Several approaches have been identified within the NHS for overcoming language barriers where verbal communication is required as part of the delivery of services in primary and secondary care. These will vary in their effectiveness or ‘adequacy’, costs, and generalisability to different locations. While at present there are no agreed national standards in services other than the criminal justice system, there are discussions underway to establish agreed levels of good practice within the NHS and interpreting profession, led by the Public Sector Advisory Group of the National Register of Public Sector Interpreters at the Institute of Linguists. In the interim, we offer a classification of current 'de facto' solutions to the problem of linguistic barriers between patient and clinician. They include:

i) Patient selection of GPs:

In some geographical areas it may be possible for patients to self select to enrol with a GP who is able to speak their language proficiently (Ahmad et al 1989). Evidence from the Bradford area suggests that patients who were less fluent “countered their linguistic disadvantage by consulting the Asian doctor who was fluent in their own language(s). Only 15% of the male patients who consulted the Caucasian doctor were poor or non-speakers of English compared with 30% of the Asian doctor’s patients.” (Ahmad et al 1989). However whilst such a strategy can be adopted by patients in some geographical locations, in others areas because of a paucity of GPs with suitable language skills (and of a suitable gender e.g. female) other strategies will need to be adopted. Furthermore one limitation of this approach is that if patients are forced to select GPs who can speak their language this could reduce a patient’s choice of physician, unless of course interpreters are available to patients should they choose to enrol with a GP who does not speak the same language. Secondly, it is now recognised that many of the Asian GPs in inner city areas are approaching retirement age and this route may not be available for much longer.

ii) Use of bilingual health care workers:

This is closely related to (i) above, but it applies more generally to other health care workers apart from GPs. The approach has been described as the “ideal option for most patients” because it removes the need for a third party to be involved in a consultation (Phelan and Parkman 1995). However, bilingual health care workers are few (Phelan and Parkman 1995) and the quality of language provision may be poor if bilingual workers have an inadequate grasp of a particular language, or if they lack the training to be sufficiently culturally sensitive. Nevertheless such provision could prove cost-effective if bilingual workers speak fluently and are well trained, particularly if they speak a language which is frequently required. It should be noted that the Race Relations Act (1976) does specifically allow employers to appoint using racial/ linguistic criteria on the basis of, for example, a genuine occupational qualification i.e. linguistic skills.

iii) Use of full-time professional interpreters or experienced sessional interpreters:

Some hospitals and health authorities use the services of either full-time trained interpreters (so-called 'linguistic model' interpreters: Turton 2003) or experienced interpreters employed on a sessional basis. These are typically either employees of NHS Trusts or, increasingly, projects and teams working for, or jointly sponsored by, Social Services Departments or outside voluntary agencies. As a general rule so long as such interpreters are adequately trained and experienced the quality of interpreting is high, with the result that many of the problems associated with inadequate communication can be avoided. Once the benefits to patients of such provision (including patient satisfaction and any improvements in health outcomes) have been taken into account, such provision is likely to prove more cost-effective than other forms of provision, except perhaps use of well trained, bilingual staff. Furthermore provision of such interpreting services can have an informing and empowering role. This may improve the health and well-being of patients of marginalised communities (Ntshona 1997).

iv) Use of tele- interpreter services:

American research has suggested that electronic simultaneous interpretation can be very effective and popular (Hornberger et al 1996). Telephone interpreter services such as ‘Language Line’ have been introduced in the NHS over recent years (Leman 1997). They can offer the advantage of 24 hour access to interpreting services. The obvious limitation of such provision is that it lacks a visual dimension. This may present difficulties for the interpreter for example if a patient wants to draw a physician’s attention to a particular part of the body. The service is also likely to be expensive if used as the main form of interpreter provision in trusts with significant numbers of patients requiring provision. Thus the use of face-to face interpreting/ advocacy is likely to be more cost-effective in such a situation.

Use of the service might be justified to provide out of hours provision, or to provide interpreter services for patients speaking very rare languages. Telephone interpreting provision might also be justified if a trust’s demand for interpreter services is very low. In these circumstances, use of services such as Language Line may prove more cost-effective than face-to-face interpreting if a trust cannot access other locally available service provision. The NHS Plan states that by 2004, tele-interpretation support will be provided nationally through NHS Direct, and a contract has recently been signed to supply a service, but at present, there is no information about quality standards or evaluation of the service, although the Audit Commission (2001) reviewed one local provider in London Borough of Hounslow.

v) Use of patient advocates:

The distinction between interpreter and advocacy (or cultural broker) roles is not clear cut, but in the UK an advocacy role clearly involves more than just interpreting (Chiu 1991). ‘Health advocates’ have been defined as those who “mediate between patients and professionals to make sure that clients are offered an informed choice of health care. If there are clinical or cultural problems they will negotiate, although ultimately they see themselves as advocates for their people.” (Parsons and Day 1992). Such a role is likely to be appropriate in contexts in which the patient is unable or unlikely to do this themselves. Moreover, advocacy can be used to “improve the access of all groups to the health service” (Chiu 1991) and to inform patient choices. Some evidence suggests that patients who are more in control (i.e. they ask more questions, make more attempts to direct conversation flow and physician behaviour) report fewer days off from work, health problems, and functional limitations because of illness, whilst rating their health more favourably during follow-up (Kaplan et al 1989). This would tend to suggest that, in so far as advocates are able to put patients in control, they may be able to improve patient well-being and health status.

Advocates may or may not act as interpreters as well. Clearly, if they perform a dual role the effectiveness of using advocates as interpreters will depend upon the extent to which they are sufficiently trained and experienced as interpreters.

