Skip to main content
NCBI home page
Sage Choice logoLink to Sage Choice
. 2022 Sep 9;60(4):626–636. doi: 10.1177/13634615221119383

Interpreter-mediated psychiatric assessments: Metacommunication as key

Orest Weber 1,✉, Jonathan Klemp 1, Florian Chmetz 1, Argyro Daliani 1, Esther-Amélie Diserens 1, Florence Faucherre 1
PMCID: PMC10504803  PMID: 36083232

Abstract

Psychiatric assessments of non-native migrant patients facilitated by an interpreter pose specific communication challenges to all participants. In this study, we developed an original interdisciplinary approach to the verbal and non-verbal practices in this triadic activity. The aim was a data-based description of challenges for clinicians and interpreters, and the identification of relevant strategies. We filmed, transcribed and translated 10 interpreter-mediated consultations focused on the psychiatric assessment of the patient. Subsequently, we submitted the consultations to clinical, interactional sociolinguistic, and interdisciplinary analyses. We identified six challenges for interpreters and clinicians engaged in psychiatric assessments: barely comprehensible and confusing speech, speech about emotions and subjective perceptions, sensitive remarks in relational terms, conclusive clinician interventions, interruptions during interpreter renditions, and non-verbal communication. Attempts by the interpreter to avoid relational offenses (protection of positive face) and to defend the participants’ autonomy (protection of negative face) play a major role in these challenges. So does an insufficient awareness of mutual needs by the clinician and the interpreter. We identified specific strategies of inter-professional metacommunication for each challenge. Clinicians and interpreters should be aware of the challenges they may face in triadic psychiatric assessments. They should take a reflexive stance towards their common practices and may consider using metacommunication tools to reach better communicational and clinical outcomes.

Keywords: assessment, challenges, coherence, face work, interpreter, investigation, linguistics, metacommunication, psychiatry, relevance, strategies

In a globalized world with linguistically and culturally superdiverse populations (Roberts et al., 2005; Vertovec, 2007), a growing number of psychiatric encounters require language interpreters (Bauer & Alegría, 2010; Searight & Armock, 2013). Interpreter-mediated consultations in psychiatry are often experienced as difficult by clinicians and interpreters (Bauer & Alegría, 2010; Bot, 2005; Searight & Armock, 2013; Singy & Guex, 2005). To support these professionals, educational offers have emerged (Bruwier et al., 2019; Leanza et al., 2017; Niemants & Cirillo, 2017; Ticca, 2018), but these offers are still insufficient and based on limited scientific knowledge.

In psychiatric assessments of new patients, the appropriate management of language barriers is crucial. Here, the term psychiatric assessment includes all the interactions which enable a clinician to understand mental health needs, reach a diagnosis, or define a clinical project (McAllister-Williams et al., 2012). Dysfunctional clinical encounters at early stages of patients’ mental health trajectories lead to inappropriate treatment, patient drop out, and disparities in access to care (Thomson et al., 2015). Therefore non-native patients need qualified interpreters collaborating efficiently with clinicians during psychiatric assessments (Leanza et al., 2017).

Early research into psychiatric assessments of non-native patients (Farooq et al., 1997; Marcos, 1979) had documented that psychiatrists are likely to lose important diagnostic and relational cues because of the way original speech is rendered by language interpreters. These studies focused mainly on the detection of interpreter errors; limited attention was paid to other interactional and relational processes in the clinical encounters. More recently, interdisciplinary studies in other settings have used interactional theory and conversation analysis to explain why phenomena labeled as interpretation errors occur (Delizée & Michaux, 2019; Gerwing & Li, 2019; Ticca, 2018; Wadensjö, 2001). These studies also offered practical solutions for clinicians and interpreters to improve outcomes.

In this study, we used interactional sociolinguistic tools in combination with clinical insights to scrutinize interpreter-mediated psychiatric assessments. We define interpreting in health settings primarily as social interaction (Wadensjö, 2001). This does not exclude a contrastive approach to original discourses and their renderings at some points of the research process (as in Bot, 2005). We submitted interaction data to interdisciplinary analysis to reveal challenges to the provision of effective mental health care. For each challenge, we determined practical strategies for interpreters and clinicians. It is our hope that this study may provide teaching material in continuing education courses for clinicians, as well as enrich the competence of public service interpreters.

Methods

Data collection

The study took place at Lausanne University Hospital, which lies in the French speaking part of Switzerland where the population is marked by a strong ethnic and linguistic diversity (Statistiques Vaud, 2020).

Ten consultations focusing on the psychiatric assessment of different non-native patients were videotaped between 2012 and 2017. Patients’ native languages were Albanian (2), Amharic, Dari, Georgian (2), Serb/Croatian/Bosnian, Somali (2) and Tamil. Four consultations—with four different patients—were part of short expert assessments offered by two psychiatrists specialized in psychosis and schizophrenia. Six consultations were encounters with six new patients of the unit for transcultural psychiatry “Psy&Migrants.” The clinical staff of this unit help asylum seekers with complex clinical and social situations access mental health care. Three clinicians attached to this unit routinely carry out psychiatric assessments that make use of interpreters; all participated in this study. All eight interpreters of the study had a federal certificate in public service interpreting, three of them also had an advanced federal diploma in this profession. None had academic qualifications in interpreting.

We filmed consultations during four periods of one to four months (two in each setting), after obtaining informed consent from both the patient and the interpreter. Patients were carefully chosen using exclusion criteria to eliminate any cases where the research process could deteriorate the therapeutic relationship. Approval was obtained by the ethics committee of the hospital (CER-VD, protocol n° 339/2012).

Transcription and translation

We transcribed video recordings according to conventions used for conversation analysis (Jefferson, 1983; Sacks et al., 1974). Speech in patients’ native languages was not transcribed in these languages. It was translated as precisely as possible into French by an interpreter and the segments of translation inserted into the transcript in a way which reflects interruptions, overlaps, and prominent prosodic phenomena (e.g., loudness).

