Prediction of surgical site infection after colorectal surgery

Abstract

Surgical site infection (SSI) after colorectal surgery is a frequent complication associated with substantial morbidity. Our objective was to identify surgical predictors of SSI in patients undergoing colorectal surgery using a retrospective case-control design. Randomly selected patients from all those undergoing colorectal surgery (2007-2013). Cases were patients who developed SSI within 30 days. Controls were patients who did not develop SSI within 30 days. Patients undergoing multiple procedures during a single surgical intervention were excluded. SSI was diagnosed according to Centers for Disease Control and Prevention definitions. The main outcome measures were SSI, surgical variables, and cumulative survival (Kaplan-Meier method). Variables considered predictors were compared using log-rank test. Of 911 patients undergoing colorectal surgery, 221 developed SSI (24.3%; 95% confidence interval, 24.0-24.6). On univariate analysis, significant risk factors for SSI were: female sex (P = .02), >72 hours preoperative stay (P = .04), open surgery (P = .08), incision class: contaminated and dirty (P = .001), and emergency procedures (P = .006). On multivariate analysis, significant independent predictors of SSI and survival were dirty surgery (hazard ratio [HR], 2.12; P = .015), contaminated surgery (HR, 1.74; P = .009), female sex (HR, 1.58; P = .003), open surgery, (HR, 1.51; P = .015) and >72 hours preoperative stay (HR, 1.48; P = .024). Dirty or contaminated surgery, female sex, open surgery, and >72 hours preoperative stay were significant predictors of SSI.