413 A Management Algorithm Based on Outcomes of Clinically Significant Delayed Bleeding After Wide Field Endoscopic Mucosal Resection of Large Colonic Lesions in a Multicenter Prospective Cohort

A Management Algorithm Based on Outcomes of Clinically Significant Delayed Bleeding After Wide Field Endoscopic Mucosal Resection of Large Colonic Lesions in a Multicenter Prospective Cohort Nicholas G. Burgess*, Stephen J. Williams, Rajvinder Singh, William Tam, Luke F. Hourigan, Joshua Butt, Simon a. Zanati, Gregor J. Brown, Michael J. Bourke Gastroenterology, Westmead Hospital, Sydney, NSW, Australia; Gastroenterology, Lyell McEwin Hospital, Adelaide, SA, Australia; Gastroenterology, Princess Alexandra Hospital, Brisbane, QLD, Australia; Gastroenterology, Greenslopes Private Hospital, Brisbane, QLD, Australia; Gastroenterology, The Alfred Hospital, Melbourne, VIC, Australia; Gastroenterology, Western Hospital, Melbourne, VIC, Australia; Gastroenterology, Epworth Hospital, Melbourne, VIC, Australia Introduction: Wide Field Endoscopic Mucosal Resection (WF-EMR) for large colonic lesions is a safe and cost effective outpatient treatment. Clinically significant (defined as requiring hospital admission) post EMR bleeding (CSPEB) following an uneventful procedure occurs in 7% and may be serious. There are few studies examining bleeding outcomes following WF-EMR and no evidence based guidelines for management of bleeding in this specific group. Aims: To examine CSPEB management and outcomes in a large cohort and propose a management algorithm. Methods: Prospective multicenter data for WF-EMR of large sessile colorectal polyps or LSTs 20 mm (June 2008-May 2012: ClinicalTrials.gov NCT01368289) was analysed. Data collection included patient and lesion characteristics, procedural events and outcomes, complications and scheduled follow up at 14 days, 4 and 12 months. Standard EMR technique was used. Peri-procedural care was standardized. Results: WF-EMR was performed on 1139 lesions (mean size 35mm, right colon 52.9%) in 1050 patients (mean age 68 years). 62 patients had CSPEB (6.1%). 42% presented at 24 hours, 24% at 24-48 hours and 34% at 48 hours. 21% had hemodynamic instability at presentation. 26% received a blood transfusion. Mean inpatient stay was 2.9 nights34 patients (55%) were managed conservatively. 27 (44%) underwent colonoscopy. One had primary embolization. Endoscopic therapy was applied in 21 patients (80% Clips alone, 10% Thermal, 5% Combination, 5% Adrenalin). Only 17 lesions were actively bleeding. 5 patients did not undergo any form of endoscopic treatment. Standardized chart review with structured interview of the responsible clinician indicated bleeding could have been conservatively managed in 10/27 casesOne patient rebled after initial conservative management. One patient with copious bleeding obscuring endoscopic views had successful angiographic embolization. The patient who had primary embolization rebled and required emergency surgery. One patient required surgery for perforation related to hemostatic clips. There was no mortality on multiple logistic regression analysis a moderate or severe event by ASGE criteria was predicted by hemodynamic instability (OR 14.7 p 0.044) and lower hemoglobin at presentation (OR 0.43 per 1.0g/dL p 0.003). Colonoscopy was predicted by hematochezia occurring hourly (OR 32.3 p 0.008), ASA grade 2 (OR 6.8 p 0.027) or 3 (OR 13.7 p 0.031), and transfusion (OR 9.7 p 0.014). Conclusion: CSPEB settles spontaneously in 55% and can be managed conservatively in up to 70%. Patients who respond to initial resuscitation should be observed, with a lower threshold for intervention in those with hematochezia hourly, ASA grade 2 or 3, transfusion, low hemoglobin or hemodynamic instability. Patients unable to be stabilized by resuscitation, or who have recurrent bleeding should have urgent intervention. 414 Gastric Residual Volume Is Trivial Soon After Polyethylene Glycol Bowel Preparation Deepak Agrawal*, Richard Robbins, Don C. Rockey Digestive and Liver Diseases, University Texas Southwestern Medical Center, Dallas, TX; Internal Medicine, University Texas Southwestern Medical Center, Dallas, TX; Internal Medicine, Medical University South Carolina, Charleston, NC Background: Split dosing colonic preparations, where part of the prep is ingested the morning of the procedure, results in a cleaner colon and better adenoma detection rate. Bowel preparative agents meet the criteria for clear liquids, and as such, a 2-hour time interval should be safe. However, this issue has been debated, and the optimal time interval between last ingestion of colon prep and sedation for colonoscopy has been controversial. Objective: The aim of this study was to determine the gastric residual volume (GRV) in patients who completed drinking colon prep 2-3 hours before moderate sedation. Methods: Patients admitted to the hospital and requiring both upper endoscopy and colonoscopy were prospectively recruited for the study. They consumed 4L of polyethylene glycol (PEG, GoLytely) prep the night before and 1L starting at 6 AM to be completed by 7AM. Esophagoastroduodenoscopy (EGD) was performed 2-3 hours after the ingestion of last dose of bowel prep. Colonoscopy was performed immediately after EGD. Gastric fluid was suctioned off and collected in a specimen collector connected to suction port on the endoscope. Patients in whom EGD could not be performed within this time frame were excluded. Results: Sixty-three patients were recruited, and 49 patients underwent EGD in a mean of 2.6 hours after ingestion of the final portion of PEG. The mean GRV was 17.3 ml (range 0-50) and the median GRV was 20 ml. Multiple patients had no residual fluid (Figure). Seven patients had diabetes, 3 proven gastroparesis, and 6 patients were taking narcotic medications. The mean gastric residual volume in this group with possible delayed gastric emptying (n 16) of patients was 13.6ml. No patient had aspiration of gastric contents during EGD. Conclusions: Patients, including those with reasons for reduced gastric motility, had minimal gastric residual volume 2-3 hours after ingestion of PEG. This suggests that PEG should be considered a clear liquid, and that a 2-hour time Characteristics of gastrointestinal defects and prior interventions in 188 patients