The AIDS epidemic will impose substantial social, political and economic burdens on the U.S. over the foreseeable future. Many have expressed interest in discovering what the potential impact on federal and state health care budgets will be if AIDS (and AIDS-related Complex--ARC) continue unchecked. Scientific knowledge concerning all aspects of AIDS epidemiology, diagnosis and treatment protocols is changing so rapidly that existing cost estimates need updating. Treatment protocols involving both investigational therapies and modifications in the use of AZT for AIDS patients and asymptomatic seropositive individuals are under active study. Many clinical experts think that substantial additional substitution of outpatient and home care for inpatient care can significantly lower the cost-per-case. Estimates of the infected population, the HIV viral transmission rates, the proportion of seropositive cases progressing to ARC and AIDS, and the mean length of viral incubation and disease survival are subject to almost continual revision. In the past eight months, we have carried out a projection of AIDS and ARC illness costs (see Appendix B). The methodological approach that we took, and which we plan to refine and validate with the proposed research, was to develop AIDS/ARC lifetime cost-per-patient estimates by constructing medical diagnosis and treatment decision tree algorithms for all major AIDS/ARC medical complications. National cost estimates were then developed by combining out epidemiologic projections on HIV seroprevalence and rates of progression to AIDS and ARC with per-patient cost estimates. The California Department of Health Services will help us to refine and extend this work by providing data from their Medi-Cal AIDS patient data base for statistical evaluation of patient care services and disease progression (see Appendix A). We intend to further validate the methodology by incorporating diagnosis/treatment modifications suggested by two physician consultants in other regions with substantial AIDS prevalence (e.g. New York and Miami or Texas) to reflect regional differences in patient mix and treatment practice. The proposed research will place computerized spreadsheets of all model calculations in the public domain. As new information become available, we, and other researchers, will be able to easily make changes or modifications in the model.