There is some evidence that would seem to suggests that advocates are able to affect the nature/ quality of service obtained by patients. A retrospective study was conducted in Hackney, East London which compared 1,000 non-English speaking women delivering at the Mothers’ Hospital in 1986 when accompanied by an advocate with 1,000 women delivering at the same hospital in 1979, and similar deliveries at a reference hospital (Parsons and Day 1992). The study identified significant differences between the 3 groups in terms of outcomes such as antenatal length of stay, and induction and mode of delivery. Rates of caesarean section rose from 11% to 17% at the reference hospital, but fell from 10.8% to 8.5% at the Mothers’ Hospital. Whilst these findings may not necessarily be causally linked to provision of advocacy services “it was considered reasonable to deduce that improved communication could have influenced clinical practice”. It was therefore suggested that health advocacy might provide a mechanism to redress adverse obstetric outcomes amongst some ethnic minorities. Health advocacy services may be able to generate similar benefits for mental health patients.

Health advocacy, as opposed to straightforward interpretation, may also provide a useful way of bridging gaps in cultural understanding. During a survey (Meadows 1992) it was noted that doctors identified a gulf between medical knowledge about a specific condition and the cultural and neighbourhood beliefs of patients relating to the cause of illness. Health advocates might be in a better position than interpreters to actually address this lack of patient understanding. This is because their role may extend to persuading patients of the value of a particular treatment approach.

Advocates may be especially appropriate in geographical areas with substantial refugee populations. They may also be particularly necessary in order to support those who have been victims of torture prior to obtaining refugee status in the United Kingdom. As a result of their experiences such individuals may well be apprehensive and not able to seek out the provision they require. In such circumstances, advocates are likely to fulfil an invaluable role.

vi) Use of volunteer and ad hoc interpreters:

Some UK hospitals use volunteer interpreters. This usually involves introducing systems to enable existing bilingual staff to provide interpreter services when required on a voluntary basis. In some hospitals voluntary provision from the local community is utilised. Whilst this approach may have the merit of flexibility it may cause difficulties if it takes staff away from another job. Registers of staff availability and language competence have also proved to be notoriously inaccurate and imprecise, and require considerable administrative costs in upkeep.

Furthermore, whilst training can be provided for volunteers, it has been reported that usually volunteers “have not received any training in interpreting” (Phelan and Parkman 1995). Moreover, volunteers who provide interpreting input on an irregular basis are less likely to acquire the same expertise as trained interpreters with considerable work experience, and thus the standard of interpreting may not be as high. Volunteer and ad hoc provision is perhaps best deployed to deal with emergency situations where a professional, trained interpreter is not available when required.

vii) Use of English speaking family members and friends as an interpreter("Bring Your Own")

Evidence suggests that this strategy should be pursued with caution, although a recent review suggests that this is the strategy used by 70% of service providers (Turton et al 2004). The use of relatives and friends, has the disadvantage that the quality of interpretation may be poor, and patient confidentiality may be violated if the patient does not want a family member or friend to have access to medical information about them. Also “family, friends, and ad hoc interpreters pose problems. They may lack sufficiently good language skills and frequently commit stereotypical errors, including omissions, additions, substitutions, or other editing, which may result in serious semantic distortions” (Woloshin et al 1995). Particular problems may arise if children are used to interpret since “both the parent and the child may be embarrassed by the problem and the information that the doctor or patient receives may thus be censored.” (Ahmad et al 1989). The use of a family’s children may also “upset the family’s social order.” (Poss and Rangel 1995). Distorted communication and confidentiality are especially problematic in cases involving issues such as sexuality and child abuse (McPake & Johnstone 2002)..

References

Ahmad, W. Kernohan, E. & Baker M. (1989), “Patients’ choice of general practitioner: influence of patients’ fluency in English and the ethnicity and sex of the doctor”, Journal of the Royal College of General Practitioners, vol. 39, pp. 153-155.

Audit Commission (2001) Community language, translation and interpretation unit London Borough of Hounslow Best Value Audit, London: Audit Commission

Chui, L. (1991), "Advocacy; the Leeds experience", Community Dental Health., vol. 8, pp. 253-256.

Hornberger JC, Gibson CD, Wood W et al. (Aug 1996) Eliminating language barriers for non-English speaking patients. Med-Care.; 34(8): 845-856.

Kaplan SH, Greenfield S, Ware JE. (1989), Assessing the effects of physician-patient interactions on the outcomes of chronic disease. Med-Care; 27(3): S110.

Leman, P. (1997), "Interpreter use in an inner city accident and emergency department", J.Accid.Emerg.Med., vol. 14:98-100, pp. 98-100.

McPake J, Johnstone R (2002) Translation, interpreting and communication support services across the public sector in Scotland Edinburgh: Scottish Executive Central Research Unit

Meadow, R. (Jun 1992); Difficult and unlikeable parents. Arch-Dis-Child. 67(6): 697-702.

Ntshona, MS. (Sep 1997), Are professional interpreters needed in the South African health care services. S-Afr-Med-J.; 87(9): 1143.

Parsons, L. & Day, S. (1992), "Improving obstetric outcomes in ethnic minorities: an evaluation of health advocacy in Hackney", Journal of Public Health Medicine., vol. 14, no. 2, pp. 183-191.

Phelan, M. & Parkman, S. (1995), "Work with an interpreter", BMJ., vol. 311, no. 26 Aug, pp. 555-557.

Poss, J. & Rangel, R. 1995, "Working Effectively with Interpreters in the Primary Care Setting", Nurse Practitioner., vol. 20, no. 12, pp. 43-47.

Turton J, De Maio F, Lane P (2003) Interpretation and translation services in the public sector (unpublished) report to the Home Office by University of East London HARP social inclusion research programme

Woloshin S, Bickell N, Schwartz L, et al. (1995), Language barriers in medicine in the United States. JAMA.; 273: 724.