There has been a rise in popularity of self-assessment risk scores for chronic conditions such as type 2 diabetes, which allow individuals to self-assess their own risk of developing a condition and take steps to reduce this risk if necessary. One such tool is the Leicester Self-Assessment Risk Score (Gray et al, 2010), which has been developed to predict the risk of type 2 diabetes in multi-ethnic populations. The success of this measure as a tool for prevention relies on the participant having a conceptual understanding of being at “high risk”, including the steps that can be taken to reduce this risk. Comparatively, little emphasis has been placed on conceptual understanding of the results when translating such documents (Weidmer, 1994). This is of particular importance in the calculation of disease risk, as subsequent behaviour change could be undermined if participants do not fully understand the concept of being at high risk.
There is a wealth of literature describing the translation of a variety of health scales (Brislin, 1970; Beaton et al, 2000). For the most part, translators have focussed on maintaining content validity and reliability between languages. Presumably, this is because these tools are most commonly used within a clinical or research setting to assess treatment effects, or to screen individuals prior to intervention. The ability of the person completing the score to interpret and understand the results is of lesser importance when used in this context.
Translating the Leicester Risk Score
Recently, Patel et al (2016) have developed and utilised a translation model to translate the Leicester tool from English into Gujarati, in order to meet the needs of communities from a multi-ethnic population in the East Midlands. The eight-stage model consisted of revision and refinement of the original version to make the language conducive to translation; development of a conceptual guidance document for translators to use; forward translation (English to Gujarati); backwards translation (Gujarati to English); reconciliation and agreement of translations; clinician review; pilot testing; and final revision and checks.
The study highlights that a focus on linguistic equivalence in the translation may be counterproductive, as words which have a similar definition in different languages can be conceptualised very differently. From the experience of translating the Leicester score, for example, the direct Gujarati translation of the word “risk” – “jokhem” – was interpreted as meaning “danger” by focus group participants. To overcome this, the authors followed an approach using a phonetic spelling of the Gujarati word, adding a detailed explanation of the meaning of the word “risk” to describe the context in which it was being used. This resulted in good understanding of the information contained in the diabetes risk score, and it allowed the majority of focus group participants to fill out the score unaided and to interpret the results.
The authors acknowledge that some people still have difficulties in completing the risk score, and there are still issues with communicating the concept of being at “high risk”. Fatalistic beliefs, low levels of health literacy and misunderstanding of risk factors for type 2 diabetes are common amongst members of the South Asian community in particular (Carr-Hill et al, 1996; Lucas et al, 2013). Low levels of literacy limit individuals’ ability to self-calculate BMI using their body weight and height, and to understand its significance in contributing to diabetes risk. Alternative strategies, including the promotion of web-based versions that can automatically calculate BMI and support from family members/healthcare professionals, may need to be considered to meet the needs of some communities.
Overall, the methods developed by Patel and colleagues promoted conceptual understanding of the risk score, and the translational model is of relevance to the development of existing self-assessed measures of disease risk. It is of vital importance to continue working on adapting methods to engage non-English-literate populations, who are often at high risk of type 2 diabetes, in screening and prevention. In not doing so, there is the potential to promote health inequality, as these tools are being widely implemented amongst the English-speaking population (NICE, 2012), thus improving rates of early detection and preventing complications in one group while failing to do so in others.
The study by Patel and colleagues can be read in full at: http://dx.doi.org/10.1155/2016/8107108