Multilingual consultations in urgent medical care

Full text

ARTICLE
Multilingual consultations in urgent medical care
Antoon Cox
a,b,c,d
and Katrijn Maryns
e
a
Department of Interpreting Studies, KU Leuven, Leuven, Belgium;
b
Department of Interpreting Studies and
Literary Studies, Free University of Brussels, Brussels, Belgium;
c
School of Medical Education, King's College
London, London;
d
Department of Translation, Interpreting and Communication, Ghent University, Ghent,
Belgium;
e
Department of Translation, Interpreting & Communication, Ghent University, Ghent, Belgium
ABSTRACT
More than half of the world’s displaced population has moved to
urban or peri-urban areas, and in Brussels, the superdiverse Belgian
and European capital, the emergency care sector provides an
important setting for analysing the multilingual challenges faced
by health practitioners. To gain a better insight in the interactional
dynamics of emergency department consultations with immigrant
patients, this paper focuses on multilingual strategies that include
‘ad hoc’ communicative solutions used in the absence of profes-
sional interpreters (lingua franca use, non-verbal communication,
medical translation software, language mediation through compa-
nions or hospital sta). Despite their eorts, the participants in our
two case-studies lacked the linguistic and interpreting subtleties
needed to perform complex linguistic-interactional tasks, and in
this way, a form of ‘false uency’ was created. Ad hoc multilingual
solutions, signicant as they are, require additional language sup-
port to avoid diagnostic insecurity. At the level of patient manage-
ment, a ‘linguistic assessment’ of patients could potentially be
integrated into the triage process, and clinicians should be trained
on how to recognise and remediate communication problems
under the specic conditions of the emergency department.
KEYWORDS
Multilingual consultations;
emergency medicine;
migrants; clinical skills;
linguistic ethnography
1. Introduction
1.1. Language and superdiversity in urgent medical care
More than half of the world’s displaced population has moved to urban or peri-urban
areas, where they blend into the poorer urban communities. In Brussels, the superdiverse
Belgian and European capital, where more than half of the people have a migration
background (Hermia and Sierens 2017), this poses substantial challenges for the organisa-
tion of language support in the healthcare sector. Immigrant patients are more likely to
use the emergency department (ED) as a rst point of entry into the healthcare system,
bypassing primary care providers (Cox and Lázaro-Gutiérrez 2016; Hampers 2004; O’Leary,
Federico, and Hampers 2003; Ramirez, Engel, and Tang 2008). In this view, the Brussels
emergency care sector provides an important setting for analysing the multilingual
challenges faced by health practitioners.
CONTACT Katrijn Maryns katrijn.maryns@ugent.be Department of Translation, Interpreting & Communication,
Ghent University, Groot-Brittanniëlaan 45, Gent 9000, Belgium
THE TRANSLATOR
https://doi.org/10.1080/13556509.2020.1857501
© 2021 Informa UK Limited, trading as Taylor & Francis Group

This article aims to gain a better insight in the interactional dynamics of multilingual ED
consultations with immigrant patients and focuses on the negotiation, use and function-
ality of multilingual resources and strategies in the interaction between patients and
clinicians. By ‘multilingual resources and strategies’ we mean the readily available or ‘ad
hoc’ communicative solutions that are used in the absence of professional interpreters,
such as lingua franca use, non-verbal communication, medical translation software or
language mediation through companions or multilingual hospital sta. We adopt a case-
study approach and concentrate on two cases in which doctors, patients and their
companions manage to mobilise complementary repertoires of verbal and non-verbal
resources and strategies to achieve challenging communicative tasks. Interactional micro-
analysis of the consultations seeks to address the implications of ad hoc linguistic support
for doctor-patient communication with the objective of optimising the multilingual
communication in ED consultations as a means to enhance patient understanding and
improve the quality of patient care.
1.2. The multilingual intricacies of ED communication
When ED clinicians attend to immigrant patients, information on previous medical and
social history, patient language skills and family relations can often not be taken for
granted (Cox and Lázaro-Gutiérrez 2016). Therefore, knowledge about the patients and
their conditions is co-constructed in situ by the patients themselves, nurses, doctors, and
anyone else involved with the patient. Given the socio-cultural and linguistic diversity of
these encounters, this local co-construction of knowledge critically depends on the
usually multilingual communication between the above-mentioned actors. This raises
important policy concerns over how the multilingual complexities that form an integral
part of these interactions constrain the eciency of healthcare provision.
It is by now well established that linguistic disparities in the ED lead to lower patient
comprehension of discharge instructions and increase diagnostic insecurity (Carrasquillo
et al. 1999; Crane 1997). The latter is evidenced by a higher rate of orders for tests (e.g.
blood samples, X-rays) (Garra et al. 2010; Ghaemmaghami and Brady 2001; Würth et al.
2018), a higher number of hospital admissions, and an increased length of stay, resulting
in additional costs (Gany et al. 2007; Garra et al. 2010; Goldman, Amin, and Macpherson
2006). Sarver and Baker (2000) found that, in the ED, language barriers reduce the number
of follow-up appointments; the fact that clinicians may be hesitant to call an interpreter
back to explain follow-up appointments could act as a psychological barrier for physicians
to give referrals (Sarver and Baker 2000). Vulnerable communication processes have also
been shown to reduce patient satisfaction with medical care (Carrasquillo et al. 1999).
In general, the medical communication literature recommends the use of profes-
sional interpreters to facilitate communication in healthcare settings (O’Leary,
Federico, and Hampers 2003; Ramirez, Engel, and Tang 2008). However, the use of
professional interpreters in the ED is often perceived as challenging, due to time,
availability and nancial constraints (Karliner, Pérez-Stable, and Gildengorin 2004;
Ramirez, Engel, and Tang 2008). Moreover, clinical sta may not be suciently
aware of the benet of professional interpreters, as they are likely to overestimate
their own and their patients’ language skills (Cox and Lázaro-Gutiérrez 2016). After
all, amid the chaos, stress and urgency of the medical encounter, there is little time
2A. COX AND K. MARYNS

