PDF下载

[1]薛家鑫丁磊徐晓妍孙谟健张茜李江峰唐笠岷.基于关键因素分析联合K-means聚类模型的DRG病组复分组路径与应用探讨[J].中国卫生质量管理,2026,33(5):40-44.[doi:10.13912/j.cnki.chqm.2026.33.5.08]
　XUE Jiaxin,DING Lei,XU Xiaoyan.Exploration on the Pathway and Application of DRG Disease Group Re-Grouping Based on Key Factor Analysis Combined with K-means Clustering Model[J].Chinese Health Quality Management,2026,33(5):40-44.[doi:10.13912/j.cnki.chqm.2026.33.5.08]

点击复制

基于关键因素分析联合K-means聚类模型的DRG病组复分组路径与应用探讨

《中国卫生质量管理》[ISSN:1006-7515/CN:CN 61-1283/R] 卷: 33 期数: 2026年5期页码: 40-44 栏目: 医疗质量出版日期: 2026-05-15

Title:: Exploration on the Pathway and Application of DRG Disease Group Re-Grouping Based on Key Factor Analysis Combined with K-means Clustering Model

作者:: 薛家鑫丁磊徐晓妍孙谟健张茜李江峰唐笠岷; 青岛大学附属医院

Author(s):: XUE Jiaxin; DING Lei; XU Xiaoyan; The Affiliated Hospital of Qingdao University

关键词:: DRG复分组; 肺恶性肿瘤; K-means聚类模型; 关键因素分析; 精细化管理; 医疗质量

Keywords:: DRG Re-Grouping; Pulmonary Malignancy; K-means Clustering Model; Key Factor Analysis; Refined Management; Medical Quality

分类号:: R197.323R734.2

DOI:: 10.13912/j.cnki.chqm.2026.33.5.08

文献标志码:: A

摘要:: 目的探索关键因素分析联合K-means聚类模型下的DRG复分组路径及其应用，为医院DRG病组精细化管控提供路径范式。方法以某综合三甲医院肺恶性肿瘤胸部大手术组为例，对2 688例病例采用非参数检验及多元逐步线性回归筛选影响住院费用的关键因素，运用手肘法确定最佳复分组数，将不可控关键因素纳入K-means聚类模型构建DRG病组复分组方案，并通过组内变异系数及组间差异性评价复分组合理性，进一步计算各复分组住院费用的波动区间及二次相对权重，总结DRG复分组方案与应用管控策略。结果有无罹患2型糖尿病、合并症或并发症数量、手术或操作数量、有无冠状动脉粥样硬化性心脏病、肺叶切除范围、有无罹患高血压等6个关键因素被纳入K-means聚类模型中，确定最佳复分组数为7个。经复分组后，住院费用组内变异系数渐小、组间差异显著，并形成DRG复分组方案与应用策略路径图。结论所建立的复分组方案有利于DRG病组内部异质性校正，在临床标准路径优化与质量控制、绩效二次分配、住院费用调控等方面具有一定应用价值。

Abstract:: ObjectiveTo explore the DRG re-grouping pathways and their applications under a combined approach of key factor analysis and K-means clustering model, providing a paradigmatic pathway for refined management and control of DRG disease groups in hospitals.MethodsTaking the group of major thoracic surgeries for pulmonary malignancies at a comprehensive tertiary-level hospital as an example, non-parametric tests and multivariate stepwise linear regression were employed to screen key factors influencing hospitalization costs from 2 688 cases. The elbow method was used to determine the optimal number of re-groups. Uncontrollable key influencing factors were incorporated into the K-means clustering model to construct a DRG re-grouping scheme. The rationality of re-grouping was evaluated through intra-group coefficient of variation and inter-group differences. Furthermore, the fluctuation ranges of hospitalization costs and secondary relative weights for each subgroup were calculated, and the DRG re-grouping scheme and application management strategies were summarized.ResultsSix key factors—presence of type 2 diabetes, number of comorbidities (or complications), number of surgeries (or procedures), presence of coronary atherosclerotic heart disease, extent of pulmonary lobectomy, and presence of hypertension—were included in the K-means clustering model, with an optimal number of seven re-groups determined. After re-grouping, the intra-group coefficient of variation for hospitalization costs decreased, while inter-group differences became significant, resulting in a pathway map for the DRG re-grouping scheme and application strategies.Conclusion The established re-grouping scheme facilitates the correction of internal heterogeneity within DRG disease groups and demonstrates practical value in optimizing clinical standard pathways, quality control, secondary performance allocation, and regulating hospitalization costs.

参考文献/References:

［1］危莉, 夏家红, 李浩, 等. 基于公立医院视角的DRG支付方式改革模拟运行结果分析［J］. 中国医院管理, 2022,42(4):75-77.［2］田博韬, 郭筱克, 陈永成. 多源流视域下我国按病种分值付费支付改革的政策动力研究［J］. 中国医院, 2023,27(12):1-4.［3］卢萍萍, 李育梅, 吴俊霞, 等. 基于决策树模型的艾滋病患者DRG优化分组探索［J］. 中国卫生统计, 2025,42(3):340-343.［4］TIAN Y, MA R, ZHAO W, et al. Comprehensive characterization of early-onset lung cancer, in Chinese young adults［J］. Nat Commun, 2025,16(1):1976.［5］王少明, 郑荣寿, 韩冰峰, 等. 2022年中国人群恶性肿瘤发病与死亡年龄特征分析［J］. 中国肿瘤, 2024,33(3):165-174.［6］周鹏飞, 杨孝光, 路强, 等. 恶性肿瘤术后化疗住院患者DRG分组研究——基于E-CHAID算法［J］. 卫生经济研究, 2023,40(6):35-39.［7］吴云, 任臻, 朱怡, 等. 基于决策树模型的恶性肿瘤化疗患者DIP分组研究［J］. 中华医院管理杂志, 2025,41(3):223-228.［8］佟朝霞, 陈宇, 孔晴宇, 等. 北京市医药分开综合改革政策对住院病组结构的影响分析［J］. 中国卫生质量管理, 2018,25(5):25-27.［9］宋菲尔, 毛燕君, 夏秋月, 等. 结核后肺疾病患者患病率和影响因素的Meta分析［J］. 中国防痨杂志, 2025,47(3):322-330.［10］魏永婷, 席祖洋, 田书梅, 等. 癌症治疗相关高血压患者血压管理的证据总结［J］. 中国护理管理, 2024,24(6):886-891.［11］张中花, 章浩然. 基于DRG的安徽某大型三甲医院病案首页质量改进研究［J］. 卫生经济研究, 2020,37(3):29-31.［12］丁涛, 郭冬. 医院信息化医保控费研究［J］. 卫生经济研究, 2021,38(10):77-79.［13］陈颖, 王峰, 孔晴宇, 等. 医改政策对医院住院费用的影响分析［J］. 中国卫生质量管理, 2019,26(3):30-32.［14］张丽敏, 杨建玲, 胡怡, 等. DRGs在医疗质量评价及医院管理中的应用［J］. 中国卫生质量管理, 2019,26(S1):53-57.

更新日期/Last Update: 2026-05-15