预测极早产儿坏死性小肠结肠炎发生风险列线图模型的建立

Establishment of a nomogram model for predicting necrotizing enterocolitis in very preterm infants

摘要:

目的 探讨极早产儿坏死性小肠结肠炎（necrotizing enterocolitis，NEC）发生的危险因素，并构建预测NEC发生风险的列线图模型。 方法 选取2015年1月至2021年12月住院的752例极早产儿为研究对象，包括2015~2020年极早产儿（建模集）654例和2021年极早产儿98例（验证集）。建模集根据有无发生NEC分为NEC组（n=77）和非NEC组（n=577），通过多因素logistic回归分析确定极早产儿NEC发生的独立危险因素，采用R软件绘制列线图模型。利用验证集的数据对列线图模型加以检验。采用受试者工作特征（receiver operator characteristic，ROC）曲线、Hosmer-Lemeshow拟合优度检验及校正曲线评估模型的效能，采用临床决策曲线评估模型的临床实用价值。 结果 多因素logistic回归分析显示，新生儿窒息、败血症、休克、低白蛋白血症、严重贫血及配方奶喂养为极早产儿NEC发生的独立危险因素（P<0.05）。建模集ROC曲线的曲线下面积（area under the curve，AUC）为0.833（95%CI：0.715~0.952），验证集ROC曲线的AUC值为0.826（95%CI：0.797~0.862），表明该模型具有良好的区分度和判别能力。校正曲线和Hosmer-Lemeshow拟合优度检验显示该模型在预测值和真实值之间的准确性和一致性较好。 结论 新生儿窒息、败血症、休克、低白蛋白血症、严重贫血及配方奶喂养是极早产儿NEC发生的独立危险因素；基于多因素logistic回归分析结果建立的列线图模型可为临床早期评估极早产儿NEC的发生提供定量、简便、直观的工具。

Abstract:

Objective To investigate the risk factors for necrotizing enterocolitis (NEC) in very preterm infants and establish a nomogram model for predicting the risk of NEC. Methods A total of 752 very preterm infants who were hospitalized from January 2015 to December 2021 were enrolled as subjects, among whom 654 were born in 2015-2020 (development set) and 98 were born in 2021 (validation set). According to the presence or absence of NEC, the development set was divided into two groups: NEC (n=77) and non-NEC (n=577). A multivariate logistic regression analysis was used to investigate the independent risk factors for NEC in very preterm infants. R software was used to plot the nomogram model. The nomogram model was then validated by the data of the validation set. The receiver operating characteristic (ROC) curve, the Hosmer-Lemeshow goodness-of-fit test, and the calibration curve were used to evaluate the performance of the nomogram model, and the clinical decision curve was used to assess the clinical practicability of the model. Results The multivariate logistic regression analysis showed that neonatal asphyxia, sepsis, shock, hypoalbuminemia, severe anemia, and formula feeding were independent risk factors for NEC in very preterm infants (P<0.05). The ROC curve of the development set had an area under the curve (AUC) of 0.833 (95%CI: 0.715-0.952), and the ROC curve of the validation set had an AUC of 0.826 (95%CI: 0.797-0.862), suggesting that the nomogram model had a good discriminatory ability. The calibration curve analysis and the Hosmer-Lemeshow goodness-of-fit test showed good accuracy and consistency between the predicted value of the model and the actual value. Conclusions Neonatal asphyxia, sepsis, shock, hypoalbuminemia, severe anemia, and formula feeding are independent risk factors for NEC in very preterm infant. The nomogram model based on the multivariate logistic regression analysis provides a quantitative, simple, and intuitive tool for early assessment of the development of NEC in very preterm infants in clinical practice.