Tue Gørgens, Australian National University

“Modelling health care costs and life expectancy: Insights from Australian individual health expenditure data”

Abstract

In recent years there has been interest in quantifying the benefits associated with the declines in morbidity and mortality experienced in many developed countries. The purpose this study is examine the feasibility of using administrative data to build a simulation model to explore the degree to which declines in the risk of major categories of disease are likely to impact on lifetime costs and life expectancy. The simulation model is based on a proportional hazard competing risks models to represent three categories of disease (vascular, cancer, and other diseases); a multinomial logit model to determine whether events are fatal and whether they occur prior to or during a hospital stay; and a generalized gamma distribution to predict hospital costs. These equations are estimated using administrative hospital and primary care use data on the population with diagnosed diabetes in Western Australia. We use the model to simulate the impact on life expectancy and lifetime costs of changing risks of each disease

 
Contact person: H.C Kongsted