首页|基于实验室检查建立儿童细菌性脑膜炎的联合诊断模型

基于实验室检查建立儿童细菌性脑膜炎的联合诊断模型

扫码查看
目的 基于实验室检查建立儿童细菌性脑膜炎的多项检验指标联合诊断模型.方法 回顾性分析 2018-2023 年惠州市中心人民医院及深圳市儿童医院收治的细菌性脑膜炎(56 例)和非细菌性脑膜炎(75 例)患儿的 16 项实验室检验指标.对细菌性脑膜炎与非细菌性脑炎患儿的人口学资料(性别、年龄)、临床表现及 16 项实验室检验指标进行单因素分析.选取单因素分析差异有统计学意义的检验指标进行ROC分析,以曲线下面积(areas under curve,AUC)>0.7 作为检验标准,根据β系数的大小对与细菌性脑膜炎独立相关的变量进行排名,利用Logistic线性回归分析将相应变量进行线性拟合,获得新的诊断模型,并选取 16 例脑膜炎患儿的数据评估模型的诊断效能.结果 细菌性脑膜炎和非细菌性脑膜炎患儿的 16 项检验指标中,12 项的差异均有统计学意义(P<0.05).其中,8 项指标的AUC>0.7,包括脑脊液白细胞计数(C-WBC)、脑脊液多核细胞占比(C-PNC%)、脑脊液多核细胞计数(C-PNC)、脑脊液蛋白(C-Pro)、脑脊液乳酸脱氢酶(C-LDH)、血降钙素原(B-PCT)、血清C反应蛋白(B-CRP)、脑脊液葡萄糖/血葡萄糖比值(C/B-Glu).通过线性拟合获得的诊断模型为:Y=-0.351×C/B Glu+0.561×C-PNC%+0.005×B-PCT+0.001×B-CRP+0.24.拟合模型诊断细菌性脑膜炎的AUC为 0.87,灵敏度、特异度均>75%.结论 由C/B-Glu、C-PNC%、B-PCT、B-CRP组成的联合模型对儿童细菌性脑膜炎具有良好的诊断价值.
The Establishment of Combined Diagnostic Model for Vacterial Meningitis in Children Base on Laboratory Tests
Objective To Establish a combined diagnostic model for multiple testing indicators of bacterial meningitis in children based on laboratory tests.Methods With the retrospective analysis of 16 laboratory test indicators of 56 cases of bacterial meningitis and 75 cases of non-bacterial meningitis admitted to Huizhou Central People's Hospital and Shenzhen Children's Hospital from 2018 to 2023,the univariate analysis was conducted on population data(gender,age),clinical manifestations,and 16 laboratory test indicators for children with bacterial meningitis and non-bacterial encephalitis.Test indicators with statistically significant differences in univariate analysis were selected for ROC analysis.With areas under curve(AUC)more than 0.7 as the test criterion,the variables independently associated with bacterial meningitis were ranked according to β coefficient.With the use of logistic linear regression analysis to linearly fit the corresponding variables to obtain a new diagnostic model,data from 16 children with meningitis were selected in the test set to evaluate the diagnostic performance of the model.Results There were significant differences in 12 of the 16 laboratory indicators between the children with bacterial meningitis and those without bacterial meningitis(P<0.05).Among them,the AUC of 8 indicators was more than 0.7,including cerebrospinal fluid white blood cell count(C-WBC),proportion of polynucleated cells in cerebrospinal fluid(C-PNC%),cerebrospinal fluid polynucleated cell count(C-PNC),cerebrospinal fluid protein(C-Pro),cerebrospinal fluid lactate dehydrogenase(C-LDH),blood procalcitonin(B-PCT),serum C-reactive protein(B-CRP),and cerebrospinal fluid glucose/blood glucose ratio Value(C/B-Glu).The diagnostic model obtained by linear fitting was Y=-0.351×C/B Glu+0.561×C-PNC%+0.005×B-PCT+0.001×B-CRP+0.24.The AUC of the fitted model for diagnosing bacterial meningitis was 0.87,and the sensitivity and specificity were more than 75%.Conclusion The combined model composed of C/B-Glu,C-PNC%,B-PCT,and B-CRP has good diagnostic value for bacterial meningitis in children.

Bacterial meningitisDiagnostic modelChildrenCerebrospinal fluid

梁燕锦、刘日阳、徐明艳、林贤彬、操德智

展开 >

汕头大学医学院 (广东汕头 515041)

惠州市中心人民医院 (广东惠州 516001)

深圳市儿童医院 (广东深圳 518038)

细菌性脑膜炎 诊断模型 儿童 脑脊液

惠州市科技计划

20220403

2024

医疗装备
国家食品药品监督管理局北京医疗器械质量监督检验中心 北京市医疗器械检验所

医疗装备

影响因子:0.339
ISSN:1002-2376
年,卷(期):2024.37(6)
  • 21