Forecast analysis of outpatient number based on seasonal ARIMA model
Zhang Juan1, Zhang Hongcheng2
1 Teaching Office of Clinical College of Affiliated Hospital of Xuzhou Medical College, XuZhou 221002, China; 2 Medical Insurance Office of Affiliated Hospital of Xuzhou Medical College
Abstract：Objective To understand hospital outpatient workload variation trend, and to provide a scientific basis for the rational allocation of health resources and the optimization of the allocation of medical and health resources. Methods According to the outpatient visit data of the hospital from January 2009 to June 2015, using PASW statistics18.0 software we established Seasonal ARIMA model. Results According to the value of the standardized bayesian information criterion (BIC) and mean absolute percentage error (MAPE) we chose ARIMA (1,1,1) × (0,1,1)12 model as the optimal model. By the white noise testing, residual autocorrelation and partial autocorrelation function figure was within the scope of confidence interval, and The Box-Ljung inspection results showed Q18 =21.863, and P＝0.111, all showing the residual error belonged to the white noisy. The statistics of Data fitting R2=0.967, predicted value was basically consistent with the increase of the actual value, and outpatient visits in the second half of 2015 and 2016 will reach 1 103 thousand and 2 460.5 thousand respectively. Conclusion Seasonal ARIMA model match perfectly outpatient number changing trend, providing basis for hospital management.
