摘要目的:探索预测儿童肺炎支原体坏死性肺炎(MPNP)发生的指标。方法:本研究为观察性研究，采用非随机抽样的方法收集2015年12月至2022年12月河北省儿童医院收治的179例肺炎支原体感染所致大叶性肺炎患儿为研究对象，根据胸部CT表现将患儿分为坏死性肺炎(NP)组67例和非NP组112例。比较2组患儿的临床资料、并发症和实验室指标，采用多因素logistic回归分析确定MPNP发生的独立危险因素，绘制受试者工作特征曲线分析独立危险因素对肺炎支原体肺炎(MPP)患儿发生NP的预测价值。结果:NP组住院天数多于非NP组[19(12，21) d比10(7，11) d]，低氧血症、应用机械通气、静脉注射免疫球蛋白比例均大于非NP组[53.7%(36/67)比37.5%(42/112)、6.0%(4/67)比0(0)、31.3%(21/67)比10.7%(12/112)]，支气管镜下改变包括黏膜糜烂、大量分泌物栓塞、支气管塑形、灌洗液脓性浑浊、支气管闭塞比例均大于非NP组[80.6%(54/67)比33.9%(38/112)、77.6%(52/67)比51.8%(58/112)、9.0%(6/67)比0.9%(1/112)、55.2%(37/67)比28.6%(32/112)、6.0%(4/67)比0(0)]，差异均有统计学意义(均 P<0.05)。NP组胸腔积液、脓胸、气胸比例均高于非NP组[61.2%(41/67)比34.8%(39/112)、7.5%(5/67)比0(0)、34.3%(23/67)比13.4%(15/112)]，神经系统、血液系统、皮疹和关节肌肉表现并发症比例也高于非NP组[16.4%(11/67)比5.4%(6/112)、7.5%(5/67)比0.9%(1/112)、16.4%(11/67)比4.5%(5/112)]，差异均有统计学意义(均 P<0.05)。NP组白细胞计数、D-二聚体(D-D)、红细胞沉降率、乳酸脱氢酶(LDH)均高于非NP组[12.6(10.6，13.6)×10 9/L比11.1(9.4，13.5)×10 9/L、2.13(1.99，2.51) mg/L比1.69(1.50，1.95) mg/L、42.5(37.8，51.6) mm/1 h比41.7(31.6，48.1) mm/1 h、429.8(389.5，467.8) U/L比349.7(308.4，396.6) U/L]，差异均有统计学意义(均 P<0.05)。D-D增高、LDH增高是MPNP发生的独立危险因素( OR＝1.460，95% CI：1.282~1.662， P<0.001； OR＝1.011，95% CI：1.004~1.017， P＝0.001)。D-D、LDH对MPP患儿发生NP均有一定预测价值，曲线下面积分别为0.851(95% CI：0.796~0.906， P<0.001)、0.764(95% CI：0.690~0.837， P<0.001)。当D-D的最佳截断值为1.795 mg/L时，敏感度为66.1%，特异度为91.0%；当LDH的最佳截断值为393.500 U/L时，敏感度为73.2%，特异度为74.6%。 结论:当MPP患儿出现D-D>1.795 mg/L、LDH>393.500 U/L时，需警惕MPNP的发生。

abstractsObjective:To identify predictive factors for Mycoplasma pneumoniae necrotizing pneumonia (MPNP) in children.Methods:It was an observational study involving 179 children with lobar pneumonia caused by Mycoplasma pneumoniae admitted to Hebei Children′s Hospital from December 2015 to December 2022 selected via a non-random sampling method.Based on chest CT manifestations, children were divided into necrotizing pneumonia group (NP group, n=67) and non-necrotizing pneumonia group (non-NP group, n=112). Clinical data, complications, and laboratory indicators were compared between groups.Multivariate logistic regression analysis was performed to identify independent risk factors for the occurrence of MPNP in children, and their potential in predicting NP in children with Mycoplasmal pneumoniae pneumonia (MPP) was determined by plotting receiver operating characteristic (ROC) curves. Results:The length of stay in NP group was significantly longer than that in non-NP group (19 [12, 21] d vs 10 [7, 11] d), while the proportions of hypoxemia (53.7% [36/67] vs 37.5% [42/112]), mechanical ventilation (6.0% [4/67] vs 0 [0]), and intravenous immunoglobulin injection (31.3% [21/67] vs 10.7% [12/112]) were significantly higher (all P<0.05). The proportions of bronchoscopic changes like mucosal erosion (80.6% [54/67] vs 33.9% [38/112]), massive secretion embolism (77.6% [52/67] vs 51.8% [58/112]), bronchial shaping (9.0% [6/67] vs 0.9% [1/112]), purulent turbidity of lavage fluid (55.2%[37/67] vs 28.6% [32/112]) and bronchial occlusion (6.0% [4/67] vs 0 [0]) in NP group were significantly higher than those in non-NP group (all P<0.05). The incidences of pleural effusion (61.2% [41/67] vs 34.8% [39/112]), empyema (7.5% [5/67] vs 0 [0]), pneumothorax (34.3% [23/67] vs 13.4% [15/112]), complications in the nervous system (16.4% [11/67] vs 5.4% [6/112]), hematological complications (7.5% [5/67] vs 0.9% [1/112]) and the involvement of rash and joint and muscle (16.4% [11/67] vs 4.5% [5/112]) in NP group were significantly higher than those in non-NP group (all P<0.05). The white blood cell count (WBC, 12.6 [10.6, 13.6]×10 9/L vs 11.1 [9.4, 13.5]×10 9/L), D-dimer (D-D, 2.13 [1.99, 2.51] mg/L vs 1.69 [1.50, 1.95] mg/L), erythrocyte sedimentation rate (ESR, 42.5 [37.8, 51.6] mm/1 h vs 41.7 [31.6, 48.1] mm/1 h), and lactate dehydrogenase (LDH, 429.8 [389.5, 467.8] U/L vs 349.7 [308.4, 396.6] U/L) in NP group were significantly higher than those in non-NP group (all P<0.05). Increased D-D ( OR=1.460, 95% CI: 1.282-1.662, P<0.001) and LDH level ( OR=1.011, 95% CI: 1.004-1.017, P=0.001) were independent risk factors for MPNP in children.Both D-D and LDH exerted the potential in predicting NP in children with MPP, with the area under the curve (AUC) of 0.851 (95% CI: 0.796-0.906, P<0.001) and 0.764 (95% CI: 0.690-0.837, P<0.001), respectively.At the cut-off value of 1.795 mg/L, the sensitivity and specificity of D-D in predicting NP in children with MPP were 66.1% and 91.0%, respectively.At the cut-off value of 393.500 U/L, the sensitivity and specificity of LDH in predicting NP in children with MPP were 73.2% and 74.6%, respectively. Conclusions:MPNP should be concerned in children with MPP who had laboratory findings of D-D>1.795 mg/L or LDH>393.500 U/L.