2143. Attempting to Add Clarity to “Indeterminates” on a Deployed Rapid Diagnostic with Antimicrobial Stewardship Program (ASP) Intervention

BACKGROUND: Rapid diagnostic testing paired with ASP intervention optimizes therapy and improves outcomes but few data guide ASP response in the absence of organism identification (ID). We describe the microbiology for organisms unidentified by Accelerate Pheno™ Gram-negative platform (AXDX) in order to inform ASP-provider team communication (PTC). METHODS: Consecutive, non-duplicate inpatient blood cultures with Gram-negative bacilli (GNB) following AXDX implementation at a single university hospital between April 2018 and March 2019 were included. Standard of care (SOC) ID and susceptibility followed AXDX. Clinical Microbiology emailed AXDX results to the ASP in real time; results were released into the EMR paired with telephone PTC or withheld after ASP review. Bloodstream Infections (BSIs) and patient outcomes for organisms labeled no/indeterminate ID by the AXDX were characterized. RESULTS: AXDX was performed on 351 blood cultures. Among 52 (15%) labeled no/indeterminate ID, SOC methods revealed: Enterobacteriaceae (40%; 9 monomicrobial with AXDX targets), anaerobes (21%), non-lactose fermenters (NLFs) other than Pseudomonas aeruginosa (21%), and fastidious GNB (10%). Frequent organisms without AXDX targets included: Raoultella planticola (4); Bacteroides fragilis, Cupriavidus spp., Haemophilus spp., Prevotella spp., Providencia spp., non-aeruginosa Pseudomonas spp., Salmonella spp. (3 each); Pasteurella multocida, Stenotrophomonas maltophilia (2 each). BSI sources were most commonly intra-abdominal (21%), central line-associated (17%), or unknown (17%). CLABSIs were associated with immune suppression and/or substance abuse in all but 1 case. BSIs without active empiric therapy included: NDM-producing Providencia stuartii SSSI; OXA-48-producing R. planticola intraabdominal infection (IAI); Pandoraea spp. CLABSI after liver transplant; enteric fever; B. fragilis, Leptotrichia wadei, and S. maltophilia, each of unknown source. In-hospital mortality occurred in 4 of these cases. CONCLUSION: When AXDX yields no/indeterminate ID, ASP chart review for possible anaerobic/IAI, unique environmental exposures, and travel history may assist in guiding empiric therapy. GNB with AXDX targets are not excluded. DISCLOSURES: All authors: No reported disclosures.