to stop and consider language. Hence, medical sta often turn to the most readily
available multilingual solutions and proceed with the consultation in a contact
language, or by using patient companions or multilingual colleagues.
This raises several challenges and concerns about how mutual comprehension is
facilitated in the absence of professional interpreters. ED communication is by denition
an intrinsically very complex discursive activity that involves gathering information,
physical examination, explanation, and planning, while building a relationship and rap-
port with the patient (Silverman, Kurtz, and Draper 2016). One of the main questions that
arises is how the use of the above-mentioned ad hoc multilingual solutions (lingua franca
use, language mediation through companions or multilingual colleagues) aects the
interactional dynamics of the consultation, and, in turn, the co-construction of clinically
relevant knowledge. These questions require an in-depth analysis of everyday ED practice.
Therefore, we adopt a linguistic-ethnographic approach to gain a better insight into the
ways ED sta manage the multilingual complexities encountered in their consultations
with patients with whom they do not, or not fully, share a language.
1.3. Research setting: linguistic diversity in a Brussels ED
We collected the data of this article in the context of a larger research project on the
dynamics of multilingual oral communication in an ocially bilingual (French – Dutch)
public hospital in Brussels (Cox 2017). The diversity of cultural and linguistic backgrounds
among patients and clinicians as well as considerable variation in socioeconomic back-
ground reect the city’s multicultural and multi-ethnic character. The ED employs 80 per-
sons of dierent ethnic and linguistic origins, including medical sta, clerks, and
maintenance and cleaning sta and about 40% of their patients hold a foreign passport
(Cerf 2012). During the day, the hospital to which the ED belongs benets from ve on-
site professional medical interpreters who cover Modern Standard Arabic, Moroccan
Arabic, Turkish, Polish, Russian and Romanian. External community interpreters cater for
other languages. The latter are, however, seldom called in.
To obtain rst-hand information on the number of immigrant patients in the emergency
department under study, we carried out a quantitative survey at triage. Over a 3-month period,
clinicians
1
working at the triage in the ED kept information on the patient’s language prole,
and on whether they worked with professional or ad hoc interpreters in the consultation. The
survey also comprised a subjective assessment by the doctor on the quality of communication
during the consultation on a 1 (‘Impossible’) to 5 (‘Excellent’) scale. The results of this survey
showed that of the 1360 patients, only 69% were native French, Dutch, or English speakers.
While French and Dutch are the two main ocial languages, English is commonly used as
a lingua franca in institutional encounters in Belgium (Maryns 2017). The remaining 31% of the
patients held a wide variety of native languages: 42 other languages were counted. The most
common languages were Arabic (11% of patients), Portuguese (5%), Spanish (3%), Polish (2%),
Romanian (2%), and Italian (1%). The remaining 7% include lesser used languages such as
Korean and Hebrew. In 28% of the consultations with non-native patients, ad hoc interpreters
were used. These where for instance family or friends who accompanied the patient or – in
exceptional cases – local medical or hospital technical sta of the same origin. In less than 2%
of the consultations with non-native-speaking patients, a professional mediator was called in.
For less than half (48%) of the interactions with non-native patients, clinicians judged the
THE TRANSLATOR 3

quality of the communication as ‘excellent’ or ‘easy’. In 15% of the interactions with foreign
language-speaking patients, the clinicians assessed the communication as ‘dicult’ or
‘impossible’.
2. Methods
This article draws on (socio)linguistic ethnographic research that examines how meanings
are interactively produced in social contexts in such a way as to relate the details of
interaction to broader social processes (Copland, Shaw, and Snell 2015; Gumperz and
Hymes 1986; Hsieh and Shannon 2005; Rampton 2007; Renz, Carrington, and Badger
2018). The ethnographic approach comprises eld notes from participant observation,
multimodal audio-recordings, and interviews (Cox 2015, 2017). We aimed to develop
a ‘thick description’ of social behaviour (Geertz 1973) on the basis of highly contextualised
case studies. A smart-pen was used (Cox 2015) to facilitate note-taking of extra-linguistic
elements such as gaze, position, appearance, and movements (Bezemer and Kress 2016).
Most of the interactions were accompanied by on-the-spot, unstructured interviews with
involved sta (Cox 2015; Firmin 2008). These interviews were recorded before and in between
the audio-recorded consultations, as the researcher also shadowed (McDonald 2005) the
clinicians in the hallways and in the PC room of the ED. As only clinicians where shadowed,
the data merely consist of interactions involving clinicians and not, for instance, social workers.
These short interviews provide additional contextual information to the recorded interactions.
The triangulation of eld notes, audio-recordings, interviews, and member checking by the
hospital sta supported the collection of a robust dataset. Member checking means that the
healthcare professionals who participated in the recorded consultations were invited to give
feedback on the analysis in which they featured.
2
The data for this study were collected in 2013–2014. A six-month immersion in the ED
has yielded a set of 129 audio-recorded clinician-patient consultations (Cox 2017). Out of
these 129 consultations, 55 include immigrant patients who do not speak any of
Belgium’s main ocial languages (Dutch or French) and 15 took place with doctors
who express themselves in non-native Dutch or French doctor. Twenty-two cases con-
cerned multiparty interactions, in other words, they involved (at least) a third person
(friends, family, sta) next to the doctor and the patient. In many cases, this third person
acted as an ad hoc interpreter for at least part of the interaction.
For this article, we have selected two consultations as illustrative case studies
(Gluckman 1961) from the larger corpus. It was a conscious decision to opt for no more
than two cases to allow for an extended microanalysis of deeply contextualised interac-
tional data. This approach aims to grasp the connections better between the participants’
(socio) linguistic and (meta) pragmatic behaviour and the wider socio-medical context of
the consultation. The two cases we have selected exemplify the extensive multilingualism
inherent in the consultations we observed and provide a signicant mixture of ad hoc
multilingual strategies that are mobilised to meet the most urgent communicative needs
of the encounter. The cases also shed an interesting light on the interplay between
contextual variables such as participant roles, repertoires, communicative and medical
goals and outcomes of the consultation. The reason why these cases, and no other similar
cases, were chosen is based on convenience sampling in that for the translation we
looked for language specialists who were familiar with both the patient’s background
4A. COX AND K. MARYNS

and the medical context. In terms of discursive organisation of the consultation, both
cases share a number of common features:
(a)the patient enters the ED with a companion who provides linguistic support;
(b)a whole range of multilingual resources are being used, including lingua franca, non-verbal
communication and the use of medical translation software;
(c)the overarching objective of each of the interactions is to establish an accurate diagnosis of
the patient’s complaint and to propose the most eective treatment or referral.
Given that our recordings cannot be made publicly available as they contain sensitive
data and require anonymisation, we transcribed and back-translated our cases to enable
analysis and share our ndings. We are, however, fully aware that processes of transcrip-
tion, translation and transliteration inevitably imply mediation and transformation of the
initial audio data (Bucholtz, 2000; Niemants, 2012). Despite this limitation, we have made
representational choices that meet our analytical needs and help provide an answer to
our specic research questions. We use a light version of the Jeersonian system, on the
basis of specic transcription conventions
3
and a spatial arrangement of the data in
columns. This visualisation includes separate columns for the participants, the original
verbal communication, back-translation/transliteration of the original utterances and
a description of physical actions. While acknowledging the limitations of our selective
visualisation of what is in fact a very complex communicative event, we do believe that it
nevertheless allows for a sucient degree of representativity of the original oral data in
light of the contextual information required for the features we analyse.
The two cases we selected for analysis were transcribed and back-translated with the
support of native language speakers of the languages involved (Punjabi and Polish)
during joint data sessions with the researcher.
4
We analysed the cases as multi-
discursive (at verbal and non-verbal levels), multilingual (involving dierent languages),
multichannel (involving dierent modes of communication), multiparty (involving doctors,
patients, companions) and multi-activity (physical examination and history-taking) events.
Interactional resources such as questioning techniques, levels of meta-communication
and rapport building were categorised. We documented how these correlated with the
interactional outcomes based on member checking with the clinicians involved.
3. Findings
3.1. Case study 1: consultation in Punjabi and English
In the rst consultation, which takes place at 1:30 am, the doctor goes to see a male patient in
the resuscitation room. The patient is sitting ‘wired’ on his bed, separated by a curtain from
other patients. The doctor is a Belgian (French-speaking) male internist. He has been working
since 8 00 from the previous morning, rst in his day shift and then, after a break from 17 00 to
20 00, in his night shift. The patient is a Punjabi-speaking man from Pakistan. Three Pakistani
men accompany the patient, two of whom actively participate in the interaction. The doctor
and the patient rely mainly on non-verbal communication during the physical examination.
For the history taking, the doctor communicates with the acquaintance in English, which
THE TRANSLATOR 5

