PRM135 - US AND LATIN AMERICA: FINDING THE BEST BASE FOR SPANISH ADAPTATIONS

Abstract

Latin America and the US are attractive locations for clinical trials, offering an enormous pool of subjects (600 million people in LATAM including Brazil and 41 million native US Spanish speakers plus 11.6 million bilingual speakers). Of 206,543 trials worldwide in 2015, the US, Mexico, and Central/South America represented 101,922. Spanish is a frequently requested language for linguistic validation. Where studies involve multiple Spanish-speaking sites, the optimal linguistic validation methodology must be decided, ideally on a case-by-case basis, depending on instrument content and the study’s practical constraints. Three methodologies are commonly used for same language, different country scenarios, described by Wild (2015): produce separate translations simultaneously for each target country; produce a translation in one country and review for each additional target country; or work with translators from each country to develop a global translation. Our objective is to examine which LATAM country is most effective as lead in Spanish harmonization or as a base for adaptation. Five linguistic validation projects were reviewed, featuring treatment satisfaction questionnaires. The sample covers all three methodologies described above and seven countries. By directly comparing translations, we established which country emerged as the most neutral base. Terminology examined covered treatment satisfaction question stems, response options and symptoms. The most neutral language base was Mexican Spanish, with 91% of its translation solutions agreeable to the other countries, closely followed by Argentina Spanish (87.5%) with Costa Rica at the bottom of the scale (41%). No matter which methodology is selected, a key facilitator in the process can be choosing the most neutral lead or base Spanish from which to create a version for other countries. We conclude that Mexico and Argentina are the optimal base countries. A study with a larger sample would be a worthwhile endeavour.