CHICAGO – Several modifiable risk factors predicted the development of anastomotic leak following elective colon resection, a large national analysis found.
“Preoperative smoking cessation, preoperative administration of oral antibiotic bowel preparation, and laparoscopic approach are modifiable factors that could reduce the risk of anastomotic leak,” Dr. Cindy Wu said at the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) National Conference.
Anastomotic leakage results in increased morbidity and mortality, yet the current literature analyzing risk factors for this complication is generally limited to retrospective studies of single institutions, she said.
To examine data from a larger sample of colectomy patients from multiple centers, the investigators used the NSQIP Participant Use Data File specifically targeted to colectomy to identify 14,848 patients who underwent elective colon resection from 2012 to 2013. Chi-square, Wald chi-square, and logistic regression analyses were performed examining patient factors (sex, race, comorbidities, smoking status, American Society of Anesthesiologists class, functional status, steroid use, and preoperative albumin), oncologic factors (chemotherapy, tumor stage, and presence or absence of disseminated cancer), and operative factors (wound class, mechanical bowel preparation, oral antibiotic preparation, surgical approach, colectomy site, surgical indication, and operative time).
In all, 3.4%, or 498 patients, experienced an anastomotic leak, which is consistent with the literature, Dr. Wu of Temple University in Philadelphia said. Of these patients, 101 required no intervention, while 272 required surgery and 125 needed percutaneous drainage. The mean age of the patients was 60.7 years and 57% were male.
In a univariate analysis, male sex (chi-square = 17.4; P less than .01), diabetes controlled with either oral medication or insulin (X2 = 9.5; P less than .01), and smoking within the last year (X2 = 20.4; P less than .01) were associated with a greater incidence of anastomotic leak.
Other risk factors that were significant in additional univariate analysis were ASA class (X2 = 23.3; P = .0001), functional status (X2 = 9.15; P = .01), 10% weight loss over the last 6 months (X2 = 5.83; P = .02), wound class (X2 = 10.8; P = .01), mechanical bowel preparation (X2 = 5.89; P = .01), lack of oral antibiotic preparation (X2 = 17.5; P less than .0001), open vs. laparoscopic/minimally invasive surgery (X2 = 60.0; P less than .0001), chemotherapy in the last 90 days (X2 = 23.1; P less than .0001), and presence of disseminated cancer (X2 = 7.41; P = .01), Dr. Wu said.
With all of these factors taken into account in multivariate analysis, independent predictors of an increased risk of anastomotic leak were male sex (odds ratio, 1.74; P = .01), tobacco use (OR, 1.73; P = .03), and lack of a preoperative oral antibiotic bowel preparation (OR, 1.79; P less than .01).
Interestingly, use of a laparoscopic technique was protective against the development of anastomotic leakage (OR, 0.54; P less than .01), she said.
The authors reported having no relevant financial disclosures.