Received: 8 November 2005
Accepted: 23 November 2005
Non-local language speakers, especially migrant populations and ethnic minority groups, often cannot communicate with their clinicians adequately to receive necessary information about their care. Members of the clinical staff often do not understand their patients’ needs, and so do not receive allrelevant information from the patient. This paper presents a report from nine European hospitals which worked together to improve communication between non-local language speakers and clinical staff by developing, implementing and evaluating professional interpreter services. The work was part of the Migrant-friendly Hospitals Project. The article outlines different strategies to improve clinical communication with migrant and ethnic minority patients such as telephone interpreting, face-to-face interpreting, intercultural mediation and written material as supportive information and presents an evaluation of their effectiveness from the perspective of the providers and thepatients in the hospitals concerned. This evaluation was based on a benchmarking design that included a pre- and a post-intervention staff survey and a post intervention patient survey. In general, evaluation results of the pre- and post-intervention staff surveys showed that the implemented measures proved to be effective for both groups of stakeholders. The paper closes with recommendations for a concertedhospital quality management response to the problem of language barriers in clinical communication.
client–providercommunication, clinical encounters, ethnic minorities, hospital setting, intercultural mediation, interpreting services, language barriers, migrants
The provision of culturally and linguistically appropriate communication in healthcare poses considerable challenges to health policy and frontline service provision alike. Hospitals in particular are consulted by an increasingly diverse clientele (Hampers and McNulty, 2002; Yeo, 2004). As a consequence, there are language barriers between specific patients and providers: patients coming to the hospital may not be able to speak and/or understand the local language, and staff may not be able to communicate in the patients’ mother tongue or a common language. This frequently leads to communication problems and misunderstandings that may negatively affect clinical encounters and treatment outcomes.
Clinical encounters are defined as client–provider communications occurring as part of diagnosis and treatment and which involve the negotiation of explanatory models between physician and patient (Pirbhai, 2003). In clinical encounters it is important that relevant clinical information is elicited from and conveyed to the patient in a correct and appropriate manner as a prerequisite for both clinical decision making and the establishment of trust between the two parties concerned. Effective communication is an essential part of successful treatment and in securing patients’ co-operation. Furthermore, it plays an important role in assessing pain and prescribing adequate medication (Cleeland et al, 1997; Brown et al, 1999).
Language barriers are a major obstacle in providing effective health services; they have adverse effects on accessibility, quality of care, patient satisfaction and patient health outcomes (Bischoff, 2003; Murphy, 2004). For example, language-related communication difficulties are associated with a higher rate of resource utilisation for diagnostic testing (Waxman and Levitt, 2000; Hampers and McNulty, 2002) and an increase in the likelihood of unnecessary invasive procedures (Bard et al, 2004); they negatively affect the continuity of care (Sarver and Baker, 2000; Yeo, 2004; Brotanek et al, 2005), and they create a barrier to the use of preventive services (Woloshin et al, 1997). Miscommunication can also lead to substantial extra costs in the provision of services: more time resources are required for routine encounters in the treatment process, leading to longer periods of stay (John-Baptiste et al, 2004) or additional organisational burden for the staff (Drennan, 1996). Moreover, delays in the treatment process or clinical decisions based on inaccurate information can result in poor treatment outcomes. In addition to threats to patients’ health and quality of life, these delays also bear the risk of actual liability costs for the healthcare institution (Bischoff, 2003).
Healthcare institutions have devised responses to these communication barriers in order to provide effective services to their linguistically diverse clientele. In the past, however, these responses have relied heavily on the ad hoc use of bilingual staff or patients’ family and friends to bridge language gaps. Such reliance carries several risks, including the likelihood of poorer healthcare due to inadequate communication and undesired health outcomes.
In recent years, however, many hospitals have seen the need to go further to ensure that quality care is provided to all patients. Professional strategies to facilitate communication between patients and providers have proven effective in improving service quality and outcomes for migrant and ethnic minority patients who face language barriers in utilising health services (Tocher and Larson, 1998; Jacobs et al, 2004), the most effective approach being the introduction of professional medical interpreters (Jacobs et al, 2004). In addition to their language competence, professional interpreters have a strong medical vocabulary and specific communication skills, as well as an understanding of the ethical issues involved in their role (Bischoff et al, 2003). While the cost of hiring interpreters or contracting their services from outside agencies may present barriers to their employment, a study in the US showed that the annual cost of interpreters per patient was compensated by significant improvements in patients’ service use, compliance and health outcomes (Jacobs et al, 2004).
The type of interpreting support provided differs significantly in terms of the healthcare systems concerned, the demographic make-up of communities, and the type of services provided (Schulze et al, 2003). For example, a hospital serving a clientele that has a high proportion of people who speak a specific language may benefit from hiring full-time interpreters. Another hospital with the same number of non-native speakers of many different languages will not be able to provide on-site service for everyone. In this case, making arrangements with freelance interpreters would seem to be the more feasible option. Hospitals that have to provide emergency services may benefit most from telephone interpreting services or may need translated material to support clinical communication.
