[Nomogram predictive model for the risk of critical pneumonia in children infected with respiratory syncytial virus].

Objective: To develop and validate a nomogram model for predicting the risk of critical pneumonia caused by respiratory syncytial virus (RSV) infection in children. Methods: A retrospective cohort study was conducted. Clinical features, underlying diseases, and laboratory data of 515 children with severe RSV pneumonia who were admitted to Shanghai Children's Hospital from January 2021 to December 2024 were collected. Based on the diagnostic criteria for community-acquired pneumonia, patients were divided into a critical group and a severe group. Multivariate logistic regression analysis was used to identify risk factors for critical pneumonia in children with RSV infection, and a nomogram model was constructed. The predictive performance of the nomogram model was evaluated using receiver operating characteristic curves and calibration curves. Results: Among the 515 children with severe RSV pneumonia, there were 299 males and 216 females, with an age of 20 (3, 36) months. The critical group comprised 207 cases (40.2%), and the severe group 308 cases (59.8%). The proportions of children with age at present ≤6 months, wheezing, obesity, premature birth, and underlying diseases in the critical group were all higher than those in the severe group (all P<0.05); the duration of fever, neutrophil-to-lymphocyte ratio (NLR), and procalcitonin levels were also higher than those in the severe group (all P<0.05). Multivariate Logistic regression analysis showed that age at presentation ≤6 months, wheezing, underlying diseases, NLR and procalcitonin (OR=3.30, 3.10, 6.57, 1.16, 1.22; 95%CI 2.07-5.25, 1.96-4.92, 3.97-10.86, 1.06-1.27, 1.05-1.41; all P<0.05) were independent risk factors for critical RSV pneumonia. A nomogram model was developed based on these variables. After bootstrap correction, the model demonstrated good discriminative ability (C-index 0.80, 95%CI 0.77-0.84). The calibration curve and Hosmer-Lemeshow test showed good consistency between the predicted and actual risks (χ²=11.64, P=0.168). Decision curve analysis showed a net benefit within the threshold probability range of 20% to 80%. Conclusions: The nomogram model constructed based on age at presentation ≤6 months, wheezing, underlying diseases, NLR and procalcitonin can effectively predict the risk of critical RSV pneumonia and has certain clinical application value.