Preoperative Risk Factors for Clostridium difficile Infection in Colectomy Patients

Background: The incidence of Clostridium difficile infection (CDI) has been steadily rising, growing in virulence, and demonstrating an increase in the severity and morbidity of the disease.  To curb the escalation of CDI, current efforts are focused on decreasing the exposure to the organism (enhanced personal hygiene and improved disinfection of surfaces) and maximizing the host resistance (reducing antimicrobial treatment).  Despite these efforts, the incidence of CDI continues to escalate without evidence of a peak or plateau. This problem is especially apparent in high-risk surgical patients undergoing bowel surgery.

Objective: The purpose of this research was to determine the preoperative risk factors of patients diagnosed with Clostridium difficile infection (CDI). 

Methods: Between August 2007 and June 2009, preoperative variables and CDI were collected for adult patients admitted for surgical colectomy (CPT codes of 44140, 44160, 44204, and 44205) from 24 hospitals in the Michigan Quality Surgical Collaborative (MSQC) Colectomy Best Practices Project.  This regional quality improvement collaborative participates in the American College of Surgeons National Surgical Quality Improvement Project and also collects additional targeted data specific to patients undergoing colectomy.  We performed bivariate analysis on 48 preoperative patient risk factors and surgical descriptors.  Variables significantly associated with CDI at p < .10 in the bivariate analysis were incorporated into a binary logistic regression model.

Results: 2274 patients underwent colectomy and fulfilled inclusion criteria.  A total of 55 patients (2.4% overall) developed a CDI per positive C. difficile toxin assays.  In crude and adjusted analysis, dyspnea (p=.033) mechanical ventilation (p=.002) and a history of a transient ischemic attack (p=.023) were significant predictors of CDI.

Conclusions: Predictors of pulmonary and neurological morbidities emerged as significant preoperative predictive variables in this cohort.  In contrast to previous single-center studies, this multi-center study did not demonstrate age or bowel preparation (with and without antibiotics) as influential predictors of CDI.  History of transient ischemic attack as a predictor for CDI may be partially explained by an emerging hypothesis attributing the use of statins with an increase in the rate and severity of the disease.  Identifying preoperative patients at highest risk for CDI can help target preventative interventions before, not in response to, infectious disease.