Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 2.
Published in final edited form as: Eur J Clin Microbiol Infect Dis. 2023 Feb 17;42(4):423–430. doi: 10.1007/s10096-023-04575-z

Post-treatment Outcomes of Ceftriaxone Versus Antistaphylococcal Penicillins or Cefazolin for Definitive Therapy of Methicillin-susceptible Staphylococcus aureus Bacteremia

Zachary A Yetmar 1, Ryan B Khodadadi 1, John Raymond Go 1, Supavit Chesdachai 1, Omar M Abu Saleh 1
PMCID: PMC13041233  NIHMSID: NIHMS2154811  PMID: 36800065

Abstract

Purpose:

Methicillin-susceptible Staphylococcus aureus (MSSA) bacteremia is associated with poor outcomes. Ceftriaxone offers logistical advantages over other standard therapies, though in vitro studies have questioned its efficacy and clinical studies of ceftriaxone in MSSA bacteremia are conflicting.

Methods:

We performed a multicenter, retrospective cohort study of adult patients who received ceftriaxone, cefazolin, or antistaphylococcal penicillins as definitive therapy for MSSA bacteremia from 2018–2019. Definitive therapy was defined as the antibiotic used in the outpatient setting. Patients were excluded if they received less than 7 days of outpatient therapy. Follow-up started on the date of definitive therapy completion. The primary outcome was 90-day treatment failure, defined as a composite of mortality and microbiologic recurrence. This was analyzed with multivariable Cox regression.

Results:

223 patients were included, 37 (16.6%) of whom received ceftriaxone. The most common ceftriaxone dose was 2 g daily (83.8%). The most common primary site of infection was skin/soft tissue (37.2%), unknown (21.1%), and catheter-related (15.2%). Twenty-six (11.7%) developed infective endocarditis. Median total duration of treatment was 31.0 days, and median outpatient duration was 24.0 days. 26 (11.7%) developed 90-day treatment failure. After adjusting for Charlson comorbidity index, duration of therapy, and use of transesophageal echocardiography, definitive treatment with ceftriaxone was associated with treatment failure (hazard ratio 2.66, 95% confidence interval 1.15–6.12; p=0.022).

Conclusions:

Among patients with MSSA bacteremia, definitive treatment with ceftriaxone was associated with a higher risk of treatment failure within 90 days as compared to cefazolin or antistaphylococcal penicillins.

Keywords: Staphylococcus aureus, bloodstream infection, ceftriaxone, cefazolin, oxacillin

Background

Staphylococcus aureus bacteremia is a common infectious syndrome with a predisposition for causing metastatic infection, including infective endocarditis (IE), osteomyelitis, and other deep-seated infectious syndromes. Mortality from Staphylococcus aureus bacteremia is high, with as much as 30% of patients dying within months of diagnosis [1]. As such, treatment is typically prolonged with most patients receiving 4–6 weeks of intravenous antibiotic therapy [2,3]. Outcomes from treatment with antistaphylococcal penicillins (ASP) and cefazolin are generally considered comparable, and use of these agents is typically recommended for treatment of methicillin-susceptible Staphylococcus aureus (MSSA) [4,5].

However, there has been recent interest in ceftriaxone for treatment of MSSA bacteremia. Ceftriaxone carries potential advantages over ASPs and cefazolin, namely convenience of outpatient dosing compared to these standard therapies. However, there are no prospective or randomized data for the use of ceftriaxone for MSSA, and retrospective studies have yielded mixed results [68]. Two recent meta-analyses found outcomes to be similar between ceftriaxone and ASPs or cefazolin [9,10], although there were few studies found in the literature and these results lacked certainty. Furthermore, in vitro data have questioned the efficacy of ceftriaxone for MSSA [11,12].

In light of this, we aimed to analyze our center’s cohort of patients with MSSA bacteremia, comparing outcomes of those who received definitive therapy with ceftriaxone to those who received ASP or cefazolin.

Methods

Study design

We conducted a multicenter, retrospective cohort study of adult patients diagnosed with MSSA bacteremia between January 2018 and December 2019 at Mayo Clinic campuses in Arizona, Florida, and Minnesota, and the Mayo Clinic Health System sites in Minnesota and Wisconsin. Inclusion criteria were age ≥18 years, receipt of at least 7 days of outpatient treatment with ceftriaxone, cefazolin, or an ASP, and follow-up at least through the end of treatment. Patients who received less than 7 days of outpatient therapy, polymicrobial bacteremia, definitive treatment with an antibiotic other than ceftriaxone, cefazolin, or an ASP, outpatient combination therapy, or chronic suppressive antibiotic therapy following the definitive course, and lack of research authorization per Minnesota statute were excluded. Data from eligible patients were manually extracted to a pre-specified data extraction form (Supplementary materials). Abstracted data included demographics, comorbid conditions, index hospitalization and diagnostic details, infectious complications, treatment details, and outcomes. Study data were collected and managed using REDCap electronic data capture tools hosted at Mayo Clinic [13,14]. Our internal institutional review board reviewed the study protocol and granted it an exempt status (#19–005199).

