Abstract: Objective To analyze and summarize the clinical manifestations and chest imaging features of necrotizing pneumonia caused by severe mycoplasma pneumoniae pneumonia in 2023, and to identify risk factors for necrotizing pneumonia. Method The clinical data and radiology data were collected on 50 children with severe mycoplasma pneumoniae pneumonia treated in Shenzhen Children's Hospital from September to November 2023. Children were examined with the chest CT scan and enhancement, and were divided into necrotizing and non-necrotizing pneumonia groups based on the enhancement of the lung parenchyma with or without necrosis. Clinical and radiology features were compared between two groups. Multivariate Logistic regression analysis was applied to analyze the indexes with significant differences in the univariate analysis. The ROC curve was used to analyze the diagnostic efficiency of different indicators and combined indicators. Result There were 25 males and 25 females among 50 children with a mean age of (6.56±3.13) years. Six cases (12.00%) had underlying diseases. Multiple lobes and segments of both lungs were involved, with partial wedge-shaped consolidation and bronchial wall thickening. Twenty-nine cases (58.00%) had a small amount of pleural effusion, with hilar lymph node enlargement in 44 cases (88.00%), pulmonary parenchyma necrosis in 28 cases (56.00%) and typical necrotizing cavities in 3 cases. All patients recovered and discharged. There was no significant difference in sex, age, hypoxemia, high fever or C-reactive protein between these two groups (P>0.05). The median hospitalization days, D-dimer, actate dehydrogenase (LDH) , alanine aminotransferase (ALT) and pleural effusion in necrotizing pneumonia group were significantly higher than those in the non-necrotizing pneumonia group (P<0.05). Multivariate Logistics regression analysis showed that LDH (OR=1.007, 95%CI 1.001-1.013) and pleural effusion(OR=35.060, 95%CI 4.523-271.777) were independent predictors of necrotizing pneumonia in children with severe mycoplasma pneumoniae pneumonia. When LDH and pleural effusion were used for joint prediction, the sensitivity was 0.778, the specificity was 0.773, and the area under the curve was 0.852 (95%CI 0.746-0.958). Conclusion Children with severe mycoplasma pneumoniae who present with pleural effusions and apparently elevated LDH should be alerted to the possibility of combination with necrotizing pneumonia. Early recognition and effective clinical intervention should be performed.

Key words: Mycoplasma pneumoniae, Necrotizing pneumonia, Computed tomography, Pleural effussion, Lactate dehydrogenase, Children