Emergency room visits for skin and soft tissue infections (SSTI) have increased dramatically. IDSA guidelines classify SSTI by diagnosis and severity. We investigated the decision-making process for site of treatment, and management in light of these guidelines.
We retrospectively analyzed medical records of 200 patients (pts) discharged from the emergency room of the VA Medical Center, Houston, with a diagnosis of SSTI. SSTIs were categorized as: (1) non-purulent (cellulitis/erysipelas or wound infection); or (2) purulent, and further stratified as mild, moderate or severe. We assessed 50 variables: demographics, co-morbidities, clinical and laboratory findings, treatment site, and treatment given. Statistical analysis used Chi-squared test.
Of 200 pts, 7 pt were excluded due to incorrect diagnosis. Of 193 remaining, 82 were treated as inpatients (IP) and 111 were treated as outpatients (OP). Among IP, 68 had nonpurulent infection (18 severe, 25 moderate, 25 mild), and 14 had purulent infection (all severe). Among 111 OP, 76 had nonpurulent infection (all mild), and 35 had purulent infection (11 severe, 15 moderate, 9 mild). Factors predictive for hospitalization (p<0.05) included: moderate-severe infection, insulin-dependent diabetes, peripheral vascular disease (PVD), congestive heart failure (CHF), alcohol use, body mass index, and fever.
Discrepancies between guidelines and practice included inappropriate hospitalization for mild disease (25 of 82 pts, 30%), inappropriate use of antibiotics, and blood cultures. Forty-one of 50 IP (82%) with mild/moderate non-purulent infection received vancomycin and/or treatment for Gram-negative rods. Of 9 OP with mild purulent infections, 8 (89%) were discharged on antibiotics. Blood cultures were drawn in 52 pts (35%) with mild/moderate infection.
IDSA guidelines regarding hospitalization for mild infection, antibiotic choice, and blood cultures were not consistently followed during management of SSTI. However, the decision to hospitalize may have been influenced by the high prevalence of underlying conditions, such as PVD, CHF, alcohol abuse, and insulin use. Our observations identify a need to validate IDSA guidelines and consider inclusion of comorbid conditions.
J. Gardner, None
H. Kamran, None
D. M. Musher, None
See more of: Poster Abstract Session