This study aimed to map the Lequesne index onto the EuroQol 5 Dimension (EQ-5D-5L) utility index for patients with knee osteoarthritis.
Baseline data from a previous randomized controlled trial were used; 461 patients were involved in the mapping development, and 230 in the validation phase. Various modeling techniques, including generalized linear models, tobit, and beta regression, were used. Factors such as age, sex, and body mass index were considered as covariates. Model selection was based on criteria such as Akaike and Bayesian information criteria, adjusted R, mean absolute error, and root mean squared error. Validation involved assessing the preselected models using mean absolute error, root mean squared error, and intraclass correlation coefficient. This study follows the Mapping Onto Preference-Based Measures Reporting Standards statement.
Five models were developed, with 2 incorporating age, sex, with or without body mass index along with the Lequesne index showing the best fit across regressions. Validation results were similar for the 3 regressions, with beta regression models exhibiting wider ranges closer to the validation data set. Intraclass correlation coefficient values were better for beta regression models. Both models tended to overpredict for lower EQ-5D-5L values and underpredict for better health status.
These mapping functions, the first of their kind, effectively translate the Lequesne index to EQ-5D-5L values in patients with knee osteoarthritis. They demonstrate satisfactory fit and precision, providing valuable tools for clinicians and researchers, particularly in situations where generic preference-based health-related quality of life instruments are inaccessible for utility derivation in cost-effectiveness studies.