Risk factors and construction of prediction model for neonatal pneumonia
10.11816/cn.ni.2025-240891
- VernacularTitle:新生儿肺炎的危险因素及预测模型构建
- Author:
Ruiyuan QIAN
1
;
Wei WU
1
;
Yajuan LU
1
;
Hongxin LI
1
Author Information
1. 常州市儿童医院新生儿科,江苏常州 213003
- Publication Type:Journal Article
- Keywords:
Neonate;
Pneumonia;
Risk factor;
Prediction model
- From:
Chinese Journal of Nosocomiology
2025;35(11):1670-1673
- CountryChina
- Language:Chinese
-
Abstract:
OBJECTIVE To analyze the risk factors of neonatal pneumonia(NP)and construct the risk prediction model for NP.METHODS A total of 60 cases of neonatal pneumonia admitted to neonatology department of Chan-gzhou Children's Hospital from Dec.2022 to Dec.2023 were selected in the study group,and another 60 healthy neonates during the same period were randomly selected as the healthy group.The risk factors of neonatal pneu-monia were summarized by logistic regression analysis,the prediction model was established,the discrimination of the model was evaluated by receiver's operating characteristic(ROC)curves,and the correction capability was as-sessed by Hosmer-Lemeshow test.RESULTS Birth 1 min Apgar score＜7 points(OR=5.930,95％CI:2.877-12.222,P＜0.001)and amniotic fluid pollution(OR=4.175,95％CI:1.251-13.935,P=0.021)were the risk factors for neonatal pneumonia,while prophylactic antimicrobial drug application(OR=0.471,95％CI:0.226-0.982,P=0.044)was a protective factor(P＜0.05).Based on the results of Logistic regression analysis,the formula of the relevant risk prediction model was established as logit(P)=20.780×1 min Apgar score-23.429 × amniotic fluid pollution+3.024 × prophylactic use of antibiotics+1.056.The results of ROC curves analysis showed that AUC of the risk prediction model was 0.774(95％CI:0.691-0.857),suggesting that the risk pre-diction model had a good discrimination,and Hosmer-Lemeshow x2=7.028,P=0.941,and there was good cali-bration degree of the model.CONCLUSION Birth 1 min Apgar score＜7 points and amniotic fluid pollution are in-dependent risk factors of NP,while prophylactic antimicrobial drug application is a protective factor,and the risk prediction model for the development of NP based on logistic regression analysis had good predictive efficiency and accuracy.