Microbiology

At least two blood culture sets were obtained from each patient. The Becton Dickinson BD BACTEC FX platform (BD Diagnostics, Sparks, MD) was used with each blood culture set consisting of 1 BD BACTEC lytic Anaerobic/F bottle and 1 or 2 BD BACTEC Plus Aerobic/F bottles using a standard incubation duration of 5 days. If blood culture bottles flagged positive on the BD BACTEC FX platform, further workup included Gram staining and subculturing onto appropriate media. The BioFire BCID2 FilmArray panel is also used for rapid identification and detection of mecA while further identification and phenotypic susceptibility testing are in process. Identification of S. aureus was performed by matrix-assisted laser desorption/ionization–time of flight (MALDI-TOF) and/or with a combination of morphological and biochemical characteristics. Isolates underwent phenotypic antimicrobial susceptibility testing by agar dilution and results were interpreted using Clinical and Laboratory Standards Institute guidelines [15].

Definitions

S. aureus bacteremia was defined as a growth of S. aureus in the blood culture. MSSA was defined as 1) absence of the mecA gene and 2) oxacillin minimum inhibitory concentration (MIC) ≤2 mcg/mL. The clinical significance of S. aureus bacteremia was determined by treating providers and ascertained by the authors. Definitive therapy was defined as the antibiotic agent used for the majority of outpatient treatment. Uncomplicated S. aureus bacteremia was defined as 1) no evidence of IE 2) no implanted prostheses 3) duration of bacteremia less than 5 days 4) resolution of fever within 72 hours of initiating effective therapy 5) no evidence of metastatic sites of infection. Patients who did not meet the uncomplicated criteria were categorized as complicated S. aureus bacteremia [3]. The primary outcome was 90-day treatment failure, defined as a composite of mortality or microbiologic recurrence within 90 days of antibiotic completion. Microbiologic recurrence was defined as culture isolation of S. aureus in the setting of a compatible clinical syndrome. Secondary outcomes included 30-day treatment failure as well as mortality and microbiologic recurrence individually within 30 and 90 days. The Charlson comorbidity index (CCI) was calculated with age-adjustment [16]. Community-acquired bacteremia was defined as a positive blood culture obtained within 48 hours of hospital admission without recent healthcare involvement [17]. Nosocomial bacteremia was defined as a positive blood culture after at least 48 hours of hospitalization. IE was defined according to the modified Duke criteria [2]. Duration of bacteremia was the number of days between the first positive and first negative blood culture. ASPs included nafcillin and oxacillin.

Statistical analysis

Continuous variables are presented as mean (standard deviation) or median (interquartile range [IQR]), as appropriate. Categorical variables are presented as number (percentage). An inversed Kaplan-Meier curve was constructed and compared utilizing a log-rank test for survival between treatment groups. A multivariable Cox proportional hazards model analysis was performed to analyze for an association between ceftriaxone use and the primary outcome after adjusting for factors that may also impact outcomes. Patients were censored at their last follow-up, at the time of treatment failure, or 90 days, whichever occurred first. The Cox regression analysis was stratified by treatment site and the proportionality assumption was confirmed by Schoenfeld residual testing. All analyses were performed using R version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Cohort characteristics

Two-hundred and twenty-three patients with MSSA bacteremia met inclusion criteria (Figure 1). Most patients were male (N=138; 61.9%) with a mean age of 63.1 years. The most common source of MSSA bacteremia was skin and soft tissue infections, while 21.1% of patients had an unknown source. Twenty-six (11.7%) were diagnosed with IE, 24 of which received definitive treatment with an ASP or cefazolin. Most patients received inpatient therapy with an ASP or cefazolin, with only 4.0% receiving ceftriaxone in the inpatient setting. Further cohort characteristics are detailed in Table 1.

Figure 1:

Figure 1:

Flow diagram showing exclusion of patients from the final cohort.

