Study On The Mode Of Payment Of Medical Expenses Based On Diagnosis Related Groups (DRGs) Model

Objective using patients with cerebral hemorrhage as the sample data from a third grade first-class medical institutions in Hengyang City, try to find the diagnosis related group of models for cerebral hemorrhage patients with hospitalization expenses and related factors, considering whether the surgery, hospital days and other factors, hospitalized patients were divided into several diagnosis related groups based on hospitalized patients’ diagnosis,, were approved cost statistics of each group, realize the scientific quantification of medical resource consumption, to provide valuable reference for the government to develop the mode of medical insurance payment policy.Method using a combination of qualitative and quantitative analysis, through literature research, statistics, application of Office2010 software and SPSS20.0 software processing the corresponding data, the establishment of case database, analyzing descriptively the general situation of data descriptive analysis, analyzing of the composition ratio of different costs of hospitalization in mining; applying BP neural network model to analyze the factors of inpatients cost impact, to investigate the demographic characteristics, clinical features of patients with variable, medical consumption variables influencing factors of hospitalization expenses of patients; and on this basis, formulation of the hospitalization expense standard interval case mix and patients using decision tree model.Results statistical analysis showed that the composition of hospitalization expenses, expenses for medicine overrepresented(40.29%), the doctor of technology capability project fee is low;the factors affecting hospitalization expenses of hospitalization expenses in total degree of influence is the largest hospital stay(0.338), whether the operation(0.222), age(0.120), the discharge mode(0.093) and sex(0.015) impact on the total cost of hospitalization is minimum;according to diagnosis related groups(DRGs) methods 442 eligible cases were divided into five groups, making each disease hospitalization expenses standard interval.Conclusion This study showed that the surgical, hospitalization days and so were the important factors of establishing the diagnosis related group for cerebral hemorrhage through the neural network model; finally cerebral hemorrhage patients were divided into 5 diagnostic groups using the decision tree model, according to the important degree of the operation, the length of hospital stay.