Study On Influencing Factors Of Hospitalization Expenses Of Patients With

ObjectiveThe data of major illness hospitalized patients in the medical institutions in Wuzhou City and four counties in 2016 were used to explore the application of the mixed linear model in the analysis of the influencing factors of hospitalization costs and to provide new analytical means for the analysis of structural health care data.Analyze the expenses of hospitalized patients with major illnesses,understand the influencing factors of hospitalization costs for patients with major illnesses in Wuzhou,and provide guidance for the cost control of inpatients with major illnesses in Wuzhou.MethodsThrough the hospital management system,data on all inpatients of 9 medical institutions in the municipal districts and four county-level administrative divisions of the city from January 1 to December 31,2016 were extracted.The contents of the data include: hospital name,hospital administrative division code,medical institution economic type,medical insurance payment method,hospitalized patient's gender,age,ethnicity,number of hospitalizations,actual length of hospital stay,current address,family register address,discharge diagnosis code,other Diagnosis of the disease code,whether the surgery,the total cost of hospitalization,and the amount of patient self-pay.Use IBM SPSS Statistics 23.0 to cleanse and process data,exclude patients whose hospitalization cost is <1000 yuan or hospital stay <3 days,determine the disease spectrum by ICD-10 code,determine whether the patient has comorbidities or complications based on other diagnoses,and use The map ABC API assists in determining whether it is remote patient data.The original hospitalization cost description uses a median(M)and interquartile range(Q).Non-parametric tests are performed on the differences between the original hospitalization expense groups for different social factors and patient factors.Perform normality test and logarithmic transformation on the original hospitalization costs,determine the fixed effect variables and random effect variables among the alternative influencing factors,establish the hierarchical structure of the zero model test data,and establish a mixed effect model to analyze the medical treatment of inpatients with severe illness in Wuzhou.Influencing factors of costs,and comparing the mixed effect model fitting with multiple linear regression models.Results(1)The basic situation of the subjects: There were 13,987 cases that met the inclusion criteria,and the average age was 52.12 years(0 to 101 years old).The average actual length of stay was 15 days(3 to 364 days).The average number of hospitalizations was 1.4.The average pay ratio is 88.9%.The gender ratio for male and female patients was 1.58.The Han nationality accounted for 99.5%,Zhuang nationality accounted for 0.02%,other ethnic groups accounted for 0.03%.82.5% of patients had complications or complications.Surgical patients accounted for 54.6%.46.7% of patients were off-site patients.New rural cooperative medical institutions accounted for 83.8% of the total,medical insurance for urban employees accounted for 11.8%,medical insurance for urban residents accounted for 3.2%,and business and other insurance accounted for 1.2%.Among the nine medical institutions,there are 8 public hospitals,1 private hospital,8 general hospitals,and 1 specialist hospital.There are 14 diseases,including circulatory diseases,tumors,damage poisoning,urogenital diseases,and digestive diseases.(2)Main factors influencing the hospitalization cost of major illnesses in Wuzhou: There were statistically significant differences between social administrative divisions,hospital economic types,types of medical institutions,patient age stratification,gender,presence or absence of comorbidities or complications,whether or not surgery,whether it is an off-site source,and the actual duration of hospitalization,hospitalization expenses,hospitalization accumulative times,insurance payment methods,and disease categories.(3)Mixed-linear model fitting results: In fixed effects,city area/county factors,medical insurance payment methods,gender,presence or absence of comorbidities or complications,surgical conditions,off-site sources,and circulatory system diseases in disease classification are the influencing factors of hospitalization expenses for patients with major illnesses.In random effects,age,actual time of hospitalization,and self-pay ratio are the influencing factors for hospitalization costs of patients with major illnesses.(4)Comparison of mixed linear models with multiple linear regression models: Mixed linear models have better residual independence.Conclusions(1)Administrative divisions,medical insurance payment methods,gender,presence or absence of comorbidities or complications,surgical conditions,off-site sources,disease classification,age,actual hospitalization time,and self-pay ratio are all influencing factors for the cost of hospitalization for major illness patients.(2)The mixed linear model can solve the problem that the traditional statistical model is inapplicable to hierarchically aggregated medical data,and retains the statistical information in the original data to the greatest extent,which can be better applied to the hierarchical structure in medical and health data.