Analysis On The Influence Factors Of Hospitalization Cost Of COPD Patients In A Tertiary Hospital Of Hefei

BACKGROUND: Chronic obstructive pulmonary disease(COPD)is a common chronic respiratory disease,which characterized by airflow limitation.Recently,the global burden of disease data show that COPD has become an important public health problems in China and the global scope and caused widespread concern.Since the high prevalence and long progression,the hospitalization cost of COPD patients often bring heavy financial burden to families and society.OBJECTIVE: To explore the constitution and main influencing factors of hospitalization cost in COPD patients,which may provide strategy to control the hospitalization cost and reduce the economic burden of families and society.METHODS: This study based on the hospitalization cost data of COPD patients in the information management system of a tertiary hospital of Hefei.We also get the influence factors which may influence the hospitalization cost.The total hospitalization cost was divided into drug fees,inspection fees,laboratory fees,materials fees,treatment fees and care fees.The constitution of hospitalization cost was analyzed by the composition ratio.Nine factors such as age,sex,region,season,length of stay,department,complication,surgery and oxygen therapy were analyzed by univariate analysis.Then multiple linear regression model analyzed the main influencing factors and constructs hospitalization cost prediction model for COPD patients.Moreover,the artificial neural network model was used to analysis the influence factors.We divided the database into training set and test set according to the ratio of 7: 3,and used artificial neural network model to predict the hospitalization cost of COPD patients.RESULTS: A total of 1022 patients data was included in this study.Drug fees was the main constitution of hospitalization cost,accounting for 47.84% of the total cost.Univariate analysis indicated that eight factors such as age,sex,region,length of stay,department,complication,surgery and oxygen therapy were related to the hospitalization cost.Multivariate analysis indicated that seven factors such as age,sex,length of stay,department,complication,surgery and oxygen therapy were related to the hospitalization cost,and constructed linear regression model using those factors.The coefficient of the model is 0.658.The order of importance is length of stay,oxygen therapy,age,complication,surgery therapy,sex,and department in linear regression model.In artificial neural network model,the order of importance is length of stay,surgery therapy,department,oxygen therapy,age,complication,season,sex,region.The coefficient of the artificial neural network model is 0.885 and 0.810 in training set and test set,respectively.The coefficient of the artificial neural network model is higher than the linear regression model,suggesting that the artificial neural network model is better than the linear regression model in predicting the hospitalization cost of COPD patients.CONCULTION: Firstly,Drug fee is the main constitution of hospitalization cost of COPD patients in the tertiary hospital of Hefei in 2015.Control of drug fee is important for control the hospitalization cost of COPD patients.Secondly,both the multiple linear regression model and the artificial neural network model identified length of stay is the most important factor for the hospitalization cost of COPD patients.Moreover,the influence factors such as surgery therapy,department,oxygen therapy,age,complication,season,sex,region also affect and can predict the hospitalization cost of COPD patients.Therefore,controlling drug fees and length of stay are important for control the hospitalization cost of COPD patients.