Endoscopy Lower.

Abstract

Journal of Gastroenterology and HepatologyVolume 31, Issue S2 p. 8-28 Supplement ArticleFree Access Endoscopy Lower First published: 05 October 2016 https://doi.org/10.1111/jgh.13516AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Assessing the effectiveness of telephoning positive FOBT patients prior to their colonoscopy on quality of the bowel preparation: a randomized controlled trial N Al-Hajjiri1, K Williams1, YH Chieng1, A Grillas1, I Valiozis1, C Rogge1, S Abey1, D Swartz1, J McDonald1, O Sharaiha1, K Holding1, M Blanchard2, R Gordon2 and T Lee1 1Department of Gastroenterology, Wollongong Hospital, Wollongong, Australia; 2Australian Health Service Research Institute, University of Wollongong, Wollongong, Australia Introduction: A good bowel preparation is critical for a successful high-quality colonoscopy. The consequences of poor bowel preparation include missed lesions and the need for a repeat colonoscopy at additional cost to the health care system. Factors affecting the quality of bowel preparation include patient factors (age, sex, comorbidity, language barrier) and type of bowel preparation. Aim: We conducted a double blind randomized controlled trial to assess the effectiveness of a dedicated endoscopy nurse telephoning patients 3 days before their colonoscopy on the quality of bowel preparation assessed using the Boston Bowel Prep Score (BBPS). Secondary measures included patient satisfaction and the type of frequently asked questions during the intervention. Methods: Participants were prospectively recruited from gastroenterology outpatient clinics based on positive FOBT test results. After obtaining written informed consent, participants were stratified by sex and randomized into 1:1 intervention or control cohorts. A calibration training session on the use of BBPS with all proceduralists took place prior to the study in order to standardize scoring. Inadequate bowel preparation requiring a repeat colonoscopy is indicated by any BBPS sub-score ≤1 out of 3 or a total score ≤ 5 out of 9. Basic demographic details, previous colonoscopy experience, language background, assistance at home, endoscopy location, and timing of the endoscopy session were recorded. The result was analyzed descriptively. The study was approved by the relevant Human Research Ethics Committee and funded by the NSW Cancer Institute. Results: Of 345 eligible patients, 305 were randomized – 141 control and 138 treatment patients who completed the study. Poor bowel preparation requiring a repeat colonoscopy was noted in 9.2% of the control group and 8.7% of the treatment group, a non-significant result (P > 0.05). Interestingly, better bowel preparation was noted in the colonoscopy naïve participants (8.5% v 15.2%) and those in the afternoon session (5.9% vs 11.9%). No statistical significance in bowel preparation quality was found between the variables age, sex, country of birth, language at home, use of interpreter and assistance at home. Conclusions: Having a dedicated nurse telephoning the participants 3 days prior to their colonoscopy to clarify any bowel preparation questions did not significantly reduce the proportion of positive FOBT patients with inadequate preparation. Service and quality improvement – residual gastric volume (RGV) after split-dose bowel preparation using polyethylene glycol (PEG): does age matter? A Alghamry1,2, H Moattar1, A Wan1,2, A Vandeleur1, ST Yerkovich1,2, Endoscopy Nurses Collaborative (ENC)1, J Thomas1,2, J Croese1,3, T Rahman1,3 and R Hodgson1,2 1Centre for Service and Quality Improvement, Department of Gastroenterology & Hepatology, The Prince Charles Hospital, Brisbane; 2School of Medicine, University of Queensland, Brisbane; 3College of Medicine and Dentistry, Division of Tropical Health and Medicine, James Cook University, Cairns, Australia Background: Prior studies suggest that pulmonary aspiration risk increases with RