Abstract
Background: Among other uses, the subjective valuation of health care is also used in economic analyses, through which we are trying to determine the relative cost effectiveness of health technologies. Since cost effectiveness studies are very demanding and take a relatively long time to give results, the parallel estimation of the values of health states would represent an additional and, above all, unnecessary burden. The values of health states are therefore calculated in a separate study. The primary objective of the study is the derivation of a value set for the EQ-5D defined health states for Slovenia while controlling contextual bias.
Methods: According to our knowledge, spatial econometrics has never before been used for the purpose of estimating the EQ-5D valuation set. A value set for the EQ-5D health states for Slovenia is presented, based on data from research carried out in 2005. For calculating the value set, a spatial econometric estimator was chosen since it gave better results than the ordinary least squares (OLS) method. To control for contextual bias, a spatial variable was included in the model. Contextual bias has not been excluded from any of the calculated EQ-5D value sets.
Results: The EQ-5D defined value set for Slovenia is logically consistent. The defined model has a very good fit. The comparison of models with and without the spatial variable showed that controlling for contextual bias by including the spatial variable improves the fit of the model and produces slightly better results.
Conclusions: Due to the good fit of model and logically consistent valuations, the value set can be used in health-related economic and population analyses. Special attention is given to the exclusion of a contextual bias from the model to improve the fit of the model and change the priority settings.