This study indicated risk factors for serious complications of ESD. Large resected tumor size was a risk factor for post-operative bleeding, while long operation time was a risk factor for perforation. Information regarding operation risk factors should be useful for planning strategies for ESD.

Background: This retrospective study aimed to determine risk factors associated with serious complications of endoscopic submucosal dissection of gastric tumors in multicenters compared between high- and low-volume centers. Methods: Between 2001 and 2010, gastric endoscopic submucosal dissection was performed in 1,190 lesions of 1,082 patients in five hospitals in Saga, three high-volume and two low-volume centers. Risk factors for serious complications were evaluated. Patients’ background characteristics were evaluated, including anticoagulants use and underlying diseases. Results: Postoperative bleeding was detected in 75 patients (6.9%), and perforation was detected in 40 patients (3.7%). Most postoperative bleeding and perforation cases were recovered with endoscopic procedures, although one case of each complication was treated by emergency surgery. Multivariate analysis indicated that risk factors for perforation were tumor location, massive submucusal invasion, endoscopists’ experience of 100–149 cases and hypertension, and that risk factors for postoperative bleeding were tumor location, resected tumor size, and scar lesion. The serious complications were not different between high- and low-volume centers. Conclusions: The present study indicated that risk factors for perforation during endoscopic submucosal dissection were tumor, endoscopist and patient related, although risk factors for postoperative bleeding were tumor related. There was no difference in complications between high- and low-volume centers.

No abstract

This study revealed that soft coagulation is as effective as hemoclipping for treating bleeding gastric ulcers. The time required to achieve hemostasis was shorter with the soft coagulation procedure.