【作者】 姚春琛；

【导师】 朱巧明；

【作者基本信息】 苏州大学 ， 软件工程， 2009， 硕士

【摘要】 随着医疗保险改革的不断深入,医保覆盖面的逐步扩大,参保人员迅速增加,医保收费系统如何响应大批量客户并发结算成为一个热点问题。同时医保基金的支出也随之快速增长,因此对医保消费的管理越来越受到各地医保经办机构的重视。但目前国内对于医疗保险基金的管理,主要依靠常规的报表统计功能,缺乏一套系统化的医保基金控制管理系统。本文在研究了基于三层体系结构的苏州工业园区医疗保险收费系统1.0版的基础上,设计与实现收费和管理软件2.0版本。针对1.0版本大批量用户并发划卡时,系统的结算组件无法响应的问题,在2.0版本中引入Tuxedo中间件平台,使用消息队列异步通信和两阶段提交机制保证大量并发请求的响应,并设计医院前置机数据同步机制,当医院和医保中心网络中断时,不影响参保人员划卡结算。同时,为了合理控制大病统筹基金的支出,本文在研究了数据挖掘的相关理论的基础上,通过分析已有医保大病报销记录的历史数据,采用了Pro-Apriori关联分析算法找出病种与使用的药品和诊疗项目之间的联系,建立了病种、药品、诊疗项目关联库。使用关联库和k-means聚类分析算法,对每个医院上年住院病案进行分析,并使用分析结果对医院信用等级进行评分,分成好、中、差三个等级,建立医院信用等级库。最后,本文提出了改进实时划卡结算流程,对信用等级中和差的医院在使用与病种无关的药品和诊疗项目时,先不予以报销。本文设计与实现的医保收费和管理系统2.0版本目前已在苏州工业园区上线运行,取得良好的效果:大批量用户并发划卡时系统结算正常,同时新增的管理方法能合理地控制大病统筹基金的支出,是医保管理系统的一个新尝试,为同类系统的设计和发展提供了理论借鉴和实现方法。更多还原

【Abstract】 With further reformation on medical insurance system, the coverage of medical insurance system is becoming wider and wider. The number of members who are subscribed to this insurance system quickly increased. It became a hot topic on how can a medical insurance charging system respond to big numbers of concurrent settlements directed from client-ends. In the meantime, the expenditure of Medical Insurance Fund grows fast. As a result, the management of medical insurance is being emphasized by related organizations in many cities. However, the management of Medical Insurance Fund still relies on statistics provided by regular forms and reports. There’s an absence of a systematic Medical Insurance Fund management and control system.The following thesis discussed the design and implementation of version 2.0 of Suzhou Industrial Park Medical Insurance Charging and Management System, which is based on a three-layer structure of the version 1.0. By introducing the Tuxedo middleware platform, using message queue asynchronous communication and two-stage submit scheme to ensure response to large amount of concurrent requests, the version 2.0 tackles the problem of settlement module fail to response to big amount of concurrent transactions in version 1.0. A scheme of data synchronizing of hospital front end processor is designed to assure in the occasion of network disconnection between hospital and medical insurance center, the settlement transaction of medical insurance members won’t be affected.In the meantime, to reasonably control the expenditure on major illness fund, based on study on data mining related theory, after analysis on history data of medical insurance reimbursement on major illness claim, Pro-Apriori related algorithm was used to find out correlations between illness type, medicine and treatment used, and built a correlated database on illness type, medicine and treatment. Using correlation database and K-means algorithm to analyze every hospital’s inpatient cases in the previous year, and use the analysis result to rate hospital’s credibility. Divide them into good, middle and bad respectively and build the hospital credibility database. In the end, the thesis provided an improved settlement transaction procedure, and system will not respond to claims from hospitals with middle or low credibility, in the cases that they try to use medicines or treatments which are irrelevant to illness identified.The Medical Insurance Charging and Management 2.0 Version discussed and implemented in this thesis is now operating in Suzhou Industrial Park. The result is satisfactory, the system maintained normal operation when facing concurrent transactions from big number of users. In the meantime, new management approach added to the system can reasonably control expenditure from major illness fund, which is a new experiment in medical insurance management, and provided not only theory reference but also detail implementation to design and development on similar system.更多还原