振幅整合脑电图对暴露于围产期感染的早产儿脑损伤诊断及预后的应用价值  

Research on the application value of amplitude-integrated electroencephalography in the diagnosis and prognosis of brain injury in preterm infants exposed to perinatal infections

作　　者：卢田甜 冯永洪 张莉[1] 林群 羊小桂 陈秀灵 蔡江云[1] LU Tian-tian;FENG Yong-hong;ZHANG Li;LIN Qun;YANG Xiao-gui;CHEN Xiu-ling;CAI Jiang-yun(Department of Neonatology,Haikou Maternal and Child Health Hospital,Haikou 570203,Hainan,China;不详)

机构地区：[1]海口市妇幼保健院新生儿科,海南海口570203 [2]海口市人民医院儿童医学部,海南海口570208

出　　处：《广东医学》2025年第3期413-418,共6页Guangdong Medical Journal

基　　金：海南省卫生健康行业科研项目(22A200322)。

摘　　要：目的探讨振幅整合脑电图(amplitude-integrated electroencephalography,aEEG)对暴露于围产期感染的早产儿脑损伤诊断及预后的应用价值,为临床提供参考依据。方法选择2022年6月至2023年6月的136例32周以下暴露于围产期感染的早产儿,根据早产儿是否发生脑损伤将其分为发生组(36例)和未发生组(100例)。收集两组早产儿的临床资料及振幅整合脑电图进行比较,分析早产儿发生脑损伤的危险因素,并以此构建脑损伤风险的预测模型,采用受试者工作特征(ROC)曲线下面积(AUC)和校准曲线评估脑损伤风险预测模型的区分度和校准度。结果多因素logistic分析显示,宫内感染、自然分娩及aEEG图形连续性、睡眠觉醒周期、波谱带宽度与下边界振幅变化均是暴露于围产期感染的早产儿发生脑损伤的独立危险因素。预测模型验证ROC曲线下面积(AUC)为0.824[95%置信区间(CI)为0.753~0.895],最大Youden指数为0.522,敏感度为0.731,特异度为0.791,对患儿发生脑损伤风险有较好的预测价值,H-L拟合优度检验可得出研究所得预测模型的预测概率与实际值的结果较一致(P>0.05)。aEEG图形连续性、睡眠觉醒周期及宽度与振幅变化评分较低的脑损伤患儿其预后较差,且与预后结局呈负相关(P<0.05)。结论aEEG监测对暴露于围产期感染的早产儿脑损伤早期诊断具有重要应用价值,并可预测早产儿远期神经系统发育状况。Objective To investigate the application value of amplitude-integrated electroencephalography(aEEG)in the diagnosis and prognosis of brain injury in preterm infants exposed to perinatal infections.And to provide clinical reference.Methods From June 2022 to June 2023,136 preterm infants,born before 32 weeks of gestation and exposed to perinatal infections were selected and included.The infants were divided into two groups:the brain injury group(36 cases)and the non-brain injury group(100 cases).Clinical data and aEEG results were collected and compared between the two groups.Risk factors for brain injury were analyzed,and a predictive model for brain injury risk was constructed.The discriminative ability and calibration of the predictive model were assessed using the area under the receiver operation characteristics(ROC)curve(AUC)and calibration curve.Results Factor analysis indicated that intrauterine infection,spontaneous delivery,aEEG pattern continuity,sleep-wake cycle,spectral bandwidth,and lower boundary amplitude variation were independent risk factors for brain injury in preterm infants exposed to perinatal infections.The predictive model showed good predictive value for brain injury risk,with an AUC of 0.824(95%CI:0.753-0.895),a maximum Youden index of 0.522,sensitivity of 0.731,and specificity of 0.791.The Hosmer-Lemeshow goodness-of-fit test indicated that the predicted probabilities were consistent with the actual outcomes(P>0.05).Infants with lower scores for aEEG pattern continuity,sleep-wake cycle,and spectral width and amplitude variation had poorer prognosis,with a significant negative correlation with long-term neurological outcomes(P<0.05).Conclusion aEEG monitoring has significant application value in the early diagnosis of brain injury in preterm infants exposed to perinatal infections and can predict long-term neurological development outcomes.

关 键 词：振幅整合脑电图 围产期感染 早产儿 脑损伤 应用价值 

分 类 号：R722.6[医药卫生—儿科] R741.044[医药卫生—临床医学]