Linguistic resources and strategies used in multilingual communication in HIV / AIDS care centres in Lesotho

This paper gives a description of multilingual practices in two HIV-care centres in Lesotho on the basis of interviews with both health care providers and some patients who make use of the services of these centres. It considers the importance of effective linguistic communication in HIV-care and the hazards posed to such communication when physicians do not share the first language of the patients and of others working in these health care facilities. It gives the insights gained in a recent study on the kinds of interventions developed to facilitate communication in such multilingual institutional settings. In one of the centres informal interpreters are appointed to assist in transfer of information during consultations; in the second centre interpreting is only casually available from bilingual staff members. Besides interpreting, participants reported engaging in a number of other mediating practices. Evidence gained from informal interpreting studies elsewhere suggests that more than literal translation is required to achieve the kind of communicative success that will ensure quality in the health care provided to a vulnerable community. This study agrees with such findings and has generated a number of suggestions for improving the management of the linguistic diversity in communication within such clinics. The paper focuses on the specific resources provided in the healthcare centres and on the strategies participants use to enhance medical communication.

1. Multilingualism in the Lesotho health care system HIV/AIDS clinics in Lesotho are sites of multilingual health care.The multilingual situation in these clinics is explained by the fact that besides the wide distribution of Sesotho, English is widely used as a lingua franca and among the local population a considerable number know Afrikaans as an additional language.Additionally, due to the reliance of the health care system on health care workers of foreign origin, referred to as expatriate personnel 1 , a range of other languages are also represented.In general care, but also in dealing with the escalating HIV/AIDS prevalence, Lesotho is dependent on a considerable contingent of such foreign medical practitioners.Recent UNAIDS statistics show that Lesotho has an estimated adult HIVprevalence of 23%, which is related to an estimated 14,000 deaths per year.In view of the strain the pandemic puts on human capital and social structures, the Ministry of Health and Social Welfare (MOHSW) engaged in several intervention programmes that culminated in a roll-out of HIV treatment in 2004.As access to treatment spread across the country, a new challenge in terms of shortage of human resources emerged in that more trained health care professionals (HCPs) are required than the system currently employs.This led to increased recruitment of expatriate physicians so that presently there are more physicians in the country who do not have Sesotho in their linguistic repertoire, than ones who do.Cohen et al (2009:3) have reported that 80% of the physicians working in Lesotho are foreigners who speak a range of different first languages (L1s).Such languages would include (e.g.) Kiswahili, Kinyarwanda, French, Spanish, Dutch and German.In many cases these physicians are not fluent in the local lingua franca which is English, and where they have started to learn Sesotho their proficiency is often quite low2 .Add to this that the patients, nursing and administrative staff at the clinics are mostly L1 speakers of Sesotho with English as a second language (L2) (see Khati 1999:3), it is apparent that communicative difficulties typically encountered in interaction between L1 speakers of different languages in multilingual communities will occur.Language discordance between speaker and hearer where the conversants have limited levels of competence in the languages that each knows, has implications for the effectiveness of communication.As will be indicated below, such difficulties include misunderstanding regarding reported symptoms or prescribed medication.Language discordance then may inhibit successful communication in healthcare, particularly between physicians and patients in HIV/AIDS care where much of the quality of care is dependent on participants understanding one another3 (Lukoschek et al. 2003;Anthonissen and Meyer 2008;Shaw et al. 2009;Bharath-Kumar et al. 2009).
Researchers considering language discordant healthcare communication elsewhere (Drennan and Swartz 2002;Harmsen et al. 2003;Meeuwesen et al. 2006Meeuwesen et al. , 2007Meeuwesen et al. , 2009;;Deumert 2010) have noted how communicative challenges related to language discordance have a negative impact on the quality of care in such contexts.The specific challenges observed in other studies include impediment to the establishment of trust and rapport between patients and health care providers (Deumert 2010:56;Van de Poel and De Rycke 2011: 71) and inaccurate diagnosis or treatment of patients (Drennan and Swartz 2002;Deumert 2010:58;Elderkin-Thompson et al. 2001:1344).
Similarly, language discordant clinical interactions in Lesotho have been found to display communicative challenges that often lead to misunderstandings regarding e.g.advice on HIVtesting, causes and treatment of HIV, prescription or sensible use of prescribed medication (see Sobane 2013).Such challenges indicate a need for some form of language intervention that would improve the effectiveness of communication and lead to patients' positive experiences of care.This entails feeling properly heard and understood by a physician, as well as assurance of achieving the desired health outcomes which includes adherence to prescribed medication, heeding calls to return for check-ups and related improved results in clinical tests.In this paper we describe and critique the language intervention resources and strategies observed to be available in two HIV/AIDS health care centres in Lesotho.The focus is on the specific resources that the health care centres provide and the strategies that the participants use to enhance communication between physicians and patients in medical consultations.

