Systematic Description of the Post EMR Defect

Overview

To date there are no available data on the utility of the endoscopic mucosal resection (EMR) defect in stratifying the risk of immediate or delayed adverse outcomes, particularly clinically significant post EMR bleeding (CSPEB). The investigators aimed to analyse the data to determine if any of these EMR defect features allow us to estimate the risk of CSPEB. This will help endoscopists to identify defects with a high risk of adverse outcomes and may translate into improved patient outcomes.

CSPEB is the most frequent serious complication after wide-field EMR of laterally spreading lesions ≥ 20mm (LSLs). There is no proven therapy for CSPEB and it remains a significant drawback of EMR. Visible vessels within the post EMR defect (PED) present themselves as logical targets for prophylactic treatment to prevent CSPEB. However, the clinical significance of these vessels is largely unknown. In the majority of studies risk factors identified for CSPEB include right colon location, lesion size and aspirin use. The investigators aimed to systematically describe and evaluate the clinical significance of the various endoscopic features of the post EMR defect PED including visible vessels. A prospective study of LSLs ≥ 20mm referred for EMR at a single tertiary referral center will be performed. Data collection includes patient, procedural and lesion characteristics. In all cases a systematic description of the PED is undertaken. The data of particular interest to this study includes: use of blood thinners, PED features including size, number and herniation of visible vessels, submucosal haemorrhage, fibrosis, fat and exposed muscle and the rate of CSPEB.These features were analysed for significant association with CSPEB. CSPEB was defined as any bleeding occurring after the completion of the procedure necessitating emergency department presentation, hospitalization or reintervention. CSPEB was compared to features of the PED to detect significant associations, using chi2 or Fisher's exact tests. Significant univariate variables will be taken forward for binomial logistic regression modelling.

Conditions