用循环系统病例探讨疾病诊断相关组统计学分组过程

目的:探讨疾病诊断相关统计学分组过程,寻找影响分组结果和费用水平的因素。方法:基于我国三甲医院7 529例循环系统疾病病案首页资料,运用决策树模型卡方自动交互检测算法寻求分组过程中的规则。结果: 数据的核对、分类变量的选择、主要诊断和手术操作的分类及剔除统计无关项等都影响分组结果和费用水平,但各因素的影响程度大小和范围不一样。结论:适宜的主要诊断分类和手术操作分类是构建DRGs的基础:分类变量宜选择与疾病严重程度有关的变量;制定费用标准时宜剔除统计无关项。

A Discuss on the Statistical Grouping Procedure for DRGs: Use Diseases of Circulatory System for Example

Purpose: To discuss the statistical grouping procedure for DRGs, and find out the influecing factors of the grouping and cost level. Method: This study bases on the data from the medical records of 7529 cardiovascular inpatients from certain three-grade hospital and use decision-making tree model to find out the rules in the grouping. Results: Datac heck, selection of grouping variables, main diagnosis and operation group, deletion of statistical insignificant items would influence the grouping results and cost level, but the influencing degree and scope of each variable is different. Conclusion: Appropriate main diagnosis grouping and operation grouping are the bases for DRGs construction; the grouping variables should be those related to the severity of disease; the statistical insignificant items should be deleted when making the standard of cost.