Communicating with people who have limited English

Increasingly, language need in any locality is diverse (many different languages are spoken) and dynamic (the languages change every few months) – often to the extent that provision fails to keep up with the demographics of the population (Baker and Eversley, 2000; Salt, 2005; Eversley, 2005). Official estimates of the burden of need for interpreting services in many areas is high and rising, especially in deprived inner-city areas (Baker and Eversley, 2000; Atkinson et al, 2001; Gerrish, 2001; DoH, 2002; McPake et al, 2002), and that access to such services is generally inadequate, especially for the most vulnerable and hard-to-reach groups (Jones and Gill, 1998; Burnett and Fassil, 2000; Atkinson et al, 2001; Gerrish, 2001; Hertog, 2001; DoH, 2002; McPake et al, 2002; Murphy, 2004).

The NHS’s commitment – and the evidence base for it
The NHS has expressed its commitment to providing an equitable service defined by need, including providing a professional interpreter to any individual if needed (DoH, 2004). However, ‘translation’ (with a focus on the words spoken) has tended to take precedence in policy circles over ‘advocacy’ (which acknowledges that a limited English speaker may also have support and access needs, and that the health professional may lack key areas of cultural knowledge or competence). Expenditure on NHS interpreting services is rising rapidly, and questions have been asked about whether or not this is a good use of resources (Adams, 2007; Jones, 2007). What is the evidence for and against the provision of such services?

Research has consistently shown that when a professional interpreter is provided for a consultation, quality of care and satisfaction with care are greater than in comparable situations when no interpreter, or an ad hoc interpreter, is provided (Karliner et al, 2007; Flores, 2005a; Green et al, 2005a). The use of ad hoc interpreters (for example, a receptionist or fellow patient) or family member interpreters is common in many healthcare situations with limited English speakers. Accuracy of translation may be low, even with professionally qualified interpreters, but even more so with lay interpreters (Cambridge, 1999; Elderkin-Thompson et al, 2001; DoH, 2002; McPake et al, 2002).

However, some research has challenged the assumption that a professional interpreter is always the best option. While family member interpreting is clearly inappropriate in certain situations (a child interpreting for a psychosexual consultation for example, or the husband for a young wife’s unexplained bruises), it may be highly valued and seen as effective in less complex or emotionally-charged situations – partly because family members may be easier for the individual to get hold of, and partly because they may shift the power balance in the patient’s favour (Angelelli, 2005; Green et al, 2005a; Green et al, 2005b; Greenhalgh et al, 2006; Robb and Greenhalgh, 2006). One qualitative study concluded that young people who interpret for family and friends ‘wished their work to be conceptualised not as merely “inappropriate and inadequate interpreting”, but as a varied contribution to the informal economy of health care that ranged from simple translation to complex mediation between families, the wider community and the healthcare system’ (Green et al, 2005b).

New technologies offer opportunities beyond conventional face-to-face interpreting, such as phone or video interpreting. Evaluation of these different models has produced mixed findings, with no model clearly best for all people in all circumstances (DoH, 2003; Jones et al, 2003; Silvera and Kapasi, 2004; Garcia et al, 2004).

The risks of poor communication with limited-English speakers
Failure to provide an effective interpreting service is increasingly seen in terms of the risk and safety agenda. For example, this may account for failure of vulnerable groups to access the services they need (why seek help when you will not understand or be understood?) as well as for a rising proportion of medical errors, such as: administering medicines to which the patient is allergic; giving inappropriate vaccinations to children; failure to offer cancer screening to eligible and at-risk individuals; dismissing early symptoms of cancer as ‘non-specific complaints’; people not being aware of what their illness is or how to take their medicines; and procedures and operations being undertaken without informed consent (Chief Medical Officer, 2001; Angelelli, 2005; Flores et al, 2005b; Green et al, 2005a; Abbe et al, 2006; De Alba and Sweningson, 2006; Greenhalgh et al, 2006; Bradshaw et al, 2007; Hunt and de Voogd, 2007; Phokeo and Hyman, 2007).

