摘要
目的:分析山西省运城市某三甲医院冠心病患者住院费用的分布状况及影响因素,为合理调控冠心病患者住院费用提供理论依据。方法:收集山西省运城市某三甲医院2019年冠心病患者的住院信息,应用描述性分析、单因素分析和4种机器学习算法(神经网络、随机森林、支持向量机、logistic回归)分析冠心病患者住院费用的构成情况和影响因素。结果:共纳入2533例冠心病患者,次均住院费用的中位数为10331.77元。住院费用构成中,卫生材料费的占比(68.05%)最高,其次为药品费(13.62%),护理费占比(2.74%)最低。4种机器学习算法中随机森林的准确率最高(71.68%),曲线下面积值为0.7083,随机森林模型中住院天数、科室、付费方式和入院途径对冠心病患者的住院费用影响较大。结论:本研究中冠心病患者住院费用构成较不合理,卫生材料费占比最高,护理费占比最低。随机森林模型的性能优于其他3种机器学习算法。应采取综合的措施控制卫生材料费,优化冠心病患者住院费用构成,以减少无效住院天数为重点,强化对可控因素的管理,降低冠心病患者的疾病负担。
Objective:To analyze the distribution and influencing factors of hospitalization expenses of coronary heart disease patients in a grade-A tertiary hospital in Yuncheng city of Shanxi province,so as to provide theoretical basis for reasonably controlling the hospitalization expenses of coronary heart disease patients.Methods:Based on the information of coronary heart disease patients in a grade-A tertiary hospital in Yuncheng city of Shanxi province in 2019,this study conducted descriptive analysis,single factor analysis and four machine learning algorithms(neural network,random forest,support vector machine and logistic regression)to analyze the composition and influencing factors of hospitalization expenses of coronary heart disease patients.Results:A total of 2533 patients were enrolled.The median of average hospitalization expenses was 10331.77 yuan.In terms of the composition of hospitalization expenses,the proportion of health material expenses(68.05%)was the highest,followed by medicine cost(13.62%),and the proportion of nursing cost(2.74%)was the lowest.Random forest had the highest accuracy(71.68%)among four machine learning algorithms,and the value of area under curve was 0.7083.In random forest model,length of stay in hospital,department,payment method and admission route were the main influencing factors of hospitalization expenses of coronary heart disease patients.Conclusion:The composition of hospitalization expenses of coronary heart disease patients in this study is relatively unreasonable.The proportion of health material expenses is the highest and nursing cost is the lowest.The performance of random forest model is better than the other three machine learning algorithms.It is necessary to take comprehensive measures to control the health material expenses,optimize the composition of hospitalization expenses of coronary heart disease patients,take the reduction of unnecessary length of stay in hospital as a breakthrough and strengthen the control of controllable factors in order to reduce the economic burden of patients with coronary heart disease.
作者
周立业
夏鑫婧
郭志飞
孙梦姣
余红梅
ZHOU Liye(School of Management,Shanxi Medical University,Taiyuan,Shanxi,030001,China)
出处
《医学与社会》
北大核心
2022年第11期116-122,共7页
Medicine and Society
基金
国家自然科学基金资助项目,编号为81973154。
关键词
住院费用
冠心病
影响因素
Hospitalization Expense
Coronary Heart Disease
Influencing Factor
