A Method for Adjusting EQ-5D-3L Health State Utilities for Cross-Country Use: Application of a Novel US-to-UK Utility Mapping Approach

Abdul Jabbar Omar Alsaleh,1 Jerome Msihid,2 Elisheva Lew,3 Charlie Nicholls4 1Health Economics and Value Assessment, Sanofi, Milan, Italy; 2Health Economics and Value Assessment, Sanofi, Gentilly, France; 3Health Economics and Value Assessment, Sanofi, Paris, France; 4Health Economics and Value Assessment, Sanofi, Reading, Berkshire, UKCorrespondence: Abdul Jabbar Omar Alsaleh, Health Economics and Value Assessment, Sanofi, Viale Luigi Bodi, 37/b, Milan, 20158, Italy, Tel +39 344 023 74 53, Email [email protected]: The EuroQoL-five dimensions-three levels (EQ-5D-3L) utility values vary across regions, hence become non-interchangeable. This study details a methodology to adjust EQ-5D-3L utility values from the published United States (US) scores for use in the United Kingdom (UK) when individual-level patient data (IPD) are inaccessible.Methods: Linear and polynomial regression models were fitted to map the US-derived EQ-5D-3L health state utility weights (HSUWs) into UK values. Model fit was evaluated via visual inspection and coefficient of determination (R2). The best-fit model was used to develop a utility-adjustment algorithm. Data for external validation of the algorithm were collected through a targeted literature review, which identified studies reporting mean EQ-5D-3L utilities from both the US and UK. The coefficients of the best-fit model were validated against the published mean US and UK utilities, and the algorithm was validated by comparing published and predicted mean UK utilities.Results: The linear model confirmed linearity between US and UK HSUWs, and was the best-fit over the polynomial model, both visually and with R2 (0.9574 vs 0.9576). Eleven publications (N = 93,625 participants) reporting the mean US and UK EQ-5D-3L utilities based on the valuation of EQ-5D-3L responses were included for validation. The linearity between the published mean US and UK EQ-5D-3L utilities was deemed valid (R2 = 0.9955) with similar slopes (1.45 vs 1.39) and intercepts (− 0.43 vs − 0.38) between the linear model and the published utility-based line. The algorithm (UtilityUK = [− 0.3813 + 1.3904 × UtilityUS]) predicted the mean UK utilities, which marginally differed from the reported mean UK utilities, confirming its validity.Conclusion: A robust method was developed to adjust US-derived EQ-5D-3L utilities and reflect UK preferences. This method helps predict and transpose utilities from the literature across countries as a supporting tool when IPD are unavailable, which is necessary when adapting cost-effectiveness models to different settings.Keywords: EQ-5D-3L, health utility, transferability, United Kingdom, United States, utility weight adjustment