Economic analyses suggest that although interpreting services is costly, they may nevertheless prove to be cost effective owing to the high impact on quality of care and reduction in error (Jacobs et al, 2004). Medic–olegal advice in the USA is often ‘pay now [for effective interpreting services] or pay later [in a lawsuit]’ (Ku and Flores, 2005). There were no available studies specifically relating to risk in diabetes, though it is not difficult to imagine how poor communication in this setting might lead to loss of limb, loss of sight, or major foetal abnormality.

More subtle than the direct risk to life and limb, but perhaps as important in the longer term, are issues of trust and power. Language and cultural barriers, even when an interpreter is present (but especially if one is not), distort communication and may undermine clinician–patient trust (Angelelli, 2005; Green et al, 2005a; Greenhalgh et al, 2006; Robb and Greenhalgh, 2006; Meeuwesen et al, 2006; Schouten and Meeuwesen, 2006). Language barriers also shift the power balance in the clinical consultation away from the patient. Typically, clinicians control the agenda in the clinical consultation since they control the allocation of time and the issuing of a prescription or referral. A vulnerable and inarticulate patient lacks the power to negotiate – a critical role of the interpreter or advocate, in addition to translating, is to get the patient’s agenda on the table (Greenhalgh et al, 2006; Robb and Greenhalgh, 2006).

Communication is not just about language
It is important not to see communication with limited-English speakers purely in terms of linguistic barriers. Health-seeking behaviour (or the absence of it) is shaped by a person’s cultural background, belief systems and identity, and by their past experiences in their country of origin. A particularly important issue here may be past experience of poorly developed and low-quality primary care services, which shape the expectation that all illness should be referred to a ‘proper’ (secondary care) doctor (Greenhalgh et al, 2006; Robb and Greenhalgh, 2006). Lack of ‘system knowledge’ often underlies poor access to services, especially among recent immigrants, refugees and asylum seekers, since such individuals often don’t know where to start looking for help when they are ill, or how to get into the system (Greenhalgh et al, 2006; O’Donnell et al, 2007).

Limited-English speakers are more likely to be poor, inadequately housed, lacking access to basic health services, and to have multiple co-morbidity (LaVeist, 2005; Sue and Dhindsa, 2006). A history of severe mental trauma is common in refugees and asylum seekers. In one recent study, for example, an interpreter from a war-torn African country estimated that over 90% of her female patients had been raped (Robb and Greenhalgh, 2006).

Limited-English speakers are also more likely to lack functional health literacy, defined by the WHO as ‘the cognitive and social skills that determine the motivation and ability of individuals to gain access to, understand, and use, information in ways that promote and maintain good health’. The concept of health literacy embraces a number of key attributes including reading and numeracy in relation to health information, the capacity to use health information, and the ability to ‘perform in the patient role’. These, in turn, rest on wider literacy and numeracy skills as well as health-related experience – the latter of which achieves two things: exposure to the medical vernacular, and building the relevant cognitive schemas within which new health information makes sense (Speros, 2005; Coulter and Ellins, 2006).

Health literacy correlates closely with healthy lifestyle choices, concordance with medication, overall cost of health care, length of stay in hospital, and outcome in a wide range of conditions (Andrus and Roth, 2002; Wilson, 2003; Speros, 2005; Coulter and Ellins, 2006; Sentell and Halpin, 2006). Many (and in some studies, all) ethnic differences in health outcome are explained by differences in health literacy (Schillinger et al, 2004; Borrell et al, 2006; Sarkar et al, 2006; Sentell and Halpin, 2006).

Organising services and developing staff
Providing an interpreter for every consultation with a limited-English speaker is a highly complex exercise in the organisation and delivery of services, and some health care organisations are more efficient and effective in this than others (Brach et al, 2005; Greenhalgh et al, 2007). Despite many studies showing that interpreters are often under-trained for their role, there is remarkably little research on the training needs of interpreters or bilingual health advocates (Angelelli, 2005). One study has shown that interpreters can be trained in disease knowledge (for example, through half-day release courses and studying videotaped consultations), and such training improves the accuracy of translation (McCabe et al, 2006).

