Examining and establishing translational and conceptual equivalence of survey questionnaires for a multi-ethnic, multi-language study

AIM: This paper is a report of techniques used to examine and establish translational and conceptual equivalence of survey questionnaires. BACKGROUND: A major concern arose about standardization of translated survey questionnaires, when preparing to evaluate differences in acute coronary syndrome presentation in European (White), Chinese and South Asian patients. METHODS: The survey questionnaires were first translated by an accredited translation company. Between July and November 2009, materials were taken to like-speaking healthcare reviewers to ensure that the clinical meaning was appropriate. Like-speaking lay reviewers were then asked to make comment about grammar; meaning and understanding of questions; and any concerns about the suitability of graphics. A key informant from each language group reviewed all comments and worked with the investigators and the translation company to create final sets of survey questionnaires. RESULTS: Readability of the questionnaires (too complex or too basic) was the most common concern. A major discrepancy between ethnic groups arose about a graphic of 'squeezing' pain. A hand grasping a balloon was considered appropriate for European and South Asian groups, while a picture of a towel being wrung out was identified as more appropriate for the Chinese. There were no negative comments about the graphics. Soliciting key informants who were highly fluent in both English and the language under study was critical to ensure that the participants' feedback was appropriately reconciled. CONCLUSION: Traditional forward-backward translation of study materials is insufficient. Translation must be accompanied by a process whereby equivalence and acceptability are also established.