Abstract
AIM: This study identified risk factors for ventilator-associated pneumonia (VAP) in mechanically ventilated (MV) children and developed a risk prediction model to guide precision nursing interventions. BACKGROUND: MV supports critically ill pediatric patients by improving oxygenation but may induce lung injury and increase VAP incidence. METHODS: We retrospectively analyzed pediatric MV patients admitted to the Pediatric Intensive Care Unit (PICU) at Shengjing Hospital, China Medical University. Independent VAP risk factors were identified using binary logistic regression, and a prediction model was developed/validated with R software. RESULTS: Absence of early enteral nutrition (EEN), duration of MV, frequency of endotracheal suctioning, central venous catheterization, and the types of antibiotics used were independent VAP risk factors (p < 0.05). The model demonstrated strong discrimination, with ROC-AUCs of 0.870 (95% CI: 0.816-0.924) and 0.761 (95% CI: 0.653-0.868) for derivation and validation cohorts, respectively. Hosmer-Lemeshow tests confirmed calibration (p=0.970 and p=0.524). CONCLUSIONS: This validated model effectively stratifies VAP risk in MV children, enabling early identification of high-risk patients and facilitating targeted nursing strategies. IMPLICATIONS FOR NURSING MANAGEMENT: The model allows rapid clinical screening for high-risk pediatric VAP cases. Interventions focused on modifiable risk factors may reduce VAP incidence.