基于AP聚类算法的诊断数据DRGs分组研究  

DRGs Grouping Research on Diagnosis Data Based on AP Clustering Algorithm

作　　者：郑艳[1] 曾红莉[1] 廖福翠[1] Zheng Yan;Zeng Hongli;Liao Fucui(Zhongnan Hospital of Wuhan University,Wuhan 430071,China)

机构地区：[1]武汉大学中南医院,湖北武汉430071

出　　处：《医学新知》2019年第5期525-528,共4页New Medicine

摘　　要：以武汉某医院心血管科的病案首页信息为样本诊断数据,采用主成分分析方法确定住院费用的主要影响因素权重,引入近邻传播算法(affinity propagation clustering algorithm,AP)对其实现细分组的分类和划分,并进行同质性和异质性评价.同时,针对诊断数据提出了逐过程分组划分流程及其管理策略,加深了DRGs分组的应用.最终通过AP聚类算法得到10个分组及相应的住院费用标准.经验证,分组具有较好的合理性和可行性,可为住院费用预算、支付、评价提供参考.Data were selected from the first page of medical record from department of cardiology of one hospital in Wuhan.Principal component analysis was used to detennine the weight of main influencing factors of hospitalization expenses.Affinity propagation clustering algorithm(AP)was introduced to classifying and grouping as well as evaluating homogeneity and heterogeneity.At the same time,the step-by-step grouping process and its management strategy according to diagnostic data deepen the application of DRGs grouping.Finally,10 groups and corresponding hospitalization cost standards were obtained by AP clustering algorithm.In conclusion,DRGs grouping based on AP clustering algorithm has good rationality and feasibility and can provide reference for budget,payment and evaluation of hospitalization expenses.

关 键 词：DRGs分组 AP聚类算法 主成分分析法 

分 类 号：R197.1[医药卫生—卫生事业管理]