Impact of Different Mapping Algorithms From Disease-Specific Measures to EQ-5D on Cost-Effectiveness Analysis in Atopic Dermatitis

OBJECTIVES: EQ-5D, a preference-based measure, is viewed as the gold standard for measuring utility in cost-effectiveness analysis (CEA). However, clinical trials frequently omit EQ-5D and instead use disease-specific measures. Several mapping algorithms have been developed for patients with atopic dermatitis to convert from disease-specific measures to EQ-5D, which have been frequently used in CEA. As multiple algorithms exist, the choice of algorithm will affect the incremental cost-effectiveness ratio (ICER). However, currently, there are no reports that evaluate and consider the use of multiple algorithms in Japan. Based on this premise, this study aims to determine the impact of different mapping algorithms on the results of CEA.

METHODS: EQ-5D-5L, Dermatology Life Quality Index (DLQI), and pruritus Visual Analog Scale (VAS) scores were extracted from 2,817 individuals diagnosed with atopic dermatitis (AD) using data from the mHealth app "kencom^®" (DeSC Healthcare Co., Ltd.). The incremental quality-adjusted life-year (QALY) and ICER for delgocitinib versus difamilast in moderate to severe AD patients were compared using utility directly measured by EQ-5D-5L and mapped utilities from DLQI and pruritus VAS. A Markov model was used, with four health states: "clear," "mild," "moderate," and "severe". The time horizon of this analysis was one year, and the analysis was from the perspective of Japanese payers. The same model and assumptions were utilized with the exception of the mapping algorithm.

RESULTS: We identified three different mapping algorithms using utilities based on the DLQI and pruritic VAS. The incremental QALY and ICER for difamilast versus delgocitinib that were derived using the directly measured EQ5D and these algorithms ranged from 0.0026 to 0.0136 and from 830,949 yen/QALY to 4,269,352 yen/QALY, respectively.

CONCLUSIONS: The choice of different mapping algorithms can substantially impact the ICER and incremental QALY in AD patient evaluation. Therefore, the use of multiple algorithms in CEA should be considered.