Research Method
Following the recommendations of Mills et al. (2008) for health research, the study on which we report here entailed case studies of the interpreting services introduced in response to language discordance between physicians and patients in two HIV care facilities.The intention was to collect detailed information on patients' and carers' experiences in the complex process of communication in HIV care across the trajectory of admission, testing, consulting and advising, prescribing and dispensing antiretroviral treatment (ART) and following up where necessary.Data was collected through direct onsite observations in two multilingual HIV/AIDS care centres in Lesotho for a period of four months from March to July in 2011, semi-structured interviews with various role players, and focus-group discussions with patients.Such clinics in Lesotho have been established in the past 10 years to provide dedicated care through counselling, testing and treatment of HIV.They also provide TB treatment to the local population living in close proximity.The two clinics differ in the size of the facilities, the numbers of patients they can treat, and staff composition.The first one, (HC-1), is a public health unit of a hospital, therefore it is an establishment with a larger infrastructure, larger patients base and more staff than its counterpart, (HC-2), which is smaller and has a smaller patient and staff population.HC-1 is a hospital owned by a Christian Association and is situated in a village around 70 km from Maseru, the capital city of Lesotho.Although the hospital has been operating for a long time, the HC-1 unit was established in 2002 to provide specialised caring for HIV positive patients.This centre was of particular interest for the study because many staff members, particularly doctors, have a variety of linguistic repertoires and cultural backgrounds.There is an established interpreting service which makes use of lay interpreters, i.e. bilingual or multilingual people with no formal training as interpreters.The second clinic, (HC-2), is a much smaller private health clinic which was opened in 2005 to serve a rural community based about 40 km from Maseru.The centre provides a complete programme of care for HIV positive patients, also running a feeding scheme for such patients.This centre was also interesting because of the multilingual nature of clinical consultations conducted there.It is served by a single expatriate doctor who does not speak Sesotho.The facility has no interpreting services, yet the patients are largely Basotho who know no or very little English.The participants were selected through purposive selection and were screened and included on the basis of willingness and availability to participate.Participants were informed that this study is interested in which languages are used in which kinds of communication within the clinics.In agreement with ethical clearance conditions, they were told how results would be used and disseminated; they were then requested to sign consent forms, and were assured that no personal references that could identify them individually would be used in the writing up of the outcomes.Care was taken to make a selection that represents all the categories of participants (patients, counsellors, nurses, physicians, administrators, interpreters) who are actively involved in HIV/ AIDS care from both health institutions.This was done with the aim of gaining perspectives on communication and real experiences of linguistic intervention from the perspective of different role players with different linguistic repertoires.The total sample from the two clinics was 45.Table 1 shows a breakdown of the categories of participants who provided data in each clinic.Interview and focus group data were collected by means of semi-structured interviews (see appendices A and B for interview schedules).Attention of staff and patients was drawn to questions on how they experience language discordant clinical interactions -whether the diversity in competences were noted, presented challenges, were easy to manage and overcome, etc.Firstly, semi-structured individual interviews were held with each participant, and then 5 different focus groups were held with patients only in groups of five.The focus groups were meant to check interview data and to elicit any additional information that patients might have omitted in interviews.Special sensitivity was required to overcome reticence of patients who presented as "out" in terms of their HIV status, but still hesitant as clients dependant on the benevolence of the healthcare centres.
The recorded data were transcribed and translated into English in preparation for analysis, using the methods of thematic analysis (Braun and Clarke 2006;Fereday 2006) and qualitative content analysis (Mayring 2000;Patton 2002;Hsieh and Shannon 2005).This entailed repeated and thorough reading of the transcripts, and identification and coding of themes.The aim here was to systematically classify data so that themes that occur repetitively, as well as the attitudes of various participants regarding these themes, become apparent.Such themes, in this study, were then topically analysed with a view to categorising findings, interpreting the status quo, recognising challenges and suggesting alternative practices.
3. The extent of language diversity in the two clinics Participants' linguistic profiles show that among staff and patients in both clinics there are no monolingual repertoires; the 39 staff and patients from the local community who participated are bilingual in Sesotho and English; among the 6 physicians most have a repertoire of at least three languages.In the physicians' consulting rooms and around the clinics the languages most widely used are the two official languages of Lesotho, namely Sesotho and English.The fact that Sesotho is the L1 of a majority of patients, nurses and administrative staff has prompted many physicians to start learning the language so that all of them have some Sesotho proficiency 4 , although none reported that they conduct consultations using the L1 of the patients.Participants exhibit varying degrees of proficiency in English, which is the lingua franca operative between L1 speakers of Sesotho and L1 speakers of languages other than English.Due to the mutual unintelligibility of the L1s of patients and physicians, and in respecting the limited understanding physicians may have of Sesotho, the language of the workplace for many Sesotho L1 staff, is English.Where patients are proficient in English it is not unusual for staff and patients who both have Sesotho as L1, to use English in the formal clinic context.It was, however, observed that as with many patients, some of the physicians have relatively low levels of proficiency in English.The majority of the physicians come from francophone countries in Africa, so that besides their indigenous African L1, they have relatively good L2 proficiency in French.This poses a communicative challenge in that sometimes physicians do not understand interpreters who use English and Sesotho, nor do such interpreters follow them properly.This appears to be related not only to differences in accent, but also in vocabulary, stylistic idiosyncrasies and differences in idiomatic expressions.This would have obvious effects on mutual understanding as in the posing and answering of questions, giving and following advice regarding lifestyle and treatment.
This mismatch of language competencies among physicians, patients and other staff creates a need for some form of language intervention (professional interpreting, informal interpreting, code switching, etc.) to facilitate communication in clinical consultations.Effective communication is crucial in the consultation between physician and HIV-patient because it is in the physician's consulting room where critical decisions are made about the clinical stage and treatment of patients.Mutual understanding is therefore of utmost importance at this point.

