Authors: Steven Y. C. Tong (1. Victorian Infectious Diseases Service, The Royal Melbourne Hospital, at the Peter Doherty Institute for Infection and Immunity, Melbourne, Australia; 2. Department of Infectious Diseases, The University of Melbourne at the Peter Doherty Institute for Infection and Immunity, Melbourne, Australia), Vance G. Fowler, Jr (3. Division of Infectious Diseases, Duke University, Durham, North Carolina, USA; 4. Duke Clinical Research Institute, Duke University, Durham, North Carolina, USA), Lesley Skalla (5. Duke University Medical Center Library & Archives, Duke University School of Medicine, Durham, North Carolina, USA), Thomas L. Holland (3. Division of Infectious Diseases, Duke University, Durham, North Carolina, USA; 4. Duke Clinical Research Institute, Duke University, Durham, North Carolina, USA)
Categories: Article, Staphylococcus aureus, bacteremia, MRSA
Source: JAMA
Authors: Steven Y. C. Tong, Vance G. Fowler, Lesley Skalla, Thomas L. Holland
Staphylococcus aureus, a gram-positive bacterium, is the leading cause of death from bacteremia worldwide, with a case fatality rate of 15–30% and an estimated 300,000 deaths per year.
S. aureus bacteremia causes metastatic infection in over one-third of cases, including endocarditis (~12%), septic arthritis (7%), vertebral osteomyelitis (~4%), spinal epidural abscess, psoas abscess, splenic abscess, septic pulmonary emboli, and seeding of implantable medical devices. Patients with S. aureus bacteremia commonly present with fever (>90%), or symptoms from metastatic infection such as pain in the back, joints, abdomen or extremities, and/or change in mental status. Risk factors include intravascular devices such as implantable cardiac devices and dialysis vascular catheters, recent surgical procedures, injection drug use, diabetes mellitus, and previous S. aureus infection. S. aureus bacteremia is detected with blood cultures. Prolonged S. aureus bacteremia (≥ 48 hours) is associated with a 90-day mortality risk of 39%. All patients with S. aureus bacteremia should undergo transthoracic echocardiography; transesophageal echocardiography should be performed in patients at high risk for endocarditis such as those with persistent bacteremia, persistent fever, metastatic infection foci, or implantable cardiac devices. Other imaging modalities, such as computed tomography or magnetic resonance imaging, should be performed based upon symptoms and localizing signs of metastatic infection. S. aureus is categorized as methicillin-susceptible (MSSA) or methicillin-resistant (MRSA) based upon susceptibility to β-lactam antibiotics. Initial treatment for S. aureus bacteremia typically includes antibiotics active against MRSA such as vancomycin or daptomycin. Once antibiotic susceptibility results are available, antibiotics should be adjusted. Cefazolin or anti-staphylococcal penicillins should be used for MSSA and vancomycin, daptomycin, or ceftobiprole for MRSA. Phase 3 trials for S. aureus bacteremia demonstrated non-inferiority of daptomycin to standard of care (treatment success of 53/124 [43%] vs 48/122 [39%]), and non-inferiority of ceftobiprole to daptomycin (treatment success of 132/189 [70%] vs 136/198 [69%]). Source control is a critical component of treating S. aureus bacteremia, and may include removal of infected intravascular or implanted devices, drainage of abscesses, and surgical debridement.
S. aureus bacteremia has a case fatality rate of 15–30%, and causes 300,000 deaths per year worldwide. Empiric antibiotic treatment should include vancomycin or daptomycin, which are active against MRSA. Once S. aureus susceptibilities are known, MSSA should be treated with cefazolin or an anti-staphylococcal penicillin. Additional clinical management consists of identifying sites of metastatic infection and pursuing source control for identified foci of infection.
In 2019, Staphylococcus aureus was the leading bacterial cause of death in 135 countries.^1^ Among multidrug resistant infections in hospitalized patients in the US in 2017, an estimated 52% were caused by methicillin-resistant S. aureus (MRSA).^2^ While the rate of endocarditis in US patients with S. aureus bacteremia has declined from over 50%^3^ in 1954 to ~12% in 2017, rates of infections involving implantable foreign bodies have increased.^4^ Despite improvements in treatment and diagnosis, 90 day mortality among patients with S. aureus bacteremia is 27.0% (95% CI, 21.5%-33.3%).^5^
A 2014 JAMA review of S. aureus bacteremia identified only 1 high-quality trial to guide antibiotic therapy.^6^ Over the past 10 years, more studies have been published about diagnostic and treatment strategies, and in 2024, the US Food and Drug Administration (FDA) issued regulatory approval of a novel antibiotic, ceftobiprole, for S. aureus bacteremia.^7^ This review will cover key aspects of the clinical management of S. aureus bacteremia, including evidence-based treatment options.
