Successful Outcomes after Implementation of a Dynamic Anitmicrobial Stewardship Program (ASP)

Background: Antibiotic stewardship has been identified as a key strategy to reduce healthcare costs and decrease emerging antibiotic resistance.  Proactive programs involving audits by a trained pharmacist (Ph) and infectious disease physician (ID) which provide feedback to prescribers have been found to be the most successful.  However, they are also the most time and resource intensive.  Maximizing the potential of the electronic medical record and other databases to generate clear and concise audit report may allow efficient real-time review of a large number of patients for potential intervention.

Objective: Assess whether computer generated audit reports can assist the implementation of a proactive ASP and have an impact on reducing antimicrobial expenditures and rates of infection with resistant organisms.

Methods: Using existing data-mining software, (Crystal Reports^TM, SAP BusinessObjects), we developed reports to capture key pharmacy, clinical, and microbiologic data on all hospital inpatients receiving antimicrobials.  The report is narrowed by the Ph who then enters further clinical data. A final report is then presented for ID review.  Written recommendations are given to providers on a standardized form and are based are based on national guidelines, local resistance patterns, and clinical experience. Number of reviews, notes written, acceptance rates, monthly/yearly antimicrobial expenditures, and tracking of nosocomial infections including MRSA, VRE, and C. difficile were followed prospectively.

Results: The ASP was launched in Jan, 2007.  Our hospital is a 583 bed facility with an average daily census of 312 and over 95,000 patient days per year.  Using the report, the Ph screens ~70-110 patients / day on antimicrobials which is then reduced to an average 18 patients for ID review. On average, 85 interventions are written per month, of which, 66% are accepted by providers.  Daily review and distribution of notes occupies ~ 40-60 minutes of ID time and 4 hours of Ph time.  Antimicrobial expenditures normalized to patient days increased 10-13% per year in the 3 years preceding ASP implementation, declined 11.2% ($207,767) in the first year after implementation, and have remained flat in the subsequent 2 years.  Furthermore, nosocomial infection rates improved.   Ph review of medical records to identify indications for antimicrobials consumed the majority of ASP time.

Conclusions: Utilization of computer-assisted audit reports allowed our ASP to screen a large number of inpatients in a time efficient manner and focus the list for ID review to the patients most amenable to intervention.  Our ASP implementation was associated with reduced expenditures of antimicrobials and reduced rates of resistant organisms.  Requiring providers to provide indications for prescribed antimicrobials would facilitate efficiency in ASP.