Surgical site infections: does inadequate antibiotic therapy affect patient outcomes?

Abstract

Background: Complicated skin/skin structure infections involve deeper soft tissues and include surgical site infections (SSIs). Inadequate antibiotic therapy (IAT) has been associated with adverse outcomes in respiratory and blood stream infections, but is seldom evaluated in SSIs. This study assessed the impact of IAT on primary outcomes of length of stay (LOS) and costs in complicated SSIs; identifying risk factors associated with receiving IAT was a secondary objective.

Methods: This retrospective cohort study of discharges from our 810-bed urban teaching hospital from Quarter 4/2004-Quarter 1/2006 identified 130 patients with complicated SSI among 298 patients with postoperative infections. Superficial infections and infections not involving the skin/skin structures were excluded. Patient characteristics, culture data, and antibiotic history were collected from charts. Inadequate antibiotic therapy was said to have occurred when a drug active against the organism cultured was not given within 24 h of culture. Multiple regression identified variables associated with LOS and increase hospital accounting costs.

Results: A total of 39 subjects (30%) received IAT; patient characteristics did not differ from those receiving adequate therapy, except that prior antibiotic use was more likely in IAT subjects (p = 0.053). Staphylococcus aureus (45% methicillin-resistant) was the most common pathogen (39%). More than one-half (60%) of the subjects received empiric vancomycin. The IAT patients experienced longer post-infection LOS and higher costs (median [25%, 75%]): 10 [6, 21] days vs. 7 [4, 11] days; p = 0.007 and $11,746 [$6,652, $28,442] vs. $7,116 [$5,210, $16,443]; p = 0.04). Longer LOS was associated significantly with Acute Physiology and Chronic Health Evaluation score, IAT, Pseudomonas infection, and sternal incisions, as were higher costs, excepting Pseudomonas infection. Inadequate antibiotic therapy was more likely in polymicrobial infections (p < 0.001), pseudomonal (p < 0.001) or enterococcal (p = 0.002) infections, and infected abdominal incisions (p < 0.001). Methicillin-resistant S. aureus was not associated with adverse outcomes, possibly because empiric therapy frequently included vancomycin.

Conclusions: Inadequate antibiotic therapy is associated with longer LOS and higher costs in complicated SSIs. Risk factors for IAT include prior antibiotic therapy, polymicrobial infection, infection with P. aeruginosa or Enterococcus spp., and abdominal incisions.