Epidemiology of Staphylococcus aureus bacteremia in adults

INTRODUCTION — Staphylococcus aureus is a leading cause of community-acquired and health care-associated bacteremia. The annual incidence of S. aureus bacteremia (SAB) in the United States is 38.2 to 45.7 per 100,000 person-years [1,2]; elsewhere in the industrialized world, the incidence is approximately 10 to 30 per 100,000 person-years [3]. Rates are higher among specific populations (such as patients on hemodialysis) (table 1). The 30-day all-cause mortality of SAB is 20 percent [4-6].

The epidemiology of and risk factors for SAB in adults will be reviewed here. The management of SAB is discussed separately. (See "Clinical approach to Staphylococcus aureus bacteremia in adults".)

EPIDEMIOLOGY — Bacteremia due to S. aureus can be classified into three categories [7]:

●Health care associated, hospital onset (ie, nosocomial)

●Health care associated, community onset

●Community acquired

In a prospective cohort study including 247 cases of SAB in Minnesota between 1998 and 2005, 23 percent were nosocomial, 59 percent were health care associated, community onset, and 18 percent were community acquired [1]. Methicillin-resistant S. aureus (MRSA) accounted for 32 percent of all cases. Outside the United States, the proportion of SAB cases due to MRSA is generally lower (table 1).

Rates of SAB over time have remained largely stable; however, the proportion due to MRSA has decreased, likely due to improved infection control practices [3]. Between 2005 and 2012, hospital-onset MRSA bacteremia decreased by 17 percent per year but did not significantly change further between 2013 and 2016. Concurrently, community-onset methicillin-susceptible SAB increased by 3.9 percent annually between 2012 and 2017 [8].

Health care associated

Hospital onset — Health care-associated, hospital-onset infection refers to nosocomially acquired infection [9].

S. aureus is a leading cause of nosocomial bloodstream infection in the United States. Among 24,000 nosocomial bloodstream infections between 1995 and 2002, S. aureus was the second most common cause (after coagulase-negative staphylococci), accounting for 20 percent of cases [10]. This trend corresponds with the increasing use of intravascular catheters and subsequent catheter-associated staphylococcal bacteremia [11]. In a prospective case-control study of hospitalized patients, the presence of a central venous catheter was the single greatest risk factor for the development of SAB [12].

Among 847 cases of SAB in a multicenter cohort, most patients had predisposing conditions including diabetes (33 percent), malignancy (26 percent), chronic kidney disease (22 percent), and immunosuppressive therapy (21 percent) [13].

Among patients who acquire health care-associated, hospital-onset SAB, approximately 20 percent develop metastatic complications, including endocarditis. The mortality rate is 20 to 30 percent [5,14].

Hospital-acquired S. aureus infections are more likely to be methicillin resistant than community-acquired infections [9,15]. Among 24,000 nosocomial bloodstream infections in the United States between 1995 and 2002, the proportion of S. aureus isolates that were due to MRSA increased from 22 to 57 percent [10]. (See 'Methicillin' below.)

Community onset — Health care-associated, community-onset infection refers to infection in an outpatient who has had recent extensive contact with the health care system. The infection must be diagnosed as an outpatient or within 48 hours of hospital admission. Examples of health care contact include [7]:

●Hospitalization in an acute care hospital for ≥2 days within the prior 90 days

●Receipt of dialysis or intravenous therapy (including chemotherapy) within the prior 30 days

●Receipt of intravenous therapy, wound care, or specialized nursing care at home

●Residence in a nursing home or other long-term care facility

In a multicenter cohort of community hospitals in the southeastern United States, S. aureus was the most common cause of bloodstream infection. Among 1470 patients with bacteremia, the majority (56 percent) were health care-associated, community-onset infections, and S. aureus was identified in 28 percent of cases [16]. The presence of an intravascular catheter is an important risk factor for these infections [7,16-18].

SAB acquired in long-term care facilities typically affects older adults and chronically ill patients. Skin or soft tissue lesions such as decubitus ulcers, diabetic foot ulcers, and wounds are common risk factors for bacteremia among these individuals. Careful bedside evaluation must be performed since important clues suggesting the presence of metastatic infection (eg, back or joint pain) are often absent in older adult patients. (See "Clinical manifestations of Staphylococcus aureus infection in adults".)

Community acquired — Community-acquired infection refers to infection in a patient who has had no recent contact with the health care system [19,20]. Patients with community-acquired SAB include injection drug users and patients with a clinically inapparent source of bacteremia (such as vertebral osteomyelitis or epidural abscess).

Patients with onset of SAB acquired in the community are likely to present with complicated infection. In one study, more than 40 percent of patients with community-acquired SAB had metastatic infection, including infective endocarditis (IE) [14]; another study noted as many as 90 percent of patients with community-acquired SAB had one or more complications [21]. In a study including more than 500 patients with SAB, IE was present in 21 percent of patients with community-acquired infection, a rate that was almost three times that seen among patients with hospital-acquired SAB [22].

RISK FACTORS — Risk factors for development of SAB include presence of a prosthetic device, injection drug use, and underlying host factors.

