Methods: A retrospective cohort study involving 12 nursing homes in the Denver metropolitan area. For residents diagnosed with a skin infection between July 1, 2013 and June 30, 2014, clinical and demographic information was collected through manual chart review. Descriptive statistics were calculated for the entire cohort, and subgroup analysis was performed for non-purulent cellulitis versus all other types of skin infections.
Results: Of 100 cases included in the cohort, the most common infections were non-purulent cellulitis (n=55), wound infection (n=27), infected ulcer (n=8), and cutaneous abscess (n=7). In 26 cases, previously published minimum clinical criteria for initiating antibiotics were not met. Most antibiotics (n=52) were initiated as a telephone order following a call from a nurse, and 41 patients were not evaluated by a provider within 48 hours after initiation of antibiotics. Nearly all patients (n=95) were treated with oral antibiotics alone. Patients with non-purulent cellulitis were less likely to be prescribed broad gram negative antibiotics compared to the rest of the cohort (9% vs 31%, respectively; odds ratio 0.22 [95% confidence interval: 0.07 – 0.68]). The median treatment duration was 7 days (interquartile range [IQR] 7-10), although 43 patients received treatment courses of ≥ 10 days.
Conclusion: Most newly diagnosed skin infections in nursing homes were non-purulent infections and were treated with oral antibiotics. Antibiotics were initiated by telephone in over half of cases, and lack of a clinical evaluation within 48 hours after starting antibiotics was common. Improved diagnosis through more timely clinical evaluations and decreasing length of therapy are important opportunities for antibiotic stewardship in nursing homes.
B. Knepper, None
W. Burman, None
P. Mehler, None
T. Jenkins, None
See more of: Poster Abstract Session