Statistics from Altmetric.com
Benda et al found that in an emergency department in the eastern United States, Spanish-speaking patients who needed language assistance received inconsistent and often non-recommended language access services across different phases of patient care.1 These ranged from emergency department staff speaking English to the patient, the provider speaking Spanish (with varying levels of proficiency) or the use of either professional or ad hoc interpreters. What happens in the hospital investigated in their study undoubtedly occurs in hospitals across the USA: linguistic minority patients are left without adequate language services, potentially reducing their access to quality healthcare. Similar scenarios exist in other countries, including the UK and Australia.2 3
In the USA, the the Affordable Care Act (ACA) requires that health care facilities that receive federal funds offer a qualified interpreter to individuals with “limited English proficiency” when the provision of an interpreter constitutes a “reasonable step” in providing “meaningful access” to a health program or activity. Under this legislation, agencies are restricted from allowing an adult or minor accompanying an individual to interpret except under very specific circumstances; and may only rely on qualified bilingual/multilingual staff to communicate directly with individuals with limited English proficiency. .4 The use of non-professional ‘ad hoc’ interpreters can cause dangerous errors in communication; in addition, using children as ad hoc interpreters, or ‘language brokers’, can cause them to suffer from emotional trauma.5 6 Similar to the the anti-discrimination measures of the ACA, the UK implemented the Public Sector Equality Duty in 2011, and public service agencies are required to have due regard to the need to eliminate unlawful discrimination and advance equality of opportunity.7 However, according to Szczepura, healthcare for minority populations is not being adequately monitored. 8
The authors of the study in the Emergency Medicine Journal suggest that machine translation might partially address the lack of interpreters in healthcare settings, but the use of non-human and even remote human interpreting services is far from ideal for language minority patients. To provide language assistance for minority language speakers, many healthcare facilities, including emergency departments, currently rely on remote interpreting services, where a trained interpreter in a remote location provides audio/visual interpreting via tablet, or audio-only services via telephone. Video remote interpreting services are superior to telephone services in that they allow the interpreter to see the provider and the patient and perceive facial expressions, gestures, bodily conditions and some components of the physical contexts of the hospital or clinic. However, language minority patients are often unsatisfied with the services provided by telephone and video remote interpreters, preferring an in-person interpreter who they can perceive as a real person. In an ongoing collaborative engagement project about language access for Spanish speakers,9 we have found that Spanish-speaking patients in Ohio described remote interpreters as distant, inaccurate and unable to convey their emotions, often referring to the tablet-based remote interpreting system as ‘la máquina’ (the machine). According to the patients, the addition of the remote interpreting device inhibited the development of personal relationships over the course of healthcare interactions.
As a linguist, I see the notion of machine interpretation as extremely risky in healthcare contexts. On the one hand, the communication that occurs in interaction is based on language that cannot be translated literally and also includes non-verbal forms of communication that may have diverse meanings in different cultures and cannot be captured by a machine translator. For example, Spanish speakers often use euphemisms to make reference to body parts or bodily practices that are considered taboo to discuss in certain Latino cultures, and words used to describe certain cultural understandings of disease may be difficult to translate. For instance, in some Mexican cultures, ‘susto’ (literally, ‘fear’ or a sense of ‘startling’), a type of illness caused by an experience of fright, is a possible cause of stomach upset, including diarrhoea. In addition to the difficulty of translating culturally specific terms, a machine cannot enact the roles of the interpreter that go beyond conveying information. Additional roles of the healthcare interpreter include the following: the ‘clarifier’, the ‘cultural broker’ and ‘the advocate’.10 It is unlikely that a machine would be able to explain cultural practices and communicative forms to the healthcare provider or advocate for a patient when a provider is being insensitive to the patients’ cultural views of healthcare or not allowing for a thorough communication process. These roles may also be more difficult for remote (rather than in-person) interpreters to enact. The use of language concordant providers is, of course, ideal, as long as the providers are trained in providing healthcare in the patients’ language. Studies have shown that the use of highly proficient language concordant providers leads to greater patient trust, more agreement, and, in some cases, better health outcomes.11
Rather than seeing linguistically marginalised patients as ‘having language barriers’, the problem should be reframed: insufficient language access services in healthcare contexts create language barriers (similarly, I prefer to avoid using the term ‘Limited English Proficiency’ because it implies a deficit perspective of the language skills of linguistically marginalised groups12). These barriers lead to health disparities because language minority patients are not provided with equal access to quality healthcare. In this way, language becomes a social determinant of health. To better understand the ways in which language minority patients experience healthcare in various settings and what kinds of challenges arise for providers who wish to offer language access in specific types of facilities, bench science is not sufficient; it is necessary to include the voices of many different stakeholders, including patients, patients’ family members, providers, interpreters and facility administrators, drawing from patient-centred and community-based research methodologies. This translational work will help researchers to understand how to apply the findings of outcomes-based research in community settings.12 For example, many of the Spanish-speaking immigrant patients who attended our stakeholder meetings in Ohio expressed that they felt that doctors in the US healthcare system paid little attention to them because they used minimal physical contact and did not always explain what was happening; this differed from medical encounters in their home countries, where physicians were more likely to physically examine them rather than relying on medical records. Inadequate language assistance in certain phases of patient care, including the physical exam, exacerbated the sense of distance that patients experienced. Patients in the US emergency department depicted by Benda et al may have experienced the inconsistency in language access services in similar ways, but this was not studied. In addition to their experiences with particular forms of language assistance, it is also important to consider patients’ experiences with the clarification and mitigation of cultural differences (such as diverse beliefs about illness, self-care practices or bereavement practices) when using different types of language assistance. More research is needed to make sense of the relationship between the inconsistent use of language assistance described in the article and the experiences and health outcomes of linguistically marginalised patients.13
There are many reasons that emergency department physicians may not provide language access even when interpreting services are available at their facility. They may feel rushed to complete medical interviews with minority language patients either because insurers do not provide additional reimbursement for the time taken to use interpreters or because they work in healthcare systems where there are pressures to decrease waiting times while seeing an increasing number of patients. The technology for remote interpreting may fail or may not be in the right location when needed in an emergency situation. In addition, providers may be swayed by patients who request permission to use family members as interpreters because the provider may wish to respect the patient’s desire. However, patients may not understand the negative effect that this practice could have on their own health outcomes and on the emotional well-being of the language brokers.
A better understanding of patient views may help to highlight the types of changes that are needed at multiple levels of the healthcare system, and not just within specific healthcare facilities to provide appropriate services. It would also increase the sense of urgency to make these changes. In an interprofessional education programme that I developed in collaboration with colleagues in the health professions, we found that students in the health professions gained a deeper understanding of the value that interpreters bring to healthcare for linguistically marginalised patients while also learning how to improve their interactions with these patients and collaborate effectively with interpreters.14 This kind of training would be beneficial in most western nations; with the large numbers of recent immigrants and migrants in these countries, emergency departments are likely to be among the first places seeing patients who need language assistance.
Contributors I am the sole author of this manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.
Patient consent for publication Not required.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.