• Adrienne G Randolph, Christian Brun-Buisson, and Donald Goldmann.
• Department of Anesthesia, Division of Critical Care, Children's Hospital, Boston, MA, USA.
• Pediatr Crit Care Me. 2005 May 1;6(3 Suppl):S19-24.

ObjectiveTo define central venous catheter-related infections in infants and children for the purpose of enrolling children in sepsis studies, for epidemiology and surveillance studies, and for clinical management.MethodsReview of the literature and consensus of experts.ResultsNo changes were made to the current Centers for Disease Control and Prevention criteria for defining local catheter infection. Because catheter tips are not available as often in children as in adults, smaller blood volumes are drawn per culture decreasing sensitivity, and antibiotics are rarely withheld, slight modifications to the existing adult Centers for Disease Control and Prevention criteria were made to increase practical use. Catheter-related bloodstream infection was categorized as definite, probable, and possible based on culture results and clinical symptoms.ConclusionsFor the purposes of enrolling patients with sepsis in clinical trials, only patients who meet criteria for definite catheter-related bloodstream infection should be categorized as having the catheter as the infection source. Because many patients suspected of having catheter-related bloodstream infection do not have positive blood culture results, which makes the confirmation of infection difficult, we recommend that these patients not be enrolled in sepsis trials. Because catheter tips are often not obtained for culture in children, the epidemiology of catheter-associated bloodstream infection (bloodstream infection in a patient who has a central venous catheter and no other obvious source of infection) is better understood than the epidemiology of confirmed catheter-related bloodstream infection in infants and children. Definitions for catheter-related bloodstream infection that compare the through-catheter and peripheral culture for time to positivity or for quantitative growth are unlikely to be falsely positive, but sensitivity requires further validation.

24 keywords...

hide…