Lead/Presenter: Christopher Graber, COIN - Los Angeles
All Authors: Graber CJ (CSHIIP, Los Angeles), Jones MM (IDEAS Center, Salt Lake City), Goetz MB (CSHIIP, Los Angeles) Madaras-Kelly K (Boise VA Medical Center, Boise) Zhang Y (IDEAS Center, Salt Lake City) Butler JM (IDEAS Center, Salt Lake City) Weir C (IDEAS Center, Salt Lake City) Chou AF (Dept of Family and Preventive Medicine, University of Oklahoma Health Sciences Center, Oklahoma City) Youn SY (CSHIIP, Los Angeles) Samore MH (IDEAS Center, Salt Lake City) Glassman PA (CSHIIP, Los Angeles)
To optimize antimicrobial use by identifying areas for improvement, we developed and pilot tested inpatient antimicrobial use (graphic) tools at eight Veterans Affairs healthcare facilities. These tools allow a facility to compare its patterns of total use and use by antibiotic class to aggregate use at VA facilities at a user-selected complexity level.
Antimicrobial stewards from eight inpatient VA facilities with a broad geographic representation participated in iterative report development and implementation, with the final product consisting of two components: an interactive web-based antimicrobial dashboard and a standardized antimicrobial usage report updated at user-selected intervals. Stewards also participated in monthly learning collaboratives. Pre-post intervention (1/2014-1/2016 vs. 7/2016-1/2018) changes in total inpatient antimicrobial days of therapy per 1000 patient days were estimated using interrupted time series. Three outcomes were assessed: all antimicrobial agents, anti-methicillin-resistant Staphylococcus aureus agents (anti-MRSA), and broad-spectrum agents predominantly used for hospital-onset/multi-drug-resistant organisms (anti-MDRO). Comparisons were made between the eight intervention sites and all other inpatient VA facilities (n = 132) as controls. We used Generalized Estimation Equation approach with Poisson distribution family to estimate the percentage difference in average monthly antimicrobial use rate between pre- and post-intervention phase for intervention and non-intervention sites, respectively.
Intervention sites had a 2.1% decrease (95% CI = [-5.7%,1.6%]) in all antimicrobial use pre-post-intervention, versus a 2.5% increase (95% CI = [0.8%, 4.1%])in non-intervention sites (p = 0.025 for difference). Anti-MRSA antimicrobial use decreased 11.3% (95% CI = [-16.0%,-6.3%]) at intervention sites versus a 6.6% decrease (95% CI = [-9.1%,-3.9%]) at non-intervention sites (p = 0.092). Anti-MDRO antimicrobial use decreased 3.4% (95% CI = [-8.2%,1.7%]) at intervention sites versus a 3.6% increase (95% CI = [0.8%,6.5%]) at non-intervention sites (p = 0.018).
Distribution of comparative reports of antibiotic use to stewards in a pilot implementation project at eight VA facilities, coupled with monthly learning collaboratives, was associated with an approximate 5% difference in trends in several measures of inpatient antimicrobial use.
An intervention that bundled individualized feedback (e.g., comparative antimicrobial use data) and learning collaborative participation proved effective, as it activated stewards to better target interventions at reducing inappropriate antimicrobial use. Dissemination of our tools through the work of the VA Antimicrobial Stewardship Task Force is underway.