Health measures are being used increasingly in clinical and research settings to monitor the health status of patients and assess the impact of service provision and interventions. But individual interpretations of health vary depending on a person's age, experience and cultural background. Thus, all these factors need to be considered when developing health instruments so that they are fit for purpose and sensitive to patients' needs. This approach is particularly important in health research where the accuracy and reliability of findings is paramount; and study results are often derived from a wide and diverse patient population.

A patient reported outcome measure (PROM) is a health measure that is administered directly to patients to seek their perspectives about the way in which aspects of their health, illness or treatment impacts on their functional status and health-related quality of life. PROMs are often presented in the form of self-reported questionnaires which may be either generic or disease-specific. Given their standardized approach to quantifying patients' perspectives, PROMs are used increasingly as an adjunct to traditional measures of clinical outcomes and as secondary end-points in clinical trials.

Communicating in a way that responds to the specific language and cultural needs of patients helps explore their personal dimensions of health and provides a more accurate rendering of their health status. In other words, sharing the same language offers a way of establishing common ground for communication and understanding that helps capture the reality of the patient's experience. Thus, in the bilingual context of Wales, it is imperative that health measures are offered in Welsh as well as English, according to individual need or preference.

Despite increasing efforts to enhance bilingual healthcare services in Wales, there are limited numbers of health measures available in Welsh. This means that the health status of some Welsh speakers may be mis-interpreted; and this can jeopardise their care management and challenge the rigour of research conducted in a bilingual context.

MI-CYM offers a database of all the recorded health measures available in Welsh, their access details and licensing arrangements. Click here to browse the site. If you have information about a Welsh language measure that is not listed on MI-CYM, or information about translations in progress, please let us know by clicking here.

Despite steady progress over recent years, we are aware that there are still many health measures in need of Welsh translation and adaptation. The Welsh Health Measures Scoping Review and Survey undertaken by LLAIS / NWORTH in 2013 gives some indication of the priority requirements, but this list is by no means exhaustive so there is still much work to be done. If you have a health measure in need of translation / adaptation OR require further information and guidance in proceeding with the work; please click here.

Where a health measure is unavailable in Welsh, it is understandable that Welsh speaking practitioners or researchers may be tempted to undertake the translation themselves, particularly in response to the increasing language needs of their client groups. However, developing a standardized translation that is meaningful, accurate and reliable is a challenging procedure which is beyond the scope of an individual practitioner, whatever their language proficiency. This is because of the need to establish strict equivalence between language versions of the instrument in order to avoid bias. This ensures that when both measures are administered amongst a cohort, any differences between scores derived from the Welsh and original measure reflect true differences amongst respondents rather than any language discrepancies that may have arisen through poor translation. There are 4 fundamental aspects of equivalence that need consideration, that is:

• Semantic equivalence, where the words and phrases used in the original measure carry the same meaning in Welsh
• Conceptual equivalence, where the concepts used in the original measure exist and carry the same meaning in Welsh
• Item equivalence, where the items in the original measure are relevant and acceptable in Welsh
• Operational equivalence, where the format of the scale and mode of administration is relevant and acceptable in Welsh.

Given these challenges, standardized procedures are required for the translation process as well as quality assessments of the newly-established Welsh language version of the measure.

In order to assure the technical, linguistic and conceptual equivalence of health measures, a number of international evidence-based protocols have been developed to guide the translation process. Many of these protocols share common elements which form a series of systematic steps for translating health measures effectively (see Translation Guidelines). These common elements include forward and back translation, consensus review and cognitive testing.

Of course, it is up to you whether you complete all the steps of the standardized approach although this often depends on the nature of the permissions agreement established. However, linguistic validation of the translated measure cannot be established until all the steps have been accomplished; and this is often required for the version to be considered the official translation of the measure for that language.

The linguistic validation of a health measure is a process whereby the translation of the measure follows a strict methodology, with quality control steps to ensure that both conceptual and semantic equivalence are achieved. This ensures that health measures are accurately translated and culturally acceptable for the target population for which they are designed.

The psychometric validation of a health measure is a process whereby the psychometric properties of a translated measure must be re-established, that is, its reliability and validity must be assessed to determine whether the scale is comparable to the original instrument. This involves the application of a series of statistical tests.

Once a measure has been translated /adapted, there is an expectation that its psychometric properties should be re-established to determine the comparability of the scale to the original instrument. As in the case of other minority language contexts, in the absence of a large population of Welsh speakers in Wales, this poses a particular challenge, especially in terms of recruiting adequate samples to establish the data. Whilst establishing the confirmatory psychometric validation of these measures is extremely valuable, there are significant resource implications. Moreover, it is likely that the rigour of the linguistic validation enables us to infer the psychometric properties of the original instrument established amongst a larger population. Thus, whilst a gold standard approach for establishing the psychometric validation of Welsh language versions of health measure may be currently beyond our reach in Wales, linguistic validation offers a step forward from ad-hoc translation and has, at its heart, the best interests of patients in enabling the availability of tools in their language of need.

You can get involved through 'Involving People'.