Feasibility, reliability, and validity of the EQ-5D-Y: results from a multinational study

Abstract

Purpose: To examine the feasibility, reliability, and validity of the newly developed EQ-5D-Y.

Methods: The EQ-5D-Y was administered in population samples of children and adolescents in Germany, Italy, South Africa, Spain, and Sweden. Percentages of missing values and reported problems were calculated. Test-retest reliability was determined. Spearman's rank correlation coefficients with other generic measures of HRQOL were calculated. Known groups' validity was examined by comparing groups with a priori expected differences in HRQOL.

Results: Between 91 and 100% of the respondents provided valid scorings. Sweden had the lowest proportion of reported problems (1-24.9% across EQ-5D-Y dimensions), with the highest proportions in South Africa (2.8-47.3%) and Italy (4.3-39.0%). Percentages of agreement in test-retest reliability ranged between 69.8 and 99.7% in the EQ-5D-Y dimensions; Kappa coefficients were up to 0.67. Correlation coefficients with other measures of self-rated health indicated convergent validity (up to r = -0.56). Differences between groups classified according to presence of chronic conditions, self-rated overall health and psychological problems provided preliminary evidence of known groups' validity.

Conclusions: Results provide preliminary evidence of the instrument's feasibility, reliability and validity. Further study is required in clinical samples and for possible future applications in economic analyses.