Interventions and strategies to achieve effective communication
The efforts to respond to language diversity and to facilitate effective communication in these clinics come in two forms, namely as institutionally driven initiatives, or as intuitive lay responses.In HC-1 the institution itself provides resources in the form of an organisationally supported interpreting service.Other kinds of communicative assistance occur, but are not organisationally provided for.In HC-2 there is no dedicated organisational framework that provides for interpreting, so that where required, interpreting relies on communicative strategies informally developed by staff and patients and initiated by either depending on the particular verbal exchange.In the following sections we present a discussion of the various interventions and strategies that were found to be used in clinic interactions where the language discordance was of a kind that threatened effective communication 5 .Such threats were found to come in the form not only of L1 competences of mutually unintelligible languages, but also of varying proficiency levels in the lingua franca.
4 Proficiencies were not tested.All physicians in this study acknowledged that they cannot use Sesotho in consultation.Some can use odd words that they have picked up, others felt that they could follow the gist of conversations between nurses and patients.As, for reasons of privacy, no consultations were observed, the researcher could not gauge levels of proficiency, nor verify claims that were made as to communicative competences.
5 "Effective communication" in this context refers to patients following the full extent of the care process -from early consultation through testing and clinical consultation to the administering of medication where that is required.It also refers to physicians being able to understand and make themselves understood by patients and nursing staff, in such a way that proper care is not compromised due to linguistic difference.

