•  
  •  
 

Abstract

Background: Tracheostomized pediatric patients are at increased risk for ventilator-associated pneumonia (VAP), and distinguishing true infection from airway colonization remains challenging. Therefore, reliable diagnostic tools are needed to support timely identification and avoid unnecessary antibiotic use. This study aimed to evaluate the diagnostic performance of CPIS, CPIS–Procalcitonin (CPIS-PCT), and the Phoenix Sepsis Score (PSS) in identifying VAP in critically ill tracheostomized children. Methods: This single-center, retrospective observational study included critically ill tracheostomized children treated in a pediatric intensive care unit between January 2019 and December 2024. Patients receiving mechanical ventilation for ≥48 hours with clinical suspicion of VAP were evaluated. Results: Eighty tracheostomized pediatric patients were included; 43 (53.7%) developed VAP. Mortality was significantly higher in the VAP group (14.0% vs. 0%, p=0.018). CPIS (AUC 0.976) and CPIS-PCT (AUC 0.973) demonstrated excellent diagnostic discrimination, while PSS showed good performance (AUC 0.865). Optimal cutoff values were ≥3 for CPIS, ≥4 for CPIS-PCT, and ≥2 for PSS. Purulent sputum or change in sputum characteristics and increased oxygen/ventilator support were identified as clinical features associated with VAP diagnosis. Conclusion: CPIS shows potential diagnostic utility for identifying VAP in tracheostomized pediatric patients. Although procalcitonin levels were elevated in the VAP group, combining procalcitonin with CPIS did not provide significant additional diagnostic benefit. These findings suggest that CPIS, in conjunction with close monitoring of clinical features, may support timely diagnosis and management; however, given the exploratory nature of this study and potential limitations regarding bias, these findings require further validation.

Download

Share

COinS