Clinical analysis

An analysis was carried out by two clinicians who were not part of the filmed consultation. Notably, the video recordings contained both the original speech in foreign language and a simultaneous voice-over by a French translation. They labeled sequences as beneficial or detrimental to the psychiatric assessment. Supplementary comments were also written down. The clinicians involved in the analysis had different theoretical backgrounds (psychoanalytic approach, systemic approach), but they mainly used non-specific theoretical and practical knowledge about psychiatric assessments.

Linguistic analysis

All sequences labeled as beneficial or detrimental were submitted to an in-depth multidimensional linguistic analysis (Weber et al., 2017). Following the discourse analytic research tradition of interactional sociolinguistics (Gumperz, 1999), the linguist undertook repeated exploratory analyses of the transcripts and the video to identify suitable theoretical and methodological tools. He subsequently reapplied these tools to each sequence. When needed, the transcripts were enriched with representations of para-verbal and non-verbal content. The following analytic tools were used:

Interdisciplinary analysis: Identifying challenges for professionals and useful strategies

During eight sessions of interdisciplinary analysis, the findings of the linguistic analysis were shared with the clinicians involved in the clinical review and discussed in depth. Three psychiatrists filmed during the study took part in these discussions. During the same period, our findings were also submitted to interpreters, external experts in interpreting, and to specialists in transcultural psychiatry, psychology and linguistics. The aim was to identify challenges for interpreters and clinicians involved in psychiatric assessments, as well as strategies to help face these challenges. Eventually, we verified that all the observations of the clinical and linguistic analyses were related to one of the identified challenges.

Results

Our study points out six challenges for interpreters and clinicians in psychiatric assessments. We present them in the following six sections and illustrate them with research data. Figure 1 contains the transcription conventions used in the cited excerpts. For each challenge, we also outline possible strategies for clinicians and interpreters that emerged during the interdisciplinary analyses.

Figure 1.

Figure 1.

Open in a new tab

Transcription conventions used in excerpts.

Barely comprehensible speech: Confusing explanations or stories, off-topic answers, neologisms

The speech of some patients can be hard to understand and consequently difficult to render in a different language. This speech may lack coherence; interpreters may reorganize discourse in a more structured and comprehensible way. Also, the relevance of that speech (Sperber & Wilson, 1989) within the larger context of the interaction may be puzzling, typically when the answer of a patient fails to show a clear connection with the question of a clinician. In all these cases, interpreters tend to verbalize their personal inferences of meaning, for instance their assumptions on how the speech of a patient could possibly answer questions previously asked. This sometimes deprives clinicians of important diagnostic material, concerning disorganized thinking—a possible sign of psychotic disorders—and the emotional state of patients, etc.

Using inferences to make sense of barely comprehensible speech is an automatism in all speakers of natural languages (Sperber & Wilson, 1989; Gumperz, 1999) and reorganizations of disorganized speech is almost unavoidable in language interpretation. We identified strategies that limit the effects of these automatisms and improve the clinical effectiveness of triadic psychiatric assessments. Clinicians and interpreters can limit original speech to shorter stretches by interrupting the patient. They can also switch to simultaneous interpreting with a lower voice volume (chuchotage) to avoid interrupting patients who need to produce longer stretches of speech to disclose important aspects of their story. In addition, metacommunication among professionals about hardly comprehensible patient speech before, during or after consultations can be of great value.

Excerpt 1

In this consultation, the patient narrates his early childhood in Somalia before he and his family left for Ethiopia. He describes the death of his sister, killed by a member of the Salafist group Harakat al-Chabab al-Mujahidin. This crime motivated the departure of his family to Ethiopia. As he verbalizes the violence of these events, the clinician asks him if he has any memories of this period:

286. C: (2 s) Ok (4 s). So it’s, it’s… Yeah, it’s, it’s, uh, it’s very violent, huh. Do you have any memories as a child of, of that period?

287. I: It’s really very dangerous stuff, do you have any memories of anything that happened, that you still remember? Do you have any memories?

288. P: Yes, uh, uh, uh, my sister, uh, yes, but, my mother, my sister, yes, like that, my mother and there was a boy, and my sister, and a little one, there are four of us, one who is even smaller. There was her and a smaller one, there were four of us altogether. They were people who came after my sister, it was difficult. They were sick. She had a lot of worries. After all she had gone through.

289. I: Yeah, he didn’t answer your question directly, but he said we were, yes I remember, there was an older brother than me and… me and… my sister, and another sister that she’s smaller, smaller than all of us, and then it’s been a while since the people who came after me, who told me that, uh ta, your mom is all alone, that she’s also sick, and all those things when I remember, well, the pain is increasing.

The discourse uttered by the patient (turn 288) in response to the physician’s question (286–7) is difficult to understand and subsequently difficult to render in a process of language interpreting. The semantic underdeterminacy (Mason, 2006) of this discourse comes mainly from the juxtaposition of short segments like “my mother,” “my sister,” “there was a boy,” “people who came after my sister.” These segments are easily understandable separately but the connections between them are not clear and this makes the entire turn highly ambiguous. The interpreter manages to render a large part of this ambiguity by reproducing a quite similar range of juxtaposed short segments of speech (289).

A second major source of underdeterminacy lies in the frequent use of personal pronouns. Namely the uses of they (“they were people who came after my sister” and “they were sick”) and she (she had a lot of worries) at the end of turn 288 allow several interpretations. In this case, the interpreter voices her inference that the “people” came after the patient’s sister, whereas the patient’s mother is sick (but why “they were sick” then?) and has worries. The clinician thus does not have access to some of the patient’s difficulties producing an easily understandable narrative, which may for instance indicate a strong emotional reaction to the topic under discussion. However, the interpreter opens her rendering with a metacommunicative comment expressing her concerns about the adequacy of the patient’s answer. According to our clinical analysis, this kind of comment is helpful for clinicians engaged in psychiatric assessments, as it conveys that the interpreter has identified comprehension issues.

