Factors involved in disruption of intestinal anastomoses.

Bowel anastomoses, as performed on 181 dogs, were studied: (1) by interposing segments of colon into small bowel and vice versa, (2) by comparing clean anastomoses to those contaminated by feces before and after suturing, (3) with and without parenteral preoperative antibiotic, and (4) with and without coaptation of an inverted serosa. All animals with a timed sacrifice as well as an unexplained death had careful autopsy. Results demonstrated no difference in the healing capacity of large (91%) versus small (92%) intestine under identical circumstances. Intraluminal bacteria were of importance only if spillage caused contamination during operation and thereby subsequent infection of the peritoneal surface of the suture line. Peritonitis preceded all 28 leaks, yet the converse never occurred. Likelihood of a complicating peritonitis (67%) and thus an anastomotic leak (24%) was significantly reduced through the preoperative administration of prophylactic cefazolin (19 and 4%, respectively). A "serosal seal" also appeared important in obviating suture line disruption. Our data emphasize the value of an inverted and serosal lined anastomosis, bowel preparatory measures, prophylactic antibiotic, and the disruptive action of local bacterial peritonitis.