基于混合效应模型的医保费用测算及监控  

The Assignment and Monitoring of Medical Insurance Cost Based on Mixed-Effect Model

作　　者：李好奇[1] 林华珍[1] 张兴凤[1] 朱玉峰[1] 张伟[2] 

机构地区：[1]西南财经大学统计学院统计研究中心,四川成都611130 [2]四川大学华西生物统计及成本效益研究中心,四川成都610041

出　　处：《数理统计与管理》2015年第6期1040-1047,共8页Journal of Applied Statistics and Management

基　　金：国家杰出青年科学基金(11125104);四川省教育厅2014年度创新团队(14TD0046);中央高校基本科研业务费专项资金(JBK141111;JBK141121)资助

摘　　要：近几年医改的一个核心内容就是医保支付方式的改革。2012年12月,人力资源和社会保障部、财政部、卫生部三部门联合出台了《关于开展基本医疗保险付费总额控制的意见》,提出在未来两年里,在所有医疗保险统筹地区实行总额预付工作。实行总额预付制的关键是如何科学合理的测算每家医院的预算总额。本文利用线性混合效应模型分别对某市684家医院的病人数,平均费用进行建模,给出每一家医院来年医保费用的合理参考,并且利用模型中的随机效应项自动识别医保费用和病人数异常的医院,为医保监管机构的监管提供科学依据。In recent years, the core content of health reform is the reform of the medical insurance payment mode. In December 2012, HRSS, Ministry of Finance and Ministry of Health, the three depart- ments jointly issued that the opinions of implementation for controlling the total cost of basic medical insurance under global budget, and put forward that the global budget should be done in all medical insurance area in the next two years. The key of global budget is how to measurement total budget of each hospital reasonably. In this paper, we made detailed analysis with the data of medical insurance for urban workers. We use Linear Mixed-Effect Model to model the number of patients and average cost of inpatients. Our model has good fitting ability and applicability. We can give a reasonable reference to next year＇s budget, and provide scientific basis for health regulators.

关 键 词：总额预付 医保费用 随机效应 

分 类 号：O212.1[理学—概率论与数理统计]