2182. Harnessing Direct-from-Blood MALDI-TOF (DBM) and Local Blood Stream Infection (BSI) Antibiogram Data to Direct Optimal Therapy for Gram-Negative Rod (GNR) Bacteremia

BACKGROUND: GNR bacteremia is common in the hospital environment. Drug-resistance challenges thoughtful use of, and timely de-escalation to, regimens that limit carbapenem use. DBM accurately provides rapid identification of bacteremia pathogens, but lacks antibiotic susceptibility testing (AST) results to aid clinicians in refining therapy. We investigated whether integrating DBM identification and local BSI antibiogram data could be used to identify carbapenem-sparing regimens with predicted efficacy of ≥ 95% prior to traditional AST results. METHODS: We identified GNR blood culture results in adults from January 1, 2018 to December 31, 2018. The first isolate per bacteremia episode per patient was analyzed. AST was performed by MicroScan. We categorized isolates into 1 of 3 groups based on the most common resistance mechanisms in our population (i.e. CTX-M ESBL, AmpC): 1] Escherichia spp. and Klebsiella spp. (EK), 2] S. marscecens, P. mirabilis, E. (Klebsiella) aerogenes, C. freundii, M. morganii, E. cloacae complex, and C. non-freundii (SPEC-MEC), and 3] P. aeruginosa (PA). We compiled susceptibility rates to common antibiotics for GNR sepsis and compared results using Chi-square. RESULTS: 381 GNR isolates were recovered in blood cultures from adult patients in 2018: EK 264 (69%), SPEC-MEC 66 (17%), PA 36 (9%), other 15 (4%). Susceptibility results in common GNR sepsis agents are shown in Figure 1 (*P < 0.05 between groups). Within each organism group, susceptibility was highest in the following scenarios: 1] EK: PTZ and MEM (P < 0.05 vs. FEP, TOB, CIP, and ATM), 2] SPEC-MEC: FEP, TOB, and MEM (P < 0.05 vs. PTZ and ATM), and 3] PA: comparably similar (> 90%) for PTZ, FEP, TOB, and MEM. CONCLUSION: Integration of DBM and antibiogram data identified species-specific carbapenem-sparing regimens that would predict ≥ 95% of patients with GNR bacteremia are on effective therapy prior to traditional AST results. In appropriate patients, this could lead to a significant reduction in carbapenem duration and avoid broad-scale carbapenem use while also optimizing time to effective therapy. This is a readily adaptable process for other centers to improve care of patients with GNR bacteremia. Our data suggest using β-lactam alternatives for GNR bacteremia, especially EK, should be done with caution. [Image: see text] DISCLOSURES: Alexander Lepak, MD, Paratek Pharmaceuticals: Research Grant; Tetraphase Pharmaceuticals: Research Grant.