Translation, cultural adaptation, and initial reliability and multitrait testing of the Kidney Disease Quality of Life instrument for use in Japan

Background: The Kidney Disease Quality of Life instrument (KDQOL(TM)) consists of 79 items: 36 asking about health-related quality of life (HRQOL) in general (the Medical Outcomes Study SF-36) and 43 asking about QOL as it is affected by kidney disease and by dialysis. Aim: Translation, cultural adaptation and initial reliability and multitrait testing of the KDQOL(TM) for use in Japan. Methods: Translation and cultural adaptation began with two translations into Japanese, two backtranslations into English, and discussions among the translators, the project coordinators in Japan, and the developers of the original (US-English) version. Focus-group discussions and field testing were followed by analyses of test-retest reliability, internal consistency, and convergent and discriminant construct validity. Results: All eight of the SF-36 scales met the criterion for internal consistency (Cronbach's alpha ranged from 0.73 to 0.92) and were reproducible (intraclass correlations between test and retest scores ranged from 0.60 to 0.82). Of the 10 kidney-disease-targeted scales, only two had alpha coefficients of less than 0.70: 'sleep' (0.61) and 'quality of social interaction' (0.35). One item on the 'quality of social interaction' scale had a very weak correlation with the remainder of that scale (r = 0.10). Eliminating that item from scoring increased the alpha coefficient of the scale from 0.35 to 0.64. All three items on the 'quality of social interaction' scale had very strong correlations with other scales. Conclusions: First, in Japanese patients receiving dialysis the SF-36 scales are internally consistent and their scores are reproducible. Second, with the possible exception of the 'quality of social interaction' scale, the Japanese version of the KDQOL(TM),can provide psychometrically sound kidney-disease-targeted data on quality of life in such patients.