Photo Credit: Dr Microbe

Study findings question the use of risk factors alone to define complicated Staphylococcus aureus bacteremia, urging physicians to weigh the whole case.

“Staphylococcus aureus bacteremia (SAB) is a common infection with a wide range of clinical manifestations, from catheter-related bloodstream infections to fulminant endocarditis or sepsis with multiple metastatic foci,” Thomas W. van der Vaart, MD, and Jan T.M. van der Meer, PhD, note. “Clinicians often distinguish between ‘complicated’ and ‘uncomplicated’ SAB. The former requires a larger dose and a longer course of antimicrobial therapy. However, there is no universally accepted definition for either.”

In 2003, Duke University published what Dr. van der Vaart and Dr. van der Meer describe as “a landmark study” that defined SAB as complicated or uncomplicated based on clinical outcomes; the study then identified risk factors for this outcome.

“Since the publication of that study, these risk factors—including community acquisition and persistent bacteremia—were increasingly used as a substitute for complicated disease itself,” they explain. “This practice was formalized in 2011 in the IDSA guidelines on the treatment of MRSA bacteremia. According to the guidelines, you can have complicated SAB based on solely having a risk factor for it, without metastatic or locally complicated disease being present. This makes risk factors for complicated SAB extremely important for clinical decision-making. However, the association between these risk factors and true complicated SAB is poorly studied.”

Study Design and Aims

The researchers conducted a multicenter prospective cohort study at seven hospitals. They classified patients with SAB as complicated or uncomplicated through adjudication (reference definition) and determined the associations and predictive values of risk factors compared with the reference definition. They also determined the accuracy of IDSA criteria (four risk factors) and the expected consequences of utilizing IDSA criteria for antibiotic use.

The criteria from the Duke study, known as the Fowler risk factors, include:

• Community acquisition;
• Persistent bacteremia;
• Skin findings indicative of systemic infection; and
• Persistent fever.

IDSA risk factors include:

• Persistent bacteremia;
• Skin findings indicative of systemic infection;
• Persistent fever; and
• Permanently implanted prosthetic material.

“We determined if patients had complicated SAB based on the clinical outcome,” Dr. van der Vaart and Dr. van der Meer explain. “A patient with infection-related mortality, relapse infection, or metastatic or locally complicated SAB was considered to have a complicated infection, regardless of the presence of risk factors. We then looked at the predictive values of those risk factors that are commonly cited in guidelines and scientific literature to identify complicated SAB.”

The authors reported their findings in Clinical Infectious Diseases.

Most Factors Weakly Associated With Complicated SAB

The study included 490 patients (median age, 68; 66.7% men). Common comorbidities included diabetes (31.8%) and chronic renal failure (27.6%), and 30.6% had a heart condition that predisposed them to endocarditis.

More than half of the patients (60.0%) had complicated SAB. In univariable analyses, only community acquisition, skin manifestations suggesting systemic infection, persistent bacteremia, and permanently implanted prosthetic material were associated with complicated SAB. In a multivariable analysis with these four risk factors, ORs were 6.8 (95% CI, 3.9-12.0) for persistent bacteremia, 2.9 (95% CI, 1.9-4.7) for community acquisition, and 2.3 (95% CI, 1.5-3.6) prosthetic material.

Persistent bacteremia had the greatest positive predictive value for complicated SAB (88%), followed by community acquisition (78%) and skin findings indicating systemic infection (77%).

“We found that the positive predictive value of many risk factors for complicated SAB was not strong enough to justify classifying a patient as having complicated SAB,” Dr. van der Vaart and Dr. van der Meer note. “The strongest association was between persistent bacteremia after 48 hours of adequate antimicrobial therapy. For other risk factors, however, this association was weaker. For example, if a patient with SAB acquired the infection in the community, the chance of them having complicated SAB was 78%. This, of course, also means that 22% of patients with community-acquired SAB would not.”

IDSA Criteria Falsely Identifies Complicated SAB

For the IDSA definition, the researchers found that “20% of all patients in our cohort would incorrectly be classified as complicated SAB because they had a risk factor present in the IDSA definition while, in fact, they had no evidence of complicated infection itself and were thus classified by us as having uncomplicated SAB,” Dr. van der Vaart and Dr. van der Meer continue.

The presence of any of the four risk factors in the IDSA definition had a positive predictive value of 70.9% and a negative predictive value of 57.5%. IDSA criteria yielded 24 false-negative (5%) and 90 false-positive (18%) categorizations of complicated SAB compared with the reference.

“Approximately four in 10 patients without any risk factors listed by the IDSA will turn out to have complicated SAB, so the absence of risk factors should not be taken as reassurance,” Dr. van der Vaart and Dr. van der Meer note (Figure). “Further, even if a patient has one of the risk factors, there is a one in three chance they will not have complicated SAB.”

Looking Ahead: Antibiotic Use and Future Research

Ultimately, the findings “call into question the use of risk factors alone to define complicated SAB,” the researchers continue.

“Clinicians should weigh the whole case to decide if a patient with SAB has a complicated disease and requires extended treatment. Risk factors can be a part of this consideration, but they should not be used to define complicated SAB alone.”

Further, the incorrect diagnosis of complicated SAB has implications for antibiotic therapy.

“Since guidelines recommend treating complicated SAB for 4-6 weeks versus 2 weeks for uncomplicated SAB, having a false positive classification of complicated SAB could lead to unnecessarily longer treatment.”

Findings related to the median and interquartile ranges for total duration of antimicrobial treatment, stratified for complicated and uncomplicated SAB, support the idea that “the patients we classified as having uncomplicated SAB, despite meeting the IDSA definition for it, were, in fact, uncomplicated, demonstrated by the overall short duration of antibiotic treatment in this group,” Dr. van der Vaart and Dr. van der Meer note.

“Future research should first confirm these findings in other settings. Our healthcare setting had low MRSA prevalence compared with other regions. With better data on the association between complicated SAB and risk factors, including biomarkers, a new definition of complicated SAB could, ideally, be constructed.”