Cellulitis is a common soft tissue infection that extends into the subcutaneous tissues. Systemic antibiotic therapy is routinely used to treat cellulitis. However, identifying a specific pathogen is often not possible on clinical grounds. Therefore, antibiotic therapy with activity against group A Streptococcus and Staphylococcus aureus is usually selected empirically. Needle aspiration of fluid from the edge of the infection is sometimes performed but yields a pathogen in less than 20 percent of patients. Blood cultures are a more common means of identifying a bacterial pathogen, but the yield seems to be rather low and the results only marginally affect treatment. Perl and colleagues performed a retrospective review to examine the usefulness and cost-effectiveness of blood cultures in patients with cellulitis.

The researchers reviewed the records of all patients with a diagnosis of cellulitis who were admitted to the hospital through the emergency department. Patients with facial cellulitis and patients who had other focal infections besides infections of the skin were excluded from the study. The records of the hospital microbiology laboratory were also reviewed to identify the number of blood cultures performed on these patients and the number of positive cultures that yielded a specific pathogen. After patients with positive blood cultures were identified, a more intensive chart review was performed to identify the clinical features of this subset of patients.

During a three-year period, 757 adult patients were diagnosed with cellulitis. Blood cultures were obtained in 553 patients (73 percent), and 710 blood samples were drawn. In some patients, more than one blood culture was drawn. The blood samples were inoculated into aerobic and anaerobic culture bottles. A specific bacterial pathogen was identified in only 11 patients (2 percent). In addition, an organism that was considered a contaminant was isolated from the blood cultures in 20 patients (3.6 percent). For comparison, the hospital laboratory processed about 14,000 blood cultures from other patients during this same period; of these cultures, approximately 11 percent produced a true pathogen.

Among the patients with positive blood cultures, group G Streptococcus was identified in five, group A Streptococcus in three, S. aureus in one, Vibrio vulnificus in one and Morganella morganii in one. Several key features that indicated bacteremia were noted among these patients when compared with patients whose blood cultures were negative. These features included an age greater than 45 years, a shorter duration of symptoms before physical examination, a higher incidence of fever, a temperature of 38.5°C (101.3°F) or higher at admission and a white blood cell count greater than 13,300 per mm³ (13.3 × 10⁹ per L) at admission.

At this institution, the cost of blood cultures was $50 each plus $50 for antimicrobial susceptibility testing. The total costs associated with negative blood cultures and those with contaminants in this study were $34,950.

The authors conclude that, in the majority of patients with cellulitis, the yield of blood cultures is very low and not cost effective. Elderly patients with acute onset of illness, high fever and a significant elevation in white blood cell count, and patients who are immunocompromised may be exceptions.

editor's note: The results of this study are not surprising considering that most cases of cellulitis are localized to soft tissue and most patients do not become bacteremic. Of interest, all of the patients in this study were admitted to the hospital although the severity of illness varied. Standard practice would normally dictate obtaining blood cultures if a patient with cellulitis is sick enough to be hospitalized for administration of parenteral antibiotics. By applying the criteria noted in this study, especially the age of the patient and the presence of fever, blood cultures could be ordered more selectively. Extrapolating this data to patients who are managed on an out-patient basis for cellulitis seems to indicate that the use of blood cultures is a waste of time and money.—j.t.k.