Comparing the mapping between EQ-5D-5L, EQ-5D-3L and the EORTC-QLQ-C30 in non-small cell lung cancer patients

Health and Quality of Life Outcomes

Research

Iftekhar Khan¹  Steve Morris²  Nora Pashayan²  Bashir Matata³  Joe Maguirre³  Zahid Bashir⁴ 

[1] Clinical Trials Unit & Department of Health Economics, University of Surrey, Guilford, UK;Department of Applied Health Research, University College London, London, UK;Department of Applied Health Research, University College London, London, UK;Liverpool Heart and Chest Hospital, Liverpool, UK;University of Birmingham, Cancer School, Birmingham, UK;

关键词: Root Mean Square Error;    NSCLC Patient;    Mapping Algorithm;    Poor Health State;    Mean Absolute Error;   

DOI  :  10.1186/s12955-016-0455-1

 received in 2015-07-19, accepted in 2016-03-23,  发布年份 2016

来源: Springer

【 摘 要 】

BackgroundSeveral mapping algorithms have been published with the EORTC-QLQ-C30 for estimating EQ-5D-3L utilities. However, none are available with EQ-5D-5L. Moreover, a comparison between mapping algorithms in the same set of patients has not been performed for these two instruments simultaneously. In this prospective data set of 100 non-small cell lung cancer (NSCLC) patients, we investigate three mapping algorithms using the EQ-5D-3L and EQ-5D-5L and compare their performance.MethodsA prospective non-interventional cohort of 100 NSCLC patients were followed up for 12 months. EQ-5D-3L, EQ-5D-5L and EORTC-QLQ-C30 were assessed monthly. EQ-5D-5L was completed at least 1 week after EQ-5D-3L. A random effects linear regression model, a beta-binomial (BB) and a Limited Variable Dependent Mixture (LVDM) model were used to determine a mapping algorithm between EQ-5D-3L, EQ-5D-5L and QLQ-C30. Simulation and cross validation and other statistical measures were used to compare the performances of the algorithms.ResultsMapping from the EQ-5D-5L was better: lower AIC, RMSE, MAE and higher R2 were reported with the EQ-5D-5L than with EQ-5D-3L regardless of the functional form of the algorithm. The BB model proved to be more useful for both instruments: for the EQ-5D-5L, AIC was –485, R2 of 75 %, MAE of 0.075 and RMSE was 0.092. This was –385, 69 %, 0.099 and 0.113 for EQ-5D-3L respectively. The mean observed vs. predicted utilities were 0.572 vs. 0.577 and 0.515 vs. 0.523 for EQ-5D-5L and EQ-5D-3L respectively, for OLS; for BB, these were 0.572 vs. 0.575 and 0.515 vs. 0.518 respectively and for LVDMM 0.532 vs 0.515 and 0.569 vs 0.572 respectively. Less over-prediction at poorer health states was observed with EQ-5D-5L.ConclusionsThe BB mapping algorithm is confirmed to offer a better fit for both EQ-5D-3L and EQ-5D-5L. The results confirm previous and more recent results on the use of BB type modelling approaches for mapping. It is recommended that in studies where EQ-5D utilities have not been collected, an EQ-5D-5L mapping algorithm is used.

【 授权许可】

CC BY   
© Khan et al. 2016