Table 1:

Characteristics of 223 patients with methicillin-susceptible Staphylococcus aureus bacteremia

Antistaphylococcal penicillin or cefazolin (N=186) Ceftriaxone (N=37) Total (N=223)
Age, years, mean (SD) 62.1 (16.9) 68.2 (15.8) 63.1 (16.9)
Male sex 118 (63.4) 20 (54.1) 138 (61.9)
Race
- Asian 3 (1.6) 0 (0.0) 3 (1.3)
- Black or African American 5 (2.7) 0 (0.0) 5 (2.2)
- White 177 (95.2) 37 (100.0) 214 (96.0)
- Unknown 1 (0.5) 0 (0.0) 1 (0.4)
BMI, kg/m2, mean (SD) 30.3 (8.1) 30.4 (7.9) 30.3 (8.0)
IV drug use 8 (4.3) 1 (2.7) 9 (4.0)
Age-adjusted CCI, mean (SD) 7.4 (4.4) 7.8 (4.3) 7.5 (4.4)
Cardiac prosthetic device 23 (12.4) 1 (2.7) 24 (10.8)
Hospital length of stay, days, median (IQR) 8.0 (5.8, 11.6) 7.3 (4.8, 10.9) 7.9 (5.6, 11.6)
Bacteremia onset
- Community-onset 74 (39.8) 14 (37.8) 88 (39.5)
- Healthcare-associated 101 (54.3) 20 (54.1) 121 (54.3)
- Nosocomial 11 (5.9) 3 (8.1) 14 (6.3)
Duration of bacteremia, days, median (IQR) 1.9 (1.3, 2.9) 1.8 (1.1, 3.0) 1.9 (1.3, 2.9)
Source of infection
- Catheter-related 31 (16.7) 3 (8.1) 34 (15.2)
- Infective endocarditis 1 (0.5) 0 (0.0) 1 (0.4)
- Gastrointestinal 3 (1.6) 2 (5.4) 5 (2.2)
- Genitourinary 7 (3.8) 1 (2.7) 8 (3.6)
- Osteoarticular 21 (11.3) 4 (10.8) 25 (11.2)
- Pneumonia 9 (4.8) 4 (10.8) 13 (5.8)
- Skin/soft tissue 67 (36.0) 16 (43.2) 83 (37.2)
- Unknown 42 (22.6) 5 (13.5) 47 (21.1)
- Othera 5 (2.7) 2 (5.4) 7 (3.1)
Infectious diseases consult 183 (98.4) 34 (91.9) 217 (97.3)
Source control 183 (98.4) 37 (100.0) 220 (98.7)
TEE performed 110 (59.1) 15 (40.5) 125 (56.1)
Complicated bacteremia 134 (72.0) 20 (54.1) 154 (69.1)
Infective endocarditis 24 (12.9) 2 (5.4) 26 (11.7)
Inpatient antibiotic
- Cefazolin 153 (82.3) 21 (56.8) 174 (78.0)
- Ceftriaxone 0 (0.0) 9 (24.3) 9 (4.0)
- Nafcillin 14 (7.5) 3 (8.1) 17 (7.6)
- Oxacillin 13 (7.0) 3 (8.1) 16 (7.2)
- Vancomycin 1 (0.5) 1 (2.7) 2 (0.9)
- Otherb 5 (2.7) 0 (0.0) 5 (2.2)
Total antibiotic duration, days, median (IQR) 32.0 (27.2, 45.0) 28.0 (16.0, 45.0) 31.0 (27.0, 45.0)
- Inpatient duration, days, median (IQR), 8.0 (6.0, 11.0) 7.0 (5.0, 11.0) 7.0 (5.0, 11.0)
- Outpatient duration, days, median (IQR) 24.0 (17.0, 35.0) 22.0 (10.0, 34.0) 24.0 (14.5, 35.0)

Data are N (%) unless otherwise specified.

Abbreviations: BMI, body mass index; CCI, Charlson comorbidity index; IQR, interquartile range; IV, intravenous; MIC, minimum inhibitory concentration; SD, standard deviation; TEE, transesophageal echocardiogram

a

Other sources of infection include thrombophlebitis (4), cardiac device pocket (2), and dialysis fistula.

b

Other inpatient antibiotics include piperacillin-tazobactam (2), cefazolin + rifampin, ceftazidime, and vancomycin + ceftriaxone.

