目的: 探讨对DRG组再细分,为制定更加合理的分组提供参考。方法: 选取湖北省十堰市某三甲综合医院和另一家二级综合医院2019—2023年消化道内镜治疗GJ1、GK1、GK2和GK3组四个ADRG组的病例;采用非参数检验和多重线性回归分析病例费用影响因素,采用决策树模型进行ADRG组的再细分,并将细分结果与国家医保局DRG2.0版分组结果比较;选取“胃和十二指肠息肉”病例,将本研究细分组结果与国家医保局DIP2.0分组结果比较。结果: 消化道内镜治疗病例分为12个DRG细分组,细分后RIV值上升,CV值下降;“胃和十二指肠息肉”病例按本研究细分组与按DIP2.0分组比较,本研究RIV值为0.4610,DIP2.0分组RIV值为0.4725,且后者的CV值普遍高于本研究结果。结论: 对消化道内镜治疗病例,在DRG分组基础上,可以以病例的主要操作类型进行再次细分,细分后组内费用集中,组间费用差异大,与DIP2.0分组方案相比,分组效能相当。本研究的分组策略能够为优化医保支付方式提供思路。

Objective: The paper explores the subdivision of DRG, providing references for the development of more reasonable grouping. Methods: Cases in four ADRG, GJ1, GK1, GK2, and GK3, of gastrointestinal endoscopic treatment in one first-class Grade-A general hospital and one second-class general hospital in Hubei Province were selected from 2019 to 2023. The influencing factors for the cost of cases were analyzed by nonparametric tests and multiple linear regression, the decision tree model was used to subdivide the ADRG, and then the subdivision results was compared with the National Healthcare Security Administration DRG 2.0 grouping results. The cases of "gastric and duodenal polyps" were selected and divided according to the subdivision scheme of this study and the National Healthcare Security Administration DIP 2.0, respectively. The grouping results of the two were compared. Results: The cases of gastrointestinal endoscopy treatment were subdivided into 12 DRG, the RIV increased, and the CV decreased. The results of the subdivision group in this study and the DIP2.0 in terms of the cases of "gastric and duodenal polyps" was compared. The RIV of this study was 0.4610, the RIV of DIP 2.0 was 0.4725, and the CV of the latter was generally higher than that of the results of the present study. Conclusion: For gastrointestinal endoscopy treatment cases, they can be subdivided again by the main operation type of the cases based on DRG. After the subdivision, the difference in expenses is not significant within a group, and there are significant differences in expenses between groups. Compared with DIP 2.0, this subgroup is equally effective. The grouping strategy of this study can provide ideas for optimizing medical insurance payment methods.