Escherichia coli (E.coli) is the most common cause of community-acquired bloodstream infections. Fluoroquinolones (FQ) and trimethoprim/sulfamethoxazole (TS) are preferred for oral step-down therapy due to high bioavailability. Antimicrobial stewardship programs commonly restrict FQ use, promoting consideration of non-FQ agents for treatment of sensitive organisms. We hypothesized that oral β-lactams would be non-inferior to FQ and TS with a primary outcome of 30-day all-cause readmission.
Methods: This was a retrospective non-inferiority study that reviewed electronic health records for patients with E.coli bacteremia from January 1st, 2016 - December 31st, 2017. Exclusion criteria included hospital acquired infections, death during hospitalization and concomitant infections. Patient demographics, Pitt Bacteremia Score, Charleston Comorbidity Index, antibiotic regimen (IV/PO), and readmission were collected. Patients were divided into two groups, oral FQ/TS verses β-lactams. A pre-trial non-inferiority margin for the primary outcome was set at 3%. Secondary outcomes included 30-day infection and E.coli readmission. Significant risk factors for readmission were entered into a multiple logistic regression model in a forward stepwise approach using SPSS.
Results: Demographics were similar between groups, 57 patients received FQ/TS and 151 received β-lactams. The 30-day-all cause readmission rate was 15.8% and 29.1%, respectively (absolute risk difference 13.3%, CI: 1-25).β-lactams were found to be inferior to FQ/TS for 30-day all-cause readmission. Infection related readmission occurred in 5.3% of patients in the FQ/TS group verses 14.6% of patients in the β-lactam group (p=0.07). E.coli accounted for 100% of the infection related readmissions in the FQ/ST group and 70% in the β-lactam group. Patients who received a β-lactam antibiotic were more likely to be readmitted then those patients treated with FQ/TS (Odds Ratio: 2.25, CI: 0.95 - 5.30, p-value=0.06).
Step-down therapy to oral β-lactams resulted in higher rates of 30-day all-cause and infection related readmissions in patients with E.coli bacteremia.
L. Avery, None