Comparing Measurement Properties of the English EQ-5D-Y 3-Level Version With the 5-Level Version in South Africa [Editor's Choice]



This study aimed to compare the performance of the EQ-5D-Y 3-level version (Y-3L) and the expanded 5-level version, the EQ-5D-Y 5-level version (Y-5L).


Children or adolescents with an acute or chronic health condition and from the general population completed the Y-5L and Y-3L self-report questionnaires. The performance of the Y-5L and Y-3L was determined by comparing feasibility, redistribution of dimension responses, discriminatory power, validity, and test-retest reliability.


A total of 550 children/adolescents completed baseline measures and 173 completed repeat measures. The ceiling effect decreased by 15% from the Y-3L to Y-5L. Informativity of dimensions improved by 0.094 on the Y-5L. There was a range of 4% to 9% inconsistent responses moving from the Y-3L to Y-5L. Convergent validity of the visual analog scale and Y-3L and Y-5L dimensions was similar and weak to moderate (r range 0.18-0.38) but similar and strong on paired Y-3L and Y-5L dimensions: Kendall tau B (range 0.69-0.80) and gamma (range 0.92-0.98). The Y-5L and Y-3L showed moderate to substantial agreement for test-retest reliability across dimensions and visual analog scale scores in stable chronic health conditions and fair agreement for the general population.


The Y-5L is a valid, reliable extension of the Y-3L for children or adolescents across health conditions and healthy children/adolescents. The expanded levels reduced the ceiling effect. The relative informativity of report across dimensions increased on the Y-5L compared with the Y-3L with retention of the evenness of reporting. The convergent validity and test-retest reliability of the Y-5L was comparable with the Y-3L.


Janine Verstraete Razia Amien Des Scott

Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on Update my browser now