[Colorectal anastomosis: manual or mechanical? A controlled multicenter study].

272 patients underwent randomly 133 hand sewn anastomosis and 139 stapled anastomosis (159 proximal and 119 distal) after elective colorectal resection for 194 carcinomas, and 78 sigmoid diverticulosis and benign tumors. Both groups were comparable with regard to mortality (3.3 p. cent), wound abcesses (3 p. cent) extra abdominal post operative complications (20 p. cent). There were however more reoperations in the stapled group (10 versus 6), and also more low intestinal anastomotic bleeding (9 versus 5). Stapled anastomosis were associated with less clinical fistulas (8 p. cent versus 12 p. cent) and only after low colorectal anastomosis (9 p. cent versus 20 p. cent). Stapling devices decrease the duration of the anastomosis by 8 mn, but this gained time does not influence the total duration of the operation. These advantages must be weighed against the risks inherent to their uses, mishaps (14 p. cent), hemorrhages (6.5 p. cent), strictures (8 p. cent), and their higher cost (15 times as much).