2026 Southeastern Residency Conference

Background 
Gram-negative bloodstream infections (GN-BSI) are associated with significant morbidity and mortality and traditionally managed with intravenous (IV) antibiotics. However, emerging evidence supports the use of oral (PO) antibiotics as step-down therapy in uncomplicated GN-BSI.1-7 Oral agents such as fluoroquinolones, sulfamethoxazole-trimethoprim, and β-lactams have demonstrated efficacy in this setting.1-2 Transitioning to a PO agent offers several benefits such as reduced cost, easier administration, and less risk of IV-associated complications, which include venous thrombosis, extravasation, and phlebitis.8-9 Despite these potential benefits, data on this practice in critically ill patients remain limited. This study aims to evaluate the safety and efficacy of IV therapy alone versus IV to PO step-down therapy in the treatment of critically ill patients with a GN-BSI. 
 
Methods 
This is a multicenter retrospective cohort analysis that evaluated adult patients with an uncomplicated GN-BSI who were initiated on appropriate empiric IV antibiotic treatment within 24 hours of their initial blood culture collection. The study included patients who were admitted to an intensive care unit (ICU) within Emory Healthcare (EHC) between May 1, 2023 and May 1, 2025. Patients were excluded if the duration of PO therapy was less than 48 hours, there was polymicrobial growth or presence of an organism other than Enterobacterales or Pseudomonas spp., hospice or comfort care was initiated within 72 hours of initial blood culture, or the organism(s) isolated was not susceptible to an available PO agent. The primary outcome was treatment failure, defined as a composite of 90-day mortality or recurrence of the same causative GN-BSI within 30 days of treatment completion. Secondary efficacy outcomes include ICU and inpatient length of stay, recurrence of GN-BSI within 90 days of treatment completion, emergence of resistance to study antibiotics, antibiotic duration, and IV to PO transition time. Secondary safety outcomes include adverse drug events leading to discontinuation or change in antibiotic therapy. Baseline demographics and outcomes were summarized with descriptive statistics while continuous data was summarized with means and standard deviations. 
 
Results 
Patients were separated into two cohorts based on antibiotic regimen: IV only (n=108) versus IV to PO (n=33). Baseline characteristics differed between groups, with patients in the IV only group demonstrating higher illness severity, depicted by higher median APACHE II (18 vs 14) and Pitt bacteremia scores (3 vs 2). The most common pathogens were Escherichia coli, Klebsiella pneumoniae, Serratia marcescens, Enterobacter cloacae, and Pseudomonas aeruginosa. Pulmonary sources were more prevalent in the IV only group, while genitourinary sources were more common in the IV to PO group. Treatment failure occurred more frequently in the IV only group compared to the IV to PO group (41.7% vs 15.1%). Recurrence of the same causative bacteria within 90 days was similar between groups (4.6% vs 6.1%), as was the emergence of resistance to study antibiotics (7.4% vs 9.1%). ICU and hospital length of stay were longer in the IV only group (13 vs 4 days and 21 vs 11 days). However, total antibiotic duration was longer in the IV to PO group (13 vs 9 days). Adverse drug events were uncommon in both groups. 
 
Conclusions 
While IV to PO step-down therapy was associated with lower rates of treatment failure, further research is needed to optimize treatment by determining appropriate drug selection and timing of transition. For critically ill patients with uncomplicated GN-BSI and adequate source control, this approach appears to be appropriate and may reduce length of stay without compromising effectiveness. These findings align with existing literature supporting PO step-down therapy as a method with comparable efficacy to continued IV therapy.