TRANSLATION AND LINGUISTIC VALIDATION – METHODOLOGICAL IMPLICATIONS WHEN THE SOURCE MEASURE IS NOT ENGLISH
Author(s)
Clayson D1, Verjee-Lorenz A1, Two R1, Gerber R2, Beaudreuil J31PharmaQuest Ltd, Banbury, Oxfordshire, United Kingdom, 2Pfizer, Inc., New London, CT, USA, 3Université Paris, Paris, France
Presentation Documents
The majority of new PRO measures are developed in English, with a small minority developed in other languages. This poses the question of how to translate and linguistically validate measures developed in languages other than English, when the resources are simply not available for translating from the source language (e.g. Hungarian) into the target language (e.g. Bengali). The URAM Scale assesses functional performance of the hand in patients with Dupuytren’s Contracture. It was developed in French and initially translated and linguistically validated into UK English. The methodology employed for this translation was the reverse of a typical English to French translation, whereby the French in-country investigator acted as the project manager and an in-house project manager (qualified in French to English translation) acted as the in-country investigator. The translation into English required consultation with the developer and two issues needed to be resolved in order to develop the UK English version. One item required alternative wording as it mentioned wash mitts, which are very rarely used in the UK, and for another item, two verbs were required to convey the meaning of a single French verb. On completion of the translation, the UK English version was then used as a source version for the translation of the URAM Scale into several other European languages. During this process the translators were asked to work from the English version but to also consider the relevance of the original French wording for the two items that required a change in English. For example, in countries where wash mitts are used, this wording was retained instead of the new English wording which used ‘flannel’. This experience highlights the importance of always considering the original development language when translating a measure using a generated English version as the source.
Conference/Value in Health Info
2011-05, ISPOR 2011, Baltimore, MD, USA
Value in Health, Vol. 14, No. 3 (May 2011)
Code
PRM49
Topic
Methodological & Statistical Research
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
Musculoskeletal Disorders