What question did this study set out to answer?

To analyze clinical data and develop an early prediction model for severe Mycoplasma pneumoniae pneumonia in children.

February 23, 2026Journal of Pediatric Infectious Diseases0 citationsOpen Access

IL-36γ and the Development and Validation of an Early Prediction Model for Severe Mycoplasma Pneumoniae Pneumonia in Children

XTXiaoDong TangSuzhou Municipal Hospital YLYing LiBeijing Tongren Hospital PYPao YuSuzhou Municipal Hospital

Key Points

To analyze clinical data and develop an early prediction model for severe Mycoplasma pneumoniae pneumonia in children.
Retrospective analysis of clinical data from 345 children with Mycoplasma pneumoniae pneumonia.
Serum IL-36γ levels measured using ELISA at admission.
Independent risk factors identified using binary logistic regression.
Nomogram prediction model constructed based on identified factors.
Model validated with an external cohort of 148 MPP patients.
Age, ESR, albumin, hsCRP, and IL-36γ identified as independent factors for severe MPP.
Nomogram model showed an AUC of 0.89 in both training and validation sets.
Model accuracy was 0.81 in training and 0.83 in validation with high sensitivity and specificity.
Calibration curves indicated good model fit in both datasets.
Decision curve analysis suggested the model's clinical applicability.

Abstract

Objective: This study aimed to retrospectively analyze clinical data from pediatric patients with Mycoplasma pneumoniae pneumonia (MPP), investigate the expression and clinical significance of serum IL-36γ, and to develop and validate an early identification model for severe MPP (SMPP). Methods: Clinical data were collected from 345 children diagnosed with MPP in our department between July 1, 2023, and December 31, 2024. Based on diagnostic criteria, enrolled patients were categorized into an SMPP group (n=145) and a non-severe MPP (NSMPP) group (n=200). Serum IL-36γ levels at admission were measured using ELISA. Independent risk factors for SMPP were identified via binary logistic regression, and a nomogram prediction model was constructed. An external validation cohort consisted of clinical data from 148 MPP patients (61 SMPP, 87 NSMPP) treated concurrently in the Xuzhou Medical University Affiliated Hospital. The model was evaluated using ROC curves, calibration curves, and decision curve analysis (DCA). Results: Age, ESR, albumin, hsCRP, and IL-36γ may serve as independent factors for the early identification of children with SMPP. Based on these five indicators, a nomogram prediction model was developed. ROC curve analysis showed an area under the curve (AUC) of 0.89 in the training set (accuracy = 0.81, sensitivity = 0.78, specificity = 0.84), and an AUC of 0.89 in the validation set (accuracy = 0.83, sensitivity = 0.80, specificity = 0.87). Calibration curves indicated good model fit in both the training and validation sets. Furthermore, decision curve analysis (DCA) suggested that the model demonstrates potential clinical applicability across a range of threshold probabilities. Conclusion: This study developed a nomogram model that showed promising predictive performance for the early identification of pediatric SMPP. Following internal and external validation, our findings suggest that this model could potentially serve as a decision-support tool to aid in risk stratification and guide timely clinical intervention, though further prospective studies are warranted to confirm its utility.

AIに質問

Bookmark

View Full Paper