Introduction
Streptococci are a diverse group of bacteria that are associated with a wide spectrum of disease manifestations, ranging from community-acquired infections to invasive infections including bloodstream infections (BSIs) and endocarditis. Intravenous (IV) antibiotics have traditionally been preferred over oral (PO) antibiotics for the treatment of BSIs. Reference Laupland and Church1,Reference Youkee, Hulme, Roberts, Daniels, Nutbeam and Keep2 However, in a survey of infectious disease physicians in the United States, the decision to transition patients from IV to PO therapy was highly dependent on the organism and source of infection. Specifically, <20% of respondents were comfortable switching to PO therapy for bacteremia caused by Staphylococcus aureus versus >80% for beta-hemolytic Streptococcus species and >50% for Enterococcus species. Reference Hospenthal, Waters, Beekmann and Polgreen3 Most data for PO stepdown in gram-positive BSIs involves Staphylococcus aureus and demonstrates conflicting results. Reference Itoh, Hadano, Saito, Myokai, Nakamura and Kurai4–Reference Willekens, Puig-Asensio and Ruiz-Camps6 Current data suggests a growing interest in PO stepdown therapy in patients who achieve clinical stability and source control with more recent randomized controlled trials demonstrating non-inferiority of PO stepdown for invasive infections such as endocarditis and osteomyelitis secondary to gram-positive pathogens. Reference Iversen, Ihlemann and Gill7,Reference Ho-Kwong, Rombach and Zambellas8 Advantages to oral therapies include shortened hospital admissions, reduced treatment costs, and avoidance of line-related complications. Reference Grau, Clarivet, Lotthé, Bommart and Parer9–Reference Cyriac and James11 Most Streptococcal spp. are susceptible to PO antibiotics but clinical evidence supporting IV to PO stepdown for bacteremia secondary to Streptococcus spp. remains limited. Reference Ramirez and Bordon12–Reference Arensman, Shields and Beganovic15 The purpose of this study was to compare clinical outcomes for patients treated with a PO antibiotic stepdown regimen to a full treatment course of IV antibiotics for uncomplicated streptococcal BSIs.
Methods
Study design
A retrospective cohort study was performed in patients treated for streptococcal BSIs from September 2019 to September 2021 at an academic teaching hospital located in Dallas, Texas. Patients ≥18 years of age were included if they had a positive blood culture for any Streptococcus spp. and received treatment with an active antimicrobial agent. Patients were excluded if they had a polymicrobial BSI, if the isolated organism was deemed a contaminant by the treatment team, if the patient had persistently positive blood cultures despite appropriate IV antibiotic therapy, or if the patient was treated for a complicated BSI, defined as a presence of endovascular or metastatic focus of infection, intravascular prosthetic device, undrained abscess >5 cm, deep tissue or bone and joint involvement, extensive burn injury, infection involving the central nervous system, or death during hospitalization. In addition, patients who received ≥16 days of antibiotic therapy were excluded on the assumption that infections requiring longer than 2 weeks (± 2 d) were less likely to be uncomplicated. Patients were also excluded if they did not complete their intended antibiotic therapy due to withdrawal of care or leaving against medical advice.
Patients were grouped by individuals that received an entire treatment course of IV antibiotics and individuals that transitioned to PO antibiotics after an initial course of IV antibiotics. Manual chart review of the electronic medical record was used to confirm patient eligibility, baseline characteristics, treatment information, and clinical outcomes. Study approval was obtained by the UT Southwestern Institutional Review Board.
Patients were identified as immunocompromised if they had a congenital immunodeficiency, solid organ or bone marrow transplantation, HIV with a CD4 cell count <200 cells/mm, Reference Hospenthal, Waters, Beekmann and Polgreen3 neutropenia (absolute neutrophil count <1000 cells/mm), Reference Hospenthal, Waters, Beekmann and Polgreen3 and patients on immunosuppressive drugs or active chemotherapy.
Outcomes
The primary endpoint was clinical failure, which was defined as a composite endpoint of BSI recurrence and/or infection-related readmission within 30 days of antibiotic completion. Reference Harris, McNamara and Paterson16 BSI recurrence was defined as a positive blood culture with the same organism identified in the index blood culture within 30 days following completion of the intended antibiotic course. Infection-related readmission was defined as readmission due to inadequate clinical improvement from the original source of infection or an adverse event related to antimicrobial therapy. Secondary endpoints included individual components of the primary outcome, hospital length of stay (LOS), and all-cause mortality at 30 days from antibiotic completion.
Statistical analysis
Descriptive statistics were used to assess all variables including frequencies and percentages for categorical data and medians and interquartile ranges (IQR) for continuous data. Baseline characteristics, primary outcome, and other secondary outcomes used χ2 or Fisher’s exact test for categorical variables and Mann-Whitney U test for continuous variables. A multivariate logistic regression of the primary outcome was made using variables that had a p value <.10 in the univariate analysis or were found to be clinically relevant from literature review. Odds ratios (OR) with 95% confidence intervals (CI) were used to summarize logistic regression results. A two-tailed p value of less than .05 was considered statistically significant.