Interpreting services
Interviews with patients and staff in both health clinics made it clear that interpreting is the most commonly used intervention in both health clinics.Patients relying on their uncertain knowledge of English, or physicians relying on their scant knowledge of either English or Sesotho, code switching and using gesture, visual materials, sketches, etc. to facilitate are very rarely observed or reported.The role of interpreters and their value in facilitating communication in medical consultations have been noted also in the studies of Drennan and Swartz (2002), Angelelli (2004), Wiener and Rivera (2004) and Fatahi et al. (2010).Fatahi et al (2010:779) observe that in a clinical interaction mediated by an interpreter with linguistic competence in both languages, quality style of interpreting and awareness of culture, positive health outcomes are achievable.This implies that not using interpreters when they are obviously needed, compromises the quality of healthcare.Wiener and Rivera (2004:94) concede to this, noting that when there are no medical interpreters, patients lack understanding of the medical encounter in general, therefore they are less satisfied and less compliant than in interpreted consultations.In another context, Drennan andSwartz (2002:1858) established that when interpreters are not used in psychiatric health institutions, psychosocial assessment is compromised and there are high chances of misdiagnosis and faulty prescriptions.
Despite their similarity in depending on interpreting in consultations, the two clinics in this study differ markedly in the ways in which they have formalised interpreting services.Even so, it has to be noted that neither of the facilities have the services of qualified interpreters, even if HC-1 does employ lay interpreters.Where such community interpreters have been appointed, they gained the position on the strength of their being fluently bilingual6 to the satisfaction of the appointing body.Thus, whether formally appointed for interpreting or informally drawn into an interpreting role, all interpreters would count as "lay interpreters" as none have had special training -neither as medical interpreters, nor as interpreters more generally.In multilingual societies such as the one in Lesotho, much informal interpreting goes on in the form of more proficient family members, friends or officials stepping in and relaying meaning to those who are not proficient enough in the dominant language.This is encountered in public spaces where the local population interacts with appointees in institutions such as the municipal offices, the police station or traffic department, in some educational contexts, and so on.Even so, the ability to interpret is valued as a relatively scarce skill, and those who do it well are often called on to assist.In spite of an enduring and pervasive need for interpreting, there are very limited opportunities for training and certification as an interpreter in Lesotho.
The two institutions differ in the ways in which community interpreting services are made available as a formal organisational feature of each clinic.In HC-1 interpreting services are a formalised, standard component of the physician's consultation.English to Sesotho interpreting is provided for every patient, regardless of whether the patient has proficiency in English or not.Patients are not asked to choose whether they would like to make use of an interpreter or not.An interpreter sits in the consultation at all times.Family members, who join the consultation when patients so prefer, do not come in as interpreters.Nurses are present as the procedure may require.Even if they could interpret, the task is assigned to the interpreter.
Interviews with interpreters show that when they establish that a patient is fluent enough in English to communicate directly with the physician, they stop interpreting but do not leave the room.They explained that this practice is justified by the possibility that the communication could become more complex and they would then be needed.
In HC-2 where there is no formal, institutional arrangement for interpreting, a practice has developed in which the physician calls in a bilingual nurse or a particular counsellor to interpret when according to his/her judgement there is a breakdown in communication in English.
Otherwise, the physician and his patients get by using Sesotho (with much code-switching between English and Sesotho), even where the physician speaks and understands Sesotho only limitedly.
Further, the two institutions differ in that different categories of staff perform interpreting duties.In HC-1 interpreting is done by (i) designated lay interpreters from the local community; (ii) administrative personnel who have no medical training; and (iii) nurses who speak the community language (Sesotho) as well as the lingua franca.In HC-2 interpreting is done "on demand" by (i) a nurse or (ii) a counsellor, depending on who is available when the physician requires interpreting assistance.There are obvious concerns about the quality of the interpreting given that even among speakers with the same L1 doctor-patient communication can be fraught and fragile.Such concerns are exacerbated when the interpreter has no specialised background in medical care or medical terminology.

Printed materials
Another intervention, besides the availability of interpreting services, that the healthcare centres introduce, is to provide printed materials for patients to read either in the waiting areas or when they have left the clinic.These printed materials are usually produced by NGOs in liaison with the Lesotho Ministry of Health and Social Welfare and are distributed for free to healthcare centres7 .The materials are informative on living with HIV, prevention of new infections, and behavioural changes related to HIV.They also serve as reminders of what was verbally imparted to patients at the clinic.The printed materials provide supplementary information about what HIV is, how it is contracted, how it can be prevented and what treatment is available.
The aim of such material is also to bridge the gap in transfer of information that may arise in the consultation room due to language discordance.These printed artefacts, mostly in the form of information leaflets, booklets and posters, are written either in Sesotho or English, and are at times published in both these languages.The use of such resources as communicative tools in multilingual clinics has been investigated and topicalised in research on multilingual health communication by a number of scholars, including Collins and Slembrouck (2006) and Moyer (2011).More specifically, research by Saal (2004), Feinauer and Luttig (2005) and Swanepoel (2005) refer to different booklets used in HIV and AIDS awareness raising campaigns in South Africa.All these studies register specific challenges regarding the production, circulation and accessibility of these materials.Often high levels of patient literacy are assumed and on proper investigation, it transpires that such assumptions are false.
Similar findings were made regarding the printed materials distributed to patients in the two clinics: in the interviews, staff confirmed that patients rarely would read them or otherwise refer to the content.No specific research has informed their user-friendliness, their alignment with patient literacies, and the ways patients are likely to actually use them.Also, there has been no study of what the information needs of visitors to the clinic are, nor how best to answer to those needs in print (or through other communicative media).
In commenting on the printed materials they came across in the clinics, patients in this study reported that they had never been asked questions that would indicate the benefit such printed matter could have; they were not asked whether they could read, liked to read, or understood English.Some to whom the written Sesotho was clear, reported that the English parts of bilingual texts were communicatively disruptive in that they did not find texts given alternatively in two languages, easy to follow.This means that staff and institutions that produce and circulate such published information, assume that all patients are literate to the extent that they can handle genres quite foreign to their everyday lives.This echoes a finding of Moyer (2011Moyer ( :1214) ) that production of and attempt to use a translation manual was premised on patients being literate, while in fact it was an inappropriate resource for limitedly-literate patients.Thus printed materials as possible forms of mediating HIV-information in the two Lesotho clinics appear not to have been helpful.
Answering a question as to whether they read the pamphlets made available to them, patients in the two clinics reported that they hardly ever read the folders as reading didn't come easily to them and they found English texts prohibiting.Some reported that they do not have the time to read; others remarked that they looked only at the illustrations, or that they selected the parts written in Sesotho and read only those.These patients in effect admitted that although they believed the information to be valuable, they often did not themselves gain from the printed matter.Van de Poel and De Rycke (2011) have recommended that a comprehensive survey be done on whether health care institutions should continue to invest in provision of these materials.The information from HC-1 and HC-2 indicates that funding put towards printed material could be better spent on other aspects of care, if not on attention to improved kinds of visual information.