Indwelling prosthetic device — Any implanted foreign body that becomes infected is a potential source for SAB. Implanted devices include vascular catheters, surgically implanted materials, and orthopedic prostheses. In a longitudinal study between 1995 and 2015 including more than 2300 patients with SAB, more than half of patients had at least one indwelling prosthetic device [23].

Intravascular catheters are the most common cause of SAB in hospitalized patients [12] as well as an increasingly important cause of community-acquired infection [7,17,18]. These devices serve as a direct conduit into the intravascular space, allowing easy access for S. aureus to enter the bloodstream. In a pooled analysis of five observational studies that included more than 3000 patients with SAB, intravenous catheters were identified as the probable focus in 28 percent of cases [24]. (See "Intravascular catheter-related infection: Epidemiology, pathogenesis, and microbiology".)

Injection drug use — Injection drug users (IDUs) who develop SAB are usually young and otherwise healthy [25-27]. IDUs have a high rate of nasal colonization with S. aureus, and the phage type of the colonizing organism typically matches that of the isolate recovered from the blood [28,29]. The combination of high S. aureus nasal colonization rate and repeated parenteral injection of nonsterile material makes IDUs prone to SAB [30,31]. (See 'Nasal colonization' below.)

Host factors — SAB incidence varies by age, gender, and ethnicity. Rates are highest at extremes of age [32]. Male-to-female ratios of approximately 1.5 have been reported consistently. In Australia and New Zealand, indigenous populations have markedly higher SAB incidence, beyond that which might be explained by differences in socioeconomic factors [3].

Diabetes mellitus is an independent risk factor for community-acquired SAB, with greatest risk in patients with type 1 diabetes, ≥10 years of diabetes, history of poor glycemic control, and individuals with other complications of diabetes [33]. (See "Susceptibility to infections in persons with diabetes mellitus".)

In a longitudinal study between 1995 and 2015 including more than 2300 patients with SAB, the most common comorbid conditions were diabetes, dialysis dependence, cancer, and corticosteroid use [23]. Of these, diabetes, cancer, and steroid use increased over time; the proportion of SAB patients with dialysis dependence decreased, possibly due to advances in vascular access methods or improvement in infection control practices [23].

Surveillance data among patients receiving hemodialysis in the United States between 2017 and 2020 demonstrated that SAB incidence was 100 times higher in adults on hemodialysis than in adults not on hemodialysis; using a central venous catheter for dialysis access was the strongest risk factor for SAB. Race, ethnicity, and social determinants of health also impacted infection rates, with Hispanic ethnicity independently associated with higher risk [34].

It is uncertain whether statin use may be protective against SAB. In a case-control study including 2638 patients with community-acquired SAB and 26,379 matched population controls, statin use was associated with a decreased risk of SAB (adjusted odds ratio 0.73, 95% CI 0.63-0.84) [35].

Host genetics may contribute to human susceptibility to S. aureus infection [36]. In addition, hereditary defects in white blood cell function or immune response (such as Job's syndrome, Chediak-Higashi syndrome, and Wiskott-Aldrich syndrome) predispose patients to recurrent staphylococcal infections [37]. (See related topics.)

Nasal colonization — Patients with S. aureus nasal colonization are at increased risk for SAB [38-40]. In one study including 3420 patients with S. aureus nasal carriage on hospital admission, methicillin-susceptible SAB was observed more frequently among carriers than noncarriers (1.2 versus 0.4 percent) [41].

The association between nasal carriage and outcome of bacteremia is uncertain. In one study including 3420 patients with S. aureus nasal carriage, the likelihood of bacteremia-related death was lower among carriers than noncarriers (8 versus 32 percent) [41]. However, another study including more than 3000 MRSA-colonized patients noted no difference in all-cause mortality between colonized and non-colonized patients who developed MRSA bacteremia [42].

Lower mortality among nasal carriers may be explained by differences in the immune response between S. aureus carriers and noncarriers; among healthy S. aureus carriers, antibody response against the colonizing strain has been observed [43].

Patients with conditions conferring increased risk for nasal colonization (such as patients with diabetes and patients who are hemodialysis dependent) are at increased risk for SAB [38-40].

HIV infection — HIV infection is associated with an increased risk of SAB (table 1). Injection drug use accounts for some of this risk; however, even in the absence of IDU, rates of SAB are higher among HIV-infected patients than among HIV-uninfected patients [44]. A CD4 count less than 100 appears to be the strongest independent risk factor among HIV-infected patients [44].

REPEAT INFECTION — Repeat infection with SAB is common [45,46]. In one cohort including more than 10,000 patients, approximately 7 percent of patients with SAB developed recurrent SAB; the risk of reinfection was defined as an episode of SAB >90 days after an initial episode of SAB [46]. The median time to recurrent infection was 15 months. Risk of recurrent SAB was associated with comorbidities including renal disease, diabetes with associated complications, severe liver disease, and paraplegia.

