373 Do Regional Antibiograms (RA) Differ from Institutional Antibiograms (IA)?

Sunday, April 3, 2011: 11:15 AM
Coronado BCD (Hilton Anatole)
Patricia Schirmer, MD , VA Office of Public Health Surveillance and Research, Palo Alto, CA
Renee-Claude Mercier, PharmD , University of New Mexico, Albuquerque, NM
Gina Oda, MS, CIC , VA Office of Public Health Surveillance and Research, Palo Alto, CA
Cynthia Lucero, MD, CIC , VA Office of Public Health Surveillance and Research, Palo Alto, CA
Ben Bau, BS , Vecna Technologies Inc, Greenbelt, MD
Clara Lewis, MPH , Vecna Technologies Inc, Greenbelt, MD
Cheryl Canody , Regional Data Warehouse, VISN 20, Sacramento, CA
Mark Holodniy, MD, CIC , VA Office of Public Health Surveillance and Research, Palo Alto, CA
Background: Monitoring antimicrobial resistance is a crucial component in developing antimicrobial stewardship programs to limit the spread of these resistant organisms.  In addition, clinical guidance for empiric antibiotic choices based on institutional patterns help clinicians decide on early, appropriate antibiotics which decreases morbidity and mortality.

Typically, institutions generate yearly antibiograms.  RAs are not created or available for comparison.  In VA, a national electronic medical record and QC PathfinderTM (QCP, Vecna Technologies, Inc.) allow us to assess how antimicrobial patterns differ on a regional/VISN (Veterans Integrated Service Network) level when compared to individual facility level.

Objective: To determine whether antibiograms differ when generated on a facility (IA) vs. VISN level (RA) and to evaluate the utility of a VISN level antibiogram.

Methods: This study was performed using data from VISN 20 (includes VA hospitals in Anchorage, AK, Boise, ID, Portland, OR, Puget Sound, WA, Roseburg, OR, Spokane, WA, Walla Walla, WA, and White City, OR) from Jan. 1, 2009 to Dec. 31, 2009.  Data was extracted for all 8 facilities from the VISN 20 data warehouse using QCP.  IA and RA were electronically generated and compared the % of isolates susceptible to each antibiotic evaluated for 3 Gram-positive and 3 Gram-negative organisms.  If there was <10% difference between each IA and the RA % susceptible, the results were considered comparable.

Results: A total of 2914 non-duplicate cultures were identified for the following organisms – Enterococcus faecalis (E. faecalis), Enterococcus faecium (E. faecium), Streptococcus pneumoniae (SP), Enterobacter cloacae (EC), Klebsiella pneumoniae (KP), and Pseudomonas aeruginosa (PA).  SP, E. faecalis, and KP had IA and RA that were comparable for all key antimicrobials tested except for erythromycin and SP.  The limited number of facilities with E. faecium isolates prevented an accurate comparison.  EC was comparable for 8/9 antimicrobials tested, and 3/5 antimicrobials tested for PA.  In addition, multiple facilities had less than 30 isolates for an organism [E. faecalis (4), E. faecium (6), SP (4), EC (4), PA (3), KP (2)], therefore these facilities are unable to generate accurate antibiograms but a RA meets this requirement.

Conclusions: IAs are the primary method of identifying trends in antimicrobial resistance.  RA was similar to IAs for the Gram positive organisms and Enterobacteriaceae evaluated, however more variability was present for PA.  Combining facility antibiotic susceptibilities generates more isolates allowing creation of expanded VISN antibiograms that meet the requirement of 30 isolates.  VISN antibiograms could provide valuable information to smaller VA healthcare facilities in a VISN that may be unable to generate their own individual antibiograms for infrequently encountered organisms.