We conducted a search for randomized clinical trials of S. aureus bacteremia antibiotic treatment published from January 1 2014 through January 25, 2025 in MEDLINE (via PubMed) and the Cochrane Central Register for Controlled Trials (Wiley). Search terms included a mix of keywords and MeSH terms representing the concepts of S. aureus, bacteremia, and antibiotics. The full, reproducible search strategies are available in Supplementary Information. Of 1624 articles initially identified by the search, there were 22 randomized clinical trials. Overall, of the 101 articles included in this review article there were 43 observational cohorts, 22 randomized clinical trials, 9 reviews, 10 systematic reviews or meta-analyses, 8 randomized clinical trial protocols, 2 laboratory studies, and 7 guidelines.
Based on data from high income countries, the incidence of S. aureus bacteremia ranges from 9.3 to 65 cases per 100,000 person-years.^8^ Risk factors include central venous catheters, implanted cardiac or other prosthetic devices, injection drug use, hemodialysis (particularly when vascular access is via central venous catheter),^9^ recent surgical procedures, and host factors such as male sex (male-to-female ratio of ~1.5), extremes of age (≤1 year and ≥70 years of age),^10^ lower socioeconomic status,^11^ diabetes mellitus,^12^ corticosteroid use,^4^ HIV infection, and S. aureus nasal colonization.^13^ In a 21-year prospective study of 2348 patients, 54.2% with S. aureus bacteremia had implanted prosthetic material (most commonly a central venous catheter or cardiac), and the proportion increased from 40% in 1995 to 54.7% in 2015.^4^ In US surveillance data from 2005–2016, persons who inject drugs (PWID) were significantly more likely to develop invasive MRSA infections than those who did not inject drugs (472.2 vs 29.0 per 100,000 person-years in 2011, rate ratio 16.3, 95% CI 15.7–16.8).^14^
S. aureus is a Gram-positive bacterium, existing as a commensal in the human nares, skin, throat and gastrointestinal tract in about 30% of people.^15^ However, S. aureus can be a virulent pathogen if it breaches the skin or mucosal barriers and accesses normally sterile sites such as the bloodstream. After entering the bloodstream, S. aureus can attach to the surface of host tissues (e.g., native cardiac valves) or implanted devices (e.g., intravascular lines, cardiac devices, prosthetic joints). Attachment is mediated by Microbial Surface Components Recognizing Adhesive Matrix Molecules (MSCRAMMs), which are surface proteins that enable S. aureus to bind to many human proteins, including fibronectin, fibrinogen, collagen, von Willebrand Factor, and platelets.^16^ After attaching to a surface, aggregates of S. aureus cells can produce a biofilm matrix of polysaccharides, proteins and extracellular DNA^17^ that protects the bacteria from detection by the human immune system. Also, after attachment, S. aureus enter a low metabolic state resulting in reduced susceptibility to antibiotics that are active against replicating bacteria.
S. aureus bacteremia may also lead to abscess formation, facilitated by clotting factors, coagulase and von Willebrand factor binding protein, which promote fibrin clots and a pseudocapsule, protecting a central bacterial aggregate from phagocytic clearance.^18–20^ If abscesses rupture, release of S. aureus may potentially lead to formation of new abscesses.
S. aureus bacteremia can present with fever alone prompting diagnostic blood cultures. Conversely, patients may present with symptoms arising from a site of metastatic infection (e.g., back pain from vertebral osteomyelitis) or a source such as a skin and soft tissue infection. Approximately 73% of patients with S. aureus bacteremia present with fever,^21^ 42% have chills, and 18% have mental status changes.^22^ Common infectious foci are osteoarticular sites in ~ 14.4%, endovascular structures in 17.8%, and pulmonary infection in 5.9%.^4^ Mucocutaneous manifestations are present in ~18% of patients with S. aureus bacteremia,^23^ and in ~ 33% of patients with S. aureus endocarditis.^24^.
