Gram-Negative Polymicrobial Bloodstream Infections and Clinical Decision Making with a Microarray Testing System

BACKGROUND: Molecular rapid diagnostic tests, such as the microarray-based Verigene Gram-Negative Blood Culture Test (BC-GN), when performed in conjunction with antimicrobial stewardship, are associated with improved clinical outcomes in bloodstream infections (BSIs). Optimal use of Verigene BC-GN to aid in clinical decision-making, however, can be hindered by a lack of confidence in results if patients have polymicrobial GN BSIs. METHODS: Blood culture data from patients tested by Verigene BC-GN from June 2015 – July 2016 from the Detroit Medical Center and University of Maryland Medical Center were retrospectively reviewed and evaluated for incidence of missed GNs in GN BSIs. Missed GNs entailed those not detected in the tested sample and those that grew in a separate complementary positive blood culture bottle collected but not tested using Verigene BC-GN. For blood culture sets with missed GNs, potential clinical significance was evaluated. RESULTS: A total of 1,003 sets of GN blood cultures were reviewed. Fifty-seven (5.6%) were determined to be polymicrobial GN BSIs by traditional microbiological culture methods. Verigene BC-GN did not identify one or more GNs in 37 cases (65% of polymicrobial GN BSIs; 3.7% of total GN BSIs.); 25 were missed by the probe in the tested culture and 12 where missed due to isolation in the complementary blood culture bottle that was not tested by Verigene BC-GN. While potentially inappropriate de-escalation of antibiotic therapy could have occurred in 18 (31.5%) of polymicrobial GN BSIs, this represented only 1.8% of total GN BSIs. CONCLUSION: Current practices utilizing Verigene BC-GN can result in the omission of undetected GNs in over half of polymicrobial GN BSIs. These misses were a combination of technology and workflow limitations. The potential for inappropriate de-escalation of antibiotic therapy in the entire cohort, however, was infrequent (<2% of total cases). DISCLOSURES: P. Lephart, Nanosphere: Grant Investigator, Grant recipient; J. K. Johnson, Nanosphere: Grant Investigator, Grant recipient