The job of many interpreters involves an inherent role conflict. The role of ‘impartial translator’ requires different competencies (and, importantly, a different identity) than the role of ‘bilingual health advocate’. The former is a contemporary role, linked to the formal division of labour in a professional bureaucracy and based on what has been called the ‘conduit model’ (Dysart-Gale, 2005) or ‘bilingual parrot’ (Greenhalgh et al, 2006). The latter is a more traditional role, linked to ‘kinship’ ties of shared cultural background, ethnicity and sometimes gender (a role that embraces both ‘witness’, ‘ally’ and ‘cultural broker’). In a study undertaken by the author’s own team, when describing positive interpreted consultations, patients often used expressions such as ‘like my daughter’ or ‘like a sister’, suggesting that the intersubjective bond based on ‘kinship’ ties is important (Angelelli, 2005; Dysart-Gale, 2005; Greenhalgh et al, 2006). Professional interpreters often experience considerable role conflict and feel pulled between the professional territory of the doctor and the ‘lifeworld’ of the patient (Angelelli, 2005; Dysart-Gale, 2005; Greenhalgh et al, 2006). In some situations (for example, end-of-life care), the appropriate balance between ‘impartial translator’ and ‘cultural broker’ shifts towards the latter (Norris et al, 2005).

The use of interpreters depends on the willingness of health professionals to work with them. The author’s team has shown that clinicians vary significantly in their willingness to use professional interpreters and in their competence in communicating through (and with) an interpreter; approximately half of all GP surgeries in London, for example, do not use the interpreting service at all (Gerrish et al, 2004; Karliner et al, 2004; Greenhalgh et al, 2007).

Interpreters and bilingual health advocates have a potentially important role to play in educating health professionals and increasing their cultural awareness, helping patients ‘navigate’ the healthcare system, and providing education and support for people with long-term illness through new models of care (Greenhalgh and Collard, 2003; Greenhalgh et al, 2005a; Greenhalgh et al, 2005b; Wu et al, 2006). A randomised trial of bilingual health advocate-led group education for diabetes through oral story telling in Newham, London, has recently been completed; this study will report shortly.

Implications for practice
There is no simple formula for effective communication with limited-English speakers, but the following recommendations are supported by evidence.

• All GP practices and PCTs should document language need, map this to burden of illness and produce aggregated, regularly updated data on this.
• Cost-constrained interpreting and advocacy services (which will probably always be unable to fully meet demand) should be targeted towards those whose need is greatest, including those with multiple comorbidity or complex social needs.
• The wide variability between GP practices in their willingness to use the available interpreting service is unacceptable. It is surely unacceptable that some GP surgeries are still allowed to say ‘we don’t have interpreters here’.
• The professional development and training of interpreters and bilingual health advocates requires attention at national policy level, but there is also much that primary care teams and PCTs can do on a relatively small scale to develop these important staff.
• New service models in which interpreters, advocates, health navigators and other staff are used in new roles, and especially in ways that meet local needs and priorities, should be explored. It is surely a priority to consider how interpreters might be used as educators to improve the cultural competence of medical, nursing and pharmacy staff, and how the NHS might link with the voluntary, NGO (non-governmental organization) and private sectors to provide effective interpreting and advocacy services in today’s mixed economy of healthcare.

Conclusion
In conclusion, contemporary diabetes care in many parts of the UK requires attention to language and communication barriers. These barriers include – but are not limited to – the task of translating what is said. The astute primary care practitioner will also question what cultural expectations and fears the patient brings, and the nature of the patient’s lifeworld. Providing interpreting services is a huge management challenge, and the professional development of the interpreting and advocacy workforce is an important aspect of this.