Risk Prediction Model for Necrotizing Pneumonia in Children with Mycoplasma pneumoniae Pneumonia

Abstract

Objective: pneumonia (MPP) and construct a prediction model. Methods: The clinical data with MPP at the Children's Hospital of Kunming Medical University from January 2014 to November 2022 were retrospectively analyzed. Eighty-four children with MPP who developed NP were divided into the necrotizing group, and 168 children who did not develop NP were divided into the non-necrotizing group by propensity-score matching. LASSO regression was used to select the optimal factors, and multivariate logistic regression analysis was used to establish a clinical prediction model. The receiver operating characteristic (ROC) curve and calibration curve were used to evaluate the discrimination and calibration of the nomogram. Clinical decision curve analysis was used to evaluate the clinical predictive value. Results: LASSO regression analysis showed that bacterial co-infection, chest pain, LDH, CRP, duration of fever, and D-dimer were the influencing factors for NP in children with MPP (P < 0.05). The results of ROC analysis showed that the AUC of the prediction model established in this study for predicting necrotizing MPP was 0.870 (95% CI: 0.813-0.927, P < 0.001) in the training set and 0.843 (95% CI: 0.757-0.930, P < 0.001) in the validation set. The Bootstrap repeated sampling for 1000 times was used for internal validation, and the calibration curve showed that the model had good consistency. The Hosmer-Lemeshow test showed that the predicted probability of the model had a good fit with the actual probability in the training set and the validation set (P values of 0.366 and 0.667, respectively). The clinical decision curve showed that the model had good clinical application value. Conclusion: The prediction model based on bacterial co-infection, chest pain, LDH, CRP, fever duration, and D-dimer has a good predictive value for necrotizing MPP.