1, p = 0.012). The mean score of the Boston Bowel Preparation Scale was similar between the two groups but there was a trend towards higher percentage of satisfactory

overall grading of bowel preparation in the split-dose group (96.9% vs. 91.4%, p = 0.056). Patients in the split-dose group were more likely to be able to complete their bowel preparation (98.4% vs. 94.2%, p = 0.07). Patients in whole-dose group were more likely to experience nausea (35.3% vs. 23.3%, p = 0.031). Although there was no significant difference in overall comfort during bowel preparation between the two groups, patients in the whole-dose group were more likely to refuse the same regime (13.7% vs. 6.2%, p = 0.042) and to want to try another regime (78.4% vs. 55.8%, p < 0.001). Conclusion: We conclude that split-dosing PEG-ELS PLX4032 group has better polyps detection rate and less side effects compared to whole-dose PEG-ELS group. Key Word(s): 1. Bowel preparation; 2. PEG-ELS; 3. Split-dosing; 4. RCT; Presenting Author: ZHEN LI Additional Authors: XIU-LI ZUO, YAN-QING LI Corresponding Author: ZHEN LI Affiliations: Shandong University, Qilu Hospital Objective: Gastric intestinal metaplasia (GIM) is a well-known premalignant lesion Dabrafenib supplier for intestinal type gastric cancer. However, present guidelines such as the updated Sydney System require multiple biopsies whereas reveal an unsatisfactory yield considering the

detection and surveillance of these lesions because of their inconspicuous endoscopic appearance. This study primarily aims at comparing the diagnostic yield of GIM by confocal laser endomicroscopy (CLE) and standard endoscopy in a high risk population. The second objective is to determine if CLE can reduce the biopsy number needed per patient for the detection of GIM in this patient specific population. Methods: Consecutive patients that were scheduled for upper CLE examinations were prospectively recruited. Enrolled subjects were randomized at a 1:1 ratio into group A (Standard white-light endoscopy)

or group B (CLE) by using computer-generated random numbers. In group A, patients received standard white-light endoscopic examination. Random biopsies Idoxuridine following the updated Sydney System (distal antrum + mid corpus + angle; greater/lesser curvature) were performed in addition to targeted biopsies of the endoscopic suspicious lesions. For patients in group B, CLE examinations were performed at endoscopic suspicious lesions and the aforementioned 5 standard areas. Biopsies were taken only in the presence of in vivo mucosal abnormalities including GIM and gastric neoplasia as determined by previously published CLE diagnostic criteria. Results: A total of 168 patients were finally analyzed in this study (85 in group A and 83 in group B). On a per-biopsy analysis, Endomicroscopy targeted biopsies significantly increased the diagnostic yield of GIM as compared to WLE and standard biopsies from 15.