Exploratory Research On The Prediction Model Of Inpatient Expenses In Lanzhou Nursing Home Based On RUG

ObjectiveThis study investigates the hospitalization costs of nursing homes in Lanzhou,analyzes the factors that affect the hospitalization costs of nursing homes,and refers to the foreign RUG payment system to establish a model for predicting the hospitalization costs of patients in nursing homes.The foundation also provides a reference for the rational use and distribution of medical resources.MethodsThis study is a cross-sectional survey method.Using the convenience sampling method,357 patients from two nursing homes in Lanzhou were selected as the total research object from April 2019 to November 2019.271 patients were randomly selected as the data for model establishment,and 86 patients were used as data for testing the feasibility and accuracy of the model.And using questionnaire survey method,the questionnaire content includes the patient's general information,Activity of Daily Living(ADL)grading level,Mini-mental State Examination Scale Level(MMSE)and cost information.Using IBM SPSS Statistics 22.0 for descriptive statistics,non-parametric test,stepwise multiple linear regression and other analyses;Using IBM SPSS Modeler 18.0 for Chi-squared Auto-matic Interaction Detector(CHAID)analysis.Results1.The results of a single factor analysis of hospitalization costs of patients in nursing homes showed that age,marital status,indwelling catheter status,wound treatment status,acute symptoms status,nasal feeding status,rehabilitation status,the Mini-mental State Examination Scale Level(MMSE)and the Level of Activity of Daily Living(ADL)had an impact on the patient's hospitalization costs(P<0.05).State Examination Scale Level(MMSE),the Level of Activity of Daily Living(ADL),and indwelling urinary catheters.3.According to the results of multi-factor analysis and combined with the prediction results of important factors of the Decision Tree,the important influencing factors will eventually be the Activity of Daily Living(ADL),the Mini-mental State Examination Scale Level(MMSE),rehabilitation status and nasal feeding status were included in the Decision Tree as the predictor variables of the grouping model,the hospitalization cost is the target variable,and through the Decision Tree's analysis,10 groups of resource utilization groupings were finally formed.The evaluation of the resulting model showed that the maximum CV value was 4.3% and the minimum was 1.1%,indicating that the internal homogeneity was good;the analysis of variance showed that the differences between the 10 groups were statistically cant(P<0.05)and the grouping was reasonable;the prediction accuracy of the new data model was 97.50%,indicating that the prediction ability was good.ConclusionWith reference to the foreign RUG payment system,combined with the degree of resource utilization of patients in Lanzhou Nursing Homes,the nursing home can be adjusted through the four variables of Activity of Daily Living(ADL)level,the Mini-mental State Examination Scale Level(MMSE),rehabilitation status and nasal feeding status.Patients were effectively divided into 10 groups.The use of Decision Trees to establish a Resource Utilization Group(RUG)cost prediction model can provide classification ideas and methods for the cost of patients in nursing homes;and lay the foundation for relevant research on the establishment of nursing home charging standards,and nursing homes can refer to this way of thinking,improving the existing payment system of nursing homes in the future,allocating and using medical resources reasonably.