ABSTRACT
BACKGROUND: Skin and soft tissue infections (SSTIs) are a common complaint in the ambulatory setting and pose a significant burden on the health care system. OBJECTIVES: We sought to determine the accuracy of ultrasound for detecting soft tissue abscesses by emergency medicine associate providers (APs). METHODS: This was a prospective observational study of adult patients with suspected SSTI in the emergency department of a rural tertiary care academic medical center. The AP performed and interpreted point-of-care ultrasound (POCUS) of the suspected infected area. Ultrasound images and interpretation were reviewed by the attending emergency physician with either rejection or agreement of the image interpretation, diagnosis, and management. If incision and drainage was performed, presence or absence of purulent drainage was recorded. RESULTS: Sixty-four patients with suspected SSTI were enrolled: 29 had POCUS-proven abscesses and 33 had cellulitis; 2 were excluded. AP clinical evaluation alone for identifying abscess revealed sensitivity of 92.3 (74.9-99.1), specificity of 67.7 (49.5-82.6), positive predictive value of 68.6 (57.0-78.2), and negative predictive value of 92.0 (81.4-100). The use of POCUS by APs in addition to clinical examination demonstrated sensitivity of 96.2 (80.4-99.9), specificity of 93.9 (79.8-99.3), positive predictive value of 92.6 (76.5-98.0), and negative predictive value of 96.9 (81.9-99.5). p values were <0.05 for test characteristics. Ultrasound results changed SSTI management decisions in 10 of 62 (16.1%) patients, with the most common change being a new incision and drainage or needle aspiration. Agreement of the POCUS interpretation between APs and attending physicians was 96.8% with a κ of 0.94 (0.85-1.00). CONCLUSION: With modest training, APs can successfully use POCUS to identify abscesses in patients in the emergency department with SSTIs. POCUS increases the ability to rule in the diagnosis and changes management in a clinically relevant number of patients with SSTIs.
