宁夏南部山区农村中老年患者住院费用影响因素分析

Analysis of Factors Influencing Hospitalization Costs of Middle-aged and Elderly Patients in Rural Areas of the Southern Mountainous Regions of Ningxia

目的:了解宁夏南部山区农村中老年患者住院费用影响因素,为合理控制医疗费用、提高宁夏南部山区农村中老年患者健康水平提出针对性建议。方法:对宁夏南部4个县抽取的53个乡镇、208个村的1802名山区农村中老年患者进行问卷调查,采用分位数回归和CHAID(Chi-squared automatic interaction detector)决策树模型分析住院费用的影响因素并建立费用分组预测模型。结果:分位数回归显示职业(β_(4)=1400.000,P4<0.05)、是否手术(β_(1)=-434.286,P1<0.001;β_(3)=-828.571,P<0.001;β_(4)=-4000.000,P<0.001;β_(5)=-12200.000,P<0.001)、缴费方式(β_(1)=-157.857,P1<0.001;β_(3)=485.300,P<0.001;β_(4)=1300.000,P<0.001;β_(5)=1866.667,P<0.100)、住院天数(β_(1)=509.286,P1<0.001;β_(2)=393.333,P<0.001;β_(3)=585.714,P<0.001;β_(4)=2800.000,P<0.001;β_(5)=7700.000,P<0.001)在住院费用的不同分位数点产生影响。CHAID决策树住院费用分组模型中,是否手术、缴费方式、住院天数是最主要的影响因素,不同决策树组合的费用差异明显。结论:职业、是否手术、缴费方式、住院天数是宁夏南部山区农村中老年患者住院费用的主要影响因素,在分位数回归模型的基础上,结合CHAID决策树模型探究住院费用影响因素更为全面,能清晰反映不同组合住院费用的高低。

Objective:To understand the factors influencing the hospitalization costs of middle-aged and elderly patients in rural areas of the southern mountainous regions of Ningxia,so as to provide targeted suggestions for rational control of medical costs,and improvement of the health of middle-aged and elderly patients in rural areas of the southern mountainous regions of Ningxia.Methods:A questionnaire survey was conducted on 1802 middle-aged and elderly patients from 53 townships and 208 villages in four counties in southern Ningxia.Quantile regression and CHAID decision tree models were used to analyze the factors influencing hospitalization costs and to develop a cost grouping prediction model.Results:Quantile regressions showed that occupation(β_(4)=1400.000,P 4<0.05),whether or not to have surgery(β_(1)=-434.286,P 1<0.001;β_(3)=-828.571,P<0.001;β_(4)=-4000.000,P<0.001;β_(5)=-12200.000,P<0.001),mode of contribution(β_(1)=-157.857,P 1<0.001;β_(3)=485.300,P<0.001;β_(4)=1300.000,P<0.001;β_(5)=1866.667,P<0.100),days of hospitalization(β_(1)=509.286,P 1<0.001;β_(2)=393.333,P<0.001;β_(3)=585.714,P<0.001;β_(4)=2800.000,P<0.001;β_(5)=7700.000,P<0.001)had an impact on different quantile points of hospital costs.In the CHAID decision tree inpatient cost grouping model,the presence or absence of surgery,payment method,and the number of days in the hospital were the most significant influencing factors,with significant cost differences between different decision tree combinations.Conclusion:Occupation,whether or not to have surgery,payment methods,and number of days in the hospital are the main influencing factors of hospitalization costs for middle-aged and elderly patients in rural areas of the southern mountainous regions of Ningxia.Based on the quantile regression model,the CHAID decision tree model is combined to explore the influencing factors of hospitalization costs more comprehensively,which can clearly reflect the high and low hospitalization costs of different combinations.