From the total cohort, 37 (16.6%) received definitive therapy with ceftriaxone. Definitive ceftriaxone treatment was most commonly dosed at 2 grams once daily (N=31; 83.8%), followed by 2 grams twice daily (N=5; 13.5%) and 1 gram once daily (N=1; 2.7%). Of those who did not receive ceftriaxone, the most common definitive antibiotic was cefazolin (N=168; 90.3%) with only 10 (5.4%) receiving nafcillin and 8 (4.3%) oxacillin. The ceftriaxone group had a lower proportion of patients with a cardiac prosthetic device (2.7% versus 12.4%), transesophageal echocardiogram (TEE) performance (40.5% versus 59.1%), complicated bacteremia (54.1% versus 72.0%), and diagnosis of IE (5.4% versus 12.9%). Nearly all patients (N=217; 97.3%) had an inpatient infectious diseases consultation.

Outcomes

Twenty-six (11.7%) patients developed the primary outcome of 90-day treatment failure (Table 2). This was comprised of mortality in 18 patients and microbiologic recurrence in 8 patients. Ten patients developed 30-day treatment failure, including 6 patients who died and 4 patients with microbiologic recurrence. Figure 2 shows the timing of treatment failure, comparing groups based on receipt of definitive therapy with ceftriaxone or ASP/cefazolin. In a site-stratified multivariable analysis, ceftriaxone was also associated with 90-day treatment failure after adjusting for CCI, total antibiotic duration, and performance of TEE (Table 3). From the 10 patients in the ceftriaxone who developed 90-day treatment failure, 9 received ceftriaxone dosed 2 grams daily and 1 received 2 grams twice daily.

Table 2:

Treatment outcomes stratified by definitive therapy

Antistaphylococcal penicillin or cefazolin (N=186) Ceftriaxone (N=37) Total (N=223)
90-day treatment failure 16 (8.6) 10 (27.0) 26 (11.7)
- 90-day mortality 11 (5.9) 7 (18.9) 18 (8.1)
- 90-day recurrence 5 (2.7) 3 (8.1) 8 (3.6)
30-day treatment failure 7 (3.8) 3 (8.1) 10 (4.5)
- 30-day mortality 4 (2.2) 2 (5.4) 6 (2.7)
- 30-day recurrence 3 (1.6) 1 (2.7) 4 (1.8)

Data are N (%).

Figure 2:

Figure 2:

Inversed Kaplan-Meier curves comparing 90-day treatment failure by definitive antibiotic between antistaphylococcal penicillins or cefazolin and ceftriaxone. The shaded regions indicate the 95% confidence interval. The dashed vertical line denotes day 30 of follow-up.

Table 3:

Multivariable Cox proportional hazards model for risk of 90-day treatment failure, stratified by treatment site

Variable Hazard Ratio (95% Confidence Interval) p value
Ceftriaxone 2.66 (1.15–6.12) .022
Age-adjusted CCI (per unit) 1.10 (1.01–1.20) .033
Total antibiotic duration (per day) 0.96 (0.93–0.99) .014
TEE performed 0.94 (0.38–2.31) .886

Abbreviations: CCI, Charlson comorbidity index; TEE, transesophageal echocardiogram

Discussion

Among patients with MSSA bacteremia who have completed definitive therapy, this study showed ceftriaxone is associated with a higher risk of 90-day treatment failure, defined as mortality or microbiologic recurrence. This association persisted despite controlling for potential confounding factors.

Several studies examining the use of ceftriaxone for MSSA in the in vitro setting have questioned its efficacy. One such study compared use of ceftriaxone with four dosing regimens: 1 g daily, 1 g twice-daily, 2 g daily, and 2 g twice-daily [11]. Only the 2 g dosing regimens achieved early reductions in bacterial burden. Additionally, only the 2 g twice-daily regimen sustained this reduction beyond 96 hours. Another study using Monte Carlo simulations found the common 2 g daily dosing to predict bacteriostasis in about half of isolates and bactericidal activity in just 17% [12]. While these in vitro effects do not always translate to clinical outcomes, it raises the question if the diminished effectiveness of ceftriaxone may be related to dosing. The most common dosing in this study was 2 g daily, and it is possible outcomes may have been different had 2 g twice-daily dosing been used. However, clinical data examining this are lacking and future study is required.

Past studies of ceftriaxone for MSSA have found mixed results [68,1822]. A study by Carr, et al. showed patients treated with ceftriaxone had a higher risk of treatment failure, defined as a composite of 8 possible endpoints, with 55% of ceftriaxone-treated patients experiencing this outcome [8]. Interestingly, this study found a longer duration of treatment to be associated with greater odds of treatment failure, as opposed to our study which found this to be protective. However, extension of parenteral therapy was included as a criterion for treatment failure, which may have influenced this analysis. Contrary to this study, a separate analysis did not show any association between ceftriaxone use and poor outcomes [7]. Notably, about one-third of patients in this study had IE, about half of which received ceftriaxone. Among this subgroup, patients who received ceftriaxone had higher 90-day mortality (14.3% versus 2.4%), though this was not statistically significant likely on account of the small number with IE.

