Background: Surgical site infection is a major cause of morbidity after colorectal resections. Despite evidence that preoperative oral antibiotics with mechanical bowel preparation reduce surgical site infection rates, the use of oral antibiotics is decreasing. Currently, the administration of oral antibiotics is controversial and considered ineffective without mechanical bowel preparation.
Objective: The aim of this study is to examine the use of mechanical bowel preparation and oral antibiotics and their relationship to surgical site infection rates in a colorectal Surgical Care Improvement Project cohort.
Design: This retrospective study used Veterans Affairs Surgical Quality Improvement Program preoperative risk and surgical site infection outcome data linked to Veterans Affairs Surgical Care Improvement Project and Pharmacy Benefits Management data. Univariate and multivariable models were performed to identify factors associated with surgical site infection within 30 days of surgery.
Settings: This study was conducted in 112 Veterans Affairs hospitals.
Patients: Included were 9940 patients who underwent elective colorectal resections from 2005 to 2009.
Main outcome measure: The primary outcome measured was the incidence of surgical site infection.
Results: Patients receiving oral antibiotics had significantly lower surgical site infection rates. Those receiving no bowel preparation had similar surgical site infection rates to those who had mechanical bowel preparation only (18.1% vs 20%). Those receiving oral antibiotics alone had an surgical site infection rate of 8.3%, and those receiving oral antibiotics plus mechanical bowel preparation had a rate of 9.2%. In adjusted analysis, the use of oral antibiotics alone was associated with a 67% decrease in surgical site infection occurrence (OR=0.33, 95% CI 0.21-0.50). Oral antibiotics plus mechanical bowel preparation was associated with a 57% decrease in surgical site infection occurrence (OR=0.43, 95% CI 0.34-0.55). Timely administration of parenteral antibiotics (Surgical Care Improvement Project-1) had a modest protective effect, with no effect observed for other Surgical Care Improvement Project measures. Hospitals with higher rates of oral antibiotics use had lower surgical site infection rates (R = 0.274, p < 0.0001).
Limitations: Determination of the use of oral antibiotics and mechanical bowel preparation is based on retrospective prescription data, and timing of actual administration cannot be determined.
Conclusions: Use and type of preoperative bowel preparation varied widely. These results strongly suggest that preoperative oral antibiotics should be administered for elective colorectal resections. The role of oral antibiotics independent of mechanical bowel preparation should be examined in a prospective randomized trial.