Anastomotic leaks after bowel resection: what does peer review teach us about the relationship to postoperative mortality?

Abstract

Anastomotic leak is a dreaded complication of intestinal surgery and has been associated with a high mortality rate. But it is uncertain exactly which patient populations are at risk of death from the leak. We sought to assess the impact of surgeon volume on leak rate and to better understand the relationship of a leak to postoperative mortality. All adult patients having a small or large bowel resection with anastomosis at a university hospital from July 2003 to June 2006 were entered into a prospectively maintained quality database; data were entered by a specially trained nurse practitioner who rounded daily with housestaff. Patients with a postoperative leak based on standardized criteria were identified. Patient characteristics, surgical procedure, and operating surgeon were noted. Overall complication and leak rates by surgeon were compared using Fisher's exact test. Individual case review by a group of peers was performed for all patients with a leak who died, to determine the relationship to mortality. Five hundred fifty-six patients underwent resection with anastomosis during the study period. There were 27 patients with leaks (4.9%), 6 of whom died. Leak rate for the highest-volume surgeons ranged from 1.6% to 9.9% (p <0.01), and overall complication rate varied from 30.5% to 44% (p=0.04). In four of six deaths, leaks occurred in very ill patients undergoing emergency procedures and appeared to be premorbid events. In only one patient did the leak appear to be the primary cause of death. The variability in leak rate by surgeons doing similar operations suggests that many leaks may be preventable. But death after a leak is most often a surrogate for a critically ill patient and was infrequently the actual cause of death.