QALYs without bias? Nonparametric correction of time trade‐off and standard gamble weights based on prospect theory

Abstract

Common health state valuation methodologies, such as standard gamble (SG) and time trade‐off (TTO), typically produce different weights for identical health states. We attempt to alleviate these differences by correcting the confounding influences modeled in prospect theory: loss aversion and probability weighting. Furthermore, we correct for nonlinear utility of life duration. In contrast to earlier attempts at correcting TTO and SG weights, we measure and correct all these tenets simultaneously, using newly developed nonparametric methodology. These corrections were applied to three less‐than‐perfect health states, measured with TTO and SG. We found considerable loss aversion and probability weighting for both gains and losses in life years, and we observe concave utility for gains and convex utility for losses in life years. After correction, the initially significant differences in weights between TTO and SG disappeared for all health states. Our findings suggest new opportunities to account for bias in health state valuations but also the need for further validation of resulting weights.