摘要
目的探寻有效方法分析大病患者住院费用的主要影响因素,为控制住院费用提供合理对策。方法样本选取2016年1月—2017年5月湖北省某市城镇居民基本医疗保险大病患者住院信息,采用K-means聚类和支持向量机进行分析。结果聚类优度检验提示将住院费用分为3类最佳,基于RBF核函数的支持向量机模型的预测准确度最高,住院费用的主要因素为主诊断疾病、住院日、医院级别、医保业务类别和医院类型。结论 K-means聚类与支持向量机模型可作为分析大病患者住院费用的有效方法,为控制住院费用提供策略。
Objective To explore methods by which the main influencing factors of hospitalization expenses of patients with serious diseases can be efficiently analyzed,and provide strategies for hospitalization expenses control.Methods Data of inpatients with Medical Insurance for Urban Residents from Jan.2016 to May.2017 were sampled in a city of Hubei.The data were analyzed by K-means clustering and support vector machine(SVM).Results Clustering goodness test indicated that the hospitalization expenses should be classified into three categories for best result;RBF-based SVM showed the highest prediction accuracy;the main influencing factors of hospitalization expenses included type of disease,length of stay,level of hospital,type of health care and type of hospital.Conclusion The K-means clustering and SVM model are effective in analyzing main influencing factors,according to which reasonable strategies for expenses control could be established.
作者
陈默
蔡苗
黄阿红
沈梦雪
吴其
张泽苗
乐虹
CHEN Mo;CAI Miao;HUANG A-hong(School of Medicine and Health Management,Tongji Medical College,Huazhong University of Science and Technology,Wuhan,Hubei,430030,China)
出处
《中国医院管理》
北大核心
2019年第5期45-47,53,共4页
Chinese Hospital Management
基金
国家自然科学基金重点资助项目(71333005)
国家社会科学基金重大项目(15ZDC037)
关键词
住院费用
聚类
支持向量机
数据挖掘
hospitalization expenditure
clustering
support vector machine
data mining
