Stepdown from intravenous (IV) to oral (PO) antibiotics is a common practice that effectively reduces healthcare costs, even for severe infections. Appropriate PO antibiotic selection and duration of therapy (DOT) for Gram-negative bloodstream infection (BSI) is understudied. Highly bioavailable agents such as fluoroquinolones (FQs) may be optimal, but serious antibiotic-associated adverse effects (AAEs) have been documented with FQ use. The objective of this study was to compare the efficacy of PO beta-lactams (BLMs) to FQs as stepdown therapy for Enterobacteriaceae (EB) BSI.
This was an IRB-approved retrospective study at a 637-bed hospital in Portland, ME. Adults started on IV antibiotics for EB-BSI, with stepdown to PO BLM or FQ were included. A sample size of 224 was calculated to assess non-inferior clinical success, with a margin of -0.15, single sided α = 0.025, β= 0.2. Secondary outcomes were microbiologic success, 30-day readmission, and AAEs. Subgroups of early (≤ 3 days IV) vs. late stepdown (> 3 days IV), and short (≤ 10 days total) vs. extended DOT (> 10 days total) were compared. Logistic regression was performed to identify predictors of clinical success.
Stepdown with PO BLM (n=84) for EB-BSI was non-inferior to PO FQ (n=140) (86.9 vs. 87.1%, mean difference 0.2%, 97.5% CI: -10.3–10.7). Microbiologic success (94.0 vs. 97.9%, p=0.16) and 30-day readmission (14.3 vs. 14.3%, p=1.00) were similar and patients were more likely to tolerate BLMs than FQs (90.5 vs. 79.3%, p=0.03). In subgroups, clinical success was comparable between early and late stepdown (86.7% vs 87.5%, p=0.86), and short versus extended DOT (88.2% vs. 86.7%, p= 0.76). Negative predictors of clinical success were uncontrolled diabetes (OR=0.36, CI: 0.15–0.84) and urinary abnormality (OR=0.39, CI: 0.16–0.94).
The results of this study suggest that PO BLMs may be a non-inferior stepdown option for uncomplicated Gram-negative BSI, with less AAEs in comparison to FQs. Early stepdown and short DOT had similar outcomes compared to late stepdown and extended DOT. Patients with uncontrolled diabetes and urinary abnormalities were less likely to achieve clinical success. These findings should be validated in a prospective analysis.
M. Wungwattana, None