The Nose Knows: Nosocomial S. aureus Bacteremia

abstracts & commentary

Synopsis: Molecular techniques demonstrated the apparent identity between nasal isolates of S. aureus and strains of this organism subsequently causing
bacteremia.

Sources: von Eiff C, et al. Nasal carriage as a source of Staphylococcus aureus bacteremia. N Engl J Med 2001;344: 11-16; Archer GL, Climo MW. Staphylococcus aureus bacteremia—consider the source. N Engl J Med 2001;344:55-56.

Staphylococcus aureus (sa) bacteremia causes substantial morbidity and mortality. No vaccine exists for prevention of this invasive disease. SA colonizes up to 80% of humans at some time during their life. That such an intimate commensal should pack such a pathogenic punch once it invades tissues and blood remains one of the great conundrums of human medicine. The havoc that SA produces in tissue has been a focus of research for more than two decades, along with the process by which the organism escapes from the nose to the blood.

The current study by von Eiff and his German infectious diseases colleagues asks a simple question: are the strains of SA that colonize the nose of patients admitted to the hospital the same strain that becomes bloodborne?

von Eiff et al, thus, undertook two studies to provide data to answer this question. The first study from 1993-1994 featured hospitalized patients in 32 German hospitals. Study patients had a nasal swab done as soon as possible after isolation of SA from blood. They collected 219 SA isolates from blood and 350 isolates from the anterior nares. Only 20 (9.1%) of the 219 bacteremic isolates were methicillin-resistant SA (MRSA), a rate certainly below what we would currently expect for U.S. hospitals now.

Isolates were typed by pulse field gel electrophoresis (PFGE). Most blood isolates (82%) had the same PFGE pattern as the nasal isolate. The association between nasal and blood isolates held true regardless of the time lag between the time the nasal swab was collected and the development of the bacteremia.

These provocative results led to the second study from 1994 through 1999 at a large 1568 patient bed tertiary-care hospital in Münster, Germany. The design of the second study was to collect nasal swabs from patients and compare the pulsed-field gel electrophoresis (PFGE) pattern of isolates subsequently obtained from the blood.

Nasal swabs were obtained from 1278 general medical patients who contributed 1640 SA isolates; only 74 (5.8%) of the nasal isolates were MRSA. Bacteremia developed in 14 study patients, only one of whom had an MRSA strain. The interval between nasal isolation and blood isolation was 6.5 days with a range of 1 day to 60 weeks! Regardless of the interval, the blood isolate PFGE matched the nasal isolate in 85.7% (95% confidence interval, 57.1-98.2%). Only two patients had a nasal strain that differed by PFGE from the blood isolates.

Bacteremia of 258 nonstudy patients did not differ from that of the study patients. The presumed origin of bacteremia in 156 patients studied with bacteremia included catheter-associated infection (46%), skin, soft tissue (27%), pneumonia (11%), urinary tract infection (10%), endocarditis (5%), and other sites (1%).

When all data are analyzed, the most cautionary conclusion is that at least 50% of patients with SA bacteremia are first colonized by an identical SA strain.

Comment by Joseph F. John, MD

These studies provide "proof of concept," for the legacy of nasal colonization as a risk for subsequent SA infection. We know that the strains inhabiting the nares will more often than not be the same ones that produce a bacteremia. Because this study uses only a macro restriction of SA chromosomal DNA, we know little about addition, deletion, or mutations that might be related to invasion and have to assume from the data of the study that nasal strains become bloodborne with their genomes relatively unaltered.

The potential clinical implication is that if certain strains can be blocked from colonizing the nose, patients may be much more unlikely to experience bacteremia with that strain. A correlate of this observation is that episodes of SA bacteremia that are classified as "nosocomial," that they occur 48 hours after admission are, in fact, quasi nosocomial since the organism that infected the patient may likely be one that habitually colonizes the patient outside the hospital. Here we see the meshing of community factors (i.e., the selection of strains of SA that preferentially colonize the nose) with nosocomial factors, as well as iatrogenic factors like indwelling IV catheters, bladder catheters, and antibiotic exposure, with both sets of factors combining to place the patient at risk for SA bacteremia.

In fact, one major finding of the study is that strains that patients carry into the hospital may persist (in their nares) for months before giving rise to invasive disease. von Eiff et al conclude that we should develop strategies to reduce systemic disease by interrupting the transmission of SA from the nose to those sites where it gains access to circulation. Furthermore, the concept of interrupting transmission as an emerging infection control strategy should be applied not only to MRSA but also to methicillin-susceptible strains as well.

The comprehensive editorial by Archer and Climo raises several issues with current strategies to limit colonization, including the development of mupirocin resistance where topical use of this agent has been widespread. The most radical inference of Archer and Climo is that we should consider screening for nasal colonization with SA in all patients admitted to the hospital. We have already advocated this strategy in patients admitted to our MICU, thus protecting them theoretically from subsequent invasive disease and protecting uncolonized patients from becoming colonized with new strains of SA. Patients are kept in isolation until their nares are shown to be free of SA, leaving the choice of decolonizing regimens to the attending physicians. The editorial emphasizes the need for a "rapid diagnostic test," though we have found that the microbiology laboratory identifies presumptive SA in overnight nasal cultures places to blood agar. I agree with the conclusion of the editorial that, "Any real reduction in the incidence of nosocomial infections with S. aureus will require widespread aggressive tactics and innovative strategies." The methods to support these strategies, I feel, are currently available. What it takes now is a willful commitment from infection control and hospital administration to implement the methodology.