28-09-2024 | Original Paper
Comparison of the EQ-5D-Y and the CHU-9D instruments in a general child population based on self-reports and proxy-reports
Published in: The European Journal of Health Economics
Login to get accessAbstract
Objective
This study utilized the EQ-5D-Y and the Child Health Utility 9D (CHU-9D) instruments to empirically investigate a general child population aged 7–8 years in China, with the aim of assessing and comparing the performance, correlation, and agreement between these two instruments. Both self-reported and proxy-reported versions of the instruments were considered in the analysis.
Methods
Data were collected from 7-8-year-old students in the second grade from four schools in Guangxi and Guiyang provinces, China. Children and their proxies independently completed their respective versions of the questionnaires, including the EQ-5D-Y, the CHU-9D, and other socio-demographic information. The psychometric properties of the EQ-5D-Y and the CHU-9D were assessed, including ceiling effects, internal consistency, and known-group validity. Spearman’s correlation coefficient, Intraclass Correlation Coefficient (ICC), and Bland-Altman plots were calculated and plotted to assess the correlation and agreement between the EQ-5D-Y and CHU-9D.
Results
A total of 369 pairs of valid questionnaires were collected from both children and proxies. Due to the study’s focus on a general child population, both EQ-5D-Y and CHU-9D yielded high utility values, with a significant ceiling effect observed, particularly in the EQ-5D-Y proxy-reported results. Compared to the EQ-5D-Y, the CHU-9D demonstrated a higher Cronbach’s alpha coefficient and better internal consistency. Both instruments also demonstrated known-group validity, distinguishing different health status groups, except for EQ-5D-Y proxy-reported results. Spearman’s correlation coefficient indicated some correlations in similar dimensions and utility values between the EQ-5D-Y and CHU-9D. The ICC of the EQ-5D-Y and CHU-9D utility values was 0.290 for self-reports and 0.383 for proxy-reports, indicating poor agreement between the two instruments. The Bland-Altman plots showed that the mean utility values obtained from EQ-5D-Y were significantly higher than those from CHU-9D.
Conclusion
The EQ-5D-Y and the CHU-9D demonstrated acceptable performance within the general child population aged 7–8 years in China, except for the EQ-5D-Y proxy-reported version. It suffered from a notable ceiling effect, poor internal consistency, as well as weak known-group validity and discriminative ability. Moreover, although there existed a certain degree of correlation between the EQ-5D-Y and CHU-9D, their utility values exhibited significant differences. Therefore, these instruments are not interchangeable in practice.