Prospective study of the serosubmucosal (extramucosal) suture technique for gastrointestinal anastomosis.

Seventy gastrointestinal anastomoses were constructed in 52 patients using a single layer of interrupted sutures placed in the serosubmucosal or extramucosal layer. Gastric anastomoses (n = 16) were performed by a partly open method, small bowel and colonic anastomoses (n = 44) by a closed method, and colorectal anastomoses (n = 10) by an open method. Wound infection occurred in two patients (3.8%) and three patients died (5.8%). Anastomotic leakage in four patients (7.7%) arose from three ileocolic anastomoses and one colorectal anastomosis. The third leak from an ileocolic anastomosis resulted in death and necessitated termination of the study on ethical grounds. There were two other postoperative deaths from cardiorespiratory causes (3.8%). The serosubmucosal or extramucosal technique is safe and effective (leak rate 1 of 26; 3.8%) when performed by the partly open or open method for gastric and colorectal anastomoses. However, when constructed by the closed method a relatively high leakage rate (3 of 44; 6.8%) occurs in anastomoses (ileocolic) generally associated with a low morbidity.