Study On Hospitalization Expenditure And Its Determinants In A County Level Hospital In Hunan Province

Objectives: Based on the dataset draw from the hospital information system for Hospitalization Expenditure in a County level hospital in Hunan province, the objectives of this study are to describe the distribution status of hospitalization expenditure for inpatients and to analyze the determinants of the hospitalization expenditure in this hospital, and to provide the evidence on the rational and effectual utilization of medical resources and on taking a measure of controlling the excessive growth of hospitalization expenditure.Methods: Hospitalization expenditure recorded in hospital information system for inpatients at a county hospital in year of 2008 was selected as a subject and data of this study. The contents of the study included socio-demographic characteristics of the inpatients, total hospitalization expenditure and the distribution feature of hospitalization expenditure. The statistic such as median, mean, standard deviation and quartile range were used to statistically describe the central tendency and dispersion tendency of the hospitalization expenditure per time respectively, and proportion and percent bar graph were used to describe the distribution of the hospitalization expenditure. The difference of the hospitalization expenditure within two groups was analyzed by Mann-Whitney U test and the differences among three and more groups with Kruskal-Willas H test, and multiple comparison was conducted with ranked data transferred from the original data. Taking the log value of the total hospitalization expenditure, medicine expenditure, diagnosing and treating expenditure and inspection expenditure as the dependent variables, the multiple linear regression model was performed to analyze the determinant of the medical expenditure, respectively, and this model were used in analyzing the influential factors of the hospitalization expenditure for inpatients with coronary heart disease, diabetes and pneumonia. All statistical methods were performed with SPSS version 15.0 and the two-side significant test level was set up asα=0.05.Results: After eliminating the missing data, a total of 20312 inpatients' data for hospitalization expenditure were involved in this study analysis dataset. In 2008, the median of total hospitalization expenditure per time in the hospital was (?)1982.99, respectively. The proportion of medical expenditure(median: (?)711.20), inspection expenditure(median: (?)451.00), diagnosing and treating expenditure(median:(?)197.00) were 40.38%, 19.89%, 12.62%, respectively. The top five diseases of the proportion of inpatients were coronary heart disease (1418 cases, account for 7.0%), diabetes (805, 4.0%), pneumonia (845, 4.2%), chronic bronchitis (541, 2.7%) and brain blood vessel accident (515, 2.5%), respectively. The results from univariate analysis showed that the median of total hospitalization expenditure per time in male((?)2004.96) was higher than that of female((?) 1951.98, Z=2.935, p=0.003); the median of total hospitalization expenditure per time in the insured patients ( (?) 3123.51) was higher than that of the self-payment inpatients((?) 1794.62, Z=29.83, p =0.000). There were significant differences of total hospitalization expenditure among clinic section( median ranged from (?)812.66 to (?)2422.34,χ~2=2212.92, p =0.000), months ((?)1712.26 to (?)2105.60,χ~2=63.222, p=0.000), and age groups ((?)845.05 to (?)2490.56,χ~2=2272.92, p =0.000), respectively. Taking the natural logarithm-transformed total hospitalization expenditure as the dependent variable is, the results of multiple linear regression analysis showed that age (b=0.012,95%CI : 0.011, 0.012), the payment way (b=0.320,95%CI : 0.0281, 0.024), the month (b=0.016,95%CI : 0.008, 0.024) were introduced into the regression model; In the multiple linear regression analysis of medicine expenditure, age (b=0.015, 95%CI : 0.014, 0.015), the payment way (b=0.556,95%CI : 0.512, 0.601), the sex (b=-0.122, 95%CI : - 0.153, -0.091) and the month (b=0.019,95%CI : 0.01, 0.029) entered the regression model; and age (b=0.015,95%CI: 0.014, 0.015), the month (b=0.051,95%CI : 0.043, 0.059), the payment way (b=0.172, 95%CI: 0.133, 0.212), the gender (b=0.047, 95%CI: 0.200, 0.075) and the clinic section (b=-0.022, 95%CI : - 0.033, -0.011) were introduced into the multiple linear regression model of inspection expenditure; Age (b=0.009,95%CI : 0.008, 0.009), the payment way (b=0.189,95%CI : 0.145, 0.233) ,the clinic section (b=-0.065,95%CI : - 0.077, -0.052), and month (b=-0.021, 95%CI: -0.030--0.011) entered the multiple linear regression model of diagnosing and treating expenditure, the results of multiple linear regression analysis showed that Age was the main determinants of total hospitalization expenditure of the coronary heart disease, pneumonia and diabetes.Conclusion: 1. The proportion of medicine expenditure, inspection expenditure, and diagnosing and treating expenditure ranked the top three position within total hospitalization expenditure, so that controlling the medicine price and the using of the large scale inspection strictly is very important to control the excessive growth of medical expenditure.2.The payment way is the important determinant of hospitalization expenditure growth, so improving the medical insurance organizational reform is the effective method to control the medical resources waste which is brought by "the third party payment".3. Age is another important determinant for growth of hospitalization expenditure. With the aging society in China, doing the public health service for the old population well is the critical way to reduce the medical expenditure.4. Age, the payment way are the determinants of hospitalization expenditure of the coronary heart disease, pneumonia and diabetes.5. With the large sample, the multiple linear regression model may be relaxed its limitations, which also can obtain the good fitting effect.