Mapping the SF-36 quality of life scale onto utility by tobit regression model

Mapping the scales which are frequently used in clinics and don't have a utility algorithm, onto utility enables to calculate utility values in the studies only such scales are used. The aim of this study is mapping SF-36, B scale, MNDI and NPDI onto utility. The utility values are estimated for each combination of aforementioned scales' scores by using OLS and Tobit methods. The models are evaluated for goodness of fit by using standard techniques such as variance explained. Th ratio of variance explained is % 35.4 and % 41.9 for OLS and Tobit regression models, respectively. Both models are determined to explain the data better than null models (p<0.001). The standard errors of estimates in Tobit model are lower than those in OLS model. A model predicting utility is obtained by generic health related quality of life scale SF-36, B scale, MDNI and NPDI, which don't have a utility algorithm.