients at end-of-life, “assuaging death anxiety through presence, explanations and reassurances” [23, p271]. Researchers also found that nurses working in a well-supported pediatric hospice environment, with manageable workloads, had increased satisfaction in care provided to linguistic minorities, higher use of interpreters, increased use of appropriate educational materials, and better overall client outcomes [37].

Use of interpreters.

It has been found that providing adequate interpretation services is effective in improving access to hospice for minority groups [34] and leads to better outcomes [31]. Hughes and colleagues reported that most centres had access to remote or in-person translation services, but that these were largely unused [35]; conversely, Green et al reported that health care providers had inadequate access to professional translation services [20]. The cost of translation services, especially for smaller organizations, is a barrier to their use [35]. The study by Green and colleagues reported that medical interpreters were used for discussions of consequence, such as when informed consent was required or a diagnosis was being discussed, and that family members were used for all other visits [19].

However, some have reported the degree of difficulty interpreters have in meaningfully translating medical terminology and jargon. Formal CMIs (or equivalent) have a difficult job that is often criticized in the literature by health care providers and families. Health care providers found medical interpreters a barrier [19]. Some had issues with interpreters who did not translate verbatim [21,26], finding interpreters who did not fully communicate the context of an issue led to an incomplete understanding of care [39]. Some studies report friction between medical interpreters and family members, there was mistrust [43], and different expectations around communication [32]. Kirby and colleagues described the role of the interpreters in end-of-life care as “emotionally taxing, linguistically difficult and highly complex” [24, p498].

Martin and colleagues found that pre-briefings between the interpreter and clinician were helpful, as they clarified expectations and discussed goals prior to meeting with the patient [27]. To that end, guidelines have been developed that provide advice to clinicians to optimize the interpreter’s role, and to help clinicians learn about the dilemmas faced by interpreters [50].

There are mixed findings on the family members as interpreters. Some individuals prefer to have their family translate [35], and their presence increases access to (free) interpretation services [17]. However, some clinicians found use of family members as interpreters problematic in that it led to missed opportunities to build rapport with family [19], less than fulsome communication [43], and inconsistent messaging as different family members may interpret differently based on opinion, family structure, and role [32].

The need for in-language palliative care resources.

Khosla and colleagues found that language-appropriate resources influenced decision-making and that individuals and their families were better equipped to make important decisions for South Asian migrants at the end of life in the United States [36]. Leonard and colleagues found that “it’s not just translating a document into a language that’s comprehendible. It’s translating a whole idea that might now make sense” [26, p8]. Lack of in-language advance care planning resources was a common barrier to members of various CALD communities in Australia [30]. Zapata and colleagues noted that education from clinicians who spoke their language and understood their culture, improved PC utilization by underserved members of the Latinx population [42]. The use of in-language printed resources in pediatric palliative was found to lead to improved experiences and satisfaction [37].

McLean and colleagues studied the impact of sharing language-concordant videos; they found that prior to watching the video, 52% of respondents had no knowledge of advanced directives. After the video, 95% of respondents felt it was important to start to organize end-of-life wishes [38]. Pan and colleagues found that after community education sessions on end-of-life care in hospice, members of the Spanish-, Chinese- and Korean-speaking communities were interested in learning more about hospice care and how it fit in with a palliative approach to care [40].