Abstract
Objectives
The EQ-5D-5L is widely used to measure adults’ health-related quality of life (HRQoL). The EQ-5D-Y-5L is a corresponding measure adapted for children/adolescents, in principle allowing HRQoL to be measured consistently from childhood to adulthood. However, little is known about how their measurement properties compare. This study investigated the relationship between EQ-5D-Y-5L and EQ-5D-5L in adolescents and compared their psychometric performance.
Methods
The Australian Pediatric Multi-Instrument Comparison Study includes a sample of 591 adolescents (aged 12-18) who completed both EQ-5D-5L and EQ-5D-Y-5L. Responses were compared descriptively and HRQoL summarized using the level sum score. Acceptability, feasibility, ceiling effects, convergence, test-retest reliability, and known-group validity were assessed overall and in subgroups defined by special healthcare needs (SHCN), mental health concerns (MHC), and age (12-13, 14-16, and 17-18 years).
Results
Ceiling effects were lower for EQ-5D-Y-5L than EQ-5D-5L. The EQ-5D-Y-5L better differentiated between adolescents with and without SHCN and MHC than EQ-5D-5L, whereas EQ-5D-5L showed better test-retest reliability in adolescents with SHCN and MHC. We found strong correlations between dimensions anticipated to be correlated. EQ-5D-Y-5L identified a higher incidence of self-reported HRQoL problems than EQ-5D-5L both overall and particularly in mental health.
Conclusions
Although both instruments are valid for measuring HRQoL in adolescents aged 12 to 18 years, EQ-5D-Y-5L had some psychometric advantages. The instruments are closely related, but differences in their descriptive systems produce differences in self-reported HRQoL. Results highlight potential discontinuities in HRQoL measured using age-specific instruments, which may be important for their use in economic models that involve transitions between age groups.
Authors
Nicole Reyes Tianxin Pan Renee Jones Kim Dalziel Nancy Devlin