With increases in global migration flows, the challenges posed to healthcare by cultural and linguistic diversity are likely to intensify. This calls for a concerted response and an expansion of the evidence base for effective strategies to address language barriers across a range of services and specific local circumstances.
TheMigrant-friendlyHospitals Project (Box 1) represented a wide spectrum of hospital services, ranging from large metropolitan university teaching hospitals to small-town community hospitals, with diverse migration situations in their catchment areas and different levels of experience in serving populations with limited proficiency in the local languages.
The project provided a range of solutions that were developed and implemented based on a combination of local needs assessments at the different sites, and a systematic review of effective interventions (Bischoff, 2003).
In the needs assessment, communication was identified as the most prominent area of concern in dealing with migrant and ethnic minority populations in clinical routines. The development of effective services to bridge the language gap between providers and diverse patient groups was therefore a central focus in the project. The emphasis in the project work under discussion here was on interpreting in clinical communication; that is to say, on conveying medical information across language barriers in spoken conversation, with the presence of an interpreter either in person or in a remote setting connected via telephone, video or other media.
Nine hospitals in Denmark, Finland, Greece, Spain, Ireland, Italy, The Netherlands, Sweden and the UK co-operated to improve clinical communication with migrant and ethnic minority patients (see Box 2). Altogether, 37.9% of the clinical staff in these hospitals, including doctors, nurses, therapists, social workers and clerical workers, reported daily contact with this patient group; another 28.5% reported contact two or three times per week. The aims of this part of the Migrant-friendly Hospitals Project were to ensure:
• the provision of professional interpreter services whenever necessary to ensure good communication between non-local language speakers and clinical staff
• that patients were informed about the language services available and how to access them
• that clinical staff were empowered to work competently with interpreters to overcome language barriers and obtain better outcomes
• the provision of education materials for patients in non-local languages.
In order to enhance clinical communication between staff and migrants, the participating hospitals implemented different measures to develop new and/or improve existing interpreting services.
Measures were planned according to the individual needs of each hospital. In some cases, no interpreting services existed, so those hospitals developed an entire plan for service development and implementation. More commonly, though, measures focused on improving existing services to increase their efficiency and effectiveness. Taking into account specific hospital structures and processes, different types of interpreting services were implemented: telephone interpreting, face-to-face interpreting, intercultural mediation, and/or written material as communication aid.
Two models of good practice that were developed and implemented in the project framework are described below to illustrate the type of specific solutions applied (see Boxes 3 and 4).
A pre- and a post-intervention staff survey and a postintervention patient survey were conducted using a benchmarking learning and evaluation design.
Hospital staff were questioned twice, once in a baseline survey in February 2004 (n = 479 in seven hospitals), and a second time in a further survey in June 2004 (n = 282 in six hospitals) after implementation of the measures. One hospital did not conduct the second staff survey because the intervention employed extended until after completion of the project.
Medical and nursing staff were asked, in a standardised written questionnaire, about their experiences of encounters with patients who had a limited command of the local language, and about how they (the staff) rated the quality of the available interpreting services. They were also asked to provide information on how they thought interpreting support should be improved. In the second survey, there was an extra question about the extent to which the individual work situation had improved as a result of project measures.
Three hospitals conducted a post-intervention patient survey (n = 42) parallel to the second staff survey in June 2004. Following a clinical encounter involving an interpreter, patients were interviewed about the attendance of the interpreter, their ratings of the interpreter’s work, the encounter with the doctor and their views as to how interpreting support should be improved.
In a third evaluation step, project managers at the hospitals answered a short questionnaire which asked them to assess the extent to which the measure could be implemented, to list both enabling factors and obstacles in the implementation process, and to provide information on the sustainability of the measures.
Analysis of the pre- and post-intervention staff surveys showed that the interventions, such as the soundstation, were effective. Comparison of the baseline and second staff surveys showed that the uptake of professional interpreting increased by 20.2%, with a concurrent decrease of nearly 10% in interpreting provided by adult relatives and friends (see Figure 1). However, a 2% increase was observed in the use of children, under the age of 18, as interpreters.
In addition, the quality of interpreting services was evaluated more positively after the interventions were introduced. Interpreters were said to be available in a timely manner and more frequently than they had been in the baseline survey (+17.5%). Furthermore, an improvementwas noted for all defined quality indicators for interpreting services, such as introduction and role explanation by interpreters (+10.8%), accurate transmission of information (+6.8%), clarification by interpreters (+7.7%), clarification of cultural beliefs (+10.5%) and the identification of patients’ further needs by interpreters (+7.3%). There was an improvement in the overall rating of interpreting services; the proportion of those judging the service to be excellent or very good increased by 21.3% and 54.9% of the staff members stated that the work situation had improved because of the interventions employed.
