NICU新生儿脐静脉置管血流感染风险预测模型的构建与验证  

Construction and validation of risk prediction model for umbilical vein catheterization bloodstream infection in neonates in neonatal intensive care unit

作　　者：苗逸群 刘文文 赵淑良 姜汇敏 李亚男 滕平[2] 王爱华[1] 张媛媛[1] MIAO Yi-qun;LIU Wen-wen;ZHAO Shu-liang;JIANG Hui-min;LI Ya-nan;TENG Ping;WANG Ai-hua;ZHANG Yuan-yuan(School of Nursing,Weifang Medical University,Weifang 261053,China;Delivery Room,Weifang Maternal and Child Health Hospital,Weifang 261011,China)

机构地区：[1]潍坊医学院护理学院,山东潍坊261053 [2]潍坊市妇幼保健院产房,山东潍坊261011

出　　处：《中国感染控制杂志》2023年第2期159-166,共8页Chinese Journal of Infection Control

基　　金：2018年潍坊市科学技术发展计划(医学类)(2018YX029)。

摘　　要：目的构建新生儿重症监护病房(NICU)的新生儿脐静脉置管血流感染的风险预测模型,并对其应用效果进行检验。方法回顾性选择某院2020年7月—2021年12月NICU行脐静脉置管的新生儿作为建模组,根据是否发生脐静脉置管血流感染分为感染组和非感染组,统计感染组新生儿血微生物培养中细菌检出情况。采用单因素及多因素logistic回归分析筛选血流感染的危险因素并纳入R语言建立预测风险的列线图模型。另选取2022年1—7月行脐静脉置管的新生儿作为验证组进行模型的外部验证。结果2020年7月—2021年12月的447例新生儿为建模组,其中感染组34例,非感染组413例,感染发病率为7.6%;2022年1—7月的225例新生儿为验证组。建模组logistic回归分析显示,置管总时间>7 d、出生体重<1500 g、穿刺次数>2次、清蛋白<35 g/L、机械通气史、经外周静脉穿刺中心静脉置管治疗史是NICU新生儿发生脐静脉置管血流感染的独立危险因素。基于回归结果建立列线图预测模型,建模组受试者工作特征(ROC)曲线下面积为0.866(95%CI:0.784~0.947),Youden指数为0.642,灵敏度为0.853,特异度为0.789;Hosmer-Lemeshow检验显示P=0.323;验证组ROC曲线下面积为0.837(95%CI:0.744~0.930),Youden指数为0.549,灵敏度为0.700,特异度为0.849;提示模型具有较好的区分度和拟合优度。结论本研究构建的模型能较好地预测NICU新生儿发生脐静脉置管血流感染的风险,可作为临床医护人员预测新生儿脐静脉置管血流感染风险的评估工具。Objective To construct a risk prediction model for neonatal umbilical vein catheterization(UVC)bloodstream infection(BSI)in neonates in a neonatal intensive care unit(NICU),and validate the application effect.Methods Neonates underwent UVC in NICU of a hospital from July 2020 to December 2021 were selected retrospectively as the modeling group and divided into the infection group and non-infection group according to whether UVC BSI occurred.Bacteria isolated from blood microbial culture of neonates in infection group was statistically analyzed.Risk factors for BSI were screened by univariate and multivariate logistic regression analysis.R language was used to construct a nomograph model to predict the risk.In addition,neonates underwent UVC from January to July 2022 were selected as the validation group for external validation of the model.Results From July 2020 to December 2021,447 neonates were selected as modeling group,including 34 in infection group and 413 in non-infection group,with an infection rate of 7.6%.225 neonates from January to July 2022 were as the validation group.Logistic regression analysis on modeling group showed that the total catheterization time>7 days,birth weight<1500 g,puncture opportunities>2 times,albumin<35 g/L,history of mechanical ventilation and peripherally inserted central catheter(PICC)treatment were independent risk factors for UVC BSI in NICU neonates.Nomograph prediction model was constructed based on the regression analysis results.The area under the receiver operating characteristic(ROC)curve of subjects in the modeling group was 0.866(95%CI:0.784-0.947),the Youden index,sensitivity,and specificity were 0.642,0.853,and 0.789 respectively.Hosmer-Lemeshow test showed P=0.323.The area under the ROC curve of the validation group was 0.837(95%CI:0.744-0.930).The Youden index,sensitivity,and the specificity were 0.549,0.700,and 0.849 respectively,which suggested that the model has good discrimination and degree of fitting.Conclusion The model constructed in this study can well

关 键 词：血流感染 脐静脉置管 预测模型 列线图 新生儿重症监护病房 

分 类 号：R181.3[医药卫生—流行病学]