脑脊液糖及脑脊液蛋白对新生儿GBS化脓性脑膜炎预后的预测价值  

Prognostic value of cerebrospinal fluid glucose and cerebrospinal fluid protein in neonatal GBS purulent meningitis

作　　者：熊星宇 赵培丽 朱慧花[1] XIONG Xingyu;ZHAO Peili;ZHU Huihua(Department of Neonatology,Kunming Children's Hospital,Kunming,Yunnan 650000,China)

机构地区：[1]昆明市儿童医院(书林院区)新生儿科,云南昆明650000

出　　处：《中国优生与遗传杂志》2023年第1期27-32,共6页Chinese Journal of Birth Health & Heredity

摘　　要：目的探讨脑脊液糖及脑脊液蛋白水平对新生儿B族链球菌(GBS)化脓性脑膜炎预后的预测价值。方法选取2018年2月至2021年7月于昆明市儿童医院新生儿科并确诊为GBS化脓性脑膜炎新生儿148例,依据预后情况分为预后不良组(n=46)和预后良好组(n=102)。利用Logistic回归分析筛选GBS化脓性脑膜炎患儿预后不良影响因素,曲线拟合分析脑脊液糖、脑脊液蛋白与GBS化脓性脑膜炎患儿预后不良的关系。采用受试者操作特征(ROC)曲线评价脑脊液糖、脑脊液蛋白对新生儿GBS化脓性脑膜炎预后的预测价值。构建预测GBS化脓性脑膜炎患儿预后不良的人工神经网络模型并验证。结果是否顺产、住院时间、休克、脑脊液白细胞异常、脑脊液蛋白为新生儿GBS化脓性脑膜炎预后不良的独立危险因素,脑脊液糖含量是保护因素。脑脊液糖、脑脊液蛋白预测新生儿GBS化脓性脑膜炎预后不良的ROC曲线下面积分别为0.778、0.796,其截断点为1.67 mmol/L、1.46 g/L。构建的人工神经网络模型的区分度和准确度均较好,具有良好的临床应用价值。结论脑脊液糖和脑脊液蛋白均是新生儿GBS化脓性脑膜炎预后不良的独立影响因素,具有较好的预测价值。本研究构建的人工神经网络模型具有较高的区分度、准确度和临床决策价值。Objective To explore the predictive value of the degree of harm and prognosis of cerebrospinal fluid sugar and cerebrospinal fluid protein levels in neonates with GBS purulent meningitis.Methods 148 newborns with GBS suppurative meningitis diagnosed in the Department of Pediatrics of Kunming Children’s Hospital from February 2018 to July 2021 were divided into two groups:Poor prognosis group(n=46)and good prognosis group(n=102).Logistic regression analysis was used to screen the influencing factors of poor prognosis in children with GBS suppurative meningitis.Curve fitting was used to analyze the relationship between cerebrospinal fluid glucose,cerebrospinal fluid protein and poor prognosis in children with GBS suppurative meningitis.The receiver operating characteristic(ROC)curve was used to evaluate the prognostic value of cerebrospinal fluid glucose and cerebrospinal fluid protein in neonatal GBS suppurative meningitis.Construct an artificial neural network model to predict the poor prognosis of children with GBS suppurative meningitis and verified it.Results Spontaneous delivery,length of stay,shock,abnormal white blood cells in cerebrospinal fluid,and cerebrospinal fluid protein were independent risk factors for poor prognosis of neonatal GBS suppurative meningitis,and glucose content in cerebrospinal fluid was a protective factor.The areas under the ROC curve of cerebrospinal fluid glucose and cerebrospinal fluid protein for predicting poor prognosis of neonatal GBS suppurative meningitis were 0.778 and 0.796 respectively,and their cutoff points were 1.67 mmol/L and 1.46 g/L,respectively.The artificial neural network model has good differentiation,accuracy,and clinical application value.Conclusion Cerebrospinal fluid glucose and cerebrospinal fluid protein are independent factors affecting the poor prognosis of neonatal GBS suppurative meningitis.The artificial neural network model constructed in this study has high differentiation,accuracy and clinical decision-making value.

关 键 词：脑脊液糖 脑脊液蛋白 新生儿 GBS化脓性脑膜炎 

分 类 号：R722.1[医药卫生—儿科]