Research And Design Of Medicare Decision Support System Based On Multilevel Model

Medicare is an important part of social security system. In recent years, with the medicare continues to develop, there are more and more risk factors in the area of medicare, and the relationships among those factors are very complicated. At the same time as the increasing volume of data, it has become an urgent need to solve the problem in the area that how to control effectively risks based on the characteristics of medicare, mine the operational rules of the area to support decision-making process of the leadership, and to promote the development and progress of medicare undertakings.The main work is as follows:The system structure of data is analyzed in medicare. The multilevel decision analysis model for medicare area is proposed, according to the correlation and level structure among data.Based on data characteristics of the medicare area, the definition and build of multilevel model are given, and the method of multilevel model parameter estimation, hypothesis testing and model extending are expounded.Medical institutions and patients are considered to be two different levels of units, using variance component model to analyze the risk factors influence extent on outpatient medical fees. The analysis results are more comprehensive and accurate than traditional regression model based on single level, according to the value of variance, covariance and intra levels correlation between two levels of units. Through an application example, the feasibility of the two-level model in the risk factors predictability of outpatient medical fees is certified.According to multilevel model, medicare decision support system based on multilevel decision analysis model is designed. The system is built on data warehouse, and the level structure of analysis data is described through data mart, and the analysis data is processed by On-Line Analytical Processing according to the characteristics of multilevel analysis, and the analysis results of multilevel decision analysis supports comprehensive, accurate and reliable help for decision-making process. Meanwhile, according to the needs of business decision-making process, data statistics, query and report generating are designed to support auxiliary decision-making. Finally, a prototype system based on the proposed model and method is applied in a practical project, and the study results of this dissertation are verified through the project.