Using stated preference discrete choice modeling to evaluate health care programs

摘要

This paper reports on the potential to use stated preference discrete choice modelling (SPDCM) in health program evaluation. Interest is developing in the translation of these techniques from business, marketing, and environmental economics to health to address the prediction of market share and the estimation of societal benefits. The paper provides an overview of current health program evaluation methods, showing that measuring success in terms of clinical effectiveness, survival, or even quality-adjusted survival may not capture important benefits. The appropriate revealed preference data to estimate demand or value benefits are rarely available in the health care context. SPDCM is based in random utility theory (RUT) and the stated preference data are obtained from choice surveys. This allows a wide range of attributes to be included and the independent effect of each to be quantified. Selected applications in the testing for the value of information provided by genetic tests, in assessing how to improve attendance rates in screening programs, in predicting the uptake of a new immunisation, and in understanding patients' preferences for medications for chronic disease are described. SPDCM has much to offer in both the evaluation of health programs and the prediction of their uptake.