# Comparing the measurement properties of the EQ-5D-Y-3L, EQ-5D-Y-5L and CHU9D in children and adolescents: a measurement property study

## TL;DR

This study tested and compared the reliability and validity of three health-related quality of life tools in Brazilian children and adolescents, finding better performance in those with musculoskeletal pain.

## Contribution

The study provides the first validation of EQ-5D-Y-3L, EQ-5D-Y-5L, and CHU9D in Brazilian children with and without musculoskeletal pain.

## Key findings

- All instruments showed better reliability and validity in children with musculoskeletal pain.
- The EQ VAS had substantial reliability for those with pain but only moderate for those without.
- More than 75% of hypotheses were confirmed for construct validity in children with musculoskeletal pain.

## Abstract

The EQ-5D-Y-3L, EQ-5D-Y-5L and Child Health Utility 9-dimension (CHU9D) are instruments that measures health-related quality of life. These instruments are widely used in children and adolescents with health conditions, however the measurement properties of the three instruments have not been tested in Brazilian children.

To compare and test the measurement properties of the EQ-5D-Y-3L, EQ-5D-Y-5L and CHU9D in Brazilian children and adolescents with and without any self-reported musculoskeletal pain.

Children and adolescents aged 8–18 years were recruited from schools in Sao Paulo, Brazil and, self-completed the EQ-5D-Y-3L, EQ-5D-Y-5L and CHU9D at baseline and after 7 days. Reliability was determined by Kappa for the dimensions and intraclass correlation coefficient (ICC) for visual analogue scale (EQ VAS). Hypothesis were developed for construct validity and tested with Spearman and Pearson correlations (adequate if > 75% of the hypotheses confirmed). Children and adolescents with and without musculoskeletal pain were compared for known-group validity.

We included 356 children and adolescents, with 51% (n = 181) reporting musculoskeletal pain. Majority were male (53%) and mean age of 11.5 years (SD: 2.9). The EQ-5D-Y-3L, EQ-5D-Y-5L and CHU9D ranged from poor to moderate reliability. Reliability of the EQ VAS was substantial (ICC: 0.81, 95% confidence interval [CI]: 0.72 to 0.87) to moderate (ICC: 0.40, 95% CI: 0.24 to 0.53) for those with musculoskeletal pain and without pain, respectively. In those with musculoskeletal pain the association was weak to moderate, with > 75% of hypotheses confirmed, when comparing EQ-5D-Y-3L and EQ-D-Y-5L with the PedsQL™ and comparing EQ-5D-Y-5L with CHU9D. All instruments were able to discriminate those with and without musculoskeletal pain.

All instruments had better measurement properties in children and adolescents with musculoskeletal pain, compared to those without for reliability and construct validity. These instruments could be used to assess health-related quality of life in Brazilian children and adolescents with musculoskeletal pain.

This study validated the EQ-5D-Y-3L, EQ-5D-Y-5L and CHU9D in Brazilian children and adolescents.

The measurement properties of the EQ-5D-Y-3L and EQ-5D-Y-5L had a better performance in children and adolescents with musculoskeletal pain compared those without musculoskeletal pain.

Clinicians and researchers can use the EQ-5D-Y-3L, EQ-5D-Y-5L, and CHU9D to measure health-related quality of life in children and adolescents.