Predictive value of a nomogram model based on complex immunoinflammatory indicators for respiratory failure in the children with severe adenovirus pneumonia

Gao Zhe; Li Jing; Yan Yang; Mao Shujiong

doi:10.3969/j.issn.1002-1949.2025.03.006

Abstract：Objective To investigate the predictive value of a nomogram model based on complex immunoinflammatory indicators for respiratory failure in the children with severe adenovirus pneumonia.Methods A study was carried out retrospectively on 202 children diagnosed with severe adenovirus pneumonia in the Third People's Hospital of Yuhang District,Hangzhou City from January 2021 to October 2024.The clinical parameters measured during their hospital stay were documented,and the children were categorized into two groups:respiratory failure group(n=104)and without respiratory failure group(n=98)based on whether respiratory failure occurred within 3 days after the onset of severe adenovirus pneumonia.Logistic regression analysis was employed to pinpoint the key risk factors associated with respiratory failure,and a nomogram prediction model was constructed accordingly.The model's accuracy was assessed by internal validation techniques such as the calibration curve and the Bootstrap method.Furthermore,the model's performance was evaluated by using the receiver operating characteristic(ROC)curve,while the clinical decision curve analysis(DC A)was utilized to investigate the practical utility of the model.Results In the children with severe adenovirus pneumonia,the frequency of respiratory failure was found to be 51.5％.Logistic regression analysis showed that several factors were independently associated with an increased risk of respiratory failure in these children,including hypoalbuminemia(OR=9.845,95％CI 1.904-50.903,P=0.006),multiple organ dysfunction score(MODS,OR=1.773,95％CI 1.113-2.825,P=0.016),confusion,uremia,respiratory rate,blood pressure,age over 65 years(CURB-65,OR=1.884,95％CI 1.160-3.060,P=0.010),acute physiology and chronic health years evaluationⅡ(APACHE Ⅱ,OR=1.293,95％CI 1.118-1.496,P＜0.001),neutrophil/lymphocyte ratio(NLR,OR=1.868,95％CI 1.271-2.744,P=0.001),and systemic immunoinflammatory index(SII,OR=1.002,95％CI 1.001-1.003,P=0.001).To assist in predicting the likelihood of respiratory failure,a nomogram prediction model was developed by using these independent risk factors.ROC curve analysis showed area under the curve(AUC)of the nomogram model in predicting respiratory failure was 0.806,with a 95％CI of 0.745-0.874,indicating strong predictive performance.Furthermore,the Hosmer-Lemeshow test indicated that the model had a good fit(P=0.089),and DC A suggested significantly clinical utility of the model.Conclusions Respiratory failure is a common occurrence in children suffering from severe adenovirus pneumonia.A nomogram model,incorporating hypoalbuminemia,MODS score,CURB-65 score,APACHE Ⅱ score,NLR and SII,has demonstrated practical utility in forecasting the likelihood of respiratory failure in such cases.Early monitoring and tailored interventions targeting at these parameters could potentially enhance the overall prognosis for affected children.