OBJECTIVES:

The linguistic validation of Clinical Outcomes Assessments (COAs) for use in global clinical studies often involves a single language for use in multiple countries (examples include Arabic, Spanish and French). In an effort to reduce costs and improve efficiency, sponsors often request a universal version of a target language that can be used in each country. This poster will discuss the risks of universal translation and provide a decision tree to help sponsors determine whether country-specific versions may be necessary.

METHODS:

An examination of past projects involving different target locales for the same language was conducted. Further, interviews with linguists were completed to analyze the differences between different locales. The information gathered was categorized based on COA type as well as item content (colloquialisms, subject matter, etc.).

RESULTS:

By analyzing past projects as well as linguists’ responses, the necessity of multiple locales versus a single, universal translation and the risks varied between different type of COAs (Clinical Outcomes Assessment). See Table 1.

CONCLUSIONS:

The primary risk of a universal translation is inadequate cultural adaptation. When deciding if a universal translation should be used, an analysis of the below factors is needed:

• COA type: Country-specific versions are more likely to be needed for PROs and ObsROs whereas ClinRos can often be adapted using a universal approach
• Cognitive assessments: testing of verbal fluency, category fluency, etc. often require adaptation for country-specific versions
• Item phrasing: the more colloquially the items are phrased, the more likely it is that country-specific versions will be necessary
• Item phrasing: the extent to which the items are related to daily life (for example, daily activities, food types consumed, medications taken, etc.)
• Developer requirements: pre-existing guidelines from developers provide details as to necessary approach