Objectives: Soft tissue infections are a common presenting complaint in the emergency department (ED). The authors sought to determine the utility of ED bedside ultrasonography (US) in detecting subcutaneous abscesses.
Methods: Between August 2003 and November 2004, a prospective, convenience sample of adult patients with a chief complaint suggestive of cellulitis and/or abscess was enrolled. US was performed by attending physicians or residents who had attended a 30-minute training session in soft tissue US. The treating physician recorded a yes/no assessment of whether he or she believed an abscess was present before and after the US examination. Incision and drainage (I + D) was the criterion standard when performed, while resolution on seven-day follow-up was the criterion standard when I + D was not performed.
Results: Sixty-four of 107 patients had I + D-proven abscess, 17 of 107 had negative I + D, and 26 of 107 improved with antibiotic therapy alone. The sensitivity of clinical examination for abscesses was 86% (95% confidence interval [CI] = 76% to 93%), and the specificity was 70% (95% CI = 55% to 82%). The positive predictive value was 81% (95% CI = 70% to 90%), and the negative predictive value was 77% (95% CI = 62% to 88%). The sensitivity of US for abscess was 98% (95% CI = 93% to 100%), and the specificity was 88% (95% CI = 76% to 96%). The positive predictive value was 93% (95% CI = 84% to 97%), and the negative predictive value was 97% (95% CI = 88% to 100%). Of 18 cases in which US disagreed with the clinical examination, US was correct in 17 (94% of cases with disagreement, chi(2) = 14.2, p = 0.0002).
Conclusions: ED bedside US improves accuracy in detection of superficial abscesses.