Medical care related laboratory-confirmed bloodstream infections in paediatrics

The aim of this survey was to describe the incidence, epidemiology, microbiology, risk factors and outcome of medical care related laboratory-confirmed bloodstream infections (LCBIs) observed during a twelve-month prospective study in a Paediatric Teaching Hospital in Turin, Italy. Inclusion criteria were clinical signs of sepsis and positivity of one or more of the following tests: blood culture, polymerase chain reaction for bacterial and fungal DNA on blood, and culture on intravascular device tips. In all, 140 episodes of sepsis were documented in 131 children: 37 (26.4%) were healthcare outpatient-associated, 91 (65.0%) healthcare-associated and 12 (8.6%) community-acquired. The overall incidence of healthcare-associated LCBIs was 13.6/1,000 hospitalized patients and incidence density 1.4/1,000 inpatient days. The overall mortality was 3.9%. Forty-seven (36.7%) episodes involved newborns and 107 (83.6%) episodes were observed in children with an indwelling central venous catheter. Coagulase-negative staphylococci (26.8%), Staphylococcus aureus (15.2%), Escherichia coli (8.7%) and Candida spp. (7.2%) were responsible for the majority of cases. 9.5% of S. aureus isolates were methicillin-resistant and 6.5% of Gram negatives were extended-spectrum beta-lactamase-producing. Incidence and epidemiology of medical care related LCBIs were similar to the existing literature data. LCBIs caused by antibiotic-resistant microorganisms were fewer and mortality rate was lower. Most of the LCBIs recorded involved newborns and oncological children.