S. aureus is an uncommon cause of urinary tract infection, particularly in the absence of urinary tract catheterization or recent instrumentation; thus, the finding of S. aureus bacteriuria should prompt consideration of underlying S. aureus bacteremia, especially in hospitalized patients and/or those with systemic symptoms.^25^ In approximately 20% of patients, the source of S. aureus bacteremia is not identified.^26^
Clinicians should ask patients diagnosed with S. aureus bacteremia about presence of indwelling cardiac devices (such as a pacemaker, implantable cardioverter-defibrillator, or cardiac resynchronization therapy device), prosthetic devices (such as joint implants), central venous catheters, recent medical procedures and injuries, history of injection drug use, use of hemodialysis, diabetes, and previous S. aureus infections.
The physical exam is guided by the propensity for S. aureus to cause infections in a broad array of anatomic sites, including endovascular, osteoarticular, and deep tissue infections.^27,28^ Joints should be evaluated for tenderness, erythema and effusions, and the spine should be assessed for tenderness due to vertebral osteomyelitis. In a cohort of 97 patients with 166 arthroplasties in place at the time of an episode of S. aureus bacteremia, 38/39 (97.4%) prosthetic joint infections presented with joint pain.^27^ Pain is also the most common symptom of vertebral osteomyelitis; in a systematic review of vertebral osteomyelitis involving 14 studies, back pain was reported in 86% of patients.^29^ Endocarditis can be suggested by cardiac murmurs, signs of heart failure such as volume overload, and embolic and vasculitic manifestations in the fundi such as Roth spots (retinal hemorrhages), conjunctival petechiae, or splinter hemorrhages, Janeway lesions (nontender macules on the palms and soles), and Osler nodes (tender nodules most common on the pads of the fingers and toes). A neurologic exam may reveal evidence of focal deficits (such as weakness) caused by septic emboli.
S. aureus bacteremia is diagnosed with growth of S. aureus in blood culture. Conventionally, Gram staining that shows gram-positive cocci in clusters and biochemical testing identify the organism in a blood culture, and antibiotic susceptibility testing is then performed on the isolate. Increasing availability of rapid molecular diagnostic tests performed on positive blood culture specimens may allow species identification within several hours and provide direct detection of antimicrobial resistance determinants, such as the presence of the mecA gene that confers methicillin-resistance in S. aureus.^30,31^ In a randomized clinical trial comparing the use of a rapid molecular diagnostic vs conventional microbiology for 89 patients with blood cultures with gram-positive cocci, use of the molecular diagnostic reduced the time to reporting of methicillin-susceptibility (median 3.9 hours from Gram stain in intervention group vs 25.4 hours in the control group; P<0.001) and time to targeted therapy for S. aureus (5 hours vs 25.5 hours; P=0.004).^32^ In a network meta-analysis of 88 studies (11 exclusively focused on S. aureus) involving 25,682 patient encounters for bloodstream infections, use of a rapid diagnostic test combined with an antimicrobial stewardship program was associated with improved mortality (OR 0.72; 95% CI 0.59–0.87) and reduced time to optimal antibiotic therapy by 29 hours compared to blood cultures alone.^33^
Despite appropriate antibiotic therapy, approximately one-third of patients with S. aureus bacteremia have persistent bacteremia.^40^ In a prospective multicenter cohort study, the 90-day mortality of patients with 2–4 days of S. aureus bacteremia following initiation of antibiotics was almost twice that of patients with only 1 day of bacteremia (39% vs. 22%).^34^ Additionally, a new metastatic focus of infection was more likely in those with delayed clearance, occurring in 10% of patients with 2–4 days of bacteremia and 22% of those with 5–7 days, compared to 6% in patients who cleared their bacteremia in a single day.^34^ Therefore, repeat blood cultures should be performed for patients with S. aureus bacteremia at intervals of 24–48 hours until blood cultures are negative.^35^
The Infectious Diseases Society of America (IDSA) MRSA guidelines^35^ define uncomplicated S. aureus bacteremia as those infections for which endocarditis is excluded, there are no implanted prostheses, follow-up blood cultures performed on specimens obtained 2–4 days after the initial set do not grow S. aureus, defervescence has occurred within 72 hours of initiating effective therapy, and there is no evidence of metastatic sites of infection. All other infections are considered complicated S. aureus bacteremia. Across different cohorts, approximately 30% of patients with S. aureus bacteremia are classified as uncomplicated.^23,36^
