The Etiology of Bloodstream Infections at an Italian Pediatric Tertiary Care Hospital: A 17-Year-Long Series

Knowledge of epidemiology is essential for guiding correct antibiotic prescription, reducing bacteremia-associated mortality, and implementing targeted infection control programs. However, only a few studies have reported on the epidemiology of bloodstream infections (BSIs) in pediatrics. We performed a retrospective analysis of all BSIs (excluding those caused by common skin contaminants) diagnosed from 2006 to 2022 in patients younger than 18 years who were treated at an Italian pediatric tertiary care hospital. Overall, 2395 BSIs were recorded, including 2207 (92.15%) due to bacteria and 188 (7.85%) due to fungi. The incidence rate (BSIs/10,000 hospital discharges, IR) of bacterial BSIs significantly increased during the study period. In particular, BSIs caused by S. aureus (including MRSA), Enterobacterales (including ESBL and AmpC producers), Enterococcus spp., and P. aeruginosa became more common. The frequency of carbapenem-resistant strains was <1% and stable over time. Conversely, there was a significant reduction in the incidence of BSIs due to S. pneumoniae. The BSIs were stratified by patient age, and S. aureus was the most frequent cause of BSIs in all age groups, while E. coli was the most frequent in the Enterobacterales family. S. agalactiae was the third most frequent cause of neonatal early-onset BSIs. The prevalence of Enterococcus spp. increased in the subgroups from 8 days to 5 years of age, while P. aeruginosa became more prevalent in children over 5 years of age. S. aureus was also the most frequent isolate in both community- and hospital-onset BSIs, followed by E. coli. The prevalence of multidrug-resistant (MDR) pathogens was very low. It was <5% for both Gram-positive (i.e., MRSA and VRE) and Gram-negative (ESBL, AmpC, and carbapenem-resistant) pathogens, and MDR pathogens were almost exclusively detected in hospital-onset BSIs. Fungi accounted for just under 8% of BSIs. C. albicans was the most frequently isolated strain, followed by C. parapsilosis. Notably, the IR of fungemia did not change significantly during the study period, in spite of an increase in the absolute number of events. The continuous monitoring of local epidemiology is essential to identify changes in the IRs of pathogens and antibiotic susceptibility and to guide antibiotic treatments, especially in the phase when antibiograms are not yet available.