Is a Minor Clinical Anastomotic Leak Clinically Significant after Resection of Colorectal Cancer?

Abstract

There are few reports comparing the variety and frequency of postoperative complications between patients with a major clinical leak requiring emergency abdominal reoperation and those with a minor leak diagnosed from clinical signs and managed expectantly without reoperation. This study examined the association between severity of leakage and 18 other postoperative complications, postoperative mortality, and length of postoperative hospital stay. Data were drawn from a comprehensive, prospective hospital registry of 1,507 colorectal cancer resections involving an anastomosis from January 1995 to December 2006. Differences were evaluated by two-tailed Fisher's exact test, Student's t-test, or Mann-Whitney U test. Leaks occurred in 54 patients (3.6%; 95% CI, 2.7% to 4.7%), comprising 21 major (1.4%; 95% CI, 0.9% to 2.1%) and 33 minor leaks (2.2%; 95% CI, 1.5% to 3.2%). Patients with a leak were significantly (p < 0.01) more likely than those without to have 11 of 18 other surgical and medical complications considered, although with few differences in complication rates between those with major and minor leaks. As compared with patients without leak, those with a leak (major or minor) had several of these complications rather than just one (p < 0.001) and greatly prolonged hospital stay (p < 0.001). Postoperative mortality was higher after major leaks than after minor leaks (4 of 21 and 0 of 33, respectively, p = 0.019). A minor leak is not trivial. Apart from the fact that major clinical leakage necessitates urgent reoperation, there were few other differences between major and minor clinical leaks in the frequency of other complications.