Informally developed communicative strategies for transfer of information
Apart from the above mentioned resources that are organisationally provided to assist multilingual communication, participants with incompatible linguistic resources have been found intuitively to develop facilitating communicative strategies.They rely on such strategies in communication between a patient and healthcare provider, where language discordance could cause misunderstanding, or even complete communicative breakdown.In the medical consultation some of these strategies are introduced by the physician, some by the patient, and others, where an interpreter is present, by the interpreter.

Non-verbal communication
One of the strategies that are regularly used in interviews among healthcare providers and patients is the use of deictic communication.For example, where a patient presents with opportunistic infections with visible physical symptoms or pain in a part of the body that can be pointed to, the patient will show the physician the symptoms or point to the ailing body part.
In some cases this non-verbal communication is solicited by the physician or interpreter, while in other cases it is volunteered by patients.In explaining the value of non-verbal communication, participants show that it is a valuable resource in assuring successful transfer of information.The physician in HC-2 describes the kinds of gestures people use as a kind of universal language that guarantees effective communication as follows: D1: ehm ... how can I say it?/ If the people I can see ...let's say we have non-verbal communication also, so it's ... non-verbal communication is quite international… The physician underscored the value of such non-verbal communication as a shared communicative tool when speakers do not share a language.
The challenge with non-verbal communication, deictic or otherwise, however, is that its use in consultation is largely limited to where the health problem has physical manifestations.This limitation was also noted in research elsewhere.For instance, in a study done in Austria, Dressler andPils (2009:1183) reported cases of limited verbal communication where they found health care workers unable to cover the range of interventions required in the rehabilitation process.Attention to aspects such as those used in psychological treatment was reported to be impossible in 50% of the cases.Similarly, Fatahi et al. (2010:777-778), in a study done in Sweden, recorded that the non-verbal mode becomes inadequate where treatment requires detailed communication, as in cases where a patient has to be instructed to periodically hold her breath, or where expected side effects of medication have to be communicated.Physicians in HC-1 and HC-2, where treatment is HIV-related, reported that when patient concerns are psycho-social, patient explanations become too complex to be handled through non-verbal communication.Participants are therefore obliged to find other means to ensure mutual understanding between physicians and patients.

Checking understanding
Another strategy used in clinical consultations and reported in interviews, is continual checking of patient understanding.The importance of mutual understanding in HIV/AIDS care, particularly ensuring that the patient understands the nature of the illness and the treatment protocol, has been highlighted in other research.For example, Watermeyer and Penn (2009: 207) note that the nature and wide distribution of HIV-infection and the complications associated with non-adherence, make it imperative that patients understand their condition and the treatment.In the Lesotho clinics, interviews with physicians indicated that patient understanding is checked at different points in the communicative event to determine whether further explanation or even repetition of information is necessary.This is how a physician in HC-1 explains their way of checking patient understanding by asking them to retell: D2: the method that we usually use… I think the the one for communications, is …I think we always ask what did they understand, for them to explain in their own terms… An illustration of such checking is given in Anthonissen (2010: 126-127) and Anthonissen and Meyer (2009:11) where, in recordings of consultations it regularly occurred that an explanation would be given of, for example, what a CD4-count is and why it is done.The explanation would be accompanied by questions such as "Can you remember what a CD4-count is? Tell me what it is."Such insistence on a show of knowledge is also reported in Watermeyer and Penn (2009:207) where pharmacists similarly required "demonstration of understanding".They found such insistence to be more effective than just asking a yes/no-question as to whether the patient has understood.At least four physicians and most of the clinical staff and interpreters in the Lesotho clinics reported frequently using such explicit and elaborate checking that patients were clear on information that the physician regarded as critical to their care.

