基于决策树模型的中医优势病种病例组合研究——以脑梗死为例

Study on Case Mix of Dominant Diseases in Traditional Chinese Medicine Based on Decision Tree--Taking Cerebral Infarction as an Example

目的:以脑梗死住院患者为例,对中医类医院中医优势病种的病例组合方案进行探索,为未来中医优势病种更好地参与按病种付费提供数据参考。方法:提取2019年北京市若干家中医类医院9055例脑梗死住院患者病案首页数据。通过多元线性回归确定分节点变量,纳入决策树模型进行病例组合并测算分组后的标准费用、病种权重等。结果:以性别、年龄、住院天数、耗材费、医院类型和医院级别等关键因素作为住院费用分节点变量纳入决策树模型,形成19个病例组合,总体分组较为合理。结论:采取措施控制影响脑梗死患者住院费用的关键因素,重点关注超标费用,制定标准费用、费用上限、超标费用、病种权重,积极探讨中医优势病种病例组合方案,为其参与按病种付费提供数据参考。

Objective:Taking inpatients with cerebral infarction as an example,the case mix scheme of TCM dominant diseases in TCM hospitals was explored to provide the data reference for better participation of TCM dominant diseases in the payment by disease in the future.Methods:The first page data of 9055 cases of inpatients with cerebral infarction in several TCM hospitals in Beijing in 2019 were extracted.The sub-node variables were determined by multiple linear regression,and incorporated into a decision tree model for case mix and calculation of the standard costs and disease weights after grouping.Results:Key factors such as gender,age,days of hospitalization,cost of consumables,hospital type and hospital level were incorporated into the decision tree model as inpatient cost sub-node variables,resulting in 19 case mixes with a more reasonable overall grouping.Conclusion:Hospitals should take measures to control the key factors affecting the hospitalization cost of patients with cerebral infarction,focus on the excess cost,develop standard cost,cost ceiling,excess cost,and disease weight,and actively explore the case mix scheme of TCM dominant diseases to provide the data reference for participation in the payment by disease.