LAUSR

Background Refractory Mycoplasma pneumoniae pneumonia (RMPP) presents a significant clinical challenge due to its potential for severe complications and long-term sequelae in children. While several risk factors have been identified, an accurate and early predictive tool to guide timely clinical intervention is urgently needed. This study aimed to identify the clinical risk factors and develop a nomogram model for the early prediction of RMPP. Methods This prospective study enrolled children diagnosed with Mycoplasma pneumoniae pneumonia (MPP) who visited The Second People’s Hospital of Changzhou from June to December 2024. RMPP was defined as persistent fever and progressive pulmonary infiltrates despite ≥7 days of standard macrolide therapy. Baseline demographic and clinical variables were assessed at admission. Independent risk factors for RMPP were identified using multivariate logistic regression and were used to construct a predictive nomogram. The performance of the nomogram model was assessed by calibration curves, area under the receiver operating characteristic (ROC) curves (AUC), and the decision curve analysis (DCA). Results A total of 210 children were included, among whom 105 were diagnosed with RMPP. The median age was 7.0 years (interquartile range, 5.0–8.5 years), and 42.4% of participants were male. No significant differences in age or sex were observed between groups (P<0.05). Multivariate analysis identified fever duration [odds ratio (OR) =2.15, P<0.001], duration of glucocorticoid use (OR =1.56, P<0.001), and YKL-40 levels (OR =1.01, P=0.001) as independent risk factors for RMPP. The nomogram incorporating these three factors demonstrated excellent discrimination with an AUC of 0.92 (95% confidence interval: 0.88–0.96). Calibration curve and Hosmer-Lemeshow test (P>0.99) indicated excellent calibration. DCA confirmed the clinical utility of the nomogram, showing net benefit across a wide threshold probability range (0.04–0.94). Conclusions The nomogram constructed based on fever duration, glucocorticoid use duration, and YKL-40 level shows promise for early prediction of RMPP in children.