Abstract

Objectives

To develop a United States EuroQol 5-Dimension Youth version, 3-Level (EQ-5D-Y-3L) value set for measuring pediatric health-related quality of life based on preferences elicited from both adults and adolescents in the general population.

Methods

The study design was informed by the EuroQol international valuation protocol and recommendations from US stakeholders. US adults (≥18 years) and adolescents (11-17 years) completed an online discrete choice experiment (DCE), with adults considering a 10-year-old child perspective. Adult respondents were subsequently invited to complete composite time trade-off (cTTO) valuation tasks via videoconferencing with trained interviewers. DCE data from adults and adolescents were weighted to reflect the relative age and gender composition of the US population. The DCE data were analyzed using a main effect mixed-logit model with 10 parameters. The cTTO data for 28 health states were estimated using Tobit models censored at −1. The latent-scale coefficients from the mixed-logit model for the DCE data were mapped onto mean, censored cTTO values to obtain an EQ-5D-Y-3L value set anchored on a 0 (dead) to 1 (full health) scale.

Results

The analytic sample included DCE data from 714 adolescents and 1669 adults, and cTTO data from 199 adults. Estimated utilities ranged from 0.079 for the worst health state (33333) to 0.976 for the health state 12111. Pain/discomfort was the most important dimension, followed by worried/sad/unhappy.

Conclusions

This US EQ-5D-Y-3L value set fulfills an important need for value-based assessment in medical interventions of pediatric health informed by US stakeholder engagement and including both adolescent and adult preferences.