How do different approaches to the economic modelling of genetic and genomic tests drive differences in results?

Abstract

In many countries, including Australia, health technologies undergo evaluation by a Health Technology Assessment body to inform decisions around public funding. These evaluations include an assessment of the cost-effectiveness (value for money) of a health intervention. Traditionally, economic evaluations were designed to assess a test or treatment for a single condition and an ‘average person’, where the costs and outcomes followed straightforward trajectories. Genetic and genomic tests, however, can assess multiple genes and produce results for multiple conditions, including variants of unknown significance and secondary findings, with dynamic impacts on costs and outcomes. Current economic evaluation methods struggle to deal with this complexity.

The overarching objective of the current research was to examine the impact of different approaches to economic modelling of genetic and genomic tests and the potential for these to lead to different public funding decisions for these tests. The focus was on two key challenges of economic methods: (1) Model Structure and (2) Selection of Outcomes. This thesis begins with a comprehensive review of economic models for genetic and genomic testing, which identified non-invasive prenatal testing (NIPT) as a case study. The subsequent stages were: (a) developing multiple model structures of NIPT for the detection of Down syndrome, and populating these models with consistent parameters to allow the impact of selected structural variations to be assessed; (b) conducting three focus group discussions with the Australian public to inform the development of a discrete choice experiment (DCE); (c) the DCE, which valued intermediate outcomes through willingness to pay estimates (WTP); and (d) extending two models developed in stage (a) by incorporating the WTP estimates derived from the DCE within both net marginal benefit and cost-utility analyses.