Clinical Characteristics, Management, and 30-Day Mortality Predictors in an 18-Year Pediatric Candidemia Cohort

Résumé fourni par la source

Pediatric candidemia is a major cause of invasive fungal infections in hospitalized children, but long-term data on epidemiology, management, and mortality predictors remain limited. We conducted an 18-year retrospective cohort study of 465 pediatric candidemia episodes at a tertiary referral center in western Turkey between 2008 and 2025. The primary outcome was crude 30-day mortality; associated factors were assessed using univariable analyses, Kaplan–Meier estimates, and multivariable logistic regression. Non-albicans Candida species predominated, with Candida parapsilosis as the most frequent isolate (46%). Central venous catheters were present in 88.4% of episodes. Crude 30-day mortality was 10.8%. Reduced survival was observed among patients without catheter removal and among those with thrombocytopenia, severe neutropenia, or immunosuppressive therapy. Among 341 episodes classified as central line-associated bloodstream infections, crude 30-day mortality differed significantly by catheter removal timing. Mortality was 4.8% with catheter removal within 72 h versus 13.1% without early removal (p = 0.022). Using a 48 h threshold, mortality was 3.1% with removal within 48 h versus 12.3% without removal within 48 h (p = 0.029). In multivariable analysis, failure to remove the catheter was the strongest independent factor associated with mortality (adjusted odds ratio, 6.63; 95% confidence interval, 2.85–15.42; p < 0.001). Antifungal resistance patterns were not consistently associated with mortality. In this large pediatric candidemia cohort, 30-day mortality was mainly associated with host vulnerability and modifiable management factors, underscoring the importance of timely source control.