乘积季节性ARIMA模型在重症监护病房医院感染发病率预测中的应用

Application of ARIMA model in predicting incidence trend of nosocomial infection of intensive care unit

目的研究乘积季节性ARIMA(p,d,q)×(P,D,Q)s模型在重症监护病房医院感染发病率预测中的应用价值。方法通过收集某医院重症监护病房(ICU)医院感染发病率,建立乘积季节性ARIMA(p,d,q)×(P,D,Q)s模型,对重症监护病房医院感染发病率进行预测。结果 ARIMA(1,1,1)×(1,1,1)s能够较好的拟合该院重症监护病房医院感染发病率情况。此模型医院感染发病率实际值与预测值的吻合程度较高,平均相对误差为0.82%。结论 ARIMA(1,1,1)×(1,1,1)s能够较好地拟合该院重症监护病房医院感染发病率情况,可为医院感染的整体监控提供参考依据。

Objective To explore the application value of ARIMA（ p,d,q） ×（ P,D,Q） s model in predicting incidence trend of nosocomial infection of intensive care unit. Methods The ARIMA（ p,d,q） ×（ P,D,Q） model was constructed according to the incidence of nosocomial infection of intensive care unit in a comprehensive teaching hospital,and the incidence rate of nosocomial infection in intensive care unit was collected. Results ARIMA（ p,d,q） ×（ P,D,Q） s model could fit the incidence of nosocomial infection of intensive care unit,and the predicted values had better accord with actual values in the model,and the mean relative error was 0. 82%. Conclusion ARIMA（ p,d,q） ×（ P,D,Q） s model can better fit the incidence of nosocomial infection of intensive care unit,it can provide reference for the overall monitoring of nosocomial infection.