Journal of Brain Research

Background: It is difficult for individuals with Alzheimer's disease (AD) to measure their quality of life via EQ-5D-5L due to their symptoms, and the caregiver proxy rating is often used. Hence, it is important to understand the validity of EQ-5D-5L by proxy rating for the patients.

Objectives: To validate the EQ-5D-5L by caregiver proxy rating and confirm its concurrent validity using the ABC dementia scale (ABC-DS) for Alzheimer's patients.

Methods: We used item response category characteristics curves (IRCCC) to investigate the responsiveness of five dimensions in five items of the EQ-5D-5L over the possible range of the QOL. We also calculated the correlation coefficient between EQ-5D-5L and ABC-DS, and obtained a regression model to estimate the expected utility score using the ABC-DS score.

Results: IRCCC indicated that three questions of the EQ-5D-5L for mobility, self-care, and usual activities accurately rated the QOL; however, the other two questions for pain/discomfort and anxiety/depression did not sufficiently contribute to the rating. EQ-5D-5L utility scores correlated well with the three-dimensional distance (TDD) scores of the ABC-DS. By using a regression model to estimate the expected utility score of EQ-5D-5L by TDD, our simulation estimated that the therapeutic effect obtained by revising a drug regimen resulted in a monetary benefit of 110 US dollars on average.

Conclusions: We confirmed the concurrent validity of EQ-5D-5L by proxy rating with the ABC-DS. Although two items of EQ-5D-5L did not contribute to estimating the QOL, the other three items did, thus reflecting the severity of AD.