Background: Children with central venous catheters and suspected bloodstream infection are often hospitalized for 48 hours to receive empiric antibiotic therapy pending blood-culture results. Continuous monitoring blood-culture systems allow for more rapid detection of bloodstream infection than previous blood-culture systems, a feature that may facilitate earlier determination of the true presence or absence of bloodstream infection and shorten empiric antibiotic therapy and duration of hospitalization.
Methods: This retrospective cohort study included children with central venous catheters who were diagnosed with laboratory-confirmed bloodstream infection after evaluation in the ambulatory care setting.
Results: Two-hundred episodes of bloodstream infection were included. The median patient age was 5.5 years. Central venous catheters were in place for a median of 80.5 days. Gram-negative bacteria accounted for 51% of infections as part of either a monomicrobial (25%) or polymicrobial (26%) infection. The overall median time to blood-culture positivity was 14 hours. The predicted probability for a culture being positive at 36 hours was 99.2% for infections caused by gram-negative bacteria and 96.6% for any infection after adjusting for age, catheter type, and recent antibiotic use. In a multivariate Cox proportional-hazards regression model, polymicrobial infections with > or = 1 gram-negative bacteria and monomicrobial infections caused by gram-negative bacteria were independently associated with an earlier time to blood-culture positivity after adjusting for age, catheter type, and recent antibiotic use.
Conclusions: The time to blood-culture positivity depends on bacterial category. Bloodstream infections caused by gram-negative bacteria are detected most quickly. Our data suggest that discontinuation of empiric antibiotic coverage may be warranted in clinically stable children with central venous catheters if the blood-culture results remain negative 24 to 36 hours after collection.