DRG subgrouping of ischemic stroke inpatients based on E-CHAID algorithm
10.16753/j.cnki.1008-2344.2025.04.010
- VernacularTitle:基于E-CHAID算法的缺血性脑卒中住院患者DRG细分组研究
- Author:
Xinlei FANG
1
;
Mingzhi HUANG
1
Author Information
1. 蚌埠医科大学第一附属医院运营管理部,安徽 蚌埠 233000
- Publication Type:Journal Article
- Keywords:
ischemic stroke;
diagnosis related groups(DRG);
decision tree
- From:
Journal of Shenyang Medical College
2025;27(4):391-395
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the influencing factors of hospitalization costs in patients with ischemic stroke,design a diagnosis related groups(DRG)subgrouping scheme,providing references for refined DRG subgrouping.Methods:Ischemic stroke patients discharged from a tertiary comprehensive hospital in Bengbu,Anhui Province from Jan 2021 to Dec 2023 were included.Through the rank-sum test and multiple linear regression analysis,the main factors influencing the hospitalization cost were analyzed.The E-CHAID algorithm was applied to construct a DRG subgrouping scheme,and the variance reduction and coefficient of variation were used to validate subgroup rationality.Results:The length of hospital stay,presence of surgical procedures,severity,and admission route were included in the decision tree model,generating 12 DRG subgroups.Both variance reduction and coefficient of variation confirmed high subgroup validity.Conclusion:The E-CHAID-based DRG subgrouping scheme for ischemic stroke inpatients is scientific and reasonable,supporting future adjustments in treatment-related grouping,hierarchical medical practices,and the fine hospital management.