Reorganising and retelling explanations
Effective communication was also found to be achieved by reorganising the original messages in a retell.This is done mainly by nurses when they take on the interpreting role, although other appointed interpreters also used this strategy.Interview data showed that interpreters mostly do this when they find physicians' explanations to be too restricted and scientifically articulated for patients to understand.The limited English proficiency of either doctor or patient added to such a need for retelling in a different register.
In other cases patient explanations are very long and apparently incoherent.For example, they give histories starting at a point in the quite distant past, narrating details which physicians find difficult to relate to the current problem.Such situations leave nurses and interpreters feeling responsible for breaking up the story, re-organising the information, on the one hand simplifying it, and on the other hand elaborating to make sure that the two parties understand each other.This means that they go beyond interpreting and produce unsolicited explanations to the party likely to have difficulty understanding the other.

Visual illustrative material support
Another strategy that physicians and interpreters report using is visual material.Often they take a piece of paper and draw a sketch for the patient.Such illustrations may, for example, be of a part of the anatomy, to indicate where a problem lies, detailing biological functions, and showing how the medication is going to help.The interviewed participants believe that the illustrations enlighten patients and that the strategy helps to explain the problem in a way that, due to the language barrier, could not be done verbally.Although illustrations are mostly done for patients by physicians, interpreters mentioned cases where they would make a drawing for a patient, thereby providing a detailed explanation not solicited by the physicians.According to physician reports in this data, illustrations are not only helpful where there is language discordance, but also where the cases are complex and terminology is too scientific to be familiar to patients.

Soliciting assistance from others
Where there is an apparent threat to communicative success, healthcare providers and patients report soliciting the intervention of another healthcare provider who is not part of the current conversation.Healthcare providers and interpreters will do this during the consultation; patients report that they will do so after the consultation.Interviews with physicians in HC-1 as well as HC-2 show that in cases of non-interpreter mediated consultations, they usually call in a nurse to interpret when patient explanations are too complex or lengthy for them to follow.Interpreters also report that they consult the nurses where they feel that they cannot explain sufficiently or where they are uncertain about the appropriate terminology.The nurses are consulted due to their shared knowledge of Sesotho, their advanced English in medical terminology compared to the informal interpreters' knowledge, and the medical qualification of nurses that does enable more clarity.
Patients report soliciting help from nurses after the consultation if they did not understand a physician or feel that the physician has not fully explained something.They also consult nurses if they have a problem that they cannot tell the physician due either to limited vocabulary, or due to issues of intimacy.A patient in HC-2 explained in Sesotho why she sometimes seeks assistance from the nurses in this way: P1: Ache nna ha nke be ke khutle ke etsa bonnete ba hore ke kopane Well I never come back (to the clinic), I make sure that I consult le batho bano ba le babeli.Mohlala, hona joale ke na le bothata bona both of them.For example, now I have this problem noo ke sa tsebeng na ke eng ene ne ke batla ho fumana thuso, ke entse hore that I don't know what it is, and I want to get help, ke kopanne le 'me X (mooki) a etse ka hohle hore tle ke fumane thuso.I consulted Mrs X (nurse), so that she can do everything to make sure that I get help.
This patient makes every effort to find someone she can talk to about her problem since she cannot discuss it with the physician.She feels confident that if she explains her problem more elaborately to the nurse, she will get the right kind of help.

