基于决策树模型的DRG优化分组探索--以新生儿疾病外科组为例

Group Exploration of DRG Optimization based on the Decision Tree Model--Taking the Surgical Group of Neonatal Diseases as an Example

目的:探索新生儿疾病外科组的DRG优化分组方案。方法:收集某儿童医院新生儿疾病外科组病例数据,采用非参数检验、相关分析和多元线性回归分析住院费用影响因素并筛选分类节点,运用决策树模型构建DRG优化分组方案。结果:是否使用有创呼吸机、是否手术治疗、出生体重、年龄、是否进入重症监护室是影响新生儿疾病外科组患者住院费用的重要分类变量,最终形成10个DRG组及相应标准费用。结论:基于决策树模型构建的新生儿疾病外科组DRG优化分组方案符合临床实际,标准费用可客观反映医疗消耗水平。

Objective To explore the DRG optimization grouping scheme in the surgical group of neonatal diseases. Methods We collected the case data of the neonatal disease surgery group in a children’s hospital. Non-parametric test, correlation analysis and multiple linear regression were used to analyze the factors affecting the hospitalization cost and select the classification nodes.We employed decision tree model to build the DRG optimized grouping scheme. Results Usage of invasive ventilator, surgical treatment, birth weight, age, and admission to the intensive care unit were the important classified variables that affected the hospitalization costs of neonatal diseases. A total of 10 DRG groups and the corresponding standard costs was yielded. Conclusion The DRG optimized grouping scheme of neonatal diseases constructed based on the decision tree model meets the clinical practice, and the standard cost can objectively reflect the level of medical consumption.