DRUG RESISTANCE

Methicillin — Methicillin resistance among bloodstream S. aureus isolates has fluctuated; overall, the rate of methicillin resistance increased until approximately 2005. This was largely due to the spread of epidemic community-associated clones such as USA300 in North America and health care-associated clones in the United Kingdom. As an example, among 24,000 nosocomial bloodstream infections in the United States between 1995 and 2002, the proportion of S. aureus isolates that were methicillin resistant increased from 22 to 57 percent [10].

Subsequent to 2005, however, reductions in rates of methicillin-resistant S. aureus (MRSA) bacteremia have been observed. In nine metropolitan areas in the United States, hospital-onset MRSA bacteremia decreased by 11 percent per year between 2005 and 2008, while health care-associated, community-onset MRSA bacteremia declined by 7 percent annually [47]. Similar trends were evident in the United Kingdom, France, and Australia [3,48].

The incidence of MRSA catheter-associated bloodstream infections in United States intensive care units also appears to be decreasing [49]. This was illustrated in a review of over 33,500 episodes of central line–associated bloodstream infections between 2001 and 2007; declines in MRSA central line–associated bloodstream infection incidence ranged from 51 to 69 percent. These findings call into question the utility of performing active MRSA surveillance rather than focusing on reducing health care-acquired infections in general. (See "Methicillin-resistant Staphylococcus aureus (MRSA) in adults: Epidemiology" and "Methicillin-resistant Staphylococcus aureus (MRSA) in adults: Prevention and control".)

With respect to health care-associated, community-onset infection, a retrospective study in 1998 found that 15 percent of community-onset SAB were due to MRSA [18]. The majority of patients with MRSA bacteremia had had contact with the health care system, including 47 percent with an indwelling catheter. Similar findings were noted in a prospective cohort study that included 186 patients with health care-associated, community-onset bloodstream infections in 2000; 52 percent were due to MRSA, and 42 percent were receiving intravenous or intravascular therapy at home or in a clinic [7]. Subsequent studies have shown that community-associated MRSA infections, including bacteremia, have become increasingly common in children and in individuals with no prior health contact [50].

In a review of data from more than 200 countries, estimates for MRSA burden in 2019 were generated using predictive statistical modeling. MRSA prevalence was highest in north Africa and the Middle East (60 to 80 percent); it was lowest in parts of Western Europe and Sub-Saharan Africa (<5 percent). Overall, bloodstream infections were the second most common syndrome causing death (after lower respiratory tract infections) and S. aureus was the second leading pathogen (after Escherichia coli) [51].

Vancomycin — Issues related to epidemiology of vancomycin resistance are discussed separately. (See "Staphylococcus aureus bacteremia with reduced susceptibility to vancomycin", section on 'Epidemiology'.)

Daptomycin — De novo daptomycin resistance is uncommon [52]. Emergence of resistance to daptomycin in patients who receive prolonged courses of therapy has been observed, particularly if their underlying infection is not amenable to surgical drainage or if it involves a prosthetic device such as a vascular graft. Among 124 patients with SAB treated with daptomycin in one trial, emergence of isolates with resistance to daptomycin was observed in 5 percent of patients [53].

Other risk factors for the emergence of drug resistance include previous exposure to vancomycin, infection with a methicillin-resistant S. aureus strain with a minimum inhibitory concentration to vancomycin of 2 mcg/mL or greater, and treatment with insufficient doses of vancomycin [53-55].

SUMMARY

●Bacteremia due to Staphylococcus aureus can be classified into three categories: health care associated, hospital onset; health care associated, community onset; and community acquired. The incidence of S. aureus bacteremia (SAB) for each of these categories has increased over the last several decades. (See 'Epidemiology' above.)

●SAB is a leading cause of bloodstream infections in the United States. This trend is due largely to the increasing use of intravascular catheters. During the past decade, the proportion of SAB cases due to MRSA appears to be declining, possibly due to improved infection control practices. (See 'Hospital onset' above.)

●Intravascular catheters are also playing an increasing role in health care-associated, community-onset SAB. SAB acquired in long-term care facilities typically affects older adults and chronically ill patients. Skin or soft tissue lesions such as decubitus ulcers, diabetic foot ulcers, or wounds are common causes of SAB in individuals who reside in these facilities. (See 'Community onset' above.)

●The presence of community-acquired SAB, particularly in the absence of an identifiable focus of infection, is an important marker for complicated staphylococcal infection. (See 'Community acquired' above.)

●Health care contact (eg, prior hospitalization) is an important risk factor for methicillin-resistant S. aureus infection, but there are numerous reports of patients with no identifiable health care contacts or other risk factors. (See 'Methicillin' above.)

●An implanted foreign body that becomes infected is a potential source for SAB. Implanted devices include vascular catheters, surgically implanted materials, and orthopedic prostheses. (See 'Indwelling prosthetic device' above.)

●Injection drug users (IDUs) who develop SAB are usually young and otherwise healthy. The combination of high skin colonization rate with S. aureus and repeated parenteral injection of nonsterile material makes IDUs prone to SAB. (See 'Injection drug use' above.)

●Inherited defects in white blood cell function or immune function predispose patients to recurrent staphylococcal infections. Patients with nasal colonization are more prone to S. aureus infection, including SAB. (See 'Host factors' above and 'Nasal colonization' above.)