基于K-means算法的病种成本聚类分析及精细化管理探究  

Study on Clustering Analysis and Refined Management of Disease Cost based on K-means Algorithm

作　　者：刘嘉慧 张萍[1] 曹瑾音[1] 张芷菁 LIU Jiahui;ZHANG Ping;CAO Jinyin;ZHANG Zhijing(The First Affiliated Hospital of Sun Yat-sen University,Guangzhou Guangdong 510080,China;不详)

机构地区：[1]中山大学附属第一医院,广东广州510080

出　　处：《卫生经济研究》2024年第8期37-40,44,共5页

基　　金：广东省卫生经济学会课题“医保支付改革下基于KMeans算法的病种分析模型及政策建议研究”(2024-WJZD-10)。

摘　　要：目的:探索构建基于“投入-产出-频次”三个维度的病种成本分析模型,完善以病种为基础、以价值为导向的医院运营管理体系,助力精细化管理水平提升。方法:以某三甲医院为样本,基于某年度病案首页数据,运用K-means算法对病种进行聚类分析。结果:将病种聚类为六类,从整体来看样本医院病种结构较好,但基层病种占比仍较大。结论:医院可以对病种进行精细化分类管理,对“重要价值病种”给予资源倾斜,对“一般价值病种”优化诊疗流程和模式,同时开展病种成本监测及分析,控制成本。Objective To construct a disease cost analysis model based on the three dimensions of"input-output-frequency",to improve the disease-based and value-oriented operation management system for hospitals,and to help improve the level of refined management.Methods Taking a tertiary hospital as a sample,based on the data of the first page from medical records in a certain year,the K-means algorithm was applied to analyze the clustering of disease types.Results The disease types were clustered into six categories.From the overall point of view,the sample hospital had a better disease type structure,but the proportion of low-value disease types was still larger.Conclusion It is necessary to refine the classification and management of disease types in hospitals,give resources to"high-value disease types",optimize the diagnosis and treatment process and mode of the"general-value disease types",and carry out the monitoring and analysis of disease cost to control the cost.

关 键 词：聚类分析 病种成本 精细化管理 

分 类 号：R197[医药卫生—卫生事业管理]