基于分位数回归模型的卫生总费用影响因素研究  

The Study of the Influence Factors of the TEH in Shaanxi Province Based on Quantile Regression

作　　者：蔡磊[1] 李艳 连肖[1] 张磊[1] 郭曼[1] 杨学文[1] 王文斌[1] 周威[1] 

机构地区：[1]第四军医大学西京消化病医院,陕西西安710032 [2]西北妇女儿童医院麻醉科,陕西西安710061

出　　处：《现代生物医学进展》2015年第30期5960-5964,共5页Progress in Modern Biomedicine

基　　金：国家自然科学基金项目(81300301);第四军医大学优秀文职人员培养基金项目

摘　　要：目的:利用分位数回归模型对近十年来陕西省卫生总费用增长的影响因素进行研究,从而为卫生行政部门配置卫生总费用以及制定卫生政策提供依据。方法:首先根据已往文献研究以及陕西地区实际情况,挑选出2003-2014年间可能影响陕西省卫生总费用的因素,然后利用逐步回归法检验自变量间的共线性,并对解释变量进行筛选,最后利用R软件对陕西省卫生总费用及其影响因素建立分位数回归模型,确定影响卫生总费用的因素。结果:陕西省国内生产总值(GDP)在10%-90%分位点上的系数检验结果均有统计学意义,且所有分位点的回归系数均为正值。结论:国内生产总值是影响陕西省卫生总费用的主要因素,其与卫生总费用的变化呈正相关,国内生产总值的增长促进了陕西省卫生总费用的增长。Objective： Quantileregression model was used to study the influence factors of total health expenditure（TEH）of Shaanxi province,so as to help the health administrative departmentsto allocatethe TEHand provide evidence for health policy-making.Metheds： First, Based on the published research and the actual situation of Shanxi Province, the factors that may affect the TEH in Shaanxi provincefrom 2003 to 2014 were picked up, then the collinearityamong theindependent variables were tested by using the method of stepwise regression, and the explanatory variable was screened. Finally, the R software was used to establishquantileregression model of total health expenditureand its influencing factors,and determine the impact factors. Results： The test results of coefficient of GDP from 10%-90% fractile aresignificant, and all the quantile regression coefficient is positive. Conclusion： GDP is the main influencing factor of the growth of total health expenditurewith the change of THE waspositively correlated, and the growth of GDP promotes the growth of THEin Shaanxi province.

关 键 词：分位数回归模型 卫生总费用 影响因素 

分 类 号：R197[医药卫生—卫生事业管理]