基于决策树模型的DIP重症细分组研究——以呼吸衰竭患者为例  

Group Study about Severe Patients of DIP Optimization Based on Decision Tree Model:A Case Study of Patients with Respiratory Failure

作　　者：杨扬[1] 史力群[1] 郑瑞强[1] 林华[1] 於江泉[1] 蔡滨[1] Yang Yang;Shi Liqun;Zheng Ruiqiang(Northern Jiangsu People's Hospital,Yangzhou,Jiangsu,225001,China)

机构地区：[1]苏北人民医院,江苏扬州225001

出　　处：《中国医院管理》2025年第2期57-61,共5页Chinese Hospital Management

基　　金：江苏省医院协会医院管理创新研究课题(JSYGY-3-2023-626);江苏省医院协会加强公立医院绩效考核工作专项研究课题(JSYGY-1-2021-PJ134)。

摘　　要：目的 探索呼吸衰竭重症患者按病种分值付费(Diagnosis-Intervention Packet,DIP)分组方案及费用标准,为优化重症病种细分组及提高医院精细化管理水平提供决策依据。方法 收集扬州市某三级甲等综合医院2023年1—9月主要诊断为呼吸衰竭的299例重症患者病案首页信息,采用单因素分析、多元线性回归分析住院总费用的主要影响因素,采用穷尽卡方自动交互检验决策树模型构建呼吸衰竭重症病例组合方案并进行各组费用测算。结果 住院日、主要手术/操作、离院方式、转归情况是影响呼吸衰竭重症患者住院总费用的重要分类变量,最终形成8个DIP组及相应的标准费用;各组内同质性高,组间异质性强,分组效果较好。结论 基于决策树模型构建的呼吸衰竭重症患者DIP优化分组方案符合临床实际,标准费用能够客观、全面反映医疗资源消耗水平,且不会大幅度增加医保基金支出负担,能够为医保部门优化该病种的细分组方案提供参考,促进医院提升医保精细化管理水平。Objective To explore the Diagnosis-Intervention Packet(DIP) group scheme and expense standards of patients with severe respiratory failure,so as to provide decision-making basis for optimizing the group of intensive diseases and improving the level of fine management of hospitals. Methods The homepage data of medical records of 299 severe patients mainly diagnosed with respiratory failure from January to September in 2023 from a Grade A general hospital in Yangzhou City was collected,and the influencing factors of total hospitalization expenses were analyzed by single factor analysis and multiple linear regression. The exhaustive chi-square automatic interaction detector algorithm for decision tree model was used to establish the case-mix scheme of respiratory failure cases and calculate the expenses of each group. Results Hospitalization day,main surgery/operation,method of leaving hospital,and patient outcomes are important classification variables affecting the total inpatient expense of severe respiratory failure patients,finally formed 8 DIP groups and corresponding expense standards,with strong inter-group homogeneity and great inter-group heterogeneity. Conclusion The optimizing group scheme of patients with severe respiratory failure based on decision tree model is consistent with clinical practice,and expense standard can objectively and comprehensively reflect the level of medical resource consumption,meanwhile,it will not greatly increase the burden of expenditure of medical insurance funds. So it can provide reference for the medical insurance government to optimize the group scheme of this disease,and also promote hospitals to improve the level of medical insurance fine management.

关 键 词：呼吸衰竭 按病种分值付费 决策树模型 

分 类 号：R197.1[医药卫生—卫生事业管理]