Results
448 patients with streptococcal bacteremia between September 2019 and September 2021 were screened for study inclusion and 155 (IV = 77; PO stepdown = 78) met eligibility for inclusion (Figure 1). The median patient age was 53 years (IQR, 43 to 64 yr), the majority of patients were male, 112 (72.3%), and only 10% had private commercial insurance. The baseline characteristics were similar between treatment groups, which included 16% immunocompromised patients (Table 1). More patients in the IV group had cirrhosis, but this was not statistically significant, (20 [26.0%] vs 11 [14.1%]; p = .07).
Figure 1. Study inclusion and exculsion.
Table 1. Patient characteristics
Note: IV, intravenous; PO, oral; IQR, interquartile range.
The majority of BSIs were caused by viridans group streptococci (61 [39.4%]), followed by S. agalactiae (35 [22.6%]) and S. dysgalactiae (33 [21.3%]). The most common source of bacteremia was skin and soft tissue (70 [45.2.3%]), followed by intra-abdominal (28 [18.1%]) and odontogenic (15 [9.7%]). The source of infection remained unknown in 13.5% of cases.
All empiric IV antibiotic regimens demonstrated antimicrobial activity against the isolated organism based on predictable susceptibility or documented susceptibility reports. In patients transitioned to PO therapy, the median duration of IV lead-in therapy was 5 days (IQR, 3.5 to 7.0 d). Most patients in the IV group received targeted therapy with a beta-lactam (72 [93.5%]). An IV to PO switch was more likely for BSIs secondary to β-hemolytic streptococci (49 [62.8%]) than for α-hemolytic streptococci (26 [33.3%]). The most common PO antibiotics used for stepdown were fluroquinolones (36 [46.2%]) followed by beta-lactams (31 [39.7%]) and linezolid (10 [12.8%]) (Figure 2). The median total treatment duration in both groups was 14 days (IV group, 14 d [IQR, 8.0 to 14.0 d] vs PO group, 14 d [IQR, 14.0 to 14.5 d]; p < .01). Of the patients that received a full course of IV antibiotics, 33% were sent home on outpatient parenteral antimicrobial therapy.
Figure 2. Selection of oral antibiotic stepdown agents.
There was no difference in clinical failure rates between IV (15.6%) and PO stepdown (15.4%) therapy (OR .99 [95% CI, .41 to 2.35]; p = .97). Clinical failure was primarily driven by infection-related readmission (12 [15.6%] vs 12 [15.4%]; p = .97), with only one patient in the IV group that had documented BSI recurrence (Table 2). Clinical worsening from the original infectious source after antibiotic completion accounted for 87.5% of infection-related readmissions. Three patients in the IV group were re-admitted for line-related complications. All-cause mortality was not different between the two groups (2 [2.6%] vs 0 [0%]; p = .25). Patients that received PO antibiotic stepdown therapy had a significantly shorter median hospital LOS by 6 days (12 vs 6 d; p < .01). The multivariate logistic regression performed on the primary outcome, clinical failure, found none of the included variables to be statistically significant (Table 3).
Table 2. Clinical outcomes
Note: IV, intravenous; PO, oral; BSI, bloodstream infection; IQR, interquartile range; OR, odds ratio; CI, confidence interval. aOne patient with BSI recurrence included in infection-related readmission.
Table 3. Multivariate logistic regression for primary outcome, clinical failure
Note: IV, intravenous; PO, oral; OR, odds ratio; CI, confidence interval.
Discussion
This retrospective cohort study of patients with uncomplicated streptococcal bacteremia found similar clinical failure rates between patients treated with IV antibiotics and those transitioned to PO antibiotics. Most failures were due to infection-related readmissions with only one case of recurrent bacteremia which occurred in the IV group. Mortality was rare and not included in our definition of clinical failure. These findings align with prior studies showing similar efficacy with PO stepdown therapy for streptococcal bacteremia, though comparisons are limited by varying definitions of uncomplicated bacteremia, clinical failure, and inclusion of variable sources, Streptococcus species, and therapeutic regimens. Notable differences in our patient population include more BSIs secondary to skin and soft tissue and intra-abdominal infections and lower rates of respiratory or line-related sources. Our study also showed a higher incidence of viridans group streptococcus compared to S. pneumoniae, which was commonly identified in prior cohorts. Reference Kang, Beuttler and Minejima17–Reference Lew, Salam and Gross20 The clinical failure rate observed in the present study most closely aligns with Kang et al., who found no difference between PO and IV groups in re-admission rates (15% vs 20%, p = .4), mortality (1% vs 4%, p = .25), or BSI recurrence (1% vs 2%, p = .65). Reference Kang, Beuttler and Minejima17 Similarly, Waked et al. reported no significant difference in clinical failure, a composite endpoint of 90 day readmission and/or mortality between PO and IV groups (18% vs 24 %, p = .23). Reference Waked, Craig, Mercuro, Wungwattana, Wood and Rokas18 Ramos-Otero et al. observed clinical failure only in the IV group (0% vs 19%, p = .001), primarily due to 30 day reinfection or new-onset sepsis. Reference Ramos-Otero, Sarangarm and Walraven19 Lew et al. reported low failure rates overall, and no difference between PO and IV groups in terms of infection recurrence (0% vs 1.6%, p = .5), infection-related mortality (.9% vs 1.6%, p = 64), or infection-related readmission (4.5% vs 3.2%, p = .74) and low line-related complications of 3.2%. Reference Lew, Salam and Gross20 Consistent with prior studies, patients who received oral therapy in our cohort had shorter hospital stays. Kang et al. reported a median LOS of 5 versus 10 days, (PO vs IV, p < .01) and other studies similarly found LOS reductions by 4–6 days. Reference Kang, Beuttler and Minejima17–Reference Lew, Salam and Gross20 In our study, patients transitioned to oral antibiotics were hospitalized six fewer days, reinforcing the established benefit of oral therapy.
