Development and Validation of Nomogram Models for Predicting Respiratory Infection Risk and Asthma Treatment Prognosis in Children with Bronchial Asthma

Xinhua Wang¹Yuan Lv^2*

• ¹The Fourth Hospital of Baotou, Baotou, China
• ²The Cancer Hospital of Baotou, Baotou, Inner Mongolia, China

Background: Bronchial asthma significantly impacts pediatric health and is often complicated by respiratory infections, resulting in varied treatment outcomes. Predictive models for these complications can provide essential insights for personalized management strategies. Objective: This study aimed to develop and validate two separate nomogram models: one for predicting the risk of respiratory infections and another for predicting the prognosis of asthma treatment in children with bronchial asthma. Methods: We conducted a retrospective cohort study involving 660 children diagnosed with bronchial asthma between January 2018 and January 2023 at The Cancer Hospital of Baotou. The participants were divided into a development set (440 children) and a validation set (220 children). Multivariable logistic regression analyses were performed to identify predictors of respiratory infections and asthma prognosis, from which we developed two nomograms. The performance of these models was assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). Results: The nomogram for predicting respiratory infections included factors such as age, frequency of asthma attacks within six months, intravenous corticosteroid use, serum total IgE, and eosinophil count, demonstrating good predictive accuracy with AUC values of 0.871 and 0.917 in the development and validation sets, respectively. The nomogram for predicting asthma treatment prognosis incorporated parental smoking history, family history of allergic diseases, and previous respiratory infections, achieving AUC values of 0.822 and 0.853 in the development and validation cohorts. Both models showed excellent agreement between predicted and observed outcomes, supported by calibration curves and DCA. Conclusion: The developed nomograms are valuable tools for predicting the risk of respiratory infections and the prognosis of asthma treatment in pediatric patients. These tools can significantly aid in clinical decision-making by considering individual patient characteristics, thereby enhancing personalized asthma management.