Department of Neurology, Children's Hospital of Chongqing Medical University, Chongqing, China.
Ministry of Education Key Laboratory of Child Development and Disorders, Chongqing, China.
J Child Neurol. 2023 Aug;38(8-9):528-536. doi: 10.1177/08830738231193217. Epub 2023 Aug 13.
We aimed to build a prediction nomogram for early prediction of poor prognosis in children with meningitis and analyzed the course of treatment and discharge criteria.
Eighty-seven pediatric patients with meningitis were retrospectively recruited from the Children's Hospital of Chongqing Medical University between June 2012 and November 2021. Univariate analysis and binary logistic analysis were used to evaluate the risk factors, and the prediction model was built.
meningitis is more common in children <3 months old in our study (86.2%). Common complications were subdural effusion (39.1%), followed by hydrocephalus (13.8%) and repeated convulsions (12.6%). The mortality rate and sequelae rate of meningitis in children was ∼10.9% and ∼6.3%, respectively. Univariate analysis showed that 13 clinical indicators were associated with poor prognosis of meningitis in children. In binary logistic analysis, risk factors were seizures ( = .032) and the last cerebrospinal fluid glucose content before discharge ( = .002). A graphical nomogram was designed. The area under the receiver operating characteristic curve was 0.913. The Hosmer-Lemeshow test showed that the model was a good fit ( = .648). Internal validation proved the reliability of the prediction nomogram.
meningitis is more common in children <3 months old in our study. The rate of complications and sequelae are high. The prediction nomogram could be used to assess the risk of poor prognosis in children with meningitis by clinicians.
本研究旨在建立预测模型，以早期预测儿童脑膜炎的不良预后，并分析治疗过程和出院标准。
回顾性收集 2012 年 6 月至 2021 年 11 月重庆医科大学儿童医院收治的 87 例脑膜炎患儿的临床资料。采用单因素分析和二项逻辑回归分析评估危险因素，并建立预测模型。
本研究中，脑膜炎患儿以<3 个月龄为主（86.2%）。常见并发症为硬膜下积液（39.1%），其次为脑积水（13.8%）和反复惊厥（12.6%）。脑膜炎患儿的死亡率和后遗症发生率分别为10.9%和6.3%。单因素分析显示，13 项临床指标与儿童脑膜炎的不良预后相关。二项逻辑回归分析显示，惊厥（ = .032）和出院前最后一次脑脊液葡萄糖含量（ = .002）是不良预后的危险因素。绘制了一个图形诺模图。受试者工作特征曲线下面积为 0.913。Hosmer-Lemeshow 检验表明该模型拟合良好（ = .648）。内部验证证明了预测诺模图的可靠性。
本研究中，脑膜炎患儿以<3 个月龄为主。并发症和后遗症发生率较高。该预测诺模图可用于临床医生评估儿童脑膜炎不良预后的风险。
