Supplementary materials: Mapping Dermatology Life Quality Index with EQ-5D health utility index score in Chinese patients with moderate to severe psoriasis

<p dir="ltr"><b>These are peer-reviewed supplementary materials for the article</b><b> </b><b>'</b><b>Mapping Dermatology Life Quality Index with EQ-5D health utility index score in Chinese patients with moderate to severe psoriasis</b><b>'</b><b> </b><b>published in the</b><b> </b><b><i>Journal of Comparative Effectiveness Research</i></b><b>.</b></p><ul><li><b>Supplementary table 1: </b>Univariate regression analysis of patient characteristics associated with EQ-5D-5L utility index score</li><li><b>Supplementary table 2: </b>Multivariable regression analysis of patient characteristics associated with EQ-5D-5L utility index score</li><li><b>Supplementary table 3:</b> Results of the developed mapping functions from OLS model. </li><li><b>Supplementary table 4:</b> Results of the developed mapping functions from Tobit model. </li><li><b>Supplementary table 5:</b> Results of the developed mapping functions from generalized linear model. </li></ul><p dir="ltr"><b>Aim:</b> To establish correlation between Dermatology Life Quality Index (DLQI) and EuroQol 5-Dimension questionnaire (EQ-5D) utility score in Chinese moderate-to severe psoriasis patients and develop mapping models for health utility prediction.<b> Materials & methods:</b> A total of 287 patients with moderate to severe psoriasis and at least one clinical visit with assessments for both DLQI and EQ-5D in a Chinese tertiary hospital were included. These patients were randomly split into a training set (n = 231) and a testing set (n = 56). Correlation analyses were performed to assess the relationship between DLQI (total score and item score) and EQ-5D utility index score. Twelve predictive models were developed using three statistical model approaches (ordinary least squares model, Tobit model and generalized linear model), incorporating various combinations of DLQI scores and patient characteristics (age, sex, education and comorbidities). Models were evaluated using root mean square error and mean absolute error (MAE). <b>Results:</b> A strong and significant correlation was found between DLQI total score and EQ-5D utility index score (r = -0.645, p < 0.001). The best-performing model, (ordinary least squares model using DLQI total score, had the lowest root mean square error (0.122) and MAE (0.076). Validation of this model in the testing set yielded a predicted utility score with an MAE of 0.072, and an MAE-to-utility ratio of 0.084, below the validation threshold of 0.1. <b>Conclusion:</b> DLQI scores can reliably predict health utility values, offering a useful tool for clinical decision-making and health economic evaluations. The model shows strong predictive accuracy and has potential applications in the management of msPsO in China.</p>