小儿重症监护室患者院内感染阴沟肠杆菌发生的时间序列分析

Time series analysis of Enterobacter cloacae nosocomial infections in children hospitalized in the pediatric intensive care unit

摘要:

目的 了解新疆医科大学第一附属医院儿科重症监护室(ICU)患者院内感染阴沟肠杆菌发生的时间分布特点,为做好院内感染控制提供参考数据。方法 收集该院2010 年1 月至2013 年12 月儿科ICU感染阴沟肠杆菌的患儿的临床资料,分月记录阴沟肠杆菌感染发生例数,使用SPSS 21.0 统计软件进行时间系列分析,得到的模型使用2014 年1~6 月的数据进行预测验证。结果 2010 年1 月至2013 年12 月儿科ICU 共发生157 例阴沟肠杆菌感染,其中2010 年33 例,2011 年35 例,2012 年37 例,2013 年52 例。时间序列拟合曲线具有代表性(R²=0.702,Ljung-BoxQ=(18)=36.021,P=0.004),显示阴沟肠杆菌在5~7 月多发。使用2014年1~6 月数据验证发现,预测值与实际值之间相差较小。结论 新疆医科大学第一附属医院儿科ICU 患儿院内感染阴沟肠杆菌发生的时间分布特点为每年5~7 月多发;预测模型具有较好的预测效果,可以用作感染预防工作的参考。

Abstract:

Objective To characterize the time distribution of the incidence of Enterobacter cloacae nosocomial infections in children hospitalized in the pediatric intensive care unit (PICU) of the First Affiliated Hospital of Xinjiang Medical University.Methods The clinical data of children with Enterobacter cloacae nosocomial infections in the PICU of the First Affiliated Hospital of Xinjiang Medical University between January 2010 and December 2013 were collected. The monthly number of cases of Enterobacter cloacae nosocomial infections was recorded, and time series analysis was performed using SPSS 21.0 software. The obtained prediction model was verified using the data from January to June in 2014.Results A total of 157 cases of Enterobacter cloacae nosocomial infections were reported in the PICU between January 2010 and December 2013, including 33 cases in 2010, 35 cases in 2011, 37 cases in 2012, and 52 cases in 2013. Time series analysis of the monthly number of cases of nosocomial infections reveals a fitted curve with a clear pattern of seasonal variation (R²=0.702, Ljung-BoxQ=(18)=36.021, P=0.004), with peaks in May, June, and July. The verification using the data from January to June in 2014 showed small differences between the predicted values and the actual values.Conclusions In the PICU of the First Affiliated Hospital of Xinjiang Medical University, the incidence of Enterobacter cloacae nosocomial infections is high in May, June, and July every year. The prediction model is accurate and can provide a reference for infection prevention.