Equity weights for economic evaluation: Using Discrete Choice Experiments in an Australian Population

Presenter: Richard Norman, University of Technology

Abstract

Economic evaluation of healthcare is conventionally based on the premise that an outcome is of equal value irrespective of who it accrues to. However, recent evidence from Tsuchiya and Dolan in the UK has suggested this does not reflect the general population view, with a majority of the general population either prioritising based solely on equality concerns, or trading between equality and efficiency [1]. To explore this issue in an Australian population, we used an online panel of 241 respondents, each of whom viewed 12 choice sets as part of a discrete choice experiment. By using experimental design principles advocated by Street and Burgess [2], the experimental design allowed for unbiased estimation of main effects and two-way interaction terms. In each choice set, respondents prioritised one of two health care programs, with different groups of potential respondents and gains in life expectancy. Using binary choice logistic regression, this paper investigates the importance of characteristics of the individual receiving an extra life year, in particular their initial life expectancy, smoking status, income and gender. We also match the characteristics of the survey respondent with those of the individuals within the experiment to identify if personal characteristics affect trade-offs made. Smoking status significantly affected the value of an extra life year, although gender and income do not. These aggregate results do however mask substantial heterogeneity (both observable and unobservable) within the sample. Various techniques for converting these results into weights for economic evaluation (such as the Hicksian compensating variation) will be discussed.

1. Tsuchiya, A. and P. Dolan, Do NHS clinicians and members of the public share the same views about reducing inequalities in health? Soc Sci Med, 2007. 64(12): p. 2499-503.

2. Street, D.J. and L. Burgess, The Construction of Optimal Stated Choice Experiments: Theory and Methods. Wiley Series in Probability and Statistics. 2007, Hoboken, New Jersey: Wiley.