| Objective Aming at the burden of patients with lung cancer,through the analysis of the basic situation of lung cancer patients,the composition of hospitalization expenses and the influencing factors of hospitalization expenses in J Hospital of Jiaozuo City,this study sums up the relevant influencing factors of hospitalization expenses of lung cancer patients,and puts forward the policy suggestions to reasonably control the hospitalization expenses of lung cancer patients,so as to provide data support and reference basis for the medical management department of Jiaozuo City,to reasonably control the growth of medical expenses and reduce the burden of lung cancer patients.Methods This study obtained the inpatient cost information of lung cancer patients discharged between January 1,2014 and December 31,2018 in J hospital in Jiaozuo City.Descriptive statistics was used to analyze the demographic characteristics and disease characteristics of patients with lung cancer,the average hospitalization expenses and the proportion of each single expense.Grey correlation analysis was used to analyze the correlation between each single expense and the total expense to determine the internal influencing factors of hospitalization expenses.Structural change analysis was used to analyze the changes of each single expense within five years.Single factor analysis is used in the test for non parametric analysis,multiple linear regression model was constructed for multi factor analysis and model validation.Kmeans clustering is used to discretize the hospitalization expenses,then Support Vector Machine model and BP neural network model were established for multi factor analysis to determine the influencing factors of hospitalization expenses,and three models were evaluated and analyzed.Results 2014 cases of lung cancer are included in this study.The results of demographic and disease characteristics analysis shows that lung cancer is mostly found in elderly male patients,most of them are clinical stage III and stage IV;new rural cooperative medical insurance,basic medical insurance for urban residents and basic medical insurance for urban employees are the main insurance;most patients are non-surgical lung cancer patients;the average length of stay is 18.00 days,which is decreasing year by year.The results of inpatient cost analysis show that the inpatient cost is distributed in a skewed way,with the per capita inpatient cost of 13769.31 yuan and increasing year by year,and the drug cost accounts for the largest proportion,accounting for 43.56%of the overall inpatient cost;the gray correlation analysis shows that the drug cost and inpatient cost are the most closely related,which are the main internal influencing factors of inpatient cost,followed by treatment cost,laboratory test cost,inspection cost and The results of structural change analysis show that the structural change of drug cost is the largest and negative,and the material cost,inspection cost and laboratory test cost are positive.Single factor analysis showed that there are statistically significant differences in hospitalization expenses among the eight independent variable groups:gender,age,type of medical insurance,clinical stage,hospitalization days,complications,operation or not,and admission route.Multiple linear regression model analysis shows that the above eight independent variables are all influencing factors of hospitalization expenses,ranking by the importance of the independent variables,the first three are the number of days in hospital,whether or not surgery is the way of admission;model test results show that the modeling effect of multiple linear regression model is not good,which is not suitable for the data analysis of this study.K-means clustering shows that it is the best to aggregate the hospitalization expenses into three categories,among which the first category has the largest number of lung cancer patients,including 1271 cases in total,548 cases in the second category and 195 cases in the third category.According to the analysis of support vector machine model,the above eight independent variables are all influencing factors of hospitalization expenses of lung cancer patients.The influencing factors are in order of importance: hospitalization days(0.473),surgery(0.251),age(0.154),medical insurance type(0.068),complications(0.043),clinical stage(0.030),admission route(0.017)and gender(0.010).The prediction accuracy of the model is 92.00%.According to the analysis of BP neural network model,the above eight independent variables are all influencing factors of hospitalization expenses of lung cancer patients.The influencing factors are in order of importance: hospitalization days(0.371),clinical stage(0.164),operation(0.132),medical insurance type(0.125),complications(0.122),age(0.051),admission route(0.033)and gender(0.029).The prediction accuracy of the model is 90.62%.Conclusion The disease burden of lung cancer patients in this study is relatively heavy,drug cost is the main internal influencing factor of hospitalization cost,hospitalization days,clinical stage,age and operation are the main external influencing factors;the effect of SVM modeling is better than BP neural network model and multiple linear regression model.Jiaozuo should carry out three-level prevention of lung cancer,increase the screening of high-risk groups,optimize the structure of hospitalization expenses,and continue to reduce the proportion of drugs.J hospital should give full play to the advantages of tumor discipline,implement clinical pathway management,promote the construction of multi-disciplinary consultation center of tumor,so as to standardize and refine the clinical treatment of lung cancer patients.Finally,the reform of medical security systems should be enhanced,and gradually exploring gradually the group payment system related to disease diagnosis,in order to control the cost of medical treatment and reduce the burden disease of patients. |
