ABSTRACT
BACKGROUND: Cost-effectiveness analysis has been recommended by many national agencies around the world as a valid methodology to improve resource allocation within the health-care system. If the preferences of the society are taken into account in such a decision-making process, it is generally recommended that these values should be elicited by using a generic health-related quality-of-life instrument, such as the EuroQol five-dimensional (EQ-5D) questionnaire. OBJECTIVES: To estimate a set of social values for EQ-5D questionnaire based on the time trade-off valuation technique for use in Chile. METHODS: A valuation questionnaire was applied to a probabilistic sample of 2000 individuals, aged 20 years or older, living in the Metropolitan region. The fieldwork took place during October to November 2008. Utility weights for 42 health states were calculated directly by the application of time trade-off. Several random effect and ordinary least-squares regression models were fitted to these valuations to predict the full set of 243 health states generated by the EQ-5D system. The best model was chosen by applying criteria of parsimony, goodness of fit, and prediction capacity. RESULTS: The selected regression model was robust and showed better predictive characteristics than others reported in similar studies conducted elsewhere. The chosen regression model showed a R(2) of 0.34, mean absolute error of 0.017, and high predictive capacity. CONCLUSIONS: This study provides an EQ-5D social value set for domestic use in Chile. Our results differ from those reported in other countries, justifying the need to perform local studies that adequately reflect societal health preferences.