The rapid aging of the U.S. population, increases in the absolute prevalence of chronic diseases, and the associated rise in the proportion of the GNP expended on medical care all indicate the need for methods to accurately forecast future health care expenditures for specific chronic diseases. Additionally, if these methods are biomedically realistic, they can be used to evaluate the economic implications of specific prevention strategies designed to reduce chronic disease incidence, prevalence, and mortality. Projection strategies that are not biomedically realistic, such as models that assume that risks for demographic subgroups do not change over time (e.g., “static component” models), though possibly accurate over the short run, are not suitable for assessing the long term effects of specific proposed health policy interventions which are designed to alter risks.In this paper we present a strategy for forecasting health care costs which is based on a model that represents the natural history of a chronic disease in terms of a preclinical state, a clinical state, case fatality rates, cures, and the implications of exogenous medical factors. Using this model we project that the treatment costs associated with respiratory cancer in the white male population of the U.S. may undergo a two‐thirds increase in real dollars over the period 1977 to 2000. About one‐half of this increase is due to a demographic shift to an older population structure, with the remainder due to higher respiratory cancer incidence rates in younger cohorts. Alteration of certain parameters of the model to simulate various interventions suggests that about three‐quarters of the cost of this disease could be eliminated, though realization of any significant part of this savings would require a lengthy phase
展开▼
