Clinical Significance of a Single Staphylococcus lugdunensis-Positive Blood Culture

Coagulase-negative staphylococci (CoNS) cause human infections that characteristically involve indwelling medical devices with an infection course that is usually indolent. In contrast, Staphylococcus lugdunensis is a species of CoNS that can cause infection of either a device or native tissue with an aggressive clinical course more reflective of infection due to Staphylococcus aureus (4, 5). The majority of infections caused by S. lugdunensis involve skin and soft tissue (8, 9). A panoply of other syndromes are caused by this organism and include endocarditis, septic shock, peritonitis, osteomyelitis, spondylodiscitis, septic arthritis, epidural abscess, brain abscess, and an array of prosthetic device (artificial joint, heart valve, ventriculoperitoneal shunt, and vascular device) infections (6, 7, 10, 11). The clinical significance of S. lugdunensis when only one of multiple blood culture sets is positive is largely undefined; this organism is a skin commensal that can cause, similarly to many other species of CoNS, contamination of blood cultures. To date, three published studies (1, 2, 13) have examined the clinical significance of S. lugdunensis in blood cultures. Ebright et al. performed a retrospective cohort study of consecutive patients with two or more blood cultures positive for S. lugdunensis at one of five Detroit Medical Center hospitals between January 1990 and May 2002 (2). A total of 21 patients were identified, and only 6 had clinically significant bacteremia. They did not evaluate, however, patients with only one set of blood cultures that grew S. lugdunensis. A retrospective analysis of patients with S. lugdunensis bacteremia at three tertiary care centers in Switzerland has also been reported (13). A total of 28 patients with S. lugdunensis bacteremia were identified, all of whom had at least two positive blood cultures and 13 of whom had endocarditis; 15 patients with one to five positive blood cultures did not have endocarditis. The clinical features of patients with only one positive blood culture were not described. In a recently published study in this journal, Choi et al. found that 63 of 1,117,598 patient admissions to seven medical centers had S. lugdunensis bacteremia, with an incidence of 5.6 patients per 100,000 admissions. Only 15 patients had clinically significant bacteremia (14 patients with 2 separate blood cultures and 1 patient with one positive blood culture and another positive culture from a “sterile” site). The clinical and laboratory characteristics and outcomes of the remaining 48 patients, however, were not described (1). None of the three studies (1, 2, 13) focused on patients with a single positive blood culture which is usually considered due to skin flora contamination, particularly in patients with no indwelling cardiovascular medical device. Does this assumption change if a CoNS isolate in one positive blood culture is identified as S. lugdunensis? In addition to its relevance to direct patient care, this issue is of interest to clinical microbiology laboratories that report blood culture contamination rates. We therefore examined outcomes of patients with a single S. lugdunensis-positive blood culture. Among 5,784 cases with a single positive CoNS blood culture at Mayo Clinic Rochester between 1 January 2001 and 24 January 2008, 29 had S. lugdunensis isolated. S. lugdunensis was detected by testing all CoNS isolates from blood cultures with ornithine decarboxylase and pyrrolidonyl arylamidase. The median patient age was 63 years (range, 22 to 91 years). Sixteen (55%) episodes occurred in males. In the initial review of medical records, positive blood culture results were not acknowledged by clinicians in 14 (48%) cases, were considered representative of contamination in 6 (21%) cases, and were considered clinically significant bacteremia in only 9 (31%) cases. The definition of clinically significant bacteremia, especially in the setting of bacteremia caused by CoNS, remains controversial and difficult to standardize. We used the criteria of Souvenir et al. (12) to retrospectively reclassify our cases. Patients were considered to have clinically significant bacteremia if they had one or more of the following: prolonged fever (temperature 38°C), hypotension (systolic blood pressure 90 mm Hg), leukocytosis or neutropenia with a left shifted differential, or disseminated intravascular coagulopathy. In addition, a major risk factor for potential infection caused by skin flora was required, including either long-term intravascular catheterization, peritoneal dialysis or hemodialysis, or extensive postsurgical infections with CoNS. Patients were considered to have indeterminate bacteremia if they had a major risk factor with minimal or transitory clinical symptoms. Blood cultures were considered contaminated if patients had an insignificant febrile episode with no significant risk factors; had significant risk factors but were shown by prior, concurrent, or subsequent blood cultures to have a septic episode with an unequivocal pathogen; or had infectious or noninfectious shock-like complications associated with an inconsistent CoNS etiology (e.g., aspiration pneumonia or acute respiratory distress syndrome, respectively). We reclassified 16 (55%) bacteremias as being caused by a possible contaminant or indeterminate bacteremia (Table 1). Bacteremia was considered clinically significant in 13 (45%) cases (Table 2), with 77% of cases associated with fever, 31% with neutropenia, and 31% with leukocytosis; these proportions were 25%, 0%, and 62%, respectively, in the contaminant or indeterminate bacteremia group. A possible source was identified in 77% of patients with clinically significant bacteremia (central venous catheter [38%], lower extremity cellulitis/ ulcer [23%], abdominal surgery [8%], and infected right femoral graft [8%]). Malignancy was identified in 54% of patients with clinically significant bacteremia and in 24% of the other group. Twenty-seven (93%) patients received antibiotics for a median duration of 14 days (range, 3 to 56 days) with a -lactam and/or vancomycin in 9 (33%) and a fluoroquinolone in 18 (67%) cases. Details regarding antibiotic administration are shown in Tables 1 and 2. Patients considered to have clinically significant bacteremia received a longer duration of treatment, with antibiotics administered for a minimum of 7, a mean of 33, a median of 15, and a range of 7 to 50 days compared to a mean of 10, a median of 10, and a range of 0 to 28 days in the contaminant or indeterminate group. In the latter group, 7 (44%) of 16 patients received five or fewer days of treatment and 2 were not treated. Echocardiography was performed in four patients (three in the clinically significant bacteremia group and one in the contami-