Mapping EQ-5D utility scores from the PedsQL™ generic core scales

Area of Science:

• Health Economics
• Pediatric Health Outcomes
• Psychometrics

Background:

• The Pediatric Quality of Life Inventory™ (PedsQL™) General Core Scales (GCS) measure health-related quality of life in children aged 2-18.
• Current limitations prevent estimating health utilities directly or indirectly from PedsQL™ GCS data.
• Health utilities are crucial for economic evaluations of healthcare interventions.

Purpose of the Study:

• To assess mapping methods for estimating EQ-5D health utilities from PedsQL™ GCS responses.
• To develop empirical algorithms for deriving health utilities in pediatric populations.

Main Methods:

• Cross-sectional survey data from 559 children (aged 11-15) in English secondary schools.
• Estimation of models using direct and response mapping approaches to predict EQ-5D utilities.
• Internal and external validation using separate datasets (n=337) and assessing predictive accuracy via Mean Squared Error (MSE) and Mean Absolute Error (MAE).

Main Results:

• Ordinary Least Squares (OLS) models using PedsQL™ GCS subscale scores, squared terms, and interactions demonstrated the best predictive accuracy.
• External validation showed minimal differences in MSE (MAE) between OLS models with (0.036, 0.115) and without (0.036, 0.114) age and gender.
• Higher prediction errors were observed for children with poorer health states (EQ-5D utility <0.6); response mapping models faced data limitations.

Conclusions:

• Developed mapping algorithms provide an empirical basis for estimating childhood health utilities when EQ-5D data is absent.
• These algorithms can inform economic evaluations of pediatric interventions.
• Algorithms are likely robust for comparable populations (11-15 year olds in secondary school); performance in other groups requires further evaluation.