用于儿童肺炎支原体大叶性肺炎的列线图系统的开发与验证

Development and Validation of a Nomogram System for Pediatric Mycoplasma Pneumoniae Lobar Pneumonia

摘要: 前言：儿童肺炎支原体大叶性肺炎是重症表现之一。因此，早期识别早期治疗对于改善预后至关重要。目的：本实验的目标是研制一种方便的、高效的、易于使用的列线图，用于识别有大叶性肺炎高风险的肺炎支原体肺炎(Mycoplasma Pneumoniae Pneumonia, MPP)儿童，进行指导和临床干预。方法：回顾性分析安徽医科大学第二附属医院2022年1月至2023年12月收治的1043例支原体肺炎病儿的临床资料，对比2022年及2023年收治MPP病儿合并大叶性肺炎占比，1043例MPP患儿按4:1的比例随机分为训练样本(836个)和内部验证样本(207个)。采用多因素logistic回归判断分析方法确定MPP患儿混合大叶性肺炎的危险变量。选定的变量用于构建列线图，通过C指数、决策曲线分析、校准曲线和受试者工作特征(ROC)曲线进行验证。结果：2023年支原体肺炎病儿数量及大叶性肺炎占比明显多于2022年，差异具有统计学意义(P < 0.05)。多因素logistic回归分析结果显示：年龄、入院前病程、热程、咳嗽时间、中性粒细胞百分数、血清总二氧化碳浓度、超敏C反应蛋白是支原体肺炎病儿并发大叶性肺炎的独立危险因素，将以上7个影响因素作为预测指标，构建训练队列支原体肺炎病儿并发大叶肺炎该风险预测的列线图模型，训练队列列线图校准曲线与标准曲线具有高度可比性。预测结果的曲线下面积(AUC)在训练队列和验证队列上分别为0.974和0.968。决策曲线分析(DCA)曲线表明，该模型具有较高的准确性。结论：本文是基于7个变量构建了列线图，用于预测MPP患儿合并大叶性肺炎情况。构建的列线图模型具有较大临床价值，可指导个体化治疗。

Abstract: Preface: Mycoplasma pneumoniae lobar pneumonia in children is one of the severe manifestations. Therefore, early identification and treatment are crucial for improving prognosis. Objective: The goal of this experiment is to develop a convenient, efficient, and easy-to-use nomogram for identifying children with Mycoplasma pneumoniae pneumonia (MPP) at high risk of lobar pneumonia, providing guidance and clinical intervention. Method: A retrospective analysis was conducted on the clinical data of 1043 children with Mycoplasma pneumoniae pneumonia admitted to the Second Affiliated Hospital of Anhui Medical University from January 2022 to December 2023. The proportion of MPP children with lobar pneumonia admitted in 2022 and 2023 was compared. The 1043 MPP children were randomly divided into training samples (836) and internal validation samples (207) in a 4:1 ratio. The multiple logistic regression analysis method was used to determine the risk variables for mixed lobar pneumonia in MPP children. The selected variables are used to construct a nomogram, which is validated through C-index, decision curve analysis, calibration curve, and receiver operating characteristic (ROC) curve. Result: The number of children with mycoplasma pneumonia and the proportion of lobar pneumonia in 2023 were significantly higher than those in 2022, and the difference was statistically significant (P < 0.05). The results of multiple logistic regression analysis showed that age, pre-admission course, fever course, cough duration, neutrophil percentage, serum total carbon dioxide concentration, and high-sensitivity C-reactive protein were independent risk factors for mycoplasma pneumoniae pneumonia complicated with lobar pneumonia in children. Using these seven influencing factors as predictive indicators, a nomogram was constructed to predict the risk of Mycoplasma pneumoniae pneumonia complicated with lobar pneumonia in a training cohort. The calibration curve of the training was highly comparable to the standard curve. The area under the curve (AUC) of the predicted results is 0.974 in the training queue and 0.968 in the validation queue. The decision curve analysis (DCA) curve indicates that the model has high accuracy. Conclusion: This article constructed a nomogram based on 7 variables to predict the occurrence of lobar pneumonia in MPP patients. The constructed nomogram model has significant clinical value and can guide personalized treatment.