Additionally, if S. aureus bacteremia is of community-onset, defined as positive index blood culture collected within 48 hours of hospital admission, there is a considerably increased risk for complicated disease.^4,37,38^ Presumably this is related to a likely longer duration of bacteremia in the community prior to commencing antibiotic treatment and thus an elevated risk of metastatic seeding. In contrast, hospitalized patients who develop a venous peripheral or central line related infection have been under close observation and have blood cultures promptly collected if they develop a fever, with empiric antibiotic therapy often quickly instituted. Thus, more recent commentary includes community-onset of S. aureus bacteremia as an important criterion for classifying patients as high risk for complicated disease.^39^
Once S. aureus bacteremia is identified, clinicians must determine both the source and potential sites of metastatic infection, including infective endocarditis. Approximately 12% of patients with S. aureus bacteremia develop endocarditis.^4^ Therefore, echocardiography should be routinely obtained for all patients with S. aureus bacteremia. Transesophageal echocardiography (TEE) is preferred but not mandatory in current IDSA guidelines.^35^ In clinical practice, transthoracic echocardiography (TTE) is usually obtained first. Whether patients with S. aureus bacteremia who do not have findings suggestive of endocarditis on TTE should undergo TEE is an area of ongoing controversy.^6^ TEE is more sensitive than TTE for detection of valvular abnormalities caused by S. aureus infective endocarditis^40,41^ and for detection of perivalvular complications.^42^ In a meta-analysis of 2807 patients with suspected infective endocarditis with TEE findings used as the reference standard, TTE had sensitivity of only 61% (95% CI, 45%–75%).^43^ However, the increased sensitivity of TEE must be balanced with its associated costs and risks, including major complications such as esophageal perforation in ~1 in 5000 patients.^44^
Several clinical predication rules have been developed to identify the need for TEE among patients with S. aureus bacteremia by quantifying the risk of endocarditis. The most accurate of these is the VIRSTA score (Table 2), which assigns points to underlying risk factors, presence of other foci of infection, severe sepsis or shock, elevated C-reactive protein, and persistent bacteremia 48 hours after the initial positive blood culture.^37^ A lower score indicates lower risk and a score <3 had a negative predictive value of 99.3% for a diagnosis of infective endocarditis in a validation study, although it classifies ~70% of patients as high risk, warranting TEE.^45^
Based on expert opinion, it is reasonable to forego TEE in patients with S. aureus bacteremia who have a VIRSTA score <3. In addition, TEE may not be required in patients without evidence of endocarditis based on clinical findings and TTE results, whose S. aureus bacteremia resolves quickly, and are being treated with prolonged antibiotic therapy for complications such as osteomyelitis, discitis, or epidural abscess.
The IDSA recommends magnetic resonance imaging (MRI) with gadolinium of the spine as the imaging modality of choice for patients with S. aureus bacteremia and back pain.^46^ Computed tomography of the chest, abdomen, and pelvis may be useful to identify unrecognized foci of infection such as abscesses or pulmonary septic emboli, particularly in patients who are not clinically improving with initial antibiotics. However, currently, there are insufficient data to recommend MRI or CT imaging as routine care for all patients.
PET/CT may be considered for the evaluation of metastatic sites of infection. A recent global survey of 2031 physicians (74% of whom were adult infectious disease specialists) found that there is wide variation by region in both PET/CT availability (range 9–78% of respondents) and use of PET/CT (13–94%) for evaluation of patients with S. aureus bacteremia worldwide.^47^
Treatment of S. aureus bacteremia requires appropriate antibiotic therapy and control of sources of infection. Clinical trials inform various aspects of S. aureus bacteremia management (Table 1, Supplementary Table 5).^7,48–65^
For patients clinically suspected to have S. aureus bacteremia (e.g., sepsis with clinically evident skin and soft tissue infection or those with a preliminary report of gram-positive cocci in blood culture), empiric antibiotic choice should be guided by local epidemiology and the individual characteristics of the patient being evaluated. Updated surveillance data on regional rates of methicillin-resistance are collated by groups such as the World Health Organisation sponsored Global Antimicrobial Resistance and Use Surveillance System (GLASS) and the Global Burden of Disease Antimicrobial Resistance Collaborators.^66,67^ Regions with very low rates (<5%) of MRSA may choose to initiate β-lactam antibiotics such as nafcillin/flucloxacillin or cefazolin. In areas with rates of MRSA >5%, such as the US, or for patients with injection drug use, recent hospitalization or surgery, the presence of prosthetic implants including central lines, long-term care facility residence, hemodialysis dependence, or prior MRSA infection, antibiotics with activity against MRSA should be initiated.
