Electronic surveillance system for monitoring surgical antimicrobial prophylaxis.

OBJECTIVES: Antimicrobial surgical prophylaxis comprises one third of all antibiotic use in pediatric hospitals and 80% of all antibiotic use in surgery. Previous studies reported that antimicrobial surgical prophylaxis is often inconsistent with recommended guidelines. An electronic surveillance system was developed to measure antimicrobial utilization and to identify opportunities to improve and monitor the administration of antibiotics for surgical prophylaxis.

METHODS: A retrospective cohort study was conducted on patients with selected inpatient surgical procedures performed from May 1999 to April 2000 at 4 US children's hospitals. International Classification of Diseases, Ninth Revision surgical procedure codes were divided into clean or unclean categories, and an electronic surveillance system was designed using antibiotic and microbiologic culture utilization data to measure appropriate antimicrobial use associated with the surgical procedure. A medical chart review was conducted to validate the electronic system.

RESULTS: Ninety percent of cases were classified properly by the electronic surveillance system as confirmed by medical chart review. Surgical antibiotic prophylaxis was not in accordance with the American Academy of Pediatrics (AAP) guidelines for almost half of all procedures. Prolonged antimicrobial administration in clean surgical procedures was the most frequent deviation from guidelines. Statistical differences between the index hospital and the comparison hospitals reflect both over- and underutilization of surgical prophylaxis with significant opportunity to improve prophylaxis for all hospitals.

CONCLUSIONS: Antimicrobial surgical prophylaxis at the children's hospitals studied is not always consistent with published AAP guidelines. This electronic surveillance system provides a rapid, reproducible, and validated tool to measure easily the efforts to improve adherence to AAP surgical prophylaxis guidelines.