Performance of spondyloarthritis-specific health utilities based on the ASAS Health Index: an ancillary analysis from the ASAS-HI validation study

Objectives: To compare the Spondyloarthritis (SpA)-specific universal utility estimation from ASAS Health Index (U-ASAS-HI) with generic utilities and to understand the contribution of health outcomes, personal- and country-level factors to the U-ASAS-HI. Methods: This was an ancillary analysis of the ASAS-HI international validation study. SpA patients who completed the ASAS-HI, 5-level EuroQol-5D (EQ-5D-5L) and Short Form-36 (SF-36) questionnaires were selected, and utilities calculated. Correlations between U-ASAS-HI and generic utility values were tested. Potential explanatory variables were evaluated in a linear multivariable mixed-effects model analysis with multilevel modelling fitted by country. Results: The number of patients included was 1425, with a mean (S.D.) age of 42 (13.5) years, with 65% male. The mean (S.D.) utility scores were for EQ-5D-5L 0.75 (0.11), for short-form 6-dimension (SF-6D) 0.72 (0.24), and for U-ASAS-HI 0.42 (0.29). Correlation between U-ASAS-HI and EQ-5D-5L and SF-6D was strong (r = 0.75 and r = 0.72, respectively; P < 0.001). The linear mixed-effects model showed that worse disease activity (ASDAS) and lower physical function (BASFI) were associated with lower U-ASAS-HI. Additionally, younger age, female gender, higher comorbidity score and symptoms of depression were associated with lower U-ASAS-HI. SpA subtype had no influence on health utility. The random effects model indicated an intercept S.D. of 0.045 with a 5.5% variance ratio between countries and the total U-ASAS-HI variation. Conclusion: The U-ASAS-HI captures the physical and mental impact of SpA, as well as personal contextual factors such as age, gender and comorbidities. The influence of the country of residence seems negligible. The U-ASAS-HI algorithm has sufficient support to be used in health economic evaluations for SpA patients across different countries.