Abstract

Objectives

The EQ Health and Wellbeing (EQ-HWB) tools have been developed to measure and value outcomes of both health and social care interventions, including those of carers, in a manner suitable for use in economic evaluation. The aim of this article is to add to the body of psychometric evidence for the performance of EQ-HWB, and its shorter version EQ-HWB-9, by assessing construct validity and reliability.

Methods

A sample of patients (n = 767) across 6 broadly defined health conditions and a sample of the general population (n = 302) completed the EQ-HWB measures alongside other measures. Convergent validity was assessed using Spearman and Pearson correlations. Known-group validity was investigated by using several self-reported variables and disease specific questions for the patient sample. Test-retest reliability was assessed by intraclass correlation coefficients and the kappa statistic.

Results

Convergent validity between EQ-HWB items and related items from EQ-5D-5L, SWEMWBS, and ICECAP-A was highest in the patient sample. At the scale level, the highest correlations of EQ-HWB summative score and other measures were observed with both PHQ-8 and GAD-7 followed by EQ-5D-5L and ICECAP-A. The EQ-HWB measures showed ability to detect differences in the defined known groups. Comparing across measures, the EQ-HWB measures had the highest standardized effect sizes for groups defined by emotional problems. The EQ-HWB measures were found to be reliable with test-retest reliability being >0.8 for both groups.

Conclusions

The results show that the EQ-HWB measures have promising psychometric properties across both the patient and general populations.