Once the S. aureus antibiotic susceptibility is determined, therapy should be tailored. For MSSA bacteremia, guidelines^68–70^ recommend using either cefazolin or an anti-staphylococcal penicillin (e.g., nafcillin, flucloxacillin), which are more rapidly bactericidal in vitro and associated with improved clinical outcomes (death and recurrent infection) compared to vancomycin for MSSA.^71–73^ Recent observational data suggest that cefazolin may be associated with lower mortality and fewer adverse effects than anti-staphylococcal penicillins for MSSA bacteremia. In a meta-analysis of 14 observational studies comparing cefazolin and anti-staphylococcal penicillins, cefazolin was associated with a lower 30 day mortality (relative risk 0.70, 95% CI 0.54 to 0.91) and less nephrotoxicity (relative risk 0.36, 95% CI 0.21 to 0.59) ^74,75^ Randomized clinical trials are currently directly comparing cefazolin to anti-staphylococcal penicillins for S. aureus bacteremia and pending results should soon inform practice.^76,77^
There are 3 antibiotics with an FDA-approved indication for the treatment of MRSA vancomycin, daptomycin and ceftobiprole. In an open label clinical trial, among 246 participants with S. aureus bacteremia, daptomycin was non-inferior to the standard of care at the time (low dose gentamicin plus either an antistaphylococcal penicillin or vancomycin) for MSSA (n=157) and MRSA (n=89) bacteremia. The primary outcome of this trial was a composite outcome of treatment success 42 days after therapy completion (53/120 [44%] vs 48/115 [42%]).^48^ In a double blinded clinical trial of 390 participants, ceftobiprole, a cephalosporin with activity against both MSSA and MRSA, was non-inferior to daptomycin for MSSA (n=293) and MRSA (n=94) bacteremia for the primary outcome of treatment success, defined as survival, bacteremia clearance, symptom improvement, no new S. aureus bacteremia–related complications, and no receipt of other potentially effective antibiotics, at day 70 (132/189 [70%] vs 136/198 [69%]).^7^ Ceftobiprole received FDA approval for S. aureus bacteremia on April 3, 2024.
Advantages of vancomycin are its availability, clinician familiarity with use, and low cost. In addition, in well-designed clinical trials, no antibiotic has been proven superior to vancomycin for treatment of S. aureus bacteremia. However, vancomycin has a narrow therapeutic window and requires drug monitoring to guide dosing and minimize the risk of kidney toxicity.^78^
Daptomycin is dosed once daily but is not always available in low and middle-income countries. Additionally, treatment-emergent daptomycin resistance in S. aureus has been reported, occurring in 7/120 (6%) patients in the daptomycin registrational trial and 3/198 (1.5%) patients in the ceftobiprole vs daptomycin trial.^7,48^
Eight randomized clinical trials assessing the addition of a second antibiotic to standard of care for S. aureus bacteremia have been published since 2016 (Supplementary Table 5). None demonstrated that combination antibiotic therapy improved clinical outcomes, including rifampin added to standard antibiotic therapy (1 trial, N=758),^54^ fosfomycin added to standard therapy (3 trials, N=397),^58,59,79^ daptomycin combined with a β-lactam for MSSA (1 trial, N=115),^60^ and β-lactams combined with vancomycin or daptomycin for MRSA (3 trials, N=452).^55–57^ The addition of fosfomycin to cloxacillin,^59^ fosfomycin to daptomycin,^58^ and β-lactams to vancomycin,^55,56^ reduced rates of persistent bacteremia, defined variously as positive blood cultures at day 3, day 5, and day 7 following trial entry, but did not improve mortality rates or treatment success, defined variously as composite endpoints incorporating mortality, microbiological relapse, and symptom resolution at different timepoints for each trial. Use of combination therapy for S. aureus bacteremia is also associated with adverse effects, such as increased kidney injury with the addition of low dose gentamicin.^80^ Adjunctive agents such as bacteriophage-derived lysins have not been proven effective when tested in sufficiently powered clinical studies.^61,62^ At this point, routine combination therapy cannot be recommended.
