Abstract
Antimicrobial stewardship can be challenging in children with bloodstream infections (BSIs) caused by Gram-negative bacilli (GNB). This retrospective cohort study explored how data elements in the electronic health record could potentially optimize empiric antibiotic therapy for BSIs caused by GNB, via the construction of customized antibiograms for categorical GNB infections and identification of opportunities to minimize organism-drug mismatch and decrease time to effective therapy. Our results suggest potential strategies that could be implemented at key decision points in prescribing at initiation, modification, and targeting of therapy.
