Predictors of secondary streptococcus pneumoniae and/or haemophilus influenzae infection in children hospitalised with RSV, and development of a clinical prediction model to inform antibiotic decisions: a retrospective cohort study based on coloniser

Background Respiratory syncytial virus (RSV) is the leading cause of acute respiratory tract infection (ARTI) in infants and young children. Co-infection with Streptococcus pneumoniae and/or Haemophilus influenzae substantially increases disease severity. However, the detection rates of these bacteria in the nasopharynx of young children vary considerably, which can lead to overestimation of bacterial infection and inappropriate antibiotic use. This study aims to develop a nomogram for predicting secondary S. pneumoniae and/or H. influenzae infection in children hospitalised with RSV, based on a rigorous distinction between colonisation and true pathogenicity. Methods We conducted a retrospective, single-centre study of children hospitalised with ARTI due to RSV between January 2023 and October 2025. According to predefined criteria for bacterial infection, 1,573 patients were identified and randomly assigned in a 7:3 ratio to a training set ( n = 1,104) and an internal validation set ( n = 469). Variable selection was performed using least absolute shrinkage and selection operator (LASSO) regression with ten-fold cross-validation. A nomogram was then constructed using multivariable logistic regression. Model performance was assessed by the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis. Result Multivariable logistic regression identified lung consolidation, RSV-CLASS, WBC, NLR, SF, and IL-6 as independent predictors. The nomogram achieved an AUC of 0.872 (95% CI: 0.841–0.898) in the training set and 0.829 (95% CI: 0.780–0.877) in the internal validation set, with an optimism-corrected C-statistic of 0.864. Calibration showed miscalibration in the internal validation set (slope 0.762, intercept −0.350), indicating a tendency to overestimate risk, and decision curve analysis indicated net benefit at threshold probabilities of 0.2–0.6. Conclusion Based on a rigorous distinction between colonising and pathogenic bacteria, this study constructed a nomogram integrating six variables that combine clinical, laboratory, and imaging findings. This nomogram offers clinicians a reference when considering antibiotics for children hospitalised with RSV and secondary S. pneumoniae or H. influenzae infection, but its current utility is primarily as a rule-out tool given the low PPV; it should not be used alone to guide antibiotic initiation. External validation is required before broader clinical implementation.