Evaluating Health-Related Quality of Life in Children: Selecting the Value-Set for the EQ-5D-Y-3L

OBJECTIVES: Child-specific value set for EQ-5D-Y is more appropriate for evaluating health-related quality of life (HRQoL) in children. Australian child-specific value-set for the EQ-5D-Y-3L is still under development. This study explores the impact of applying EQ-5D-Y-3L (‘Y’) value sets from nine countries and Australian EQ-5D-3L value set on the assessment of inter-rater gap between child-self and proxy reports of HRQoL.

METHODS: An online sample of 845 dyads, comprising children aged 6-10 years and their parents from across Australia completed the respective self and proxy versions of the EQ-5D-Y-3L independently. HRQoL values were derived using the EQ-5D-Y-3L and the international ‘Y’ value sets. Analyses were stratified by age, gender, and health condition. Inter-rater gap was explored for dyadic mean differences using paired t-tests and the Intraclass Correlation Coefficient (ICC) for assessing agreement.

RESULTS: The highest HRQoL values were reported using the Indonesian value set (child=0.95, proxy=0.96), while Slovenia reported the lowest (child=0.87, proxy=0.89), following the Australian adult value set (child=0.85, proxy=0.86). Compared to child-self reports, proxies significantly overestimated HRQoL values across all ‘Y’ value sets, a discrepancy not observed with the Australian adult value set. Girls and children without health conditions reported significantly lower HRQoL than proxy estimates across all value sets. Proxies reported significantly higher HRQoL for older children (8-10 years) across ‘Y’ value sets, but not with the adult value set. Inter-rater agreement was mostly consistent across all value sets (ICC=0.62 to 0.71).

CONCLUSIONS: The choice of value set might influence the assessment of child-proxy reported HRQoL. Child-specific value sets offered more consistent outcomes than adult value set, advocating the use of any available ‘Y’ value set for assessing child HRQoL when a country-specific version is unavailable, rather than relying solely on the country-specific adult EQ-5D-3L value set. This also highlights a cultural consensus on the valuation of child health.