Validation of the EQ-HWB-9 in a mental health sample and an investigation of modifications to items

BACKGROUND: The EQ Health and Wellbeing (EQ-HWB) is a new, generic instrument designed to evaluate quality-of-life across health, public health, and social care settings. The short form comprises nine items (EQ-HWB-9) and validation across diverse populations and contexts is required. We aimed to investigate the validity of the EQ-HWB-9 in an international sample of adults experiencing poor mental health who downloaded a meditation app. We further examined the impact of four item-level modifications on psychometric performance, including investigating a potential ordering effect for the 'activities' item (hypothesised in prior studies) and three minor changes to response options. METHODS: The current study was embedded in a larger trial examining engagement with meditation via a free, downloadable app. Participants were randomised to complete the original (2022) and modified (2024) experimental version of the EQ-HWB-9. Psychometric evaluation included analyses of item distribution, known group and convergent validity, and responsiveness to change. RESULTS: There were no differences in demographic characteristics between the EQ-HWB original and modified versions at baseline (n = 865) or follow-up (n = 130). All psychometric tests supported the validity of the EQ-HWB-9 in this population. We found an ordering effect for the activities item, where the activities item showed a greater level of difficulty and a wider distribution over response options when asked before the mobility item, rather than after. There were no observable differences between versions for the other modifications. CONCLUSIONS: These findings add to growing literature supporting the EQ-HWB-9 as a suitable instrument for measuring quality-of-life across a range of settings. When the 'activities' item was presented first, as in the modified version, participants appeared to interpret the item more broadly, in line with the developers' intentions. Accordingly, our results support changing the item order of the first two items. Other modifications had little impact on outcomes, suggesting that further qualitative research will be required to inform decisions about their inclusion. Results from this study provide support towards finalisation of the instrument. The development of country-specific value-sets is now critical to support its application in economic evaluation.