Priorities and Prospect Theory

Whose preferences are to be used for cost-effectiveness analysis? It has been recommended that community preferences for health states are the most appropriate ones for use in a reference case analysis. However, critics maintain that persons are not able properly to judge a health state if they have not experienced the condition themselves. This problem is analyzed here in the framework of Prospect Theory. It can be argued that the differing reference points of patients and the general public are responsible for deviating results. In addition, we argue that risk attitudes with respect to health-related quality of life are an indicator of reference points. If patients and the general public refer to the same reference point, i.e., they have the same risk attitude, the hypothesis is that deviations no longer significantly differ. Evaluations of the health condition of tinnitus by 210 patients and 210 unaffected persons were compared. The Time Tradeoff and Standard Gamble methods were applied to elicit preferences. Risk attitude was measured with the question of whether participants would undergo a treatment that could either improve or worsen their health condition, both with an equal chance (five possible answers between “in no case” and “in any case”). Affected persons indicated significantly higher values for tinnitus-related quality of life according to the Standard Gamble method. The difference between Time Tradeoff values was less dramatic but still significant. In addition, nonaffected persons are more risk-averse than affected persons. However, differences in evaluations are not significant considering single risk groups (e.g., those who answered “in no case”). Prospect Theory is a reasonable framework for considering the question of whose preferences count. If this result can be generalized for other diseases as well, it allows the mathematical combination of “objective” evaluations by the general public with the illness experience of patients. These evaluations should be weighted with patients' risk attitudes, i.e., community preferences can be used if they are corrected for risk attitudes. Copyright Springer-Verlag Berlin Heidelberg 2002