Mention of emotions and subjective perceptions

In psychiatric assessments, clinicians are particularly attentive to patients’ emotions and subjective experiences. Clinicians’ access to these elements depends on the precise rendering of patients’ speech with emotional or subjective contents. The rendering of clinicians’ questions or comments focused on these aspects is also important. It is not uncommon that interpreters’ renderings attenuate or omit emotional or subjective elements and make the discourse more factual. Interpreters should be aware of the need for high precision in this area when collaborating in psychiatric assessments. Clinicians can help them with targeted metacommunication on this topic.

Excerpt 2

During this consultation, the patient expresses that thinking about her family makes her sad. She also explains that she was not able to say goodbye before she left her country. The clinician wants to know whether the patient was aware of what was going on emotionally and relationally when she left. She does this in a way that emphasizes her interest for the patient’s subjective perception of this particular moment of her past:

157. C: Mhm, mhm … Mhm (4 s) And you, did you realize what you were doing? That you were leaving your family, the country. Was that clear to you?

158. (11:53) I: Was it very clear to you that you were leaving the country, were you ready or did you just have to leave?

The first part of the interpreter’s rendering of the clinicians question is very close to the original. The major difference lies in the end of the interpreter’s intervention, where she offers the patient two candidate answers (Pomerantz, 1988) connected by “or”: “were you ready” or “did you just have to leave?” The second answer, the one selected by the patient, offers her the possibility to present herself as someone who was only reacting to a context which forced her to leave, whereas the clinicians question originally framed the patient as someone who is taking action (turn 157: “what you were doing”). According to the clinical and linguistic analyses, the adjunction of these candidate answers may have blocked the health professional’s access to introspective clinical material she was seeking. From a sociolinguistic perspective, the protection of the patient’s negative face (Goffman, 1955; Brown & Levinson, 1987) may play a major role in practices of this kind. In turn 158, the way the question is put by the interpreter is less threatening for the negative face of the patient—i.e., for her intimate, personal territory—than the original question. With the possibility of selecting “I just had to leave,” the patient can go past the clinicians aim to investigate her subjective and intimate experience at this sensitive moment of her life.

A strategy for interpreters to deal with this kind of difficulty in a more transparent way is to disclose their embarrassment with clinician questions if they perceive them as particularly intrusive. Clinicians can encourage this kind of metacommunication during pre-consultation briefings and by building a relationship of trust and mutual respect with interpreters (Leanza et al., 2017).

Sensitive remarks in relational terms

Interpreters also sometimes omit or mitigate elements creating or reinforcing relational tensions within the triad when the emotions and subjective perceptions are not central aspects of the discussion. This general trend may deprive clinicians of important cues regarding the therapeutic relationship (as in the example below), but also of the possibility of using confrontation as a therapeutic or diagnostic tool. Clinicians and interpreters should support each other and create a working alliance within which these sensitive elements can be reproduced or at least topicalized at a metacommunicative level.

Excerpt 3

At the beginning of this clinical interview, the psychiatrist tells the patient that they will go on with the patient’s life story. The clinician subsequently asks the patient if she has important issues she would like to address before resuming this narrative. The patient raises a concern with the apartment that she has just occupied. The clinician responds that this problem should be discussed with Anna Pozzi (code name), the social worker of the mental health facility. This is how the interpreter renders this response and how the patient reacts:

55. I: For this question, you have to ask Anna Pozzi, the social worker.

56. P: = I want to see her.

57. I:  = She’s the one who can answer these questions, she’s the one who knows.

58. P:  = But what can I ask you?

59. I: Ok, then

60. P:  = What can I ask you?

61. I: Okay

62. P:  = Thank you

63. I: So, I’ve understood this. But, if for this question, I’ve got to ask Anna Pozzi, what questions can I ask you?

A short discussion between the patient and the interpreter (56–62) follows the rendering of the psychiatrist’s question by the interpreter. It represents a monolingual talk-back sequence (Li, 2015): the interpreter talks back to the patient in their native language instead of rendering her initial response to the clinician. The interpreter eventually summarized the content of this discussion in the clinician’s language in turn 63. The first contribution of the patient (turn 56: “I want to see her”) is not part of this summary, even though it seems particularly sensitive on a relational level, since it is likely to be interpreted as “I want to see her (the social assistant), and not you (the psychiatrist).” Our clinical analysis identified this kind of mitigation practice on the part of the interpreter as limiting the process of psychiatric assessment: the filmed psychiatrist understands that the patient challenges the usefulness of her contribution, but the interpreter’s rendition diminishes the strength of this challenge.

We can read this kind of interpreters’ communicational behavior as face work. Interpreters take care of the clinician’s positive face, i.e., of the positive social image of self the psychiatrist is able to project (Goffman, 1955; Brown & Levinson, 1987). In excerpt 3, the interpreter reduces the strength of the patient’s attack at the clinician’s image of “useful professional.” The patient herself seems aware of the face threat contained in her intervention, since she uses repair work in turn 62 (“Thank you”). Interestingly the interpreter does not include this in her rendering. She rather uses her personal means, such as framing this troubled piece of interaction as a misunderstanding (beginning of turn 63), to reduce the potential of tension or conflict contained in the patient’s words.

In this kind of situation, interpreters should feel that the working alliance within the triad is solid enough to bear relational challenges. Comprehensive pre-consultation briefings between clinician and interpreter on this very aspect may be helpful, especially when the professionals do not know each other and cannot rely on a pre-existing trustful relationship. Self-reflection for interpreters during consultations is also crucial, with reflexive thoughts such as the following: “This answer could offend the doctor, but it is important for her to have access to this aspect of the patient’s behavior; I will try to let the doctor deal with this answer, even if it takes me some energy to let relational tensions affect our peaceful meeting.”

