Objectives An array of government white papers and scholarly works have raised concerns that a purely utilitarian (QALY-based) approach to health prioritisation is ethically inadequate. To accommodate this, various severity criteria have been suggested and attempted operationalised in e.g. Norway, the Netherlands, Sweden, and recently the UK. However, what severity is remains elusive, and is an ongoing topic of debate.

Some empirical research has attempted to identify how the severity of disease plays a role, in addition to cost-effectiveness, when people make priority decisions. The definition of severity in these studies varies, but in most cases does not adequately quantify health state utility values and severity or rely on abstract numeric representations. These practices allow for misinterpretation.

This study aims to investigate whether people divert from QALY-maximizing strategies in priority setting DCE tasks based on individual-level TTO values for the states used in comparisons.

Methods Data collection is about to start. 500-600 participants will first be administered 10 EQ-5D-5L health states for valuation using a R/Shiny-based EQ-VT-equivalent cTTO task, with dynamic state selection to ensure substantial variation in elicited values. Using the same EQ-5D-5L health states, respondents will then be presented with a set of discrete choice tasks with varying degrees of discrepancy between utility maximisation and severity. The severity component will have different operationalisations. This way, we know the utility values associated with each health state without relying on a numeric representation of utilities.

Results The data collection will be completed by Q1 2022.

Discussion We hypothesise an aggregate inclination towards concern for the worse off, sacrificing some utility maximisation, and expect substantial between-respondent heterogeneity, both in the presence and strength of preferences for concerns other than utility-maximisation. Evidence of such inclinations may be informative when operationalising severity criteria in health priority processes.