PNS246 EXPLORING THE PERFORMANCE OF USING DISCRETE CHOICE EXPERIMENTS FOR ELICITING HEALTH STATE UTILITIES IN CHINA: EVIDENCE FROM SF-6D V2

J. Wu,S. Xie,X. He,G. Chen,J. E. Brazier
DOI: https://doi.org/10.1016/j.jval.2019.04.1601
IF: 5.156
2019-01-01
Value in Health
Abstract:To explore the acceptability, consistency, and accuracy of eliciting health states utilities in the Simplified-Chinese version of SF-6D V2, by using DCE and DCE with survival-duration (DCETTO), as compared to classical TTO approach. A representative sample of general population in Tianjin, China, was recruited through cross-sectional face-to-face interviews. All respondents completed 8 TTO-tasks, and were randomly allocated to complete either 10 DCE- or 10 DCETTO-tasks, with the order of TTO and DCE/DCETTO been randomized. TTO data was estimated using fixed-effect OLS model. DCE and DCETTO data were estimated using conditional logit models, with latent utilities anchored using mapping method for DCE, and using "survival-duration" coefficient for DCETTO. Acceptability was assessed by completion rates and self-assessed difficulties on understanding/answering. Consistency was observed by the monotonicity of model coefficients. The accuracy of two DCE approaches was evaluated by investigating differences between predicted state values and observed TTO state values using intraclass correlation coefficient (ICC), mean absolute difference (MAD) and root mean squared difference (RMSD). A total of 503 respondents (53.7% males; mean age 45.4 years, range from 18-86 years) participated in this study, with comparable demographic-characteristics between DCE-group (N=252) and DCETTO-group (N=251). The completion rates were 91.6% for TTO, 94.4% for DCE, and 95.4% for DCETTO, respectively, while no significant difference was observed in self-assessed difficulties among three methods. The estimated coefficients for each attribute were monotonic for TTO; however, non-monotonicity was observed in two attributes for two DCE approaches. Comparing with observed TTO values, DCE (ICC=0.872; MAD=0.1339; RMSD=0.1710) and DCETTO (ICC=0.873; MAD=0.1620; RMSD=0.2154) were less accurate as compared to TTO (ICC=0.938; MAD=0.1003; RMSD=0.1311). Both two DCE approaches performed higher completion rates; however, their prediction were less accurate relative to TTO approach when comparing to the observed TTO values. The issue of non-monotonicity of coefficients from two DCE approaches remains a concern.
What problem does this paper attempt to address?