Linguistic and Psychometric Validation of the Cognition Bolt-On Version of the Japanese EQ-5D-5L for the Elderly

Background: The need for a cognition bolt-on version of the EQ-5D, which would capture cognitive impairment by adding a dimension to the generic instrument assessing health status, has been increasing in Japan. Objective: To develop a cognition bolt-on version of the 5-level EQ-5D (EQ-5D-5L+C), we linguistically validated a cognition dimension and psychometrically validated the EQ-5D-5L+C. Methods: Following linguistic validation of the cognition dimension, psychometric validation of the EQ-5D-5L+C proxy version utilized anonymized data collected from nursing home residents between October 2021 to April 2022. The validity, reliability, and sensitivity to change were evaluated. Results: Data from 254 participants, including the finalized Japanese EQ-5D-5L+C proxy version, were analyzed for the psychometric validation. Mean (+/- standard deviation) age and Mini-Mental State Examination (MMSE) scores were 87.14 +/- 7.29 years and 15.76 +/- 8.46, respectively. The correlation was strongest between the cognition dimension and MMSE scores (r(s) = -0.640). Test-retest reliability was good in the cognition dimension in both baseline and two-time points (3 months: k = 0.644; 6 months: k = 0.656). Although a correlation between changes in the cognition dimension and those in the MMSE score from baseline was weak (3 months: r(s) = -0.191; 6 months: r(s) = -0.267), a correlation with changes in the MMSE score was higher when the cognition dimension was added compared to the EQ-5D alone (3 months: r(s) = -0.142 versus r(s) = -0.074). Conclusion: The Japanese EQ-5D-5L+C proxy version developed is a valid tool that captures health status including cognitive function, with a consideration for an over-time assessment. The benefits in adding the cognition dimension to the EQ-5D-5L to assess health state were suggested.