Corrective measures after the consultation
The strategies reported in 4.2.1 to 4.2.5 above are mostly used in consultation with the physician in attempts to solve communication problems as they arise.However, many cases of communication breakdown only become apparent after a patient has left the consultation room.For example, when a patient talks to a nurse or other staff member after the consultation, it may become clear that something has either not been made clear enough to the doctor or has not been properly understood by the patient.Many such misunderstandings become apparent in the prescription notes.Staff who are from the local community reported that some communication failures arise not simply due to language diversity, but also due to the fact that physicians are not familiar with the Lesotho guidelines for HIV care since they are of expatriate origin.In the following excerpt a Sesotho L1 nurse explains a problem she once spotted: N1: … a le 3 years ebe ho thoe o fuoa Streptomycin.a 3 year old patient was given Streptomycin.Ae, liguideline tsa naha ha li cho joalo, ngoana eo No, the country guidelines don't say that, when this child ngoana ha thibana litsebe o tla tseba tseba ho cho hore o thibane litsebe?goes deaf, will s/he be able to say her/his ears are blocked?Ha foufala o tla tseba ho cho hore ha a bone ke streptomycin?When s/he goes blind, will s/he be able to say that s/he is blind due to the use of Streptomycin?This nurse's account is her way of indicating that the physician had prescribed medication that is not age appropriate, and which is not compliant with the Lesotho guidelines of care.Her concern is not only that a young patient may not be able to tell her of difficulties with hearing or vision and would not even realise that it is a side-effect of the medication; she is indignant about the physician's ignorance of critical local medical protocols.The physician may not have understood the nurse or parent when they gave the age of the child, or may not have understood guidelines for the use of the medication.Such a misunderstanding was noticed by a nurse who was not in the consultation, but was there when the prescribed medication was being dispensed.
The nurses were generally outspoken about frustrations they experienced in situations like this where they noticed what appeared to them to be bad medical practice related to language discordance.They felt concerned that ultimately they are the ones responsible for ensuring that a patient gets good health care.They believed that miscommunication between doctors and patients, even where interpreters were present, often leads to misdiagnosis and incorrect prescriptions.Equipped with medical knowledge and experience in their field, they report on four ways in which they react and try to remedy what they consider to be physicians' "mistakes".These are: a) sending the patient back to the physician with a note which points out the mistake and asks for a correction of the diagnosis and/or prescribed treatment; b) changing the prescription by writing in a "correction", and then asking the patient to go back to the physician for a signature; c) diagnosing the clinical problem as they themselves see it, and sending the patient to the laboratory for tests to confirm their diagnosis.If confirmed, they then prescribe and dispense medication for the problem; and d) simply ignoring the physician's transcription and dispensing medication they deem to be appropriate for the problem.
The narrative of these corrective measures shows a degree of confusion as to the roles standardly ascribed to nurses and physicians.Some nurses take authority that typically lies in the domain of the physician.Most striking is the fact that in some cases physicians are not consulted, while in others they are simply asked to endorse the changed prescriptions with a signature.On the one hand, when nurses take on a role for which they are not trained, it could be considered inappropriate, if not dangerous; on the other hand, these nurses seem to play a valuable role in the clinics because they are able to pick up mistakes and rectify them before patients leave the clinic.

Discussion
The efforts towards effective communication made by all participants in the two Lesotho clinics that provided data reflect that communication in these contexts is fraught with obstacles to comprehension.This study has highlighted the significance of communication in achieving quality health care to vulnerable patients, as has been indicated by others, such as Shaw et al. (2009) and Lukoschek et al. (2003).Effective communication ultimately is positively linked to the achievement of quality health care and positive health outcomes.Past research positively links effective communication with several indicators of quality health care and positive health outcomes such as patients' comprehension of diagnosis and treatment (Lukoschek et al. 2003;Roberts et al. 2005); patient satisfaction with the healthcare experience (Gallagher et al. 2005;Schouten et al. 2009); and improved self-care and adherence to treatment (Rivadeneyra et al. 2001;Harrington 2004).
Both institutions prove to be aware of this and therefore try to address communicative difficulties that arise due to language diversity.Although both institutions do try to address difficulties that arise due to language diversity, the lack of a clear guiding policy in this regard, and limited resources allocated to securing support such as reliable interpreting services, are responsible for the absence of such services in HC-2 and for the weaknesses of using untrained community interpreters in HC-1.This explains many instances of inadequate care even when the HCPs are working with best intentions.Research elsewhere points to the use of professional interpreters as a valuable contribution to health care in linguistically diverse settings.Such interpreters are able to foster cultural appropriateness in the discourses of physicians and patients (Rosenberg et al. 2007) and they are also able to provide the required emotional support for patients (Hsieh and Hong 2010).Kale and Syed (2010) therefore view the use of professional interpreters as a desirable way to overcome language barriers.The absence of professional interpreters in the two clinics in Lesotho could account for some of the reported instances of misunderstanding and impairment of patient satisfaction.The use of nurses as suitable interpreters has been suggested and is reported to be successful elsewhere.Evidence from the data we have does not convincingly support this as an ideal remedy.
Insufficient attention to the role of language in quality care becomes clear in the grumbling of nurses in our study who are occasionally called on to interpret while they are not trained or appointed for such a task.When untrained interpreters are used, there is a high risk of poor quality interpreting.Although this study did not specifically test such services, the reports of participants give us no reason to believe that the interpreting is flawless.Nevertheless, considering the pervasive shortage of trained interpreters, and the cost attached to instituting a highly dependable interpreting service, one has to recognise that for the foreseeable future the clinics will remain dependent on largely self-trained, informal interpreters.In such circumstances, the contribution of the appointed local interpreters in HC-1 has to be acknowledged.Similarly, the contribution of healthcare staff such as nurses and counsellors who add an interpreting function to their daily tasks, as happens in HC-2, needs to be acknowledged.

