Discounting and independence in preferences between health sequences

Abstract

Decision analysts and cost-effectiveness analysts in the health domain often use discounted Quality-Adjusted Life Years (QALYs) as a measure of utility. These analysts commonly assume a positive discount rate: health outcomes in the distant future receive less weight than health outcomes in the near future. In Experiment 1, however, we found that most people actually use negative discounting when choosing between health sequences. Experiment 2 cast doubt on the idea that negative discounting can be explained by desires to maximize savoring and minimize dread. Experiments 3 and 4 gave strong support for the independence assumption of the discounted QALY model, and also replicated the tendency toward negative discounting for health sequences. Taken together, these findings suggest that the general structure of the discounted QALY model is descriptively accurate, but the specific use of a positive discount rate for health sequences should be reconsidered.