Conclusive clinician interventions

On several occasions, conclusive comments added at the end of a question–answer sequence are not reproduced in renderings, even though they appear important for clinical and relational dynamics. These conclusive comments are usually uttered by the clinician. Some comments valorize aspects of the patients’ life situations and capacities, or the quality of the doctor–patient (–interpreter) collaboration. Other conclusive comments left out in renderings could have prevented or partly solved tensions; for instance when the clinician signals that they have heard and/or accepted the patient’s point of view in case of disagreement. The rendering of all these elements deserves more attention from interpreters and clinicians. Interpreters should monitor their practices with a special focus on conclusions until rendering conclusive comments becomes an automatic skill. Clinicians can facilitate conclusive comments by pausing before asking their next question.

Excerpt 4

At the end of this psychiatric consultation, the clinician, the patient, and the interpreter arrange the next appointments. The clinician makes sure that information is passed on, taking into account that the patient is illiterate:

764 C: But, but how do you do it? Because I don’t think you can read this (C points at the card with the appointments). H-, how do you, uh, know what day to come?

765 I: You can’t read. How did you know about this appointment today.

766 P: Here, I know, the dates of the days. The days, the dates and then the time.

767 I: I, I can understand the, the, the dates of the appointments and the time.

768 C: Okay, okay, great ((rising intonation, C smiles)). And, and who was it who told you how to get here?

769 I: How did you know this address?

In this excerpt, the beginning of turn 768 “okay, okay, great” was not rendered by the interpreter. At least two factors are likely to explain this non-rendering. First, many interpreters tend to focus above all on questions and answers. Psychiatric assessments are speech events (Gumperz, 1999) belonging to the interview-type and interpreters seem sometimes to see their main function in providing usable patient answers to clinician questions. Second, interpreters may not render conclusive comments like “okay, okay, great” because they do not consider this needed. Words like “okay,” borrowed by many languages throughout the world, are likely to be understood by the patient. In addition, the prosodic (here a rising intonation) and non-verbal cues (here a smile) accompanying these closing comments convey a large part of its key meaning even without voicing the verbal material by the interpreter (Mason, 2012). However, in cases like excerpt 4, the positive message would be stronger if this verbal material was rendered too, here especially the word “great.” Since our clinical analysis has highlighted the importance of establishing solid therapeutic alliances with patients during psychiatric assessments, it seems sensible to use all available means for this.

Interruptions during interpreter renditions

When other members of the triad interrupt interpreters while they are delivering a prolonged message, this often results in significant changes of the structure of the original speech, for example, reduction, simplification, and/or clarification. Clinicians and interpreters should be aware of this risk and help each other limit clinically problematic outcomes (e.g., limited access to the coherence of speech for clinicians; see also the first section, “Barely comprehensible speech”). They can achieve this when interpreters use post-consultation debriefing to tell clinicians, who have repeatedly asked new questions before the completion of their renditions, that this may result in a lower precision of their work. Consequently, clinicians can be more attentive to this aspect of their verbal behavior in the following consultations.

Excerpt 5

The patient in this clinical encounter reports that he is often flooded with memories of war and aggression. The clinician asks him if these memories also emerge at night as nightmares. The patient says yes and explains that he only sleeps a few hours a night. He then adds that he has never used marijuana—probably as a sleeping medication—since he has arrived at the asylum seeker residence where he currently lives and that he practices sports. The clinician wants to know which sports he practices and this is what the patient answers:

323 P: I was, for 9 years, a professional athlete, a basketball player and in kick-boxing but I like, all sports, cycling, everything, everything, everything. I was a professional athlete.

324 I: uh, athletics

325 C: mhm

326 I: for 9 years

327 C: in, in Yugoslavia then?

328 I: [So… basketball]

329 P: [Former Yugoslavia], Junior Champion of the former Yugoslavia.

330 C: mhm

331 I:  = basketball and… combat sports,

332 C:  = mhm

333 I:  = kick-boxing but mostly athletics, I was uh… in the junior professional team.

In this troubled sequence of interpreter-mediated interaction, the elements expressed by the patient undergo slight modification in the process of interpretation. The patient’s statement that he was a professional athlete is less clear in the rendering and his title of Junior Champion disappears. These elements are both summarized in “I was in a junior professional team.” Our clinical analysis showed that this modification limits the psychiatrists’ access to elements of the patient’s speech pointing at a strong narcissistic personality. The linguistic analysis, in turn, reveals that the interpreter is interrupted repeatedly in this sequence. She therefore has to memorize new elements before she manages to deliver the end of the patient’s first turn. The psychiatrist is the first to interrupt, when he asks if this was in Yugoslavia (turn 327). The interpreter tries to go on with her rendering and to delay the delivery of this question (328). The patient has some knowledge of the doctor’s language and engages immediately in answering and adding new information (329). This troubled pattern of interaction—specifically, how turns at talk are taken and assigned (Li, 2015; Sacks et al., 1974)—adds considerable complexity to the interpreter’s work and this is certainly one of the main reasons why clinically relevant elements escape.

Non-verbal communication

Some gestures and mimics of the consultations are not immediately understandable for all members of the triad and require some form of treatment by the interpreters. They interpret these gestures or mimics with verbal means or they replace them with gestures or mimics that are understandable for the recipients. The quality of the interaction can be improved if the interpreters identify and process these gestures in the best possible way. Clinicians can reinforce these useful practices thanks to metacommunicative comments.

