Abstract

Objectives

In conventional quality-adjusted life-year maximization, each unit of health is valued equally. It overlooks distributional preferences, such as whether health improvements should be concentrated on a smaller number of individuals or dispersed across a larger population. It is plausible that these preferences follow an S-shaped pattern, in which marginal social value of health gains initially increases before diminishing after a threshold. This review aims to identify and evaluate the methods used in empirical studies to elicit S-shaped preferences.

Methods

Searches were conducted in PubMed, PsycINFO, and EconLit in August 2024 and October 2025, with 1 round of citation chasing for 4 key articles. Studies were included if they elicited preferences on concentration and dispersion of health gains among otherwise identical individuals. Data synthesis was conducted narratively.

Results

Thirteen studies published between 1994 and 2025 were included, collectively involving 6799 participants. All studies used a Person Trade-Off (PTO) approach, but 2 also used Time Trade-off. Five key methodological themes emerged: whether health gains should be measured in life-years or health-related quality of life, the threshold between concentration and dispersion, perspective, framing effects, and specification of the social welfare functions.

Conclusions

This review highlights methodological challenges associated with eliciting preferences for concentration and dispersion. Although there is a consensus on the use of PTO, other heterogeneity such as pool size or perspective and sparse evidence on the threshold effect warrants further research.