Using clinical information to project federal health care spending.

Complications from chronic illnesses often do not emerge for many years. Current federal cost projection methods are constrained by ten-year cost estimates, which capture increases in near-term intervention costs but not changes in long-term costs. Current methods also cannot easily capture the cost implications of changes in disease progression. Type 2 diabetes is a prime example of a chronic illness with long-term health and cost consequences. We present results from an epidemiologically based model that projects federal costs for diabetes under alternative policies, and we discuss the potential changes in the federal budget process needed to capture the full impact of these interventions.