Objective To establish an optimal model suitable for predicting the outpatient prescriptions with SARIMAX in a hospital,so as to provide a reference for refining the allocation plan of manpower,drugs and materials in outpatient pharma-cy.Methods The weekly outpatient prescriptions data of a hospital and meteorological data(mean temperature,air quality in-dex)from January 2017 to June 2023 were collected and divided into the training set and the validation set.In the training set,the optimal SARIMAX model was determined mainly with Akaike information criterion(AIC),supplemented by coefficient of de-termination(R2),parameter significance test,and Ljung-Box Q statistics.The performance test of the optimal model was carried out in the validation set.Results The optimal SARIMAX model was SARIMAX(1,1,2)(2,1,0)52,which included mean temperature with a lag of order 0.And the model parameters were statistically significant(P＜0.05),with an information criteri-on AIC of 3 716.34,a mean absolute percentage error of 8.64％,and a residual sequence of white noise.In the validation set,prediction value was ideally fitted to the actual value,with a mean absolute percentage error of 12.28％.Conclusion SARIMAX model is suitable for the early warning and prediction of outpatient prescription volume in the hospital,which can provide the ba-sis for managers to make decisions and promote the fine management of outpatient pharmacy.
关键词
SARIMAX模型/门诊处方量/预警预测
Key words
SARIMAX model/outpatient prescription volume/early warning and prediction
维普
万方数据