Study design and setting

This was a multicenter, retrospective cohort study of adult patients treated for uncomplicated Enterobacterales bacteremia at a large health system between September 1, 2021, and September 1, 2024. The health system included one tertiary care academic medical center, 3 community teaching hospitals, and 1 community hospital. The study protocol was approved by the Institutional Review Board with a waiver of informed consent.

Inclusion and exclusion criteria

Eligible participants were adults 18 years or older with uncomplicated Enterobacterales bacteremia (Figure 1) who had definitive source control, if applicable, and achieved clinical stability within 72 hours of appropriate IV antimicrobial initiation (Figure 2). Definition of clinical stability was developed utilizing The Consensus Statements for Best Practices for the Management of Uncomplicated Gram-Negative Bloodstream Infections criteria. ^{Reference Heil, Bork and Abbo13} Therefore, we defined clinical stability as achievement of at least 3 of 4 criteria: Temperature ≤38°C, HR <90/min, RR <20/min, hemodynamic stability (SBP > 90 mmHg). Patients additionally had to be eligible for transition to oral antibiotics (Figure 2), have a final antibiotic susceptibility report available for the infecting organism, and be discharged alive without hospice or palliative care.

Figure 1.

Definition of uncomplicated Enterobacterales bacteremia.

*Clinical stability defined as achievement of at least 3 of 4 criteria: Temperature ≤38°C, HR <90/min, RR <20/min, hemodynamic stability (SBP > 90 mmHg).

Figure 2.

Definitions of eligibility criteria.

Patients were then excluded for the following reasons: never transitioned to oral therapy during inpatient admission, had polymicrobial infection with non-Enterobacterales organisms, received less than 7 days of appropriate antimicrobial therapy, transitioned back to IV antibiotics, received inappropriate antimicrobial therapy based on dosing chosen or susceptibilities, transferred from an institution outside of the health system, or were pregnant or incarcerated. Patients were also excluded if they were immunocompromised as defined by having received any solid organ transplant, expected prolonged neutropenia (ANC <500 cells/mL), CD4 cell count <200 cells/mL, or currently on chronic corticosteroids and/or immunomodulator therapy. ^{Reference Heil, Bork and Abbo13} Exclusions are outlined in Figure 3.

Figure 3.

Screening and enrollment.

Patients were separated into two groups based on the timing of their transition from IV-to-oral antibiotics: an early transition group, defined as patients transitioned on day 3 or 4 of antibiotic therapy, and a late transition group, defined as patients transitioned on day 5 or later of antibiotic therapy.

Data collection

Patients were identified from blood cultures positive with any of the three predefined isolates within the study time frame and digitally abstracted from the electronic health record (EHR) system, Epic. Data pertaining to patient demographics, clinical presentation, and outcomes were manually collected from the EHR. Laboratory results, microbiological findings, antibiotic treatment regimens, International Classifications of Disease (ICD) codes used to calculate the Charlson Comorbidity Index (CCI) were also extracted digitally from Epic and confirmed through manual patient chart review. Standardized data collection forms were utilized to ensure consistency and completeness of data abstraction, and a separate process was implemented to ensure the de-identification of all patient information to maintain confidentiality. All data was collected by the primary investigator, and sub-investigators audited 10% of charts throughout data collection to verify the accuracy of information digitally extracted and manually obtained from patients’ EHR.

Study outcomes

The primary outcome of the study was treatment failure within 30 days of completing the prescribed course of oral antibiotics. This was a composite end point defined as the occurrence of one or more of the following: the need for additional antimicrobial therapy, infection-related readmission to the hospital, microbiological relapse, or all-cause mortality. The need for additional antibiotic therapy was defined as a broadening in coverage of the patient’s current antimicrobial regimen or the initiation of a new antibiotic. Microbiological relapse was defined as recurrent bacteremia with the same organism. Secondary outcomes included the incidence of each component of the composite primary outcome, total hospital length of stay measured in days from admission to discharge, and the total duration of antibiotic therapy. Duration of antibiotic therapy was measured as the sum of IV and oral antibiotic days. Antimicrobial days were counted using the total number of calendar days in which a patient receives any amount of a specific antimicrobial agent as outlined by NHSN. ¹⁴

Statistical analysis

Using 80% power and an alpha of 0.05, it was estimated that 125 patients would be needed in each group to detect a 10% difference in treatment failure. A greater difference in treatment failure was assumed compared to what Rieger et al observed due to differences in timing of IV-to-oral antibiotic transition and treatment failure definition between studies. ^{Reference Rieger, Bosso, MacVane, Temple, Wahlquist and Bohm6}

Statistical analysis was performed using IBM SPSS Statistics 28.0.1. Categorical variables were analyzed using Pearson χ ² or Fisher’s Exact tests, as appropriate. Continuous variables were analyzed using the Mann–Whitney test for non-parametric data. Statistical significance was set as a P value <.05. To address potential confounding variables, clinically relevant risk factors for treatment failure and risk factors that had a bivariate analysis P value <.2 were considered for multivariate logistic regression.