|
| 基于决策树模型HC25病组DRGs细分组及住院费用的影响因素研究 |
| Study on the DRGs sub-grouping of patients with HC25 disease group and the influencing factors of hospitalization expenses based on decision tree model |
| 投稿时间：2023-04-12 |
| DOI：10.3969/j.issn.1000-0399.2024.04.020 |
| 中文关键词: 胆囊切除术 决策树 疾病诊断相关分组 住院费用 |
| 英文关键词: Cholecystectomy Decision tree Disease diagnosis related grouping Hospitalization expenses |
| 基金项目:陕西省创新能力支撑计划(编号:2022KRM194) |
|
| 摘要点击次数: 623 |
| 全文下载次数: 490 |
| 中文摘要: |
| 目的 构建基于决策树模型HC25病组患者DRGs细分组的研究及住院费用的影响因素。方法 收集西安交通大学第一附属医院2020年1月至2022年6月依据CHS-DRG分组方案进入HC25(胆囊切除术,不伴并发症或合并症)病组患者的病案首页数据(n=3 844),通过单因素和多因素分析了解影响住院费用的主要因素,通过建立决策树模型进行病例组合,分析分组效果及不同分组的费用标准。结果 正态性检验结果显示:住院费用呈偏态分布(χ2=0.197,P<0.001);单因素方差分析结果显示:住院费用在不同年龄、性别、手术方式、患者来源、医疗付费方式、入院病情危重情况、术前时间、住院时间、是否有再住院计划之间,差异有统计学意义(P均<0.05);多重线性逐步回归分析结果显示,影响住院费用显著性差异的因素从大到小依次为:住院时间、手术方式、入院病情危重情况、性别、术前时间、患者来源、再住院计划(P均<0.05);采用CHAID共生成了7个DRGs分组,第一层分类节点变量是手术方式,第二层是术前时间、住院时间,第三层是入院时病情危重情况;不同DRGs组合的住院费用比较,差异有统计学意义(χ2=2 474.595,P<0.001),组合DRGs(1)的住院费用最高,权重最高;DRGs(7)的住院费用最低,权重最小;所有DRGs病组中超标费用患者共170例(4.42%),各组超标费用比例均<10%。结论 基于决策树模型对HC25病组的细分组效果合理有效,能较客观反映医疗资源消耗水平。 |
| 英文摘要: |
| Objective To study the DRGs grouping of patients with HC25 disease and the influencing factors of hospitalization expenses based on decision tree model.Methods From January 2020 to June 2022, collected the first page data of medical records of patients who were admitted to the HC25(cholecystectomy, without complications or complications) group were collected according to the CHS-DRG grouping scheme in a tertiary hospital in the First Affiliated Hospital of Xi'an Jiaotong University(n=3 844). The main factors affecting hospitalization costs were analyzed through univariate and multivariate analysis, and the cases were combined by establishing a decision tree model to analyze the grouping effect and cost standards.Results The results of normality test showed that the hospitalization expenses were skewed(χ2=0.197, P<0.001). Univariate analysis showed that there were significant differences among patients in different ages, genders, surgical methods, patient sources, medical payment methods, severity of admission, preoperative time, length of stay, and whether there was a rehospitalization plan(P<0.05). The results of multiple linear stepwise regression analysis showed that the factors influencing the significant difference of hospitalization expenses from large to small were: length of stay, mode of operation, severity of hospitalization, gender, preoperative time, source of patients, and re-hospitalization plan(P<0.05). A total of 7 DRGs were generated using CHAID, on the first level, the classification node variable was the operation mode, the second level is the preoperative time and hospital stay, and the third level is the critical condition at admission. The hospitalization expenses of different DRGs combinations were statistically different(χ2=2 474.595, P<0.001). The hospitalization expenses and weight of DRGs combination(1) were the highest; DRGs(7) had the lowest hospitalization cost and weight; there were 170 patients(4.42%) in all DRGs disease groups with excessive expenses, and the proportion of excessive expenses in each group was less than 10%.Conclusions Based on the CHAID, the sub-grouping effect of HC25 is reasonable and effective, which can objectively reflect the level of medical resource consumption. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
