目的 探讨肺炎支原体肺炎（Mycoplasma pneumoniae pneumonia, MPP）患儿并发塑型性支气管炎（plastic bronchitis, PB）的预测因素，并构建PB发生的列线图预测模型。 方法 以2023年1月—2024年6月徐州医科大学附属医院住院治疗的MPP患儿为研究对象进行回顾性分析，通过简单随机抽样方法按照7∶3的比例随机分为训练集（562例）和测试集（240例）。训练集患儿中，根据患儿是否发生PB分为PB组（70例）和非PB组（492例）。对相关变量进行Spearman秩相关分析排除共线性，通过单因素分析及LASSO回归分析筛选变量，构建MPP患儿并发PB的列线图预测模型。采用受试者操作特征曲线分析评估预测模型的区分度，校准曲线分析评估预测模型的拟合度，决策曲线分析评估预测模型的临床价值。 结果 与非PB组比较，PB组支气管镜检查前的病程、发热时间、发热峰值、有家族过敏史比例、有过敏史比例、胸腔积液比例，以及白细胞计数、中性粒细胞百分比、C-反应蛋白、降钙素原、纤维蛋白原、D-二聚体、谷草转氨酶、谷丙转氨酶、肌酸激酶、乳酸脱氢酶、免疫球蛋白A、白细胞介素-6水平显著升高，淋巴细胞百分比显著降低（均P<0.05）。LASSO回归分析结果显示，胸腔积液、降钙素原、D-二聚体、乳酸脱氢酶是MPP患儿发生PB的主要预测指标。基于这些指标构建的列线图模型具有较好的区分度（训练集和测试集的受试者操作特征曲线的曲线下面积分别为0.852、0.830），模型的拟合度和获益率良好。 结论 基于胸腔积液、降钙素原、D-二聚体、乳酸脱氢酶4个指标建立的列线图预测模型对MPP患儿PB的发生有较好的预测效能。

Objective To investigate the predictive factors for plastic bronchitis (PB) in children with Mycoplasma pneumoniae pneumonia (MPP) and to establish a nomogram prediction model for PB occurrence. Methods A retrospective analysis was conducted on children with MPP hospitalized at The Affiliated Hospital of Xuzhou Medical University from January 2023 to June 2024. The patients were randomly divided into a training set (n=562) and a validation set (n=240) at a ratio of 7:3 using simple random sampling. In the training set, patients were categorized into a PB group (n=70) and a non-PB group (n=492) based on the occurrence of PB. Spearman correlation analysis was performed to exclude collinearity among variables, followed by univariate analysis and LASSO regression to identify predictive factors. A nomogram prediction model for PB in children with MPP was constructed. The discriminative ability of the model was assessed using receiver operating characteristic (ROC) curve analysis, model calibration was evaluated with calibration curves, and clinical utility was appraised through decision curve analysis. Results Compared with the non-PB group, the PB group exhibited significantly longer disease duration prior to bronchoscopy, prolonged fever duration, higher fever peaks, higher proportions of patients with a family history of allergy and personal allergy history, and a higher proportion of patients with pleural effusion, as well as significantly elevated levels of white blood cell count, neutrophil percentage, C-reactive protein, procalcitonin, fibrinogen, D-dimer, aspartate aminotransferase, alanine aminotransferase, creatine kinase, lactate dehydrogenase, immunoglobulin A, and interleukin-6, along with a significantly lower lymphocyte percentage (all P<0.05). LASSO regression analysis identified pleural effusion, procalcitonin, D-dimer, and lactate dehydrogenase as major predictive factors for PB occurrence in children with MPP. The nomogram model based on these factors demonstrated good discriminative ability (area under the ROC curve: 0.852 in the training set and 0.830 in the validation set), with satisfactory calibration and clinical benefit. Conclusions The nomogram prediction model based on pleural effusion, procalcitonin, D-dimer, and lactate dehydrogenase provides effective predictive performance for the occurrence of PB in children with MPP.