A nomogram for predicting macrolide resistance-associated 23S rRNA mutations in children with mycoplasma pneumoniae pneumonia

Macrolide resistance-associated 23S rRNA mutations are common in children with Mycoplasma pneumoniae pneumonia (MPP). However, limited access to molecular testing and long turnaround times in routine clinical practice hinder the early identification of children at high risk of harboring these mutations. This study aimed to develop and validate a nomogram for predicting macrolide resistance-associated 23S rRNA mutations in children with MPP. This retrospective multicenter cohort study included hospitalized children with MPP who underwent testing for macrolide resistance-associated 23S rRNA mutations. Patients were randomly divided into a training cohort and an internal validation cohort in an 8:2 ratio, while an independent cohort from another tertiary hospital served as an external validation cohort. Candidate predictors were selected using least absolute shrinkage and selection operator (LASSO) regression to minimize the risk of overfitting, followed by multivariable logistic regression to develop the prediction model. A nomogram was constructed based on the final predictors. Model performance was assessed in terms of discrimination, calibration, and clinical utility using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA), respectively. A total of 282 hospitalized children with MPP who underwent testing for macrolide resistance-associated 23S rRNA mutations were included, comprising 225 patients in the training cohort and 57 in the internal validation cohort. An additional 165 patients from an independent center formed the external validation cohort. Four variables were identified as independent predictors of these mutations: C-reactive protein (CRP), lactate dehydrogenase (LDH), D-dimer, and pulmonary consolidation on chest radiographs. The model demonstrated good discrimination, with AUCs of 0.870, 0.843, and 0.826 in the training, internal validation, and external validation cohorts, respectively. Calibration curves showed good agreement between predicted and observed probabilities, while DCA indicated favorable clinical utility across a wide range of threshold probabilities. We developed and validated a parsimonious nomogram for predicting macrolide resistance-associated 23S rRNA mutations in hospitalized children with MPP. The model showed good discrimination, calibration, and clinical utility in the study population, and may facilitate early individualized risk stratification before molecular testing results become available. Further validation is warranted in more diverse real-world populations, particularly among children with mixed respiratory infections.