Mapping EQ-5D-5L utility scores from the severe asthma questionnaire in the SHARP CRC registry

Abstract

BACKGROUND: Severe asthma imposes a substantial burden on individuals' health-related quality of life (HRQoL), which might not be fully captured by generic instruments. The Severe Asthma Questionnaire (SAQ) is a validated disease-specific HRQoL instrument capturing the unique lived experience of people with severe asthma. However, the SAQ does not generate preference-based utility values required for health economic evaluations. This study aimed to develop and validate mapping algorithms from the SAQ to the EuroQol (EQ)-5D-5L, enabling estimation of utility values when EQ-5D data are unavailable. METHODS: We used baseline and 6-month data from the longitudinal, multicentre SHARP Burden of Asthma study, including adults with severe asthma across 7 European countries. Direct mapping (ordinary least squares and beta-mixture regression models) using the UK value set and indirect mapping (ordered probit regression models) were developed with baseline data to predict EQ-5D-5L utility values from SAQ scores. Performance was assessed via mean absolute error (MAE), root mean square error (RMSE), and R(2), with 6-month data used for validation. RESULTS: Data from 327 patients were included. The beta-mixture model using SAQ subscales marginally outperformed the other models in the development set (MAE = 0.090, RMSE = 0.129, R(2) = 0.704), whereas in the validation set the ordered probit regression model showed the best predictive performance (MAE = 0.106, RMSE = 0.151, R(2) = 0.589). CONCLUSIONS: Direct mapping with beta-mixture regression showed good overall performance, aligning with prior studies. Indirect mapping may be advantageous in settings requiring flexibility across EQ-5D-5L value sets. These findings demonstrate that EQ-5D-5L utility values can be reliably estimated from the SAQ, supporting use of the instrument in health-economic analyses.