Excerpt 6

This excerpt is located at the end of the first consultation of a psychiatric assessment in several encounters. The clinician asks the patient, a young man from Eastern Africa, questions in order to fill in his administrative file, namely his address, telephone number, and civil status:

554 C: Mhm. (2 s.) And you are single?

555 I: You are not married?

556 P: I was there (3 s.) after 4 months

557 I: Were you married?

558 P: We were, 4 months, but I don’t know, we got married, after (hand gesture: the right hand open starts very close to the right shoulder and moves upwards away from it)

559 I: (s) You split up (hand gesture: joins her open hands in front of the chest and moves them away from each other upwards)? (s) Back home I got married and we stayed 4 months, and then we couldn’t go on, and that’s when we split up (same gesture with both hands).

An apparently standard administrative question seems to put the patient in an awkward situation. The interpreter has to word the question twice to receive an answer she is sure to understand correctly. A major element, the splitting up between the patient and his wife, is only non-verbally communicated (hand gesture in turn 558). However, the gesture used is obviously ambiguous for the interpreter, who checks her understanding by reformulating it to the patient verbally and non-verbally (559). In this process she uses a gesture which is comprehensible for the patient and the physician, and which she can use again as she renders the patient’s answer in the physician’s language. She thus provides the clinician with the information that non-verbal communication played a crucial role in this patient’s answer (Mason, 2012). The only possibly useful element which is lost is that the patient has never worded his separation. However, this limitation was not mentioned during the clinical analysis, unlike the positive contribution of the interpreter in helping the clinician to understand the patient’s gesture.

Discussion

Strengths and limits of the study

This in-depth interdisciplinary analysis of interpreter-mediated psychiatric assessments has produced new insights into interpretation processes and the collaboration between clinician and interpreter. In comparison with previous work (Farooq et al., 1997; Marcos, 1979), our study emphasizes the communicational and relational dimensions affecting diagnostic and therapeutic cues. Thanks to a detailed discursive analysis, the findings reach far beyond an inventory of clinically problematic “interpreter errors.” These phenomena are: a trend, in interpreters, to produce coherent and relevant patient answers to clinicians’ questions and to neglect elements that do not fit the question–answer pattern (see also Delizée & Michaux, 2019; Mason, 2006); attempts to save interlocutors’ positive and negative face (see also Mason & Stewart, 2001); interruptions of the interpreter during renditions (see also Li, 2015); and the management of non-verbal communication (see also Mason, 2012). The same phenomena also affect the opportunity to build a therapeutic alliance within the triad.

A major strength of our findings is their immediate practical relevance. They suggest that interpreter-mediated psychiatric assessments may be looked at as sensitive interactional trajectories within a complex of intersecting branches. At intersections, clinicians and interpreters make conscious or unconscious choices that lead the exchange toward different clinical outcomes. Reflexive distancing and more conscious choices at some embranchments could lead to optimized psychiatric assessments. In professionals’ reflections on triadic interactions, conceptual tools such as interpreter roles (interpreters as language conduits, cultural brokers, co-diagnostician, etc.; Delizée & De Ridder, 2016) can be helpful; they may also appear abstract. In contrast, our findings provide conceptual models connecting directly practical clinical goals (e.g., to get access to disorganized thinking) with practical linguistic and interactional techniques (e.g., short turns, chuchotage, avoidance of the voicing of inferences), but also with elements of the larger social and psychological context (e.g., face work, overwhelming emotions) that explain why doing the right thing at the right moment is not always easy (Delizée & Michaux, 2019; Mason, 2009; Ticca, 2018).

Our study has several limitations. Although the number of consultations in our study is sufficient to outline relationships between discourse and clinical reasoning (Bot, 2005; Cox, 2020; Gumperz, 1999), we could not analyze the full range of patient assessments occurring regularly in psychiatry. For instance, interpreter-mediated assessments with older people with possible dementia may bear different challenges. The study design also had some limitations due to financial and human resource constraints. We were unable to analyze the recorded sequences from the perspective of the interpreters and patients. These groups should be more actively involved in future work.

Conclusion

Regarding the interpreters’ and clinicians’ practice in consultations, our findings highlight that optimizing interpreter-mediated psychiatric assessments does not solely rely on the interpreters. Interpreters need appropriate language and interpretation skills to manage interventions in sensitive mental health contexts (Van Vaerenbergh, 2020). However, clinicians must also gain knowledge on how to facilitate adjusted practices when collaborating with interpreters in triadic psychiatric assessments and more generally in mental health encounters (see also Delizée & Michaux, 2019; Leanza et al., 2015). Courses addressed at both audiences (ideally in joint sessions) should convey basic knowledge, offer the possibility to test and train self-reflection, as well as present useful practical strategies. Our educational practice shows that interpreters and clinicians are interested in courses pertaining to useful strategies for mental health consultations; experiences of other scholars and teachers carrying out similar work reinforce this finding. Research results and data sets like those presented in this article offer meaningful didactic tools for these courses (Weber, 2020; Weber & Faucherre, 2020). Notably, some of this content is not specific to psychiatry and rests on the broader literature. For example, strategies for the interpretation of hardly comprehensible speech and efficient ways of learning about routines and goals of institutions are classical components of courses and handbooks for public service interpreters (see Gentile et al., 1996; Skaaden, 2019).

Our findings show that inter-professional metacommunication about the goals of clinical encounters is a core competence to teach in courses for clinicians and interpreters. Practical modalities of collaboration should be included in the course. With regards to psychiatric assessments, three topics of metacommunication between clinicians and interpreters appear fundamental.

