785. Assessment of risk factors for inappropriate empiric antibiotic therapy in patients with Gram-negative sterile site infection complicated by sepsis or septic shock
Session: Poster Abstract Session: Treatment of Resistant Infections - Clinical Analyses
Thursday, October 5, 2017
Room: Poster Hall CD
Background: The association between the administration of inappropriate empiric antibiotic treatment (IEA) and an increased risk of hospital mortality has been consistently reported. This study explores pre-infection risk factors for IEA in patients with Gram negative (GN) sterile site infections complicated by sepsis or septic shock.

Methods: Retrospective cohort study at Barnes-Jewish Hospital (2010 – 2015). Risk factors including history of hospitalization, receipt of intravenous antipseudomonal antibiotics, and isolation of GN organisms 90 days prior to admission were collected. Process of care variables assessed included ICU admission, durations of mechanical ventilation, central venous catheter, and urinary catheter insertion, and antibiotic days prior to isolation of a GN pathogen(s). IEA was defined as receipt of antibiotic therapy that lacked in vitro activity against the identified pathogen(s) within the 24 hours of the culture being obtained. Multivariable logistic regression analysis (MVLRA) risk factor modeling that included IEA as the dependent outcome variable was conducted.  

Results: 855 consecutive patients with first episode sepsis or septic shock were included. Compared to patients receiving appropriate empiric therapy (n = 715), variables significantly associated with IEA (n = 140) within 90 days prior to admission included recent hospitalization (23.1% v. 34.3%, p = 0.005), mean days of meropenem (0 v. 2.1, p = 0.010) and piperacillin-tazobactam (0 v. 1.6, p <0.001) therapy, and isolation of a GN organism(s) (8.4% v. 20.0%, p <0.001). Prior to isolation of the GN pathogen(s), median hospital (0 v. 6 days, p < 0.001) and ICU (0 v. 0 days, p < 0.001) length of stay, as well as the median duration of CVC dwell time (10 v. 17 days, p = 0.050) was associated with IEA. MVLRA identified isolation of a GN pathogen (AOR 3.432 95%CI 2.024-5.820, p <0.001) and days of piperacillin-tazobactam therapy (AOR 1.149 95%CI 1.061-1.245, p <0.001) in the 90 days prior to admission, as well as Charlson score (AOR 1.097 95%CI 1.023-1.177, p + 0.010) as independent risk factors for IEA. 

Conclusion: Consideration of risk factors prior to admission and prior to collection of a sterile site specimen appear to be critical when making empiric antibiotic decisions targeting GN pathogens in patients with sepsis or septic shock.

Scott Micek, PharmD, Pharmacy Practice, St. Louis College of Pharmacy, St. Louis, MO, Nicholas Hampton, PharmD, BJC Center for Clinical Excellence, St. Louis, MO and Marin Kollef, MD, Washington University School of Medicine, St. Louis, MO


S. Micek, Merck: Grant Investigator , Research grant

N. Hampton, None

M. Kollef, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 4th with the exception of research findings presented at the IDWeek press conferences.