Conclusion
This study accentuates a need for a political will and clear policy towards managing language diversity at an institutional level, not only in hospitals and clinics, but also in other public spaces where the multilingualism of the community is obvious and thus contains potential communicative risks.In the meantime, in anticipating appropriate policy formulations, there are shorter term interventions that could make a considerable difference to service provision in the clinics.Such interventions would include (i) workshops that raise awareness of the functions of language in providing quality health care; (ii) short training courses to those already functioning as interpreters; (iii) guidance to otherwise appointed staff in how to manage the triadic communication between (e.g.) physician, patient and interpreting nurse; as well as (iv) clear guidelines to nurses on how and where to report perceived misdiagnosis or faulty prescriptions in a safe and supportive manner.
Where printed materials are given to patients as supplementary information, the value of printed information is not denied.However, this is likely to be more effective in medical care if it were developed with due consideration to existing advice on user-friendliness, language preferences and literacy experience of the intended readers.The data in this study shows no evidence that such consultation or surveys similar to those of e.g.Feinauer & Luttig (2005) and Swanepoel (2005) have been conducted in Lesotho.
Language discordance has shown some resourcefulness among healthcare providers and patients.Their ability to use a number of conversational strategies such as checking patient understanding, reorganising messages and enhancing explanations by visual illustrations (see discussion in section 4.2 above) to manage the flow of information and to achieve communicative goals was noted.However, such resourcefulness alone cannot be the only or primary recourse.The levels of efficiency of speakers' communicative creativity have not been investigated and documented.They still need to be used in conjunction with other regulated interventions such as interpreting.
Interestingly, although a number of the expatriate physicians in this study have acquired some communicative Sesotho, none of the participants suggested that more, or more intensive, language courses should be introduced.The possibility that physicians should learn Sesotho (or improve their English proficiency) was not suggested as a solution to the communicative difficulties that they recognised or encountered themselves.
In view of these findings, it is recommended that the already existing language intervention resources be enhanced to improve their effectiveness.The ideal for these two HIV clinics would be to make sure that interpreting services are available and that informal interpreters are at least minimally trained in medical interpreting and mediation technique, in medical terminology and in basic care considerations.In the longer term, translation and interpreting services should be professionalised and incorporated into the staff cadres of the Ministry of Health and Social Welfare in Lesotho.State contributions towards in-service training and certification for community interpreters, are strongly advised.Where nurses are used in interpreting, some compensatory token that recognises the skill, should be offered.
This study has illuminated some of the linguistic issues that characterise HIV/AIDS care in the particular contexts in Lesotho.It is recommended that research in this area should be intensified to provide practical guidelines to the HCPs on how to approach communication in these multilingual clinics.The study also identified a need for a clear language policy with regard to the incorporation of non-Sesotho speaking physicians in HIV/AIDS care.As a contribution to existing knowledge on how various role players manage language diversity in healthcare clinics, this study has described what happens in Lesotho where there are no language policies guiding language services and support in a healthcare system that is linguistically diverse in terms not only of the local population, but also as a result of heavy reliance on expatriate physicians.It emphasises the need for evidence based policies in order to create a linguistically accessible and equitable healthcare system that benefits all communities, particularly those dependent on state care.
At the moment every health care centre makes its own decisions depending on available resources and there is no clear policy on communicative procedures that have to be followed.While this study has done introductory work, it is necessary to undertake large-scale projects that would allow collection of data from a more diverse range of clinics in order to get a more general picture of the prevalent linguistic and communicative practices and what communicative interventions are appropriate, if not imperative, in HIV/AIDS care.Such research can inform policy formulation, and so enter the protocol of organisations that currently have developed their own structures, but could do with more official support.

Table 1 :
Categories of participants