Approximately 30% of patients have S. aureus bacteremia for longer than 3 days despite appropriate antibiotics.^7,34,56^ Persistent S. aureus bacteremia is associated with increased mortality,^34^ and should prompt investigation for and control of sources of infection. For patients with persistent bacteremia, clinicians may also consider switching antibiotics or adding antibiotics, although there are no randomized clinical trial data to provide guidance in such situations. Options include adding agents such as ertapenem to cefazolin^81^ or fosfomycin to anti-staphylococcal β-lactams for MSSA^59^, and adding cefazolin,^56^ fosfomycin,^58^ ceftaroline^57^ or ceftobiprole^7^ to vancomycin or daptomycin for MRSA.
Low-risk, uncomplicated MSSA and MRSA bacteremia is typically treated with a 2 week-course of antibiotics. Patients with high-risk, complicated MSSA and MRSA bacteremia requires treatment for 4 to 6 weeks or longer.^35^ These recommendations, provided in the IDSA MRSA treatment guidelines, are largely based on observational data.^82^
Guidelines such as the 2011 IDSA MRSA treatment guidelines have recommended prolonged durations of intravenous antibiotic therapy for S. aureus bacteremia.^35^ However, the Partial Oral Treatment of Endocarditis (POET) trial published in 2018 randomized 400 patients with infective endocarditis (87 had MSSA) who were clinically stable (afebrile for >2 days, C-reactive protein had dropped to <25% peak value, white blood cell count < 15 x 10^9^/L, no sign of abscess formation on echocardiography performed within 48 hours of randomization, and had received at least 10 days of parenteral antibiotics) to use of a combination of 2 oral antibiotics vs continuation of intravenous antibiotics for the remainder of the treatment course.^64^ Participants received a median of 17 days of pre-randomization intravenous antibiotics. Among the 87 patients with MSSA endocarditis, the primary outcome of mortality, unplanned surgery, relapse or embolic events, occurred in 3/47 (6.4%) and 3/40 (7.5%) allocated to oral and intravenous therapy respectively. While the study was insufficiently powered to draw definitive conclusions about use of oral antibiotics for patients with S. aureus endocarditis, the point estimate of treatment effect for oral vs intravenous therapy was similar for those with S. aureus infections (odds ratio 0.84, 95% CI 0.15–4.78). Limitations of the POET trial included the absence of MRSA infections, the requirement for dual oral antibiotic therapy, and more frequent outpatient follow-up than is practical in routine clinical practice.^83^ The SABATO trial randomized 213 patients with low-risk S. aureus bacteremia to oral antibiotics after 5–7 days of intravenous antibiotics vs continuing intravenous antibiotics, with both groups completing 14 days of antimicrobial therapy.^72^ Patients were not enrolled in this trial if they had complicated bacteremia (deep seated focus of infection, septic shock, prolonged bacteremia [positive blood culture obtained >72 hours after start of appropriate antibiotic therapy], fever in the prior 2 days), or had an intravascular catheter that was not removed, a history of S. aureus bloodstream infection within the preceding 3 months, injection drug use, severe immunodeficiency or severe immunosuppression, or presence of a prosthetic heart valve or deep-seated vascular graft.^65^ Of the 213 participants, there were 16 MRSA and 197 MSSA infections. The rates of failure, defined as a composite of relapsing S aureus bacteremia, deep-seated infection with S. aureus, or death attributable to S aureus bacteremia, were similar in the oral antibiotic group (14/108 [13%] and intravenous antibiotic group 13/105 [12%]). Rates of drug-related serious adverse events were low in both groups (3/107 [2.7%] in the oral antibiotic group vs 0/103 [0%] in the intravenous group). The European Society of Cardiology 2023 Infective Endocarditis guidelines indicate that oral antibiotic treatment should be considered in patients satisfying the POET trial eligibility criteria.^69^ The WikiGuidelines for infective endocarditis support switching to oral antibiotic treatment for infective endocarditis, including that caused by S. aureus.^70^ Guidelines from the American Heart Association and IDSA have not been updated since the publication of the POET and SABATO trials. The Staphylococcus aureus Network Adaptive Platform trial provides details about potential oral antibiotic options and dosing recommendations within the protocol of an ongoing clinical trial for patients with S. aureus bacteremia.^84,85^
Source control is a critical component of S. aureus bacteremia treatment. Procedures may include incision and drainage of abscesses, debridement of infected tissue, and removal of implanted prosthetic material. Early source control improves outcomes; in a cohort of 884 US patients with S. aureus bacteremia, shorter time to source control procedure (median 1 day vs ≥3 days) was associated with earlier clearance of bacteremia and lower mortality, with each additional day of bacteremia associated with a relative risk of death of 1.16 (95% CI, 1.10–1.22, p<0.0001).^86^