Several limitations to our study are secondary to the retrospective design including susceptibility to confounding variables. A multivariate logistic regression analysis was performed and found no variables including age, cirrhosis, source, and species classification, were statistically associated with clinical failure, but sample size may have been too small to detect a difference. This cohort also did not capture post-discharge data such as medication adherence, external readmissions, or mortality events, but the patient population we serve primarily follows at our institution.
The study was not designed to provide guidance for total duration of therapy nor duration of IV therapy lead-in prior to oral transition. The average time to oral stepdown in our study was 5 days with an average cumulative duration of 14 days, which was similar to other studies. Reference Kang, Beuttler and Minejima17,Reference Ramos-Otero, Sarangarm and Walraven19 In light of increasing data supporting shorter durations of antibiotic therapy, future studies are needed to determine the shortest effective duration for uncomplicated streptococcal BSIs. Reference Spellberg and Rice21
Given significant variability amongst selected antimicrobial agents and dosing strategies, we were unable to evaluate the impact of drug selection on clinical outcomes. Fluroquinolones were more frequently prescribed than oral beta-lactams in our cohort. The ideal agent for oral stepdown therapy in streptococcal bacteremia has not been well studied. Agents such as fluoroquinolones, linezolid, and clindamycin may be favored over oral beta-lactams given higher bioavailability. However, these agents may be unnecessarily broad and have higher risks for collateral damage and adverse effects. Reference Floris22 On the other hand, Streptococcus spp. have low minimum inhibitory concentrations to beta-lactams; however, oral bioavailability varies amongst agents and there are limited data supporting optimal dosing and correlation to pharmacodynamic target attainment. Reference Hale, Snyder, Ahern, Eliopoulos, Ricotta and Alston23 In a study comparing oral therapies for non-staphylococcal gram-positive BSIs, the authors found no difference in clinical failure between high bioavailability agents (ie, fluroquinolones and linezolid) and low bioavailability agents (ie, beta-lactams). Reference Quinn, Sebaaly, Patel, Weinrib, Anderson and Roshdy14 Similar findings were reported in a study comparing oral fluoroquinolones versus beta-lactams for uncomplicated streptococcal BSIs. Reference Arensman, Shields and Beganovic15
In summary, our study found that in hospitalized patients with uncomplicated streptococcal BSIs, oral antibiotic stepdown therapy demonstrated similar outcomes to a continued treatment course with IV antibiotics. Important considerations prior to oral stepdown for bacteremia include clinical stability, adequate source control, and selection of an effective and highly bioavailable agent. Reference Hale, Snyder, Ahern, Eliopoulos, Ricotta and Alston23–Reference Shelburne and Musher24 Adequate investigation for need of source control and endovascular source should be evaluated for streptococcal bacteremia, especially in species with higher risk of endocarditis such as S. gallolyticus, S. sangunis, S. gordonii, S. mitis/oralis, S. mutans, and S. cristatus. Reference Chamat-Hedemand, Dahl and Ostegaard25 Infectious diseases consultation has been shown to improve clinical outcomes in guiding optimal therapy as well as providing recommendations for source control. Reference Fourre, Zimmermann, Senn, Aruanno, Guery and Matthaios26 The results of this study are hypothesis-generating and may assist in stewardship efforts to reduce prolonged courses of IV antibiotics, avoid central line placement for outpatient access, and help facilitate earlier transitions of care in patients necessitating antibiotics for uncomplicated streptococcal BSIs.
Acknowledgements
Anusha Govind, MD, assisted with review of the final manuscript.
Author contributions
All listed authors have contributed significantly to this publication. J.L., J.O., N.M., B.P., W.W. participated in study design, data analysis, and manuscript development. L.B. participated in statistical analyzation and manuscript editing.
Financial support
None.
Competing interests
All authors: No reported conflicts.
Open access
