Test-Retest Reliability of EQ-5D-5L Valuation Techniques: The Composite Time Trade-Off and Discrete Choice Experiments

Abstract

ObjectivesTo explore the test-retest reliability of the composite time trade-off (C-TTO) and discrete choice experiment (DCE) used in the Indonesian five-level EuroQol five-dimensional questionnaire (EQ-5D-5L) valuation study. MethodsA representative sample aged 17 years and older was recruited from the Indonesian general population by stratified quota sampling with respect to residence, sex, and age. Trained interviewers conducted computer-assisted face-to-face interviews using the EuroQol valuation technology. Each respondent valued 10 health states using C-TTO and 7 pairs of health states in a DCE exercise. The retest interview was conducted after 2 weeks by the same interviewer. The Wilcoxon matched-pairs signed-rank test, intraclass correlation coefficient, and multilevel regression were applied in comparing the C-TTO test and retest data. For DCE, the analysis of proportions was used. ResultsA total of 226 respondents with characteristics similar to the Indonesian population completed the retest interview. For C-TTO, 82 (95.3%) of 86 health states had no significant mean value differences between test and retest. The mean value of the second test was statistically significantly higher than that of the first test by 0.042. For DCE, 72.5% of responses were identical. DCE retest showed a different pattern concerning the relative importance of the dimensions, whereas the C-TTO remained the same. ConclusionsC-TTO is stable over time, whereas in DCE the relative values of the dimensions shift. The results support the use of the C-TTO, in particular the Indonesian EQ-5D-5L value set, and suggest a critical examination of the reliability of DCE results over time.