Assessment of risk factors for neonatal bacterial meningitis and establishment of a clinical prediction model
10.3760/cma.j.cn113903-20240827-00589
- VernacularTitle:新生儿细菌性脑膜炎危险因素评估及临床预测模型的建立
- Author:
Guanchu CHEN
1
;
Kun CHENG
1
;
Shuyang HOU
1
;
Yuan HUO
1
;
Jianming TANG
1
;
Fangping ZHAO
1
;
Weiyang LI
1
;
Hongxia GAO
1
Author Information
1. 甘肃省妇幼保健院（甘肃省中心医院）新生儿科，兰州 730050
- Publication Type:Journal Article
- Keywords:
Bacterial meningitis;
Lumbar puncture;
Neonate;
Risk factors;
Nomograms
- From:
Chinese Journal of Perinatal Medicine
2025;28(4):313-319
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the risk factors and construct a nomogram prediction model for neonatal bacterial meningitis (BM).Methods:A retrospective cohort study was conducted on 1 228 neonates who underwent lumbar puncture for cerebrospinal fluid examination in the Department of Neonatology at Gansu Provincial Women and Child Healthcare Hospital from December 2019 to February 2024. The subjects were randomly divided into a training cohort and a validation cohort at a ratio of 7∶3 using a computer program. Rank sum test or Chi-square tests were used to compare differences between the two cohorts. The subjects were divided into BM and non-BM groups based on the presence or absence of BM. Multivariate logistic regression analysis (forward stepwise regression method) was used in the training cohort to identify risk factors for BM. The area under the receiver operating characteristic curve (AUC) and the Hosmer-Lemeshow goodness-of-fit test were used to assess the discrimination and calibration of the model, respectively. Calibration curves were used to evaluate the accuracy of the model and to construct the nomogram. Internal validation was performed using the Bootstrap resampling method. Decision curve analysis was used to assess the clinical utility of the model. Results:Among the 1 228 neonates, 151 (12.3%) had BM. The training cohort included 859 neonates, of whom 106 (12.3%) had BM and 753 (87.7%) did not. The validation cohort included 369 neonates, of whom 45 (12.2%) had BM and 324 (87.8%) did not. The results of the multivariate logistic regression analysis in the training cohort showed that sepsis ( OR=4.446, 95% CI:2.583-7.653), convulsions ( OR=3.749, 95% CI:1.930-7.280), high maximum body temperature ( OR=2.027, 95% CI:1.636-2.513), and elevated C-reactive protein ( OR=1.007, 95% CI:1.003-1.012) were independent risk factors for BM, while greater gestational age at birth ( OR=0.946, 95% CI: 0.898-0.995) and higher hemoglobin levels ( OR=0.990, 95% CI:0.981-0.998) were protective factors for BM (all P<0.05). Based on these findings, a nomogram prediction model for neonatal BM was constructed and validated for accuracy. The AUC values of the nomogram model in the training and validation cohorts were 0.796 (95% CI: 0.750-0.843) and 0.781 (95% CI: 0.700-0.862), respectively. The Hosmer-Lemeshow goodness-of-fit test showed P>0.05 in both cohorts. The clinical decision curve analysis demonstrated good net benefit across most threshold ranges. Conclusions:Sepsis, convulsions, high maximum body temperature, and elevated C-reactive protein increase the risk of neonatal BM. The nomogram model constructed based on these factors, combined with gestational age and hemoglobin levels, provides a reference value for predicting the risk of neonatal BM.