A first topic lies in the treatment of patients’ speech by interpreters when speech is difficult to understand. The two professionals may agree that it is not the interpreter’s role to clarify ambiguous patients’ utterances, that interpreters should remain as close as possible to the original wording and preserve ambiguities wherever possible (see excerpt 1). Interpreters should also inform clinicians about their difficulties of understanding, particularly when they feel that their renderings have been improved for comprehension. However, metacommunication on this topic may jeopardize patients’ trust in professionals if it takes place within consultations. Patients may understand some of the language they do not speak and non-verbal or para-verbal features may make sense across languages (Mason, 2012; Wadensjö, 2020). Interpreters and clinicians should be aware of this and postpone metacommunication if needed. Clinicians and interpreters may also decide to use semi-simultaneous interpreting (chuchotage). This technique might help interpreters convey more easily the coherence and relevance of the patients’ speech. To our knowledge, studies of the use of chuchotage in mental health are lacking. Chuchotage may present some risks if interpreters lack training in simultaneous interpreting; it should be abandoned if patients show signs of distrust.

A second key topic of metacommunication between clinicians and interpreters pertains to the handling of tensions within the triad. Since the behavior of patients in interpersonal relationships is a major aspect of the diagnostic and therapeutic processes, triadic interactions must be handled carefully to allow tensions to emerge (see excerpts 2–5). Clinicians can help interpreters by explaining that irritating or disrespectful behavior on the side of patients, as well as challenging questions or comments on the side of clinicians, can be essential: they provide insight into the way patients handle emotions and function in relationships. Thus, interpreters should convey these elements to their fullest in their renderings, without mitigation, despite the relational discomfort they cause. Additionally, clinicians and interpreters may also agree on ways to signal unintended cultural missteps on the part of the clinician, which may upset or inhibit the patient, for example, excessively direct questioning on taboo topics. The overall atmosphere in consultations should allow interpreters to voice concerns about the relational dynamics within the triad.

The third major topic of metacommunication lies in interruptions of the interpreter during renderings. Interpreters can sensitize clinicians to the fact that on-the-spot interruptions lead to renderings that are less close to the original discourses (see excerpt 5).

In practice, inter-professional metacommunication can take place during pre-consultation briefings, within consultations, and in post-consultation debriefings. Clinicians should remain cognizant of power disparities between themselves and the interpreters when they introduce and manage metacommunication (Leanza et al., 2008; Weber, 2020), as interpreters depend on clinicians for their work opportunities. While interpreters may have more training and experience of interpreter-mediated psychiatric consultations than psychiatrists, they may refrain from sharing their knowledge spontaneously, since this might challenge the clinicians’ expertise.

Interpreters may also be hesitant to implement good practices, such as rendering fuzzy patient speech in a fuzzy way, if they are unsure that clinicians actually promote these practices. They may be concerned by a misconception by the psychiatrist that they are unable to formulate clear renderings because of lacking language skills or because of convoluted communication patterns (e.g., indirect and overpolite) in their cultures of origin. These issues can be addressed by metacommunication between professionals.

Metacommunication about clinical needs, communicative needs, and working modalities should take place as early on as possible and as often as needed throughout clinician–interpreter collaborations. This may seem difficult to achieve in a typically busy daily practice, or conflict with the protection of consultation time. Our results show that it is worth it for the clinician to “lose” a few minutes with the patient so that an agreement can be articulated with the interpreter on common goals and practices. Metacommunication makes triadic encounters more comfortable and more efficient for all participants. This can also be explained to patients if needed.

Acknowledgements

We would like to thank Appartenances, the employer of the interpreters involved in the study, for the fruitful collaboration, and the Swiss Federal Office of Public Health for the provided financial support. We are particularly grateful to all participants of the filmed consultations. Many thanks to Camille Bohner and Sam Musy for transcription tasks, to the interpreters involved in the translation of the data, and to Melissa Riat for her help during clinical analysis. Thanks also to Yvan Leanza, Anne Delizée, Nicolas Chambon, Véronique Traverso, Anna-Claudia Ticca and Vanessa Piccoli for offering us opportunities to discuss our methodology and results with peers at various stages of our project.

Biography

Orest Weber, PhD, is a senior research manager at the Department of Psychiatry at Lausanne University Hospital (Community Psychiatric and Liaison Psychiatric Services) and a private lecturer at Lausanne University. Dr. Weber’s research focuses on health communication with migrants and other vulnerable and minority groups. He is currently in charge of a project developing cultural competency and sensitivity education for clinicians taking care of End-of-Life patients and their caregivers. His publications pertain to misunderstandings in multilingual medical settings, interpreter-mediated interactions in health care, clinical interactions with elderly persons, and conceptual and practical issues of cultural competency and sensitivity training.

Jonathan Klemp, MSc, is a clinical psychologist at the Department of Psychiatry at Lausanne University Hospital and a researcher at the Institute of Psychology at Lausanne University (Faculty of Social and Political Sciences). His current clinical practice involves adolescents and young adults at the Child and Adolescent Psychiatric Service (Unite DEPART) at Lausanne University Hospital. From 2019–2020 he worked at the Community Psychiatric Service (Unité Psy&Migrants) as a research assistant. He has co-authored publications on interpreting in mental health.

Florian Chmetz, MSc, is a psychologist. He worked as a research assistant at the Department of Psychiatry at Lausanne University Hospital (Unité Psy&Migrants) and the Agalma Foundation. He co-authored publications on the consolidation of memory and on interpreting in mental health settings. Currently he works at the Clinique La Métairie in Lausanne.

Argyro Daliani, MD, is a psychiatrist. She worked at Lausanne University Hospital (Unité Psy&Migrants) until October 2021. Her clinical domains of expertise encompass migrant patients with complex mental health conditions and therapeutic groups. She co-authored publications on interpreting in mental health. She now works in a private practice in Heraklion, Greece.

Esther-Amelie Diserens, MSc, was a clinical psychologist at the Department of Psychiatry in Lausanne. She worked for 15 years with migrant patients and has extensive experience in transcultural care in mental health and in interpreter-mediated consultations. She published work on frequent users of emergency services and co-authored publications on the use of interpreters in mental health.

