Validation of Electronic Antibiograms for the Veterans Affairs Palo Alto Health Care System

Background: Antibacterial resistance has been on the rise in hospitals and communities worldwide, including the Veterans Affairs (VA) hospitals.  Effective monitoring of antibacterial resistance is paramount to limiting spread of resistant organisms.  In addition, clinical guidance of empiric antibiotic choices based on institutional patterns helps clinicians decide earlier on appropriate antibiotics for their patients.  Early and appropriate antibiotic choices help decrease morbidity as well as mortality.  A pillar of the Veterans Health Administration’s Healthcare-Associated Infections and Influenza Surveillance System (HAIISS) is to develop an antibiotic stewardship program that electronically provides institutional antibiograms as well as advises clinicians on appropriate antibiotic choices for various organisms.

Objective: To illustrate the preliminary validation process of the generation of institutional antibiograms using the QC Pathfinder™ (QCP) electronic program.  

Methods: Culture data for select gram-negative organisms was pulled from the VA’s electronic medical record (EMR) as well as from QCP.  Data was compared to insure QCP extracted all positive cultures for each gram-negative organism.  Based on Clinical and Laboratory Standards Institute (CLSI) guidelines, institutional antibiograms were generated using QCP and compared to manually generated antibiograms following the same guidelines.

Results: In a sample of 2,416 culture isolates (not excluding duplicates) from January 1, 2008 to December 31, 2008, 96% of cultures were appropriately extracted from the VA’s EMR using QCP and could be automatically separated by source of culture and location of culture collection.  QCP successfully generated institutional antibiograms excluding duplicates in accordance with CLSI guidelines. Some of the culture results that were not appropriately extracted from the VA’s EMR were due to failure of data transfer to QCP on several discrete days and in no consistent pattern.

Conclusions: QCP effectively generates institutional antibiograms of gram-negative organisms for the Palo Alto VA hospital from the VA’s EMR.  Next steps will be to validate QCP institutional antibiograms for a wider assortment of organisms as well as VAs nationwide.  QCP is undergoing modifications for allowing the user to determine which isolates are being used to generate the antibiogram as well as exploring improved extraction techniques to avoid the rare days of missed culture data.  QCP offers a valid and efficient use of the VA’s EMR to generate institutional antibiograms.