Risk factors for adverse events of colorectal endoscopic submucosal dissection: a systematic review and meta-analysis

Colorectal endoscopic submucosal dissection (ESD) is already an established treatment for superficial colorectal tumors. However, its technical difficulty and high adverse events rates, compared to endoscopic piecemeal mucosal resection, are a concern to some specialists and have probably contributed to discouragement in its widespread adoption. The debate mentioned above stimulated us to perform a systematic review aiming to identify risk factors for colorectal ESD-related adverse events. We conducted this study following the Preferred Reporting Items for Systematic Reviews and Meta-analyses statement and registered in the PROSPERO (University of York) international database (CRD42016042625). We searched MEDLINE, EMBASE, Cochrane Library and LILACS for the publications focused on risk factors for colorectal ESD-related adverse events from inception until April 2020. We included a total of 22 qualified studies in this analysis. We found that fibrosis had an odds ratio (OR) for perforation of 2.90 [95% confidence interval, (1.83-4.59)], right colon location of 2.35 (1.58-3.50), colonic location of 2.20 (1.44-3.35) and larger size of 2.17 (1.47-3.21), as well as one protective factor, the endoscopist experience OR = 0.62 (0.45-0.86). For bleeding, we considered rectal location a risk factor [OR = 3.55 (2.06-6.12)]. Through the several meta-analyses that we performed in this article, we could summarize the main risk factors for perforation and bleeding on colorectal ESD. Therefore, we provide insightful information for clinical judgment on regions where colorectal ESD is already widespread and help in the learning process of this challenging technique.