Psychometric performance of the EQ Health and Wellbeing short (EQ-HWB-S) in a UK population sample

Objectives

This study assessed the psychometric performance and construct validity of the EQ Health and Wellbeing Short (EQ-HWB-S), using a validated measure, the EQ-5D-5L, as a comparator.

Methods

The experimental version of the EQ-HWB-S was compared with the EQ-5D-5L to assess the psychometric performance of the measures. Data was drawn from the valuation stages of the Extending the QALY (E-QALY) project (UK general population, n=429) and the EQ-5D-5L UK valuation pilot study (UK general population, n=248). Construct validity was assessed based on convergent validity, using Spearman correlations and Pearson correlations. Known group validity was assessed by estimating effect sizes to assess the ability of the EQ-HWB-S and EQ-5D-5L to discriminate between known groups based on ‘healthy’ status, presence of a long-term condition, health and life satisfaction, age and employment status. The degree of agreement in utility values across instruments was also evaluated using Bland-Altman plots.

Results

Strong associations (rs ≥0.5, P<.001) were found between conceptually overlapping dimensions and the utility scores of the EQ-HWB-S and EQ-5D-5L. The instruments performed comparably in discriminating between known groups including healthy vs unhealthy groups (based on the VAS≥80), long-term condition (vs no long-term condition), above vs below average health and life satisfaction and employed (vs unemployed and long-term sick).

Conclusion

The EQ-HWB-S performs favourably with utility values successfully discriminating between groups where differences are expected. Convergence between the EQ-HWB-S and EQ-5D-5L is evident, especially between conceptually overlapping dimensions.