Time series analysis of per diem expense of inpatients with chronic kidney disease in a tertiary referral hospital in Shanghai

Jun-rui LONG,Chan-juan SHAN,Jiu-sheng WANG,Chang-lin MEI,Lin-ping XIONG
DOI: https://doi.org/10.16781/j.0258-879x.2018.01.0029
2018-01-01
Abstract:Objective To study the tendency of per diem expense and drug cost ratio of inpatients with chronic kidney disease (CKD) and the top three diseases (chronic renal failure, nephrotic syndrome, and primary glomerulonephritis) in a tertiary referral hospital in Shanghai from 2011 to 2016, so as to predict the per diem expense and drug cost ratio of inpatients in 2017. Methods We analyzed the per diem expense, drug cost ratio and other general situation of inpatients with CKD, and conducted the modeling research by the expert modeler in SPSS 23.0 software time series analysis module. Firstly, the per diem expense and drug cost ratio of inpatients with CKD from Jan. 2011 to Dec. 2015 were used to build model, and then the per diem expense and drug cost ratio of the inpatients in 2016 predicted by the model were compared with the real data of the inpatients in 2016 and the values of the model were evaluated. Secondly, the per diem expense and drug cost ratio of inpatients with CKD in 2017 were predicted. Finally, according to the same principle, the per diem expense and drug cost ratio of inpatients with chronic renal failure, nephrotic syndrome and primary glomerulonephritis in 2017 were predicted. Results The per diem expense, per capita hospitalization expense and per capita drug expense of inpatients with CKD increased annually from 2011 to 2016, and the expenses were (1506.89±1003.34) yuan, 8252.51 yuan and 2009.88 yuan in 2016, respectively, with the growth rates being 62.38%, 41.62% and 14.44%, respectively. The drug cost ratio of the inpatients decreased annually, and the ratio reached 32.33% in 2016. The time series models established by the per diem expense and drug cost ratio of inpatients with CKD were Winters additive model and simple seasonal model, respectively. According to the time series model, the predicted per diem expense and drug cost ratio in 2016 were consistent with the real values in 2016. The per diem expense and drug cost ratio of inpatients with CKD in 2017 predicted by the time series model were about 1600 yuan and 30%, respectively. The predicted per diem expenses of inpatients with chronic renal failure, nephrotic syndrome and primary glomerulonephritis in 2017 were about 1700 yuan, 1800 yuan and 1600 yuan, respectively, and the predicted drug cost ratios were about 30%, 35% and 10%, respectively. Conclusion The hospital income structure is constantly optimized, and a real-time disease cost monitoring information platform should be established. The research based on the trend of single disease expense needs to be further studied.
What problem does this paper attempt to address?