Rapid Testing using Verigene® Microarray in Combination with Antimicrobial Stewardship Intervention in Gram-negative Bacteremia
Methods: Clinical isolates from adult patients at the University Maryland Medical Center with GN bacteremia from January 2012 to June 2012 were included. Blood culture bottles were spiked with clinical isolates and processed by Verigene® BC-GN. AST reviewed charts along with Verigene® result and recommended antibiotics using derived algorithms. The time interval included in analysis was time of Gram stain report until 48 hours after susceptibility results. The intervention group’s (Verigene® with AST) antibiotic recommendation was theoretically implemented at 3 hours from actual Gram stain report, and compared with the control group’s actual antibiotic administration times obtained from the chart, using Student's t-test.
Results: A total of 117 isolates were tested, demonstrating 97.6% sensitivity, 99.6% specificity and overall concordance rate of 95.7% (112/117) for organism identification. Half of the Enterobacteriaceae isolates resistant or intermediate to ceftriaxone were detected as CTX-M (5/10) and all of the MDR A. baumannii were detected as OXA (7/7). The intervention group had a significantly shorter mean duration to both effective (1.4 vs 6.6 hours, P < 0.01) and optimal (6.7 vs 47.7 hours, P < 0.01) antibiotic therapy. Using proportional hazards regression, we found that the intervention group was significantly more likely to receive timely optimal antibiotic therapy compared to controls (hazard ratio 1.5, P = 0.01).
Conclusion: Our study demonstrated a potential decreased time to both effective and optimal antibiotic therapy in GN BSI using combined intervention of rapid testing and AST recommendation. Prospective studies are needed to further validate this strategy.
J. T. Bork,
E. Heil, None
R. Badamas, None
J. K. Johnson, Nanosphere: Investigator, Research support
Bio-Fire: Investigator, Research support
OpGen: Investigator, Research support