Florence Faucherre, MD, was the medical Director of the transcultural psychiatry unit (Unité Psy&Migrants) at Lausanne University Hospital until October 2021. Dr. Faucherre’s research focuses on mental health services for asylum seekers and other ethnic populations. Her publications focus on mental health services, with an emphasis on conceptual and methodological issues with minority populations, risk behaviors, and disparities in healthcare delivery. Currently she works in a private practice in Lausanne.

Footnotes

Author note: OW and FF developed the research questions and design with some help of FC. OW and FC collected the data and coordinated the project. FF, AD and EAD realized the clinical analysis. They ranged also among the filmed psychiatrists. OW was in charge of the linguistic analysis. JK and OW carried out the final coding of the corpus and prepared the excerpts. OW wrote the article with contribution from JK. All authors read and commented on the paper and approve its content.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this work was supported by the Federal Office of Public Health (Switzerland), grant number 339/2012.

References

  1. Bauer A. M., Alegría M. (2010). Impact of patient language proficiency and interpreter service use on the quality of psychiatric care: A systematic review. Psychiatric Services, 61(8), 765–773. 10.1176/ps.2010.61.8.765 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bot H. (2005). Dialogue interpreting in mental health. Rodopi. [Google Scholar]
  3. Brown P., Levinson S. (1987). Politeness: Some universals in language use. Cambridge University Press. [Google Scholar]
  4. Bruwier N., Delizée A., Michaux C., de Brouwer S. (2019). Evolution du service de traduction et interprétariat en milieu social bruxellois (SeTIS Bruxelles): De la reconnaissance à la valorisation des compétences chez l'interprète. FITISPos International Journal, 6(1), 28–45. 10.37536/FITISPos-IJ.2019.6.1.205 [DOI] [Google Scholar]
  5. Cox A. (2020). Venturing into interdisciplinary research as an applied linguist: Challenges and opportunities. In Weber O. (Ed.), L’interprétariat dans la santé: études linguistiques de terrain et activités de formation: actes du colloque de Lausanne, 13 mars 2020 (pp. 13–37). Université de Lausanne (BIL n.31). [Google Scholar]
  6. Delizée A., De Ridder P. (2016). L’interprète-kaléidoscope ou le questionnement identitaire de l’interprète en santé mentale. Un état de la question. Université de Mons. [Google Scholar]
  7. Delizée A., Michaux C. (2019). The negotiation of meaning in dialogue interpreting. On the effects of the verbalization of interpreters’ inferences. Translation, Cognition and Behavior, 2(2), 263–282. 10.1075/tcb.00029.del [DOI] [Google Scholar]
  8. Farooq S., Fear C. F., Oyebode F. (1997). An investigation of the adequacy of psychiatric interviews conducted through an interpreter. Psychiatric Bulletin, 21(4), 209–213. 10.1192/pb.21.4.209 [DOI] [Google Scholar]
  9. Gentile A., Ozolins U., Vasilakakos M. (1996). Liaison interpreting: A hand-book. Melbourne University Press. [Google Scholar]
  10. Gerwing J., Li S. (2019). Body-oriented gestures as a practitioner's window into interpreted communication. Social Science & Medicine, 233, 171–180. 10.1016/j.socscimed.2019.05.040 [DOI] [PubMed] [Google Scholar]
  11. Goffman E. (1955). On face-work: An analysis of ritual elements in social interaction. Journal for the Study of Interpersonal Processes, 18(3), 213–231. 10.1080/00332747.1955.11023008 [DOI] [PubMed] [Google Scholar]
  12. Gumperz J. (1999). On interactionnal sociolinguistic method. In Sarangi S., Roberts C. (Eds.), Talk, work and institutional order: Discourse in medical, mediation and management settings (pp. 453–471). Mouton de Gruyter. [Google Scholar]
  13. Jefferson G. (1983). Issues in the transcription of naturally-occurring talk: Caricature versus capturing pronunciational particulars. Tilburg Papers in Languge and Literature, 34, 1–12. [Google Scholar]
  14. Leanza Y., Bischoff A., Meyer B. (2008). Community interpreter’s power: The hazards of a disturbing attribute. Curare, 31(2–3), 211–220. [Google Scholar]
  15. Leanza Y., Boivin I., Moro M.-R., Rousseau C., Brisset C., Rosenberg E., Hassan G. (2015). Integration of interpreters in mental health interventions with children and adolescents: The need for a framework. Transcultural Psychiatry, 52(3), 353–375. 10.1177/1363461514558137 [DOI] [PubMed] [Google Scholar]
  16. Leanza Y., Brisset C., Rocque R., Boilard A. (2017). Challenges to and recommendations for working with a community interpreter in mental health. A Canadian perspective. In Jacobs E., Diamond L. (Eds.), Providing health care in the context of language barriers: International perspectives (pp. 56–70). Multilingual Matters. [Google Scholar]
  17. Li S. (2015). Nine types of turn-taking in interpreter-mediated GP consultations. Applied Linguistics Review, 6(1), 73–96. 10.1515/applirev-2015-0004 [DOI] [Google Scholar]
  18. Marcos L. R. (1979). Effects of interpreters on the evaluation of psychopathology in non-English-speaking patients. The American Journal of psychiatry, 136(2), 171-174. 10.1176/ajp.136.2.171. [DOI] [PubMed] [Google Scholar]
  19. Mason I. (2006). On mutual accessibility of contextual assumptions in dialogue interpreting. Journal of Pragmatics, 38(3), 359–373. 10.1016/j.pragma.2005.06.022 [DOI] [Google Scholar]
