Research On The Prediction Of Children’s Hospital Outpatient Amount Based On ARIMA Model

Objective: To establish ARIMA model, accurately predict medium-term and short-term outpatient amount of children’s hospital, provide premise conditions for hospital outpatient management decision and offer a reference basis for the reasonable arrangement for hospital outpatient service resources, especially the human resources.To analyze the correlation of daily outpatient amount and air quality index, maximum temperature,minimum temperature as well as the difference in temperature on the same day with transfer function model, analyze the related factors affecting outpatient amount,and provide the reference for the pathogenic factors of children disease for hospital outpatient management and medical expert researches.Methods: The study used descriptive method to analyze the data of the Affiliated Children’s Hospital of Soochow University from 2007 to November 2014; the medium-term and short-term outpatient amount was respectively predicted by using ARIMA model. The monthly outpatient amount data from 2007 to 2013 were used for medium-term prediction; the hospital outpatient amount between January and June 2014 was predicted by applying ARIMA product seasonal model; the error rate was calculated. The transfer function model was used for short-term prediction and for analyzing the correlation of daily outpatient amount and air quality index, maximum temperature,minimum temperature as well as the difference in temperature on the same day; 41 weeks of outpatient amount data from February 2, 2014 to November 15, 2014 were selected to predict the weekly outpatient amount from November 16 to November22, 2014; the error rate was calculated; the accuracy of the model was tested.Results:In the medium-term prediction model,the hospital outpatient amount showed seasonal change trend,with the cycle of 12;it was predicated that the models of outpatient amount of internal medicine department,the total outpatient amount of internal medicine department,the total outpatient amount of department of surgery,and the total outpatient amount from January to June 2014 were respectively2.1 ARIMA模型ARIMA(0,0,1)(0,1,1)12,ARIMA(1,0,0)(0,1,1)12,ARIMA(2,0,0)(0,1,1)12 and ARIMA(1,0,0)(0,1,1)12;the error rates of half year were 11.01%,3.68%,9.05%and0.01%respectively;the prediction was comparatively accurate.In the short-term prediction outpatient amount,the transfer function model was used for analysis,with the cycle of 7;the maximum temperature,minimum temperature and air quality had obvious correlation with outpatient amount,to be specific,the total outpatient amount of internal medicine department and the emergency amount of internal medicine department were significantly correlative with the maximum temperature;the total outpatient amount of department of surgery was significantly correlative with maximum temperature and air quality;the total outpatient amount was significantly correlative with maximum temperature and minimum temperature.But the difference in temperature on the same day was not correlative with all kinds of outpatient amount.The model expression of outpatient amount of internal medicine department,the total outpatient amount of internal medicine department,the total outpatient amount of department of surgery,the emergency amount of internal medicine department and the total outpatient amount were respectively:ARIMA(0,1,1)(0,1,1)7,ARIMA(0,1,2)(0,1,1)7,ARIMA(0,1,1)(0,1,1)7,ARIMA(0,1,4)(1,0,1)7 and ARIMA(0,1,2)(0,1,1)7;the error rates of one week were respectively 2.92%,3.89%,3.56%,20.1%and 4.32%.The prediction of emergency amount of internal medicine department had greater error rate,while other prediction effect was good;proper model could not be established for outpatient amount of department of surgery and emergency amount of department of surgery.Conclusions: For the product seasonal model and transfer function model, the prediction of children’s hospital outpatient amount based on ARIMA model was feasible and it could provide accurate mid-term and short-term prediction. Maximum temperature, minimum temperature and air quality were significantly corrective with outpatient amount. The accurate prediction of outpatient amount could provide reliable basis for the hospital to scientifically arrange outpatient medical resources and improved the level of hospital management.