Analysis showed high levels of patient satisfaction with the performance of hospital interpreters: 78% of patients strongly agreed that, with the interpreter’s help, they understood all the medical information (the other 22% agreed); 83% strongly agreed that the interpreter assisted them in relaying important (lifestyle) information to the doctor (17.1% agreed). In general, ratings concerning the doctor’s performance were also very high: patients felt that doctors took time to talk with them (70.7% strongly agreed/29.3% agreed), showed sensitivity to their cultural beliefs and needs (68.3%/ 29.3%), encouraged them to ask questions (68.3%/ 29.3%) and worked together with the interpreter to make sure they understood everything (67.5%/30.0%).On the other hand, patients were less satisfied with direct doctor–patient communication: only 31.7% strongly agreed (14.6%agreed) that the doctor talked directly to them rather than to the interpreter when giving or asking for information, but 48.8% disagreed and 4.9% even strongly disagreed.
The interventions can be categorised in four groups:
• development/improvement of face-to-face interpreting
• development/improvement of telephone interpreting
• development/translation of written material as supporting communication
• implementation of intercultural mediation.
It was not possible to identify one specific measure as a model of best practice. In fact, the combination of the local situation, involved partners and the concrete measures seemed to be the determining factor for the acceptance, feasibility and the effectiveness of interventions.
Implementing the interventions turned out to be a difficult task. Local project managers had to adjust their strategies, make compromises and demonstrate considerable flexibility. Four interventions were implemented completely, eight to some extent, and two not at all. Two-thirds of the interventions were difficult to implement because of factors such as lack of acceptance and utilisation of the service, lack of sufficient time, and logistics. Only four could be easily implemented and these are described in Box 5.
From the project managers’ point of view, the aims of the project were very worthwhile. All of the interventions that were implemented will be continued. Three have become an established part of the hospitals’ service, four will be adapted and then included in service routines, and five will even be extended.
The evaluation confirmed existing international knowledge and experience about improving the quality of interpreting services in hospitals (Jacobs et al, 2004; Bischoff et al, 2003). Hospital co-operation within a scientifically supported benchmarking approach proved to be feasible and effective. Despite the huge diversity of healthcare systems, hospital and departmental structures, and migrant population characteristics, all partner hospitals had to handle very similar problems.
Comparability of data and experiences was, on the one hand, assured by a central benchmarking evaluation of procedures with questionnaires, guidelines and documentation sheets as standardised tools, and there was a central analysis of survey data on the basis of detailed documentation of the diverse evaluation procedures. On the other hand, a high degree of flexibility was allowed concerning the individual design of procedures for conducting the surveys, for selecting target groups and for planning and implementing the measures.
In order to ensure the viability of the benchmarking group and to support the hospitals’ processes of learning from one other, a co-ordinated schedule for central project steps was necessary, and there was a clear need for information about the context of each partner’s work in order to allow for a determination of where each partner ranged in the group of hospitals. The design, the services of the project’s expert consultant as a practical advisor, and scientific support and co-ordination all proved to be essential for the success of the benchmarking. Box 6 presents an account of some of the concrete difficulties faced, taking as an example the pilot hospital in Spain.
The experiences of nine pilot hospitals across Europe show that the central steps in setting up interpreting services in a hospital setting are:
1 obtaining both symbolic and practical support from management
2 establishing central co-ordination of interpreting services, and hiring an interpreting co-ordinator or appointing staff from within the hospital to control the budget, track volume increase and assure quality of service
3 choosing appropriate interpreting resources, for example, employee language bank, telephone interpreters or face-to-face interpreters, either from external interpreting agencies or from in-house professional clinical interpreters
4 training for interpreters and clinical staff
5 marketing of the service to increase awareness and visibility
6 asystematic approach to the production of policies and guidelines and to the translation of written materials.
Further, the outcomes of the project reveal the central prerequisites for the successful and sustainable implementation of measures to improve the quality of clinical communication with migrant or ethnic minority patients. There are several preconditions that have to be met in order to improve clinical communication for migrants consistently:
• linguistically adequate clinical communication has to be integrated into a hospital’s general policy on diversity. As an important step, the needs and assets of all stakeholders – including both users (patients, relatives, community) and providers (staff) – should be monitored. The outcomes of implemented interpreting services, as well as the structures and processes that influence access to interpreting services and their quality, must be monitored
• services and processes have to be sustained by becoming mainstream and not relying on local champions alone. In other words, an organisational development process should be initiated, supported and monitored by those responsible for hospital leadership, overall management and quality management. It is necessary to invest in both the medical interpreters’ and clinical staff ’s developing skills and competences regarding the effective performance and use of interpreting services
• adequate political and managerial will and funding have to be assured. Adequate resources – working time, financial resources, and qualification – must be provided if changes are to be realised. By providing legal and financial support and by formulating organisational aims, health policy must provide a framework in order to make the development of interpreting services on the hospital level relevant and feasible.
Special thanks to the many partners in the pilot hospitals of the Migrant-Friendly Hospitals project that implemented the interventions and were most helpful in evaluation. Representing their teams, the authorswould like to thank the subproject co-ordinators Antonio Chiarenza, Ulises Penayo, Fiona McDaid, Adrian Manning, Jette Ammentorp, Anne-Mette Rasmussen, Minna Pohjola, Marja-Leena Pilkkinen, Ines Garcia-Sanchez, Carmen Fernandez Guerra and Georgia Vasilopolous. Special thanks is also expressed to Lourdes Sanchez, our expert consultant from Boston, USA.\