  20. Mason I. (2009). Role, positioning and discourse in face-to-face interpreting. In de Pedro Ricoy R., Perez I., Wilson C. (Eds.), Interpreting and translating in public service settings: policy, practice (pp. 52–73). Pedagogy St. Jerome. [Google Scholar]
  21. Mason I. (2012). Gaze, positioning and identity in interpreter-mediated dialogues. In Baraldi C., Gavioli L. (Eds.), Coordinating participation in dialogue interpreting (pp. 177–195). Benjamins. [Google Scholar]
  22. Mason I., Stewart M. (2001). Interactional pragmatics, face and the dialogue interpreter. In Mason I. (Ed.), Triadic exchanges: Studies in dialogue interpreting (pp. 51–70). St. Jerome. [Google Scholar]
  23. McAllister-Williams R. H., Cousins D., Lunn B. (2012). Clinical assessment and investigation in psychiatry. Medicine, 40(11), 568–573. https://doi.org/10.1016/j.mpmed.2012.08.014 [Google Scholar]
  24. Niemants N., Cirillo L. (2017). Dialogue interpreting. Research, education and professional practice. In Niemants N., Cirillo L. (Eds.), Introduction: Teaching dialogue interpreting. Research-based proposals for higher education (pp. 2–25). Benjamins. [Google Scholar]
  25. Pomerantz A. (1988). Offering A candidate answer: An information seeking strategy. Communication Monographs, 55(4), 360–373. 10.1080/03637758809376177 [DOI] [Google Scholar]
  26. Roberts C., Moss B., Wass V., Sarangi S., Jones R. (2005). Misunderstandings: A qualitative study of primary care consultations in multilingual settings, and educational implications. Medical Education, 39(5), 465–475. 10.1111/j.1365-2929.2005.02121.x [DOI] [PubMed] [Google Scholar]
  27. Sacks H., Schegloff E. A., Jefferson G. (1974). A simplest systematics for the organization of turn-taking for conversation. Language, 50(4), 696–735. 10.1353/lan.1974.0010 [DOI] [Google Scholar]
  28. Schegloff E. A. (1987). Some sources of misunderstanding in talk-in-interaction. Linguistics, 25(1), 201–218. 10.1515/ling.1987.25.1.201 [DOI] [Google Scholar]
  29. Searight H. R., Armock J. A. (2013). Foreign language interpreters in mental health practice: A literature review and research agenda. North American Journal of Psychology, 15(1), 17–38. [Google Scholar]
  30. Singy P., Guex P. (2005). The interpreter’s role with immigrant patients: Contrasted points of view. Communication and Medicine, 2(1), 45–51. 10.1515/come.2005.2.1.45 [DOI] [PubMed] [Google Scholar]
  31. Skaaden H. (2019). Invisible or invincible? Professional integrity, ethics, and voice in public service interpreting. Perspectives: Studies in Translation Theory and Practice, 27(5), 704–717. 10.1080/0907676X.2018.1536725 [DOI] [Google Scholar]
  32. Sperber D., Wilson D. (1989). La pertinence: Communication et cognition. Minuit. [Google Scholar]
  33. Statistiques Vaud. (2020, September 25). Retrieved October 26, 2020, from https://www.lausanne.ch/portrait/carte-identite/population.html
  34. Thomson M. S., Chaze F., George U., Guruge S. (2015). Improving immigrant populations’ access to mental health services in Canada: A review of barriers and recommendations. Journal of Immigrant & Minority Health, 17(6), 1895–1905. 10.1007/s10903-015-0175-3 [DOI] [PubMed] [Google Scholar]
  35. Ticca A. C. (2018). The interpreter’s activity between complexity and simplification in psychotherapy sessions. In Anderson L., Gavioli L., Zanettin F. (Eds.), TAIL: Translation and interpreting for language learners: A volume of teaching/learning activities (pp. 57–71). Bologna Studi AItLA. [Google Scholar]
  36. Van Vaerenbergh L. (2020). Ethics and good practice in interpreting multimodal communication in (mental) healthcare. In Hohenstein C., Lévy-Tödter M. (Eds.), Multilingual healthcare: A global view on communicative challenges (pp. 311–331). Springer Fachmedien Wiesbaden. [Google Scholar]
  37. Vertovec S. (2007). Super-diversity and its implications. Ethnic and Racial Studies, 30(6), 1024–1054. 10.1080/01419870701599465 [DOI] [Google Scholar]
  38. Wadensjö C. (2001). Approaching interpreting through discourse analysis. In Gile D., Dam H. V., Dubslaff F., Martinsen B., Schjodlager A. (Eds.), Getting started in interpreting research: Methodological reflections, personal accounts and advice for beginners (pp. 185–198). Benjamins. [Google Scholar]
  39. Wadensjö C. (2020). Involvement, trust and topic control in interpreter-mediated healthcare encounters. Communication and Medicine, 15(2), 165–176. 10.1558/cam.38681 [DOI] [Google Scholar]
  40. Weber O. (2017). Migration et communication médicale: difficultés de compréhension dans les consultations centrées sur des douleurs chroniques. Lambert-Lucas. [Google Scholar]
  41. Weber O. (2020). Entretiens avec interprètes en santé mentale: un projet multidisciplinaire d’analyse des pratiques et de formation. In Weber O. (Ed.), L’interprétariat dans la santé: études linguistiques de terrain et activités de formation: actes du colloque de Lausanne, 13 mars 2020 (pp. 81–103). Université de Lausanne (BIL n.31). [Google Scholar]
  42. Weber O., Daliani A., Diserens E.-A., Chmetz F., Faucherre F. (2017). La recherche interdisciplinaire sur l’interprétariat dans les services publics : expériences tirées d’une étude exploratoire dans le cadre d’investigations psychiatriques. Dragoman – Journal of Translation Studies, 7, 38–59. [Google Scholar]
  43. Weber O., Faucherre F. (2020). Training interpreters for their involvement in mental health interventions: The development of a teaching kit. Cahiers de Rhizôme, 75, 99–109. https://doi.org/10.3917/rhiz.075.0099 [Google Scholar]

Articles from Transcultural Psychiatry are provided here courtesy of SAGE Publications

RESOURCES