Translating the EORTC CAT core and the QLQ-C30 to the EQ-5D-5L in patients with metastatic breast cancer: A comparison of direct and indirect mapping algorithms

Journal article

Background To enable the use of different non-preference-based patient-reported outcome measures to derive utility values for health economic evaluations in oncological trials, this study developed direct and indirect mapping algorithms for estimating the EQ-5D-5L utility index via the German value set from the EORTC CAT Core and the QLQ-C30 in metastatic breast cancer patients. Methods We included 1,839 observations from 878 patients with metastatic breast cancer from the PRO B study. We compared direct mapping algorithms, including adjusted limited dependent variable mixture models (ALDVMM), Tobit regression, ordinal least squares regression, and adjusted beta regression, while indirect mapping employed a generalized ordered logit model. Visualization was used to assess model performance across the entire distribution, while quantitative evaluation was performed using mean absolute error (MAE), root mean squared error (RMSE), and mean prediction bias. Results Among the direct algorithms, adjusted beta regression demonstrated the best performance. It had the lowest MAE of 0.07–0.08 and RMSE of 0.11–0.13, a mean prediction bias of -0.004, close to zero. The indirect mapping model also performed well, with a mean prediction bias of 0.04 and MAE of 0.07, showing performance comparable to the preferred direct mapping algorithm for both the EORTC CAT Core and the QLQ-C30. Conclusions This study developed and validated robust direct and indirect algorithms for estimating the EQ-5D-5L utility index from the EORTC CAT Core and the QLQ-C30 based on the German tariff. In particular, using this indirect mapping algorithm, the EORTC CAT Core and QLQ-C30 can be translated into quality-adjusted life-years, facilitating health economic evaluations across different country tariffs. Trial registration DRKS (German Clinical Trials Register) DRKS00024015. Registered on 15 February 2021, https//drks.de/search/de/trial/DRKS00024015

Authors: Gebert, P; Hage, AM; Fischer, F; Klapproth, CP; Grittner, U

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Springer
• Journal: European Journal of Health Economics
• Volume: 27
• Issue: 2
• Pages: 435-451
• Publication date: 2025-08-21
• DOI: 10.1007/s10198-025-01824-0
• EISSN: 1618-7601
• ISSN: 1618-7598
• Language: English
• Keywords: Cancer; Algorithm; Mathematics; Regression; Mean squared error; Regression analysis; Internal medicine; Medicine; Statistics; Breast cancer