预测极早产儿早发型败血症发生风险的列线图模型的构建

Establishment of a nomogram model for predicting the risk of early-onset sepsis in very preterm infants

摘要:

目的 探讨极早产儿早发型败血症（early-onset sepsis, EOS）发生的危险因素，并构建预测EOS发生风险的列线图模型。方法 回顾性选取2020年1月—2022年12月在郑州大学第一附属医院出生并入住新生儿科的344例极早产儿，按7∶3的比率随机分为训练集（241例）和验证集（103例）。训练集根据是否发生EOS分为EOS组（n=64）和非EOS组（n=177）。采用多因素logistic回归分析筛选极早产儿EOS发生的危险因素，利用R语言构建列线图，并由验证集进行验证。分别采用受试者操作特征曲线（receiver operating characteristic curve, ROC曲线）、校准曲线和决策曲线分析评价模型的区分度、校准度和临床净收益。结果 多因素logistic回归分析显示，胎龄、产房气管插管、羊水粪染、生后首日血清白蛋白水平和绒毛膜羊膜炎是极早产儿EOS发生的独立危险因素（P<0.05）。训练集ROC曲线的曲线下面积为0.925（95%CI：0.888~0.963），验证集ROC曲线的曲线下面积为0.796（95%CI：0.694~0.898），表明模型的区分度良好。Hosmer-Lemeshow拟合优度检验表明模型拟合度良好（P=0.621）。校准曲线分析和决策曲线分析提示模型的预测效能和临床应用价值较高。结论 胎龄、产房气管插管、羊水粪染、生后首日血清白蛋白水平和绒毛膜羊膜炎与极早产儿EOS的发生独立相关；根据这些因素构建的极早产儿EOS发生风险的列线图模型有较高的预测效能和临床应用价值。

Abstract:

Objective To identify risk factors associated with early-onset sepsis (EOS) in very preterm infants and develop a nomogram model for predicting the risk of EOS.Methods A retrospective analysis was conducted on 344 very preterm infants delivered at the First Affiliated Hospital of Zhengzhou University and admitted to the Department of Neonatology between January 2020 and December 2022. These infants were randomly divided into a training set (241 infants) and a validating set (103 infants) in a 7:3 ratio. The training set was further divided into two groups based on the presence or absence of EOS: EOS (n=64) and non-EOS (n=177). Multivariate logistic regression analysis was performed to identify risk factors for EOS in the very preterm infants. The nomogram model was developed using R language and validated using the validating set. The discriminative ability, calibration, and clinical utility of the model were assessed using receiver operating characteristic (ROC) curve analysis, calibration curve analysis, and decision curve analysis, respectively.Results The multivariate logistic regression analysis revealed that gestational age, need for tracheal intubation in the delivery room, meconium-stained amniotic fluid, serum albumin level on the first day of life, and chorioamnionitis were risk factors for EOS in very preterm infants (P<0.05). The area under the ROC curve for the training set was 0.925 (95%CI: 0.888-0.963), and that for the validating set was 0.796 (95%CI: 0.694-0.898), confirming the model's good discrimination. The Hosmer-Lemeshow goodness-of-fit test suggested that the model was well-fitting (P=0.621). The calibration curve analysis and decision curve analysis demonstrated that the model had high predictive efficacy and clinical applicability.Conclusions Gestational age, need for tracheal intubation in the delivery room, meconium-stained amniotic fluid, serum albumin level on the first day of life, and chorioamnionitis are significantly associated with the development of EOS in very preterm infants.The nomogram model for predicting the risk of EOS in very preterm infants, constructed based on these factors, has high predictive efficacy and clinical applicability.