Sun Xiaowei, Jing Rui, Li Yang
Department of Pediatrics, Qilu Hospital, Shandong University, No.107, West Culture Road, Lixia District, Jinan City, Shandong Province, 250000, China.
Department of Pediatrics, Weifang People's Hospital, No.151 Guangwen Street, Kuiwen District, Weifang City, Shandong Province, 261000, China.
BMC Pediatr. 2025 Jan 4;25(1):3. doi: 10.1186/s12887-024-05349-y.
Purulent meningitis (PM) is a commonly encountered infectious condition in newborns, which unfortunately can result in infant mortality. Newborns with PM often present nonspecific symptoms. The success of lumbar puncture, an invasive test, relies on the operator's expertise. Preterm infants pose diagnostic challenges compared to full-term babies. The objective of this study is to establish a convenient and effective clinical prediction model based on perinatal factors to assess the risk of PM in very preterm infants, thereby assisting clinicians in developing new diagnostic and treatment strategies.
This study involved very preterm infants (gestational age < 32 weeks) admitted to the Qilu Hospital of Shandong University from January 2020 to December 2023. All included infants underwent lumbar puncture. We gathered comprehensive data that included information on maternal health conditions and the clinical features of very preterm infants. The PM was diagnosed according to the diagnostic criteria. This study conducted data analysis and processing using R version 4.1.2. A stepwise regression method was applied for multivariate Logistic regression analysis to select the best predictors for PM and to develop a predictive model. Differences were considered statistically significant at P < 0.05.
This study enrolled a total of 201 preterm infants, including 117 boys and 84 girls. The gestational age was 28.7 ± 1.7 weeks, and the weight was 1166.2 ± 302.7 g. Ninety infants were diagnosed with PM, while 111 did not have PM. The influencing factors include birth weight, PCT within 24 h after birth, cesarean delivery, and premature rupture of membranes. These were used to construct a risk prediction nomogram and verified its accuracy. The Brier score was 0.157, the calibration slope was 1.0, and the concordance index was 0.849.
We developed and validated a personalized nomogram to identify high-risk individuals for early prediction of purulent meningitis in very preterm infants. This practical predictive model may help reduce unnecessary lumbar puncture procedures.
化脓性脑膜炎（PM）是新生儿常见的感染性疾病，不幸的是可导致婴儿死亡。患有PM的新生儿常表现出非特异性症状。腰椎穿刺作为一种侵入性检查，其成功与否依赖于操作者的专业技能。与足月儿相比，早产儿在诊断方面存在挑战。本研究的目的是基于围产期因素建立一个方便有效的临床预测模型，以评估极早产儿患PM的风险，从而协助临床医生制定新的诊断和治疗策略。
本研究纳入了2020年1月至2023年12月在山东大学齐鲁医院住院的极早产儿（胎龄<32周）。所有纳入的婴儿均接受了腰椎穿刺。我们收集了全面的数据，包括母亲健康状况和极早产儿的临床特征信息。根据诊断标准诊断PM。本研究使用R版本4.1.2进行数据分析和处理。采用逐步回归法进行多因素Logistic回归分析，以选择PM的最佳预测因素并建立预测模型。P<0.05时差异具有统计学意义。
本研究共纳入201例早产儿，其中男117例，女84例。胎龄为28.7±1.7周，体重为1166.2±302.7g。90例婴儿被诊断为PM，111例未患PM。影响因素包括出生体重、出生后24小时内的降钙素原（PCT）、剖宫产和胎膜早破。利用这些因素构建了风险预测列线图并验证了其准确性。Brier评分0.157，校准斜率1.0，一致性指数0.849。
我们开发并验证了一种个性化列线图，用于识别极早产儿早期化脓性脑膜炎的高危个体。这种实用的预测模型可能有助于减少不必要的腰椎穿刺操作。
