Testing the Valuation of the EQ-5D-Y-5L in Adults and Adolescents: Results From a 5-Country Study and Implications for the Descriptive System

OBJECTIVES: The EQ-5D-Y-5L (Y-5L) is a new health-related quality-of-life instrument for children and adolescents. Value sets for the Y-5L are planned. This article aimed to test the ability of adult and adolescent respondents to differentiate the ordinal levels of the Y-5L in valuation tasks and to explore the characteristics of stated preferences for the Y-5L between adults and adolescents. METHODS: We collected latent-scale discrete choice experiment data via an online survey of adults (≥18 years) and adolescents (12-17 years) in Australia, Canada, China, The Netherlands, and Spain. A D-Efficient design consisting of 192 choice pairs was grouped into 16 blocks of 12 choice tasks per respondent. We used mixed-logit models to analyze the data and incremental dummies to represent movements from a less-severe level to its consecutive more-severe level. RESULTS: We did not observe preference inversions in adults or adolescents (ie, no statistically significant positive coefficients on the incremental dummies). Adults showed similar preferences for the Y-5L in terms of dimension importance: Pain/Discomfort was considered the most important dimension in all countries except for China; Looking After Myself and Usual Activity were the least important dimensions. In contrast, Mobility was considered the most important dimensions by adolescents in Canada, Spain, and China. CONCLUSIONS: Adults could differentiate between the Y-5L level labels in valuation tasks, whereas more randomness was observed in adolescents' choices. Observed differences between adult and adolescent stated preferences for the Y-5L raise questions about how these preferences should be reflected in cost-effectiveness analysis.