Abstract
OBJECTIVES: The EQ-5D-5L is a multi-attribute utility instrument recommended by many health technology assessment agencies. This study aimed to develop an EQ-5D-5L value set for Singapore. METHODS: A 'lite' version of the EuroQol Research Foundation's EQ-5D-5L valuation protocol, which was designed to value a total of 91 health states using a composite time trade-off (cTTO) method, was followed. Five hundred members of the general public in Singapore were quota-sampled and invited to a personal interview face-to-face or via Zoom. All participants completed 20 cTTO tasks administered using the EuroQol Valuation Technology (EQ-VT) program. Cross-validation analysis was performed to identify the best-performing model for estimating the values of all the 3,125 EQ-5D-5L health states. RESULTS: A 20-parameter main-effects model with two two-way interaction terms outperformed other models in the cross-validation analysis. The value set estimated using this model ranges from - 0.851 (for state 55555) to 1.000 (for state 11111), with pain/discomfort and anxiety/depression dimensions associated with the greatest disutility. CONCLUSIONS: We developed an EQ-5D-5L value set based on the health preferences of Singaporeans. We recommend EQ-5D-5L users in Singapore to use this value set and encourage a more systematic and dedicated methodological effort to understand interaction effects and potential non-linearities in the valuation of multi-attribute health descriptive systems.