serves as a contact language to obtain more information about the patient’s medical history
and the nature of the pain.
5
Excerpt 1 is from the physical examination, where the doctor communicates directly
with the patient. The doctor, who predominantly takes the oor, is using all sorts of non-
verbal means. He draws on gestures and facial expression, proxemics (using space),
paralanguage (voice quality, pitch, volume and other prosodic features such as rhythm,
intonation and stress) and haptics (using the sense of touch). The patient, who is sitting
‘wired’ on the bed is rather quiet.
The interaction is carried by the non-verbal physical action of the doctor (by pointing,
showing, demonstrating, touching) and supported verbally in English. The verbal input is
restricted to some elementary English phrases. However, it is precisely this complemen-
tarity of verbal and non-verbal resources, viz. the patient’s passive knowledge of English
and the doctor’s non-verbal physical cues, that enables mutual understanding and rap-
port building (‘very strong’ is clearly understood by the patient), two of the main
preconditions for ecient treatment. It can be argued that in this specic phase of the
examination, and under the given interactional conditions, doctor and patient can man-
age without the mediation of an interpreter. The immediacy of the interaction between
the doctor and the patient even proves highly benecial for the examination in this phase.
However, as soon as the doctor proceeds with the history taking and the further ques-
tioning of the patient, linguistic support becomes essential for the doctor to be able to
detect and recognise the symptoms of the disease, as can be seen in excerpt 2.
Excerpt 1. Chest pain: physical examination.
Languages Original verbal interaction Description of physical action
1 DOC English Show me teeth (.) Like this . . . Ok . . . (. . ..) DOC shows his teeth by opening his lips
DOC holds PAT’s head; looks at teeth
2 The tongue (.)
Ok (.) Good (.) Ok (.)
PAT sticks out tongue
DOC holds PAT’s head; looks at tongue
3 Give me the hand (.)
Shake, shake, strong (.)
Good (.) Oh (.) Very strong (.)
Ok (.) perfect*// (. . .)
Doctor reaches out his hand
PAT gives DOC a hand
PAT squeezes DOC’s fingers
Laughter
4 It’s eeeuhh totally normal he/ so =
Excerpt 2. Chest pain: history taking.
Languages Original verbal interaction Back translation from Punjabi to English
1 DOC English OK (.) Tell me (.) So (.)
What* is the problem?
2 COM English Actually he is not =
=the problem is (.) ((incomprehensible))
3 PAT Punjabi Siir nu badi dard hondi hay, te dosra na medy day
wich na badi jai gas jai uthdi ay. Itni takleef hoi
keh mein hil he nahin skya
I have a headache. I also have gas [in the
stomach] I had even so many problems
that I could not move
4 DOC English What does he =
(Continued)
6A. COX AND K. MARYNS

As he starts the history-taking in English, the doctor implicitly allocates an interpreter
role to the companion by choice of language. The companion, instead of interpreting the
question for the patient, takes up the role of the patient’s spokesperson. He starts
immediately by answering the question on the patient’s behalf, without consulting him
rst (turn 2). The patient’s passive knowledge of English enables him to interfere in the
turn-taking process and to take the oor (turn 3). As the patient proceeds by answering
the doctor’s question in Punjabi, the companion is now allocated the role of language
mediator. His renditions, however, diverge considerably from the way in which profes-
sional interpreters are supposed to intervene according to their code of conduct. The
companion edits the patient’s answers and leaves out essential information. Not only
does the companion refrain from interpreting what the patient said before the doctor’s
metapragmatic intervention (turn 4), he also leaves out the fact that the patient is
suering from nausea. This omission of signicant information – nausea being one of
the symptoms of a heart attack
6
–remains undetected in the consultation, which gener-
ates an illusion of understanding or false uency. Leaving out such crucial information,
however, can be of high clinical consequence. Despite the patient’s proactive engage-
ment in the interaction (oor taking) and despite the doctor’s multilingual management
(assigning an interpreting role to the companion), back translation of the data indicates
that assumed understanding may have been much higher than actual understanding.
This gives rise to inconsistencies of potential clinical consequence. In the feedback inter-
view, when we asked the doctor’s view on how he perceived the communication, he
Excerpt 3. Feedback from the doctor on the consultation.
Researcher Tu as obtenu l’information que tu voulais ? Did you get the information you wanted?
DOC C’était pas terrible (.) c’était pas facile (.) It was not great (.) It was not easy (.)
Researcher Pourquoi ? Why ?
Doc Mais je sais pas, il (companion) faisais la
traduction, (.) mais, je ne sais (.)
I don’know – he did translate, (.) but I don’t know(.)
J’ai l’impression que ça passait pas trop (.) ou
bien il traduit pas tout, (.) ou bien il
traduisait pas. (.)
I have the impression that the message did not pass (.) or
that he did not translate everything (.) or that he did not
translate at all.
La communication n’était pas top quoi The communication wasn’t great you know
Alors le patient ne comprenait pas tout, je ne
sais pas quoi . . .
Or maybe the patient did not understand everything, I am
not sure . . .
Excerpt 2. (Continued).
Languages Original verbal interaction Back translation from Punjabi to English
5 PAT Punjabi =Hun ess time sir nu thodi jai normal bachat ay
leiken jehda na
At this time head pain is a bit better, but
that.
6 COM Punbjabi Medy ich jalan ay? Burning stomach?
COM English Actually(.) you know (.)
He = have headache (.)
8 DOC English ok
9 COM English He have a big* headache (.) you know (.)
Like when he stand up (.)
he have to go sit down//
Like he is he is not feeling good*
THE TRANSLATOR 7

