2190. Influence of Microbiological Culture Results on Antibiotic Choice for Veterans with Hospital-Acquired Pneumonia

BACKGROUND: Respiratory specimens help inform the treatment of hospital-acquired pneumonia (HAP), permitting clinicians to ensure effective and, ideally, narrow-spectrum antibiotic therapy. Here, we examine changes in antibiotic regimens to treat HAP based on the antibiotic susceptibility of pathogens recovered from respiratory samples. METHODS: At a single Veterans Affairs (VA) Medical Center, we identified veterans hospitalized between October 2014 and September 2018 with HAP, defined as a clinical respiratory sample obtained >48 hours after admission and corresponding clinical signs and symptoms. Exclusion criteria were death, transfer to hospice care or discharge within 48 hours of sample collection or admission from an outside hospital. For each specimen, we assessed timestamps for collection, Gram stain, identification of organisms and results of susceptibility testing. We used the antibiotic spectrum index (ASI) to assess changes in antibiotics given to patients during hospitalization and at discharge. RESULTS: Between October 2014 and September 2018, 70 veterans met our inclusion criteria and experienced 73 episodes of HAP. Their mean age was 66.2 years (±9 years) and 47 (67%) had chronic pulmonary disease. All-cause mortality at 30-days after specimen collection was 14%. The median time from specimen collection to Gram stain result was 0.8 days (interquartile range (IQR) 0.1–1.9) and to antibiotic susceptibility results was 2.4 days (IQR 1.5–3.3). The most common bacteria recovered were Enterobacteriaceae (20 isolates), Pseudomonas aeruginosa (11 isolates), Streptococcus spp. and Staphylococcus aureus (8 isolates each); colonization with Candida spp. was frequent (26 isolates). Vancomycin and piperacillin–tazobactam were the most common antibiotics on day 0 (24%, 22%, respectively) and day 3 (21%, 13%, respectively). Compared with the day of sample collection (day 0), the ASI score was lower at day 3 in 23 (32%) and higher in 21 (29%) cases. CONCLUSION: The high proportion of escalation and de-escalation of antibiotics suggests that results of bacteria identification and susceptibility testing influence therapeutic decisions, emphasizing the importance of obtaining respiratory samples to inform treatment of HAP and improve antibiotic stewardship. DISCLOSURES: All authors: No reported disclosures.