Study design: Systematic review of cross-cultural adaptation of the Oswestry Disability Index (ODI).
Objective: The aim of this study was to evaluate the translation procedures for and measurement properties of cross-cultural adaptations of the ODI.
Summary of background data: The ODI is the most commonly used questionnaire to determine the outcome of low back pain, and has been translated into many other languages, such as Danish, Greek, and Korean, and adapted for use in different countries.
Methods: PubMed, the Cochrane Library, Medline, and EMBASE were searched from the time they were established to January 2015. Studies related to cross-cultural adaptation of the ODI in a specific language/culture were included. Guidelines for the Process of Cross-Cultural Adaptation of Self-Report Measures and Quality Criteria for Psychometric Properties of Health Status Questionnaire were used for assessment.
Results: This study included 27 versions of ODI adaptations in 24 different languages/cultures. Only the Danish-Danish adaptation employed all six of the cross-cultural adaptation processes. Expert committee review (three of 27), back translation (eight of 27), and pretesting (nine of 27) were conducted in very few studies. The Polish-Polish (two) adaptation reported all (nine of nine) the measurement properties, whereas the Traditional Chinese-Taiwan and Hungarian-Hungarian adaptations reported six of them. Content validity (16/27), construct validity (17/27), and reliability (22/27) were determined in a relatively high number of studies, whereas agreement (three of 27), responsiveness (12/27), floor and ceiling effects (six of 27), and interpretability (one of 27) were only determined in some studies.
Conclusion: We recommend the Traditional Chinese-Taiwan, Simplified Chinese-Mandarin Chinese, Danish-Danish, German-Swiss, Hungarian-Hungarian, Italian-Italian, and Polish-Polish (two) versions for application, but Traditional Chinese-Hong Kong, French-Swiss, Japanese-Japanese (two), Polish-Polish (two), Tamil-Indian, and Thai-Thai versions may need more research. Furthermore, supplementary tests for the adaptations are necessary, especially for assessing agreement, responsiveness, and interpretability.
Level of evidence: 1.