节点文献
基于回归树模型的肝癌住院费用病例组合研究
Case-Mix Study of Patients with Liver Cancer Based on Decision Tree
【摘要】 目的探索适合南通市肝癌患者的病例诊断组合相关组(DRGs)及费用标准,为相关部门决策提供依据。方法用多重线性回归方法对2006-2013年以肝癌为主要诊断的6882例病案数据进行分析,筛选分类节点变量,运用决策树模型进行病例分组,用绝对值法和相对权重法对住院费用的标准进行测算。结果治疗方法、入院途径、伴随病、转归、并发症、首次入院、性别等作为肝癌病例分类节点的关键变量,共形成17个DRGs组合和相对应住院费用标准。结论病例分组结果合理,住院费用标准和病种权重可作为相关部门参考。
【Abstract】 Objective The study explored the case-mix diagnosis-related-groups of patients with liver cancer in Nantong and developed related health expense standards,so as to provide evidence for related departments,decisions. Methods According to 6882 inpatients w ith the principal diagnosis liver cancer from 2006 to 2013 in Nantong tumor hospital,multiple regression model w as employed to screening the classification node variables,and the regression tree model w as applied to set up the casemix diagnosis-related-groups. The absolute value method and relative w eights w ere used to calculate the standard. Results The main classification node variables w ere therapies,admission from emergency setting,prognosis,concomitant diseases,complication,first hospitalization,and gender. 17 diagnosis-related-groups and related health expense standards w ere established. Conclusion Case-mix diagnosis-related-groups are reasonable,and the health expense standards and related case w eight could make references for related departments.
【Key words】 Diagnosis-related-groups; Regression tree; Liver cancer; Hospitalization expense;
- 【文献出处】 中国卫生统计 ,Chinese Journal of Health Statistics , 编辑部邮箱 ,2017年05期
- 【分类号】R197.323;R735.7
- 【被引频次】11
- 【下载频次】374