Feasibility and validity of the EQ-5D-3L in the elderly Europeans: a secondary data analysis using SHARE(d) data

Abstract

Purpose: To determine feasibility and validity of the EQ-5D-3L in the elderly European population. Methods: Secondary data analysis based on the study of health, ageing, and retirement in Europe (SHARE) to determine the percentage of missing items for EQ-5D dimensions and EQ VAS, and to demonstrate convergent/divergent validity with measures included in the SHARE survey. Known-groups validity was tested using literature-based hypotheses. Correlation coefficients and Cohen’s f are reported. Results: Missing values were below 3% across all EQ-5D dimensions and gender strata, slightly increasing with age. Individuals’ responses to each EQ-5D dimension were related to their ratings of other measures in expected directions. The EQ VAS and all EQ-5D dimensions (except anxiety/depression) moderately to strongly correlated with physical [e.g. number of limitations in activities of daily living (ADL): r = 0.313–0.658] and generic measures [CASP (control, autonomy, self-realization, pleasure)-19 scale, self-perceived health, number of symptoms: r = 0.318–0.622], while anxiety/depression strongly correlated with the EURO-D scale (r = 0.527). Both EQ-5D dimensions and EQ VAS discriminated well between two [or more] groups known to differ [e.g. anxiety/depression discriminated well between persons classified as depressed/not depressed using the EURO-D scale, f = 0.51; self-care differentiated best between individuals without and with 1 + ADL limitations, f = 0.69]. Sociodemographic variables like gender, education, and partner in household were hardly associated with EQ VAS scores (f < 0.25). Conclusion: With item non-response of less than 3%, good discriminatory, and construct properties, the EQ-5D-3L showed to be a feasible and valid measure in the elderly Europeans.