The purpose of this study was to evaluate risk factors for failure of antibiotic treatment within 30 days for uncomplicated skin and soft tissue infections of outpatients treated in a Veteran’s Affairs hospital.
A retrospective chart review of outpatients between January 1, 2006 to July 1, 2015 with an ICD-9 code of cellulitis or abscess were included in the analysis. The primary outcome was success versus failure of the antibiotic with failure defined as another antibiotic prescribed or hospitalization within 30 days for the original indication. Significant factors in the bivariate analysis were built into a multivariate logistic regression analysis to determine odds of failure.
A total of 293 patients were included in the final analysis; 24% of which failed within 30 days. Cellulitis was found in 62%, abscess in 23%, and mixed abscess and cellulitis in 15%. Obesity (BMI > 25kg/m2) was identified in 83% of the population. A BMI greater than 40 kg/m2 was identified in 16% of the population. In the subset of patients that failed oral antibiotics, they were found to have an elevated mean BMI of 34.2 kg/m2 (p=0.0098) and an elevated mean Charlson Comorbidity Index of 1.9 (p=0.048). Additionally, the patients that failed were found to have an increased prevalence of diabetes at 48% (p=0.026) and heart failure at 16% (p=0.027). There were no statistically significant differences in antibiotic regimens or treatment duration in the group of patients that failed versus those patients that did not fail. Using multivariate logistic regression, BMI and heart failure were determined to be significant predictors of antibiotic failure. An elevated BMI increased the unit odds of failure by 1.1 (1.01-1.1) for every 1 kg/m2unit increment. A diagnosis of heart failure increased the odds of failure by 2.6 (1.1-5.8).
Patients with an elevated BMI and heart failure were found to have increased odds of failure, defined as hospitalization or additional antibiotics within 30 days for skin and soft tissue infection. Further studies are needed in the subset of obese patients to determine solutions for the unacceptably high failure rates.