Econometric methods for estimating health utility and a catalogue of preference-based scores for chronic diseases

Quality-adjusted life years (QALYs) are the most common measure of health outcomes used in cost-effectiveness analysis (CEA). To calculate QALYs researchers often need to have health related quality of life (HRQL) measures for chronic conditions. In this thesis we use the EQ-5D index as a measure of HRQL and develop a catalogue of preference-based weights (health disutility scores) for chronic diseases.; The EQ-5D index is censored from the top; thereby Tobit is the most appropriate econometric method to use in the regression analysis. However, the Tobit estimates may be inconsistent since we reject the hypothesis of normality and homoskedasticity of errors. To find an estimator that is robust to non-normality, heteroskedasticity and censoring problems we perform a Monte Carlo simulation and compare the performance of Tobit, logistic-Tobit, extreme value type I (EVT-I)-Tobit, ordinary least squares (OLS) and censored least absolute deviation (CLAD) methods. We also adjust for heteroskedasticity in the maximum likelihood estimators. Based on the results of the Monte Carlo simulation we choose the heteroskedasticity adjusted logistic-Tobit estimator to develop the catalogue of HRQL measures for chronic conditions.