Purpose: This study compared outcomes between end-to-end anastomosis and other anastomotic configurations after intestinal resection for patients with Crohn's disease by using meta-analytical techniques.
Methods: Comparative studies published between 1992 and 2005 of end-to-end anastomosis vs. other anastomotic configurations were included. Using a random effects model, end points evaluated were short-term complications and perianastomotic recurrence of Crohn's disease. Heterogeneity was assessed and sensitivity analysis was performed to account for bias in patient selection.
Results: Eight studies (2 prospective, randomized, controlled trials; 1 nonrandomized, prospective; 5 nonrandomized, retrospective studies) reported on 661 patients who underwent 712 anastomoses, of which 383 (53.8 percent) were sutured end-to-end anastomosis and 329 (46.2 percent) were other anastomotic configurations (259 stapled side-to-side, 59 end-to-side or side-to-end, 11 stapled circular end-to-end). Anastomotic leak rate was significantly reduced in the other anastomotic configurations group (odds ratio (OR), 4.37; P = 0.02) and remained significantly lower in studies comparing only side-to-side anastomosis vs. end-to-end anastomosis (OR, 4.37; P = 0.02) and studies including only ileocolonic anastomosis (OR, 3.8; P = 0.05). Overall postoperative complications (OR, 2.64; P < 0.001), complications other than anastomotic leak (OR, 1.89; P = 0.04), and postoperative hospital stay (weighted mean difference, 2.81; P = 0.007) were significantly reduced in the side-to-side anastomosis group when considering studies comparing only side-to-side anastomosis vs. end-to-end anastomosis. There was no significant difference between the groups in perianastomotic recurrence and reoperation needed because of perianastomotic recurrence.
Conclusions: End-to-end anastomosis after resection for Crohn's disease may be associated with increased anastomotic leak rates. Side-to-side anastomosis may lead to fewer anastomotic leaks and overall postoperative complications, a shorter hospital stay, and a perianastomotic recurrence rate comparable to end-to-end anastomosis. Further randomized, controlled trials should be performed for confirmation.