displayed uncertainty about the quality of the information he had obtained from the
patient. The excerpt demonstrates how in settings of urgency such as the ED, physicians
often have to make do with uncertain diagnostic information (Cox and Lázaro-Gutiérrez
2016). Cases like this explain why the ED has been referred to in the literature as ‘the
natural laboratory of error’, where clinical decision-making is often based on little evi-
dence and is therefore particularly raw and fragile (Croskerry 2003).
3.2. Case study 2: Consultation in Polish and French with translation technology
The second case study discusses a consultation with a female patient and her female
companion, both from Poland. The doctor, a male French-speaking internist, is very tired
as he has been working an exhaustive 12-hours shift. As the patient enters the ED after 20
00, she is not admitted through triage. Before the physician enters the examination booth,
a nurse, who has just taken the vital signs of the patient, briefs him. The patient has
diculty breathing, and she is coughing up blood from time to time. On top of the
physician’s prior knowledge on the patient’s vital signs, the nurse utters a candidate
diagnosis of tuberculosis (TB). The physician has no prior information on the patient’s
social, linguistic, or medical background. Neither does he know who the woman is
accompanying the patient and whether she is a valid source of information on the
patient’s medical history. Both the patient and her companion seem quite uent in
French in casual talk, but when the doctor’s questions become more detailed and
technical, their repertoire in French turns out to be limited. During the examination
stage, the doctor performs a review of the respiratory system in combination with
a dierential diagnosis of TB. In the selected excerpts, he mostly tries to characterise
the manifestations associated with the patient’s main complaint (coughing), notably
sputum and breathlessness.
7
In excerpt 4, the doctor wants to nd out whether the patient is coughing a lot,
whether the cough is dry or produces sputum or phlegm, whether the patient is coughing
up blood and, if this is the case, whether the blood is pure or mixed with saliva.
Excerpt 4. Tuberculosis: history taking without UDR.
Original in French
Back translation from French to
English
Description of physi-
cal action
1 DOC et donc (.) vous toussez aussi
beaucoup? = = = C’est ça?
And so you are also coughing a lot?
Is that it?
2 PAT non pas beaucoup tousser/ No, not much coughing
3 DOC = = d’accord/ Okay
3 PAT = = et tousser tout sec* And coughing all dry
5 DOC = = sec Dry
6 PAT = = Sec* ((confirming tone)) Dry
7 DOC
= =
quand vous toussez c’est sec? When you cough, it is dry?
8 PAT = = ouais Yes
9 DOC = = d’accord/ (. . .) Donc vous crachez* pas? [il]
n’y a pas de crachats? Cracher? [Vous]
comprenez ‘cracher’?
Ok. So you do not spit? There is no
sputum? To spit? Do you
understand ‘to spit’?
10 PAT Cacher*? To hide?
(Continued)
8A. COX AND K. MARYNS

The patient answers eectively to the doctor’s rst question about the frequency of the
coughing. She even expands on her answer by adding that her cough is dry (turn 3). This is
followed by multiple conrmation and accuracy checks (turns 4–8), probably because the
dryness of her coughing is seemingly at odds with her earlier assertion that she was coughing
up blood. It is important for the doctor to clarify this issue, as sputum (spitting while coughing)
is a key symptom of TB. He therefore continues exploring whether or not the patient is spitting
as she coughs, this time through two subsequent closed-ended questions (turn 9). The
negative polarity form of these questions – formulated as negative declarative statements
with rising intonation – displays the doctor’s expectation of a negative answer. The doctor then
seems to realise that the patient might not understand the word ‘cracher’ and uses both verbal
(meta-communication in turn 9; repetition with emphasis in turn 11) and non-verbal meaning-
making strategies (immitating a ‘spitting’ action in turn 13). When she tries to answer this
question, the patient has word-nding diculties – she seemingly cannot nd the word for
‘blood’ – and this is where the companion comes in, rst in Polish with the patient (turn 17),
then in French, completing the patient’s answer (turn 19). In response, the doctor has to nd
out whether she is coughing up pure blood, or rather a mix of blood and saliva (turn 20). This is
an important distinction as a combination of blood and saliva is an indication of a chest
problem and thus possibly TB. Vomiting blood is more likely to reect problems originating in
the digestive system. As the companion hesitates (turn 21), the patient takes over, this time
supporting her answer by taking out a napkin on which bloodstains are visible (turn 22). The
doctor now expresses his understanding and satisfaction with her answer (turn 23).
Despite their proactive display of understanding and non-understanding, the doctor,
who needs absolute assurance about the symptoms, decides to re-do part of the patient’s
history taking. He does so with the linguistic support of the Universal Doctor Speaker Web
(UDR) application.
8
UDR is a relatively novel tool that is not yet widely used in hospital
settings. The application oers a structured menu of pre-formulated medical questions,
Excerpt 4. (Continued).
Original in French
Back translation from French to
English
Description of physi-
cal action
11 DOC CR**acher (with emphasis) To spit?
12 PAT (. . .) cracher To spit
13 DOC = = comme ça Like this DOC demonstrates
how to spit
14 PAT = = non No
15 DOC Non No
16 PAT = = eh oui un peu, un peu des petits*
morceaux (. . .) rouges/
Ehm yes, a little, some small red
pieces.
17 COM = ((in Polish: incomprehensible))
18 PAT = ah c’est aujourd’hui aussi c’est ave[c]- Ah it is today as well it is with
19 COM = avec sang/ (. . .) with blood
20 DOC c’est que du sang?(.) ou c’est salive/ avec du
sang?
Is it only blood ? or is with blood
with saliva?
21 COM = avec (.) = avec des (.) With . . . with . . .
22 PAT = avec des petits morceaux With little pieces PAT shows a napkin
in which she has
coughed up
23 DOC (.) d’accord/ OK
THE TRANSLATOR 9

potential answers, and statements that were translated into dierent languages and
could be read aloud. For instance, the respiratory system section features a screen with
respiratory system-related questions and multiple-choice answer options in French, and
the corresponding translations in Polish. Upon clicking a question, the application reads it
aloud in Polish. Essentially, the app is a one-directional tool: It provides the doctor with
translations of usual medical questions, but it cannot provide translations of the patient’s
answers, as these are typically less structured or predictable. The use of closed-ended
multiple-choice questions can accommodate this problem to some extent by allowing the
patient to convey information through simple yes/no answers. The doctor is not yet
familiar with the app. He constantly switches back and forth between the front stage
frame of the consultation (talking to the patient) and the backstage frame of the applica-
tion (talking to the researcher). The doctor is now standing next to the patient who sits
straight on the examination table. As the doctor has several indicators to consider
pulmonary TB as a highly probable diagnosis, everybody in the room puts on a mask.
Excerpt 5. Tuberculosis: history taking with UDR.
Languages Original interaction
Back translation from Polish/
French to English
Description of physical
action
1 UDR Polish Czy odksztuszasz wydzielinę
śluzową podczas kaszlu?
Do you cough up phlegm when
you cough?
2 PAT French (.) euh (.) Oui / Euh . . . yes
3 DOC Polish = = Tak? Yes ?
4 PAT French = = Oui/ (.) Oui/ Yes, yes
5 DOC French = = Oui**? Yes?
6 PAT French = = oui/ Yes?
7 DOC French Ah*! Ah!
8 COM French = = et avec sang// And with blood
9 PAT French = = Avec* sang* With blood
10 DOC French = = Ouais c’est ça// (.) Yes, that’s it
11 UDR Polish Jak wygląda wydzielina śluzowa? What is the phlegm like? (.)
12 PAT French
Polish
(0:56)
Euh (.) C’est (.) comment/
Jest gęsta
Euh () it is (.) like,/
It is thick ((dense or viscid))(. . .
.)
PAT and COM address
each other
13 COM Polish Jakiego koloru? = Which colour? DOC and RES look at
the tablet
14 PAT Polish Jest gęsta It is thick
15 PAT French = liquide avec (.) avec sang Liquid with . . . with blood
16 DOC French = = avec sang, ok With blood, OK.
17 PAT French = = liquid liquid// Liquid, liquid
18 DOC French = = oui, c’est ça. Ok Yes, that’s it. Ok.
19 PAT French Liquid, liquid// (. . . .) Liquid, liquid
20 UDR Polish Z krwią? Bloodstained?
21 PAT French = = oui* Yes
22 COM French = oui* (. . .) Yes
23 UDR Polish Z krwią? Jest biała? Bloodstained? (. . . .)
Whitish? (.)
24 PAT French Non* c’est rose// No, it is pink
25 DOC French = = rose Pink
26 PAT French = = rosé Pink
27 DOC French = = ok Ok
10 A. COX AND K. MARYNS