These studies have also used a variety of different outcomes to assess effectiveness of ceftriaxone, often a composite of different endpoints denoted as “treatment failure”. The study by Carr et al. incorporated 8 different 90-day outcomes including extension of parenteral therapy, incomplete antibiotic course, relapse, readmission, unplanned surgery, planned oral antibiotic suppression, loss to follow-up, and mortality [8]. Hamad et al. defined treatment failure as mortality, infection-related readmission, or microbiologic recurrence within 90-days [7]. Other definitions of treatment failure have included persistent bacteremia after source control, progression of infection on therapy, microbiologic relapse, and mortality [22], infection-related readmission or lack of improvement in clinical, laboratory, or radiographic markers [6], and treatment not meeting expectations of outpatient parenteral antibiotic therapy [20,21]. Timing of endpoints have also been variable, ranging from 7 days to 6 months [68,18,22]. These differences in interpretation of what constitutes treatment failure have contributed to heterogeneity between studies, making between-study comparison difficult. Future research would benefit from standardization of the definition of treatment failure as an objective, clinically relevant, and easily measurable endpoint.

It is important to note that studies examining ceftriaxone have largely used oxacillin susceptibility testing to extrapolate ceftriaxone susceptibility for S. aureus. While there was a previous concern regarding the correlation between these tests, recent studies have shown very high concordance between oxacillin and ceftriaxone susceptibility testing [23,24].

One important source of potential confounding is socioeconomic status. Daily ceftriaxone offers many logistic advantages over other standard parenteral options. For patients with normal renal function, cefazolin is typically administered three-time-daily while outpatient ASPs require either more frequent administration or use of a continuous infusion pump. Access to continuous infusion mechanisms and resources for home administration of cefazolin is often dictated by insurance coverage and financial means. Additionally, time off from work and other responsibilities may limit frequent attending an outpatient infusion center multiple times per day. These aspects preferentially disadvantage patients with lower socioeconomic status, measures of which have been associated with poor outcomes from infections [2527]. Indeed, in our centers, cefazolin and ASPs are preferred agents, and ceftriaxone is typically only used when limited by these logistic and socioeconomic barriers. We were unable to control for these factors in the present study, though these should be considered in future studies.

This study has several limitations of note. It was conducted retrospectively and is susceptible to multiple sources of bias from this study design. Additionally, it was not randomized and there are likely residual confounders in spite of controlling for some relevant factors in multivariable analysis. There were relatively few patients who received ceftriaxone as definitive therapy, which may have limited our analysis. Not all deaths included in the primary outcome are directly infection-related; however, it is often difficult to disentangle complications of significant infections and unrelated issues contributing to mortality. We also only enrolled patients and examined outcomes after completion of definitive therapy for MSSA bacteremia, and these data do not necessarily apply to more immediate outcomes that occur during treatment, such as development of metastatic complications or early mortality. There is also a potential for selection bias as this study only included patients who had survived to complete definitive therapy and were able to discharge from the hospital prior to therapy completion. There were very few patients with infective endocarditis who received ceftriaxone, and we were unable to evaluate if these patients specifically had differences in outcomes based on definitive therapy choice. Finally, most patients treated with ceftriaxone received cefazolin or an ASP during their early, inpatient therapy, which may have influenced outcomes as well.

In conclusion, this study found definitive treatment with ceftriaxone to be associated with a higher risk of mortality or infection recurrence among patients with MSSA bacteremia. Given the heterogeneity in the literature and questions regarding optimal ceftriaxone dose, the uncertainty regarding efficacy of ceftriaxone for MSSA bacteremia will likely require prospective, randomized study. In the interim, these data do not support ceftriaxone for MSSA bacteremia and other options such as cefazolin and ASPs should be considered instead.

Acknowledgements

Preliminary results of this work were presented at IDWeek 2022 (abstract #738).

Funding

This project was supported by Grant Number UL1 TR002377 from the National Center for Advancing Translational Sciences (NCATS). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Footnotes

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Ethics Approval: Our internal institutional review board reviewed the study protocol and granted it an exempt status (#19–005199).

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

RESOURCES