Urinary tract infection after colon and rectal resections: more common than predicted by risk-adjustment models.

Abstract

Urinary tract infections (UTIs) are the most common hospital-acquired infections in the United States. We hypothesized that the risk of UTI after colorectal surgery exceeds the risk after other gastrointestinal operations. We used National Surgical Quality Improvement Program (NSQIP) data from 2005 to 2008 to compute rates of UTI after colorectal resections and other gastrointestinal and general surgery. We used a validated multivariate UTI prediction model to compare risk-adjusted rates of UTI by type of procedure. Then we identified predictors of UTI after colorectal resection using stepwise logistic regression models. Crude UTI rates were significantly greater after colorectal resection (4.1%) than after other gastrointestinal (1.8%) or nongastrointestinal procedures (1.2%; all p < 0.001). Even using standard risk-adjustment from the NSQIP, rates of UTI were significantly higher after segmental colectomy (2.8%; 95% CI 2.5% to 3.2%), total colectomy (3.5%; 95% CI 2.9% to 4.3%), proctectomy (3.5%, 95% CI 3.1% to 4.2%), proctocolectomy (4.6%, 95% CI 3.8% to 5.7%), and abdominoperineal resection (5.6%, 95% CI 4.6% to 6.8%) than after noncolorectal gastrointestinal surgery (2.6%, 95% CI 2.2% to 2.9%). Factors associated with UTI after colorectal resections include age, sex, functional status, American Society of Anesthesiologists class, and resection type. Colorectal resections incur substantial risk of postoperative UTI, exceeding rates predicted by the NSQIP model. Because of their patients' unanticipated high incidence of UTI, surgeons with a specialty interest in colorectal surgery risk being flagged as "high outliers," particularly if they perform many rectal resections. A simple set of risk factors discriminates 10-fold differences in the rate of UTI after colorectal resection.