Miller and colleagues conducted their study in adults as well as children. Patients were included if they had either a discrete skin abscess or cellulitis, or both. They were excluded if they had one of the following:
- perirectal, genital, or hand infection
- a human or animal bite at the site of infection
- temperature of 38.5° C or higher
- morbid obesity
- prosthetic device at the site of infection.
The remaining patients then were stratified into one of 2 groups:
- those with a larger abscess (>5 cm in maximum diameter in adults, proportionally smaller in children) and/or cellulitis
- those who had a smaller abscess.
This study by Miller and colleagues focuses only on the patients in the former group.
Details of the trial
All discrete abscesses were incised and drained, and patients were randomly assigned to either:
- clindamycin, 300 mg 3 times daily for 10 days
- trimethoprim-sulfamethoxazole, 2 single-strength tablets orally twice daily for 10 days.
The primary endpoint was clinical cure at 7 to 10 days after completion of antibiotic therapy.
The study enrolled 524 patients—264 in the clindamycin group and 260 in the trimethoprim-sulfamethoxazole group. Approximately 30% of the patients were children. One hundred sixty patients (30.5%) had a discrete abscess, 280 (53.4%) had cellulitis, and 82 (15.6%) had both an abscess and cellulitis. An incision and drainage procedure was performed in 44.5% of patients. Slightly more than 50% of patients had a microbiological culture.
The most common organism isolated was S aureus (217 of 524 patients, or 41.4%), of which 167 (77%) were methicillin-resistant S aureus (MRSA). Of the 217 isolates identified as S aureus, 27 (12.4%) were resistant to clindamycin, and only one (0.5%) was resistant to trimethoprim-sulfamethoxazole.
Of the 466 patients who were fully evaluable, the rate of cure was 89.5% in the clindamycin group (95% confidence interval [CI], 85.2–93.7) and 88.2% in the trimethoprim-sulfamethoxazole group (95% CI, 83.7–92.7). The difference in the observed rate of clinical cure was not statistically significant.
Eleven of 15 patients in the clindamycin group who had clindamycin-resistant isolates were cured, compared with 77 of 84 patients with susceptible isolates (73.3% vs 91.7%; P = .06). At 1 month after treatment, cure rates remained similar. The overall rates of adverse effects in the 2 groups were similar, at 19%. No patient developed Clostridium difficile-associated diarrhea.
Skin infections can be life-threatening
Skin and skin-structure infections are common—and can influence the decision of when to perform a cesarean delivery, how to prepare the skin before surgery, and where to place the surgical incision. In some patients, these infections can be quite debilitating, even life-threatening. When a discrete abscess (furuncle, carbuncle) is present, the most likely organism is S aureus, and the majority of strains are MRSA. When cellulitis is present, S aureus is less likely, and the dominant organisms are usually streptococci, particularly Streptococcus pyogenes.
Abscesses of any size need incision and drainage and, in most cases, systemic antibiotic therapy. When cellulitis without a discrete abscess is present, the key to treatment is antibiotic therapy.
WHAT THIS EVIDENCE MEANS FOR PRACTICE
For uncomplicated skin and skin-structure infections in immunocompetent women, oral clindamycin and oral trimethoprim-sulfamethoxazole are equally effective; both achieve cures in approximately 90% of patients.
Given that more strains of S aureus were resistant to clindamycin, trimethoprim-sulfamethoxazole may be the preferred agent. It also is less expensive and, in theory at least, less likely to cause drug-induced diarrhea.
Affected patients need to be followed closely because recurrences are common and, in isolated instances, serious complications such as sepsis can develop.
—Patrick Duff, MD
Share your thoughts on this article! Send your Letter to the Editor to firstname.lastname@example.org. Please include your name and the city and state in which you practice.