Indwelling intravascular catheters should be promptly removed in patients with S. aureus bacteremia. In a study of 324 patients with catheter-associated S. aureus bacteremia retention of intravascular catheters was associated with increased risk of hematogenous complications such as septic arthritis or endocarditis (RR 2.28, 95% CA 1.22–4.27, p=0.011).^87^ In another study of 299 patients with central catheter-associated S. aureus bacteremia delayed intravascular catheter removal (>3 days) was associated with higher rate of S. aureus bacteremia relapse (12.7 vs 4.7%, p=0.017).^88^
Similarly, cardiac device removal is generally recommended for patients with S. aureus bacteremia.^89^ In a cohort of 5325 US patients with S. aureus bacteremia and an indwelling cardiac device, in-hospital mortality was lower among patients whose device was removed (5.6% vs 16.4%, aOR 0.31, 95% CI 0.21–0.44).^90^
For patients who have S. aureus bacteremia and a prosthetic joint, management should be individualized. The decision about whether to remove the prosthetic joint depends on many factors, including the timing of S. aureus bacteremia after joint implantation, whether infection occurred through hematogenous route or during the surgical procedure, surgical expertise, and patient comorbidities.
Based upon a systematic review and meta-analysis of 341 studies that included 536,791 patients, the estimated mortality of patients with S. aureus bacteremia was 10% at 7 days, 13% at 2 weeks, 18% at 1 month, 27% at 3 months, and 30% at 1 year.^5^ In a 2020 cohort of 31,002 patients in the US Veterans Health Administration hospitals, the 5-year mortality rate after S. aureus bacteremia was 61%.^91^ Key predictors of mortality are increasing age, comorbidities (such as heart failure, alcohol use disorder, malignancy, immune suppression, and/or hemodialysis dependence), and disease severity at presentation.^91,92^ In a pooled analysis of 3395 adult patients with S. aureus bacteremia, crude 90-day mortality was 29.2%. However, having an unidentified infective source was associated with higher mortality of 48.7% (adjusted hazard ratio for 90-day mortality 2.92; 95% confidence interval 2.33 to 3.67, p < 0.0001).^26^ Multiple study have reported MRSA bacteremia is associated with increased mortality compared to MSSA bacteremia,^91,92^ although this may be confounded by the older age and comorbidities of patients acquiring MRSA bacteremia.^93^ While increased duration of bacteremia of even a single additional day has been associated with increased mortality in observational studies,^34^ interventions such as combination antibiotic therapy in clinical trials that have reduced the proportion of participants with persistent bacteremia have not improved mortality.^56,58,59^
Infectious diseases consultation for patients with S. aureus bacteremia has been associated with improved patient outcomes in observational studies.^91,94^ In a study that included 31,002 patients with S. aureus bacteremia, 15,360 (49.5%) received infectious diseases consultation during their hospitalization. At 5 year follow-up, infectious diseases consultation was associated with improvement in the composite outcome of all-cause mortality or recurrence of S. aureus bacteremia (adjusted hazard ratio, 0.71; 95% CI, 0.68–0.74; P < .001).99 Importantly, the benefit of infectious diseases involvement is primarily observed with direct patient care at the bedside,^95^ and was not seen in a small randomized clinical trial of a telehealth consultation model.^96^
This review has limitations. First, there is limited high-quality evidence to guide treatment recommendations for S. aureus bacteremia. Second, the heterogeneity of S. aureus bacteremia means that recommendations are unable to cover all circumstances. Third, relevant articles may have been missed.
Several completed or actively recruiting randomized clinical trials involving patients with S. aureus bacteremia have not yet been published. Summarized in Supplementary Table 6, these trials involve antibiotic choice,^76,77,97,98^ duration,^99,100^ and route,^84,101^ as well as novel therapeutics and diagnostics.
S. aureus bacteremia has an incidence of 10–30 per 100,000 per year, case fatality rate of 15–30%, and causes 300,000 deaths per year worldwide. Empiric antibiotic treatment should include vancomycin or daptomycin which are active against MRSA, Once the S. aureus susceptibilities are known, MSSA should be treated with cefazolin or an anti-staphylococcal penicillin. Additional clinical management consists of identifying sites of metastatic infection and pursuing source control for identified foci of infection.