Sandia National Laboratories and Kaiser Permanente (Southern California Region) are teaming to develop a prototype computer-based model for analyzing health care delivery systems. The ultimate goal is produce a model and a methodology that can be extended to cover all the essential aspects of a large health care organization and to be able to predict the effect of medical and administrative policies on the performance of that organization. The initial two-year feasibility study will model a subset of Kaiser Permanente's operations to understand the critical technical and administrative issues.
Health care has been identified as a major national concern for economic competitiveness. Although as an industry, health care is nearing revenues of one trillion dollars per year, the use of information management technologies to understand and control these costs is in its infancy. Sandia and Kaiser Permanente will combine the technologies of advanced information modeling, object oriented software design, distributed computing (both LAN and WAN), and expert database systems into an integrated health care management tool. The tool will provide top down simulation of processes and procedures along with their outcomes (both at the patient level and aggregated across all patients) and costs. For example, the effect of a change in the preferred practice for cholesterol screening will effect the distribution of outcomes for individual patients, the need for screening facilities of the health care provider, and the need for other facilities and resources for the provider based on results (or lack thereof) of the screening. Information modeling and object oriented software design will be used to capture and simulate the health care system. The distributed computer system will be the system already existing and planned by the health care provider that includes both same site and between site communications plus a heterogeneous mix of workstations. The expert database systems will be used to gather and check information residing across this system, both to look for similarities between individual patient cases, aggregate data for epidemiological studies, and identify and study data trends.