Mapping oral health related quality of life to generic health state values

Background: A summary utility index is useful for deriving quality-adjusted life years ( QALY) for cost analyses or disability weights for burden of disease studies. However, many quality of life instruments provide descriptive profiles rather than a single utility index. Transforming quality of life instruments to a utility index could extend the use of quality of life instruments to costs analyses and burden of disease studies. The aims of the study were to map a specific oral health measure, the Oral Health Impact Profile to a generic health state measure, the EuroQol, in order to enable the estimation of health state values based on OHIP data. Methods: Data were collected from patients treated by a random sample of South Australian dentists in 2001 - 02 using mailed self-complete questionnaires. Dentists recorded the diagnosis of dental conditions and provided patients with self-complete questionnaires to record the nature, severity and duration of symptoms using the EuroQol (EQ-5D) and 14-item version of the Oral Health Impact Profile (OHIP-14) instruments. Data were available from 375 patients ( response rate = 72%). A random two-thirds sample of patients was used in tobit regressions of EQ-5D health state values estimated using OHIP- 14 in a model with categories of OHIP responses as indicator variables and in a model with OHIP responses as continuous variables. Age and sex were included as covariates in both models. The remaining one-third sample of patients was used to test the models. Results: The OHIP item 'painful aching in mouth' was significantly related to health state values in both models while 'life less satisfying' was also significant in the continuous model. Mean forecast errors relative to the mean observed health state value were higher when fitted to the categorical model (17.4%) compared to the continuous model (15.2%) ( P < 0.05). Conclusion: The findings enable health state values to be derived from OHIP- 14 scores for populations where utility has not or cannot be measured directly.