Principal Investigator: Douglas Wolf
Active Dates: 2009 – 2013
Funding Source: National Institute on Aging
Description:
An issue that has received a great deal of attention in recent years is the relative importance of age and time to death in determining health care costs. It is well-known that health care costs rise sharply at the end of life, especially in the year before death. This has led many to the view that population aging, insofar as it is a consequence of increased life expectancy, might not produce proportionate increases in Medicare costs, because the costs associated with the extra years lived may simply be experienced later in life. However, all studies of the age-versus-time-to-death question have used a model in which, conditional upon measured attributes, there is a single population-level pathway of expected Medicare costs. Yet there are strong reasons to suppose that there are several prototypical mean pathways of Medicare costs as individuals approach death. We will build a short-term Medicare-cost forecasting model that recognizes the existence of a set of distinctive, but unobserved, cost-trajectory types. Specifically, this project will:
- Estimate an integrated model of mortality and Medicare costs—both in the aggregate and within major categories such as inpatient care, outpatient care, SNF usage and so on—using a generalization of latent-class trajectory models and applying longitudinal data from the National Long Term Care Survey and the Health and Retirement Study, both of which have been linked to continuous Medicare claims records; and
- Produce new estimates of the relative effects of age and time to death on health care costs, while controlling for diagnoses and selected service-use indicators.