住院新生儿医院感染发生季节ARIMA模型预测

Prediction of nosocomial infection rate and occurred season in neonatal inpatients by ARIMA model

目的研究医院住院新生儿医院感染发生情况,为有效防控医院感染提前预警。方法通过回顾性调查方法,收集某大型教学医院住院新生儿医院感染数据,利用乘积季节ARIMA模型进行预测。结果 ARIMA(1,0,0)(0,1,1)12为最优模型,R2为0. 659,BIC最小为-2. 115。数据统计显示,该医院新生儿科2016-2017年住院患儿医院感染发生率在1. 9%～3. 3%范围。根据ARIMA模型预测,2018年新生儿住院患儿医院感染发生率范围在1. 7%～2. 4%,2019年新生儿住院患儿医院感染发生率范围在1. 5%～2. 2%。结论 ARIMA能够较好地预测新生儿住院患儿医院感染发生情况,能为医院感染监测和制定防控措施提供参考。

Objective To study the incidence of nosocomial infection in neonates in hospital for the effective prevention and control of nosocomial infection early warning. Methods The data of nosocomial infection of neonates in a large teaching hospital were collected by retrospective investigation,and the product seasonal ARIMA model was used for prediction.Results ARIMA( 1,0,0)( 0,1,1) 12 was the optimal model,R2 was 0. 659,BIC was-2. 115. Statistics show that the incidence of nosocomial infection in the neonatal department of the hospital ranged from 1. 9% to 3. 3% in 2016-2017.According to the prediction of ARIMA model,the incidence of nosocomial infection in newborn inpatients ranges from 1. 7%to 2. 4% in 2018,and 1. 5% to 2. 2% in 2019. Conclusion ARIMA can better predict the incidence of nosocomial infection in newborn hospitalized children,and can provide a reference for nosocomial infection monitoring and the development of prevention and control measures.