After having located the symptom ‘toux’ (meaning ‘cough’) in the UDR module (see
Figure 1 for a visual illustration), the doctor can now tap on several questions in French,
which are then read aloud in Polish. The patient’s somewhat hesitant answer to the UDR
question urges the doctor to use the Polish word ‘tak’ (meaning ‘yes’) in a questioning
tone (turn 3). The doctor’s linguistic accommodation not only serves as a conrmation
check but also helps in building rapport and making the patient feel more comfortable
talking through the app. In response, the patient conrms in a more condent tone
(turn 4), followed by an expression of relief that they understand each other. Interestingly,
as in excerpt 3, it is the companion who draws the attention to the fact that the phlegm is
bloodstained (turn 8), repeated and conrmed by the patient. The doctor then switches
back to UDR, this time tapping on an open question about the quality of the phlegm (turn
11). The patient seems to struggle to answer the question in French and switches to Polish
to consult the companion. The patient and the companion, however, seem to point to
dierent qualities when describing the phlegm. While the patient refers to its density
(turn12), the companion urges her to ne-tune her answer, drawing attention to its
colour, once more alluding to the fact that it is bloodstained (turn 13). The patient
includes both dimensions in her French answer, with extra emphasis on ‘liquid’ (probably
because the doctor only repeated ‘avec sang’ in turn 16). However, in her self-translation
of the Polish term ‘gęsta’ (thick, dense) into French ‘liquide’, some of the subtleties of the
original (the density of the liquid) tend to get lost, without this being detected by the
doctor. The doctor once more switches back to the UDR application before moving on to
the next topic. The interaction in turns 23–27 illustrates how UDR fulls a conrmatory
and reassuring function in the meaning-making process.
In sum, excerpt 5 demonstrates how the use of UDR facilitates the acquisition of
detailed and technical information about the features of the patient’s phlegm, especially
when they use closed-ended questions. Due to its one directionality, however, the app is
less suitable for elicitation of open-ended responses, as it cannot provide translations of
Figure 1. Screenshot of universal doctor module on attributes of symptom ‘cough’. Source: Universal
Doctor©. Note: This screenshot shows an illustration of the English interface of UDR. During the
consultation under study, the doctor was using the French interface.
THE TRANSLATOR 11

the patient’s Polish answers. Although the patient understands the questions, she has to
fall back on her partial knowledge of French to formulate her answers and this is where
inaccuracies may occur that tend to pass unnoticed. Nevertheless, although in this
context the usability of UDR as a translation tool is limited, it enhances the overarching
goal of establishing rapport with the patient. As compared to the initial stages of the
consultation in which UDR was not used, the patient and the companion now seem to
speak in a more relaxed way. The observed mutual linguistic accommodation, by which
the patient accommodates the doctor by speaking French, and the doctor accommodates
the patient by repeating phrases in Polish read aloud by UDR, has proven to produce
a reassuring and stabilising eect on the consultation. This is conrmed at the end of the
consultation, when the pneumologist who takes over the patient’s care, checks with the
doctor if he has obtained all the necessary information from the patient (excerpt 6).
Also during member checking afterwards, the doctor conrmed that using the app
helped him to obtain the information he needed to double-check his diagnosis.
4. Discussion
4.1. Complementary multilingual strategies in emergency consultations
Interactional micro-analysis of the data has revealed that, despite the multilingual chal-
lenges encountered in these consultations, the participants did their utmost to get by with
a bare minimum of readily available communicative resources before seeking professional
language assistance, which is dicult to nd in the given circumstances (urgency, night
shift). Instead, they could be seen to mobilise a whole range of ad hoc multilingual strategies
(non-verbal communication, lingua franca, companion as ad hoc interpreter, UDR) accord-
ing to their most urgent communicative needs. These multilingual solutions, despite
obvious limitations, could be seen as mutually supporting and supplementing each other
in revealing potential interfaces between the speakers’ communicative repertoires. In the
physical examination of the Punjabi patient, mutual understanding and rapport were
achieved on the strength of the participants’ non-verbal cues (pointing, showing, demon-
strating, touching). This was supplemented by their mutual linguistic accommodation (the
patient’s passive understanding of English and repeated comprehensibility checks).
Similarly, in the TB case, it could be argued that as isolated solutions, neither lingua franca
French nor UDR would suce to meet the high communicative demands of the consulta-
tion. When aggregated, however, their verbal and non-verbal strategies turned out to be
mutually enforcing, enabling an active display of understanding and alignment, reassurance
and better rapport between the clinician and the patient.
Excerpt 6. Tuberculosis: conversation between the doctor and the pneumologist.
PNE On a, on a tous de qu’on a besoin. Do we have all (the information) we need.
DOC Oui. Yes.
PNE D’acoord. Ok.
12 A. COX AND K. MARYNS

4.2. High discursive demands and false uency
At the same time, the data conrm that while such a rich accumulation of ad hoc resources can
facilitate the negotiation of meaning under the given conditions of urgency and immediate
intervention, these resources may fall markedly short of giving an accurate and sophisticated
response to the communicative challenges in these consultations. Despite their interactional
eorts, the participants in the observed consultations clearly lacked the linguistic and inter-
preting subtleties needed to perform the more complex linguistic-interactional tasks. Some of
these tasks, however, turned out to be of potential diagnostic signicance. This was evident in
the consultation with the Punjabi patient. The companion, who was assigned the role of ad
hoc interpreter, left out a signicant part of the patient’s anamnesis, which remained unde-
tected in the further consultation. In this way, a form of ‘false uency’ was created. This false
uency was partly due to the largely invisible gaps in the companion’s interpreting and
deontological competence. It is by now well established in medical interpreting research
that when ad hoc interpreters are used, the quality of the renditions cannot be guaranteed, nor
can it be monitored (Flores 2006; Meeuwesen et al. 2012; Meyer 2012; Meyer and Bührig 2014).
Apart from the fact that ad hoc interpreters are more inclined to leave out specic and more
technical parts, they tend to adopt the role of the third participant in the interaction, speaking
on behalf of the primary participants (Rosenberg, Seller, and Leanza 2008). By doing so, they
threaten to disrupt the immediate connection between the doctor and the patient. In this way,
communication becomes so fragmented and imprecise that it seriously hampers the diag-
nostic process.
Still, the need for more sophisticated forms of mediation tends to be underestimated in
this setting. Due to the urgency of the situation, clinicians tend to underinvest in assessing
the communicative context at the outset of the consultation. In this way, they put very
high discursive demands and added pressure on themselves and on the other participants
in the consultation. Not only do they require the necessary multilingual skills and
exibility to switch between languages, express linguistic subtleties and anticipate their
interlocutors’ ways of speaking, they also assume dierent coordinating and metaprag-
matic responsibilities in the consultation. In the case of the Punjabi patient for instance,
the physician could be seen to combine the role of implicit coordinator of the consultation
by choice of language (Goman 1981; Wadensjö 1998b), coordinating both hearer and
speaker roles (the companion being attributed the role of ‘listeners as responders’) and
that of the multilingual manager of the consultation, coordinating at what point multi-
lingual support is required by attributing an interpreting role to the companion.
4.3. A contextual approach to language mediation
Given the particularly high complexity of the analysed discursive activities, we argue that ad
hoc multilingual solutions, signicant as they are, require additional language support to avoid
diagnostic insecurity. After all, the multilingualism encountered in consultations with immi-
grant patients correlates with a complex interplay of contextual variables including participant
roles and expectations, linguistic-communicative repertoires and interactional goals and out-
comes. In public service interpreting research, it is now generally accepted that professional
language mediation involves more than linguistic conversion alone (Angelelli 2004; Baraldi
and Gavioli 2012; Hsieh 2006; Wadensjö 1998a). Several contextual variables, such as the
THE TRANSLATOR 13

interpersonal relations between the participants, the available communicative potential,
cultural factors, expectations and agendas, can inuence the variability in levels of under-
standing. It is precisely this local context of the observed communicative practices that calls for
a contextual approach to language mediation in its broadest sense as an activity involving
‘interactional management and achievement of a complex relationship between renditions
and consideration for participants’ responsibilities, needs and perspectives’ (Baraldi and Gavioli
2016, 52).
A contextual approach to language mediation moves beyond strictly normative con-
ceptions of interpreted interaction and promotes a more nuanced understanding of the
dierent interaction roles and pragmatic needs of the participants in the encounter (Hsieh
2008; Hsieh, Ju, and Kong 2010; Johnson et al. 1999; Meyer 2012; Rosenberg, Seller, and
Leanza 2008). Instead of suggesting either/or responses to the encountered multilingual
needs (either with or without professional interpreter), it is worth considering the full
potential of multilingual and multimodal solutions that are already there, while reinfor-
cing this potential with professional language support. Such an approach implies
a maximum harnessing of the multilingual prociencies of all the participants: by switch-
ing between direct and mediated interaction, it aims to combine the best of dierent
multilingual solutions to ensure more accurate understanding.
5. Implications for practice
At the level of patient management, a ‘linguistic assessment’ of patients can potentially be
integrated into the triage process. After all, despite the urgent and unpredictable nature
of care delivery in the ED, patients may have to spend a considerable time in the waiting
room between intake, treatment, and discharge. If communication with the patient is
dicult in the intake process, this can be included in the patient’s record. In this way,
necessary steps can be taken to identify the patient’s language repertoire and, if needed,
to call in the services of a professional interpreter. Moreover, and in line with the ndings
of a study conducted by Roels et al. (2015) on the uses and eects of community
interpreting in Flanders, we argue that also at the level of the government and local
authorities, there is a need for improved management of the allocation of multilingual
support. This should involve a well-considered use of professional interpreters, ‘alongside
other instruments or strategies for bridging language barriers’ (Roels et al., 2015, 163).
Clinicians should be trained on how to recognise and remediate communication problems
under the specic conditions of the ED. These skills should be linked to the clinical reasoning
cycle as to facilitate the transfer from theory to practice. Such training should be based on real-
life examples from the medical eld, as in the cases presented in this article. These can be
oered via online learning modules that can be accessed by physicians in quieter moments in
the ED, allowing them to learn while remaining at the disposal of potential patients. In
addition, further research is needed to examine the use and impact of automated translation
technology such as Google Translate or DeepL on the communication with foreign language-
speaking patients. In order to address the specic needs of the linguistically vulnerable
(illiterate people and speakers of lesser used languages), it is important to also develop further
insights on the potential of ASR-based apps that allow for speech-to-speech translation.
Third, as EDs are dealing with an increasing linguistic, cultural, and socioeconomic hetero-
geneity, diversity management (Gray, Hilder, and Stubbe 2012; Lundin, Hadziabdic, and Hjelm
14 A. COX AND K. MARYNS

2018) should be integrated in the existing ED healthcare paths. On the basis of our ndings,
we argue that ecient and reliable patient handover should, besides clinical information and
nursing issues, include a transfer of information about the patient’s linguistic needs, the lack of
which may seriously aect care. We recommend further research in this area to address the
implications of language variation in patient management, both in doctor–patient interaction
and between health professionals in the clinical handover process.
Research involving human participants
For the collection of the data considered in this study, ethics approval was obtained as per
the requirements of the hospital ethics board (Comité Local d’Ethique Hospitalier – O.M.
007; Centre Hospitalier Universitaire CHU Saint-Pierre, Rue Haute 322, 1000 Bruxelles;
approval number AK/12-10-81/4181).
Informed consent
We designed an informed consent procedure in compliance with the World Medical
Association Declaration of Helsinki Ethical Principles for Medical Research Involving Human
Subjects. We developed audio-recorded versions of the IC in a wide range of dierent
languages. For the translations, the collaboration of eld specialists was sought. During the
daytime, the hospital beneted from ve on-site professional medical interpreters, covering
the major foreign languages encountered among patients (Classical Arabic, Moroccan Arabic,
Turkish, Polish, Russian and Romanian), and from external community interpreters catering
for Pashtu, Dari, Urdu, Mandarin and Portuguese. The majority of these interpreters origi-
nated from the same countries as the patients they interpreted for and were thus familiar
with the general background of potential study participants (Marshall 2007). This was crucial
for communication, as diculties are known to arise in goal-oriented talk (such as IC
negotiation) when interlocutors do not share the same background (Gumperz 1982).
Notes
1. The ED in question has a triage system in which both doctors and nurses assess the initial
complaints of the patients.
2. The medical interactions studied in our analysis were all of a triadic nature, involving a doctor,
a patient, communication tools and the patient’s companion. The analysis was enriched with
data from immediate feedback interviews and member checking sessions with the involved
doctors. For practical and privacy reasons, however, it was not possible to perform member
checking with the patient and the patient’s companion.
3. Transcription conventions:(.)short pause(. . ..)long pausetextback translation of the original
interactiontextUDR transcription(())researcher’s meta-comments*increased pitch//nal fall=-
latching, overlaps
4. The spoken utterances in Punjabi were transcribed into Latin script and translated into
English. The Polish utterances by the patient and the companion were translated into
English. The utterances in Polish by UDR were available in Polish in the app.
5. For the back translations of the utterances in Punjabi, we consulted two Punjabi and Urdu
speakers from Pakistan: A Punjabi and Urdu-speaking man who is familiar with both the
Belgian and Pakistani health contexts and a Punjabi speaker who has the experience of being
an immigrant himself in both Belgium and Sweden.
THE TRANSLATOR 15

6. Other symptoms include chest discomfort, shortness of breath, lightheadedness, sweating,
anxiety, and heart palpitations.
7. For the back translations, we consulted two Polish speakers: The rst, who has a PhD in
pharmaceutical sciences, is familiar with health care in both Belgium and Poland and has
experience with interpreting in healthcare settings; the second has an extensive experience
in intercultural communication.
8. At an earlier stage in this eldwork project, the researcher had informed the doctor about the
existence of the UDR application and they had discussed the possibility of piloting it together.
Disclosure statement
No potential conict of interest was reported by the authors.
Notes on contributor
Antoon COX is a postdoctoral researcher at the Brussels Institute of Applied Linguistics at Vrije
Universiteit Brussel and a visiting scholar at the School of Medical Education at King’s College
London. He is a lecturer in Spanish interpreting at KU Leuven and a trainer of community inter-
preters. His research focuses on interpersonal communication in stressful multilingual settings such
as the emergency department and asylum reception centres. He has authored and co-authored
papers in Patient Education and Counselling (2017), Monti (2015), and New Voices in Translation
studies (2015).
Katrijn MARYNS is an assistant professor in the Department of Translation, Interpreting and
Communication at Ghent University, Belgium. Her linguistic-ethnographic research examines multi-
lingual discursive practices and linguistic inequality in institutional contexts of globalisation, with
a particular focus on asylum and migration. She has published on diverse aspects of the Belgian
asylum procedure, including the role of narrative construction, credibility assessment, interpreting
and lingua franca interaction. She is the author of ‘The asylum speaker: Language in the Belgian
asylum procedure’ (Routledge 2006), editor (with Philipp Angermeyer, York University) of the book
series ‘Translation, Interpreting and Social Justice in a Globalised World’ (Multilingual Matters), and
she has published in various international peer-reviewed journals (Applied Linguistics, Language in
Society, Journal of Sociolinguistics, Language & Communication).
ORCID
Antoon Cox http://orcid.org/0000-0002-2261-1721
Katrijn Maryns http://orcid.org/0000-0002-8919-3050
References
Angelelli, C. V. 2004. Medical Interpreting and Cross-cultural Communication. Cambridge: Cambridge
University Press. doi:10.1017/CBO9780511486616.
Baraldi, C., and L. Gavioli, Eds.. 2012. Coordinating Participation in Dialogue Interpreting. Vol. 102.
Amsterdam: John Benjamins Publishing Company. doi:10.1075/btl.102.
Baraldi, C., and L. Gavioli. 2016. “On Professional and Non-professional Interpreting in Healthcare
Services: The Case of Intercultural Mediators.” European Journal of Applied Linguistics 4 (1): 1.
doi:10.1515/eujal-2015-0026.
Bezemer, J. J., and G. R. Kress. 2016. Multimodality, Learning and Communication: A Social Semiotic
Frame. London: Routledge, Taylor & Francis Group.
Bucholtz, M. 2000. „The Politics of Transcription„. Journal of Pragmatics 32 (10): 1439–1465.
doi:10.1016/S0378-2166(99)00094-6
16 A. COX AND K. MARYNS

Carrasquillo, O., E. J. Orav, T. A. Brennan, and H. R. Burstin. 1999. “Impact of Language Barriers on
Patient Satisfaction in an Emergency Department.” Journal of General Internal Medicine 14 (2):
82–87. doi:10.1046/j.1525-1497.1999.00293.x.
Cerf, D. 2012. “Le Bon Usage du Service des Urgences. Presented at the Séminaires des Services
d’Urgences IRIS, Séminaires des Services d’Urgences IRIS.” Séminaires des Services d’Urgences
IRIS. Brussels: IRIS Public Hospital Network.
Copland, F., S. Shaw, and J. Snell, Eds.. 2015. Linguistic Ethnography: Interdisciplinary Explorations. 1st
ed. London: Palgrave Macmillan. doi:10.1057/9781137035035_1.
Cox, A. 2015. “Ethnographic Research on Ad Hoc Interpreting in a Linguistically Diverse Emergency
Department: The Challenges of Data Collection.” New Voices in Translation Studies 12: 30–49.
Cox, A., and R. Lázaro-Gutiérrez. 2016. “Interpreting in the Emergency Department: How Context
Matters for Practice.” In Mediating Emergencies and Conicts, edited by F. M. Frederici, 33–58.
London: Palgrave Macmillan.
Cox, A. (2017). The Dynamics of (mis)communication in language-discordant multi-party consultations
in the Emergency Department. PhD dissertation. Vrije Universiteit Brussel.
Crane, J. A. 1997. “Patient Comprehension of Doctor-patient Communication on Discharge from the
Emergency Department.” Journal of Emergency Medicine 15 (1): 1–7. doi:10.1016/s0736-4679(96)
00261-2.
Croskerry, P. 2003. „The Importance of Cognitive Errors in Diagnosis and Strategies to Minimize
Them„. Academic Medicine 78(8): 775–780.
Flores, G. 2006. “Language Barriers to Health Care in the United States.” New England Journal of
Medicine 355 (3): 229–231. doi:10.1056/nejmp058316.
Gany, F., L. Kapelusznik, K. Prakash, J. Gonzalez, L. Y. Orta, C. Tseng, and J. Changrani. 2007. “The
Impact of Medical Interpretation Method on Time and Errors.” Journal of General Internal Medicine
22 (Suppl 2): 319–323. doi:10.1007/s11606-007-0361-7.
Garra, G., H. Albino, H. Chapman, A. J. Singer, and H. C. Thode. 2010. “The Impact of Communication
Barriers on Diagnostic Condence and Ancillary Testing in the Emergency Department.” The
Journal of Emergency Medicine 38 (5): 681–685. doi:10.1016/j.jemermed.2009.01.004.
Geertz, C. 1973. The Interpretation of Cultures: Selected Essays. New York: Basic Books.
Ghaemmaghami, C. A., and W. J. Brady. 2001. “Pitfalls in the Emergency Department Diagnosis of
Acute Myocardial Infarction.” Emergency Medicine Clinics of North America 19 (2): 351–369.
doi:10.1016/s0733-8627(05)70188-7.
Firmin, M. 2008. “Unstructured Interview.” In The SAGE Encyclopedia of Qualitative Research Methods,
edited by L. Given, 906-908. Thousand Oaks, CA: SAGE Publications. doi:10.4135/9781412963909.n475.
Gluckman, M. 1961. “Ethnographic Data in British Social Anthropology*.” The Sociological Review 9
(1): 5–17. doi:10.1111/j.1467-954X.1961.tb01082.x.
Goman, E. 1981. Forms of Talk. Philadelphia, PA: University of Pennsylvania Press.
Goldman, R. D., P. Amin, and A. Macpherson. 2006. “Language and Length of Stay in the Pediatric
Emergency Department.” Pediatric Emergency Care 22 (9): 640–643. doi:10.1097/01.
pec.0000227865.38815.ec.
Gray, B., J. Hilder, and M. Stubbe. 2012. “How to Use Interpreters in General Practice: The
Development of a New Zealand Toolkit.” Medical Care 3: 6. doi:10.1071/hc12052.
Gumperz,, and D. Hymes. 1986. Directions in Sociolinguistics: The Ethnography of Communication.
Oxford: Blackwell.
Gumperz, J. J. 1982. Language and Social Identity. Cambridge: Cambridge University Press.
Hampers, L. C. 2004. “Cross-cultural Issues in Pediatric Emergency Medicine.” Clinical Pediatric
Emergency Medicine 5 (2): 73–75. doi:10.1016/j.cpem.2004.01.002.
Hermia, J.-P., and A. Sierens. 2017. Belgen En Vreemdelingen in Het Brussels Gewest, Van Bij Hun
Geboorte Tot De Dag Van Vandaag. Brussels: BISA (Brussels Institute for Statistics en Analysis).
Hsieh. 2006. “Understanding Medical Interpreters: Reconceptualizing Bilingual Health
Communication.” Health Communication 20 (2): 177–186. doi:10.1207/s15327027hc2002_9.
Hsieh. 2008. ““I Am Not a Robot!” Interpreters’ Views of Their Roles in Health Care Settings.”
Qualitative Health Research 18 (10): 1367–1383. doi:10.1177/1049732308323840.
THE TRANSLATOR 17

Hsieh,, and Shannon. 2005. “Three Approaches to Qualitative Content Analysis.” Qualitative Health
Research 15 (9): 1277–1288. doi:10.1177/1049732305276687.
Hsieh, E., H. Ju, and H. Kong. 2010. “Dimensions of Trust: The Tensions and Challenges in
Provider—Interpreter Trust.” Qualitative Health Research 20 (2): 170–181. doi:10.1177/
1049732309349935.
Johnson, M., C. Noble, C. Matthews, and N. Aguilar. 1999. “Bilingual Communicators within the
Health Care Setting.” Qualitative Health Research 9 (3): 329–343. doi:10.1177/
104973299129121893.
Karliner, L. S., E. J. Pérez-Stable, and G. Gildengorin. 2004. “The Language Divide.” Journal of General
Internal Medicine 19 (2): 175–183. 10.1111%2Fj.1525-1497.2004.30268.x.
Lundin, C., E. Hadziabdic, and K. Hjelm. 2018. “Language Interpretation Conditions and Boundaries
in Multilingual and Multicultural Emergency Healthcare.” BMC International Health and Human
Rights 18: 1. doi:10.1186/s12914-018-0157-3.
Marshall, P. 2007. Ethical Challenges in Study Design and Informed Consent for Health Research in
Resource-poor Settings. Geneva: World Health Organization.
Maryns, K. 2017. “The Use of English as Ad Hoc Institutional Standard in the Belgian Asylum
Interview.” Applied Linguistics 38 (5): 737–758.
McDonald, S. 2005. “Studying Actions in Context: A Qualitative Shadowing Method for
Organizational Research.” Qualitative Research 5 (4): 455–473. doi:10.1177/1468794105056923.
Meeuwesen, L., E. Ani, F. Cesaroni, J. Eversley, and J. Ross. 2012. “Interpreting in Health and Social
Care: Policies and Interventions in Five European Countries.” In Inequalities in Health Care for
Migrants and Ethnic Minorities, edited by D. Ingleby, A. Chiarenza, W. Devillé, and I. Kotsioni,
144–157. Vol. 2. Leuven: Garant.
Meyer, B. 2012. “Ad Hoc Interpreting for Partically Language-procient Patients in ‘Coordinating
Participation in Dialogue Interpreting.” In Coordinating Participation in Dialogue Interpreting, edited
by C. Baraldi and L. Gavioli, 99–114. Amsterdam: John Benjamins. doi:10.1075/btl.102.05mey.
Meyer, B., and K. Bührig. 2014. “Interpreting Risks. Medical Complications in Interpreter-mediated
Doctor-patient Communication.” European Journal of Applied Linguistics 2 (2): 233–253.
doi:10.1515/eujal-2014-0010.
Niemants, N. 2012. „The Transcription of Interpreting Data„. Interpreting 14(2): 165–191. doi:10.1075/intp
O’Leary, S. C. B., S. Federico, and L. C. Hampers. 2003. “The Truth about Language Barriers: One
Residency Program’s Experience.” Pediatrics 111 (5): e569–e573. doi:10.1542/peds.111.5.e569.
Ramirez, D., K. G. Engel, and T. S. Tang. 2008. “Language Interpreter Utilization in the Emergency
Department Setting: A Clinical Review.” Journal of Health Care for the Poor and Underserved 19 (2):
352–362. doi:10.1353/hpu.0.0019.
Rampton, B. 2007. “Neo-Hymesian Linguistic Ethnography in the United Kingdom.” Journal of
Sociolinguistics 11 (5): 584–607. doi:10.1111/j.1467-9841.2007.00341.x.
Renz, S. M., J. M. Carrington, and T. A. Badger. 2018. “Two Strategies for Qualitative Content Analysis:
An Intramethod Approach to Triangulation.” Qualitative Health Research 28 (5): 824–831.
doi:10.1177/1049732317753586.
Roels, B., et al.2015 “Equal Access to Community Interpreting in Flanders: A Matter of Self-Reective
Decision Making?” THE INTERNATIONAL JOURNAL FOR TRANSLATION & INTERPRETING
RESEARCH 7 (3): 149–65.
Rosenberg,, R. Seller, and Y. Leanza. 2008. “Through Interpreters’ Eyes: Comparing Roles of
Professional and Family Interpreters.” Patient Education and Counseling 70 (1): 87–93.
doi:10.1016/j.pec.2007.09.015.
Sarver, J., and D. W. Baker. 2000. “Eect of Language Barriers on Follow-up Appointments after an
Emergency Department Visit.” Journal of General Internal Medicine 15 (4): 256–264. doi:10.1111/
j.1525-1497.2000.06469.x/full.
Silverman, J., S. Kurtz, and J. Draper. 2016. Skills for Communicating with Patients. 3rd ed. London:
CRC Press. doi:10.1201/9781910227268.
Wadensjö, C. 1998a. Interpreting As Interaction. London: Routledge.
18 A. COX AND K. MARYNS

Wadensjö, C. (1998b). The Social Organisation of Remembering in Interpreter-mediated Encounters.
Second International Conference on Interpreting in Legal, Health, Educational and Social Services
Settings, Vancouver, Canada May 19-23.
Würth, K. M., S. Reiter-Theil, W. Langewitz, and S. Schuster. 2018. ““Getting By” in a Swiss Tertiary Hospital:
The Inconspicuous Complexity of Decision-making around Patients’ Limited Language Prociency.”
Journal of General Internal Medicine 33 (11): 1885–1891. doi:10.1007/s11606-018-4618-0.
THE TRANSLATOR 19