Predictive analysis of outpatient visits to a grade 3, class a hospital using an ARIMA model
Statistical forecasting has been an indispensible tool for the analysis of hospital-related information that can provide objective evidence for hospital management strategies(Xiong, 2002, Li, 2012). The ability to forecast trends in hospital outpatient visits, by accurately evaluating the dynamic changes in outpatient visits and fitting these data to a rational statistical model, is extremely useful for the appropriate allocation of human, financial and material resources and for future planning (Yuan et al., 2005). Over the past three decades, the autoregressive integrated moving average (ARIMA) model (also known as the BoxJenkins model) has been a popular linear model for time series forecasting, and it is now widely used in the field of econometrics(Shen, 2014, Wang et al., 2014). In the field of bio-medicine, ARIMA has been used mainly in the forecasting of infectious diseases(Guan, 2013, Zou et al., 2013), although it has also found preliminarily application in the prediction of outpatient volume(Zhou et al., 2011a, Liang et al., 2006, Xiang and Chen, 2009, Zhou et al., 2008, Zhou et al., 2011b). In order to obtain objective information that would facilitate the planned expansion of the outpatient building of a grade 3, class A (first-class) general hospital in China, the present study was designed to collect the data for outpatient visits during the last three years, and establish an ARIMA model of the outpatient visits by time series forecasting, using Predictive Analytics Software (PASW). The effect of the model was evaluated by forecasting future outpatient visits during the subsequent two years.