Qual Life Res. 2025 Oct 23. doi: 10.1007/s11136-025-04082-y. Online ahead of print.
ABSTRACT
PURPOSE: To examine whether ratings on each of the five dimensions of the EQ-5D-5L vary across ethnic groups after adjusting for health conditions and other relevant covariates.
METHODS: We applied the Hierarchical Ordered Probit (HOPIT) model to the EQ-5D-5L responses of 2,642,805 participants in nine pooled cross-sectional waves of the General Practice Patient Survey in England. The model separated latent health, determined by health conditions, from reporting behaviour, which was assumed to be influenced by socio-demographic characteristics, recent healthcare use, and healthcare access.
RESULTS: We find systematic differences in rating behaviour between five ethnic groups. Across all five dimensions, respondents from White ethnic backgrounds were most likely to choose middle severity levels. Respondents from Asian and Black ethnic backgrounds were more likely than those from White backgrounds to report the less severe levels for Anxiety/Depression and Usual Activities. Respondents from Other ethnic backgrounds not classified within the four main categories were more likely to choose the lowest and highest severity levels for Anxiety/Depression.
CONCLUSION: Our results suggest important heterogeneity in self-rating of health and highlights the importance of adjusting for reporting heterogeneity in applications of this measure. Where such adjustments are not feasible, alternative approaches such as sensitivity analyses to test the impact of differential reporting should be considered.
PMID:41128830 | DOI:10.1007/s11136-025-04082-y
Recent Comments