2450. Antibiotic Treatment for Carbapenem-Resistant Enterobacteriaceae (CRE) and Outcomes in Veterans With Spinal Cord Injury/Disorder (SCI/D)

Background A total of 282,000 people (17% veterans) in the United States have SCI/D. Infection is a significant source of morbidity and the leading cause of death in this population. Due to frequent healthcare contact and antibiotic use, SCI/D is associated with high risk of multidrug-resistant infections, including CRE. CRE are resistant to most antibiotics and associated with high mortality. The objective of this study was to describe antibiotics used for CRE infection and clinical outcomes in veterans with SCI/D. Methods This retrospective cohort used national VA data of veterans with SCI/D and active CRE infection (per documentation in the health record) from 2011 to 2013. CRE was defined as resistant to a carbapenem and third-generation cephalosporin. Antibiotics were described by empiric/definitive and monotherapy/combination therapy. Clinical outcomes included clinical failure/improvement, microbiological resolution, mortality and readmission in 30 days/1 year. SAS was used for analysis with significance at P ≤ 0.0125 due to multiple comparisons. Results Ninety-two CRE infections (62% K. pneumoniae) were identified in 87 patients, most often in urine cultures (58.7%). Carbapenems (20.7%) were used most frequently for CRE treatment. Combination therapy was used more often than monotherapy (empiric 56.3%, definitive 69.0%). Definitive combinations consisted of carbapenems/polymyxins (16.7%) or carbapenems/aminoglycosides (13.3%). Clinical outcomes for definitive monotherapy vs. combination, respectively, were: clinical failure (29.6% vs. 46.7%), improvement 1–10 days (48.2% vs. 33.3%), and 11–30 days (70.4% vs. 53.3%); microbiological resolution (48.2% vs. 38.3%); mortality at 30 days (22.2% vs. 30%), 90 days (22.2% vs. 41.7%), 1 year (25.9% vs. 51.7%) and readmission at 30 days (11.1% vs. 10%) and 1 year (37% vs. 30%). No significant differences in outcomes were identified for monotherapy vs. combination therapy or susceptible vs. nonsusceptible treatment. Conclusion For CRE treatment in the SCI/D population, carbapenems were the most widely used drug class; combination therapy was used most frequently. No improvements in clinical outcomes were found for combination therapy as either empiric or definitive treatment or for susceptible vs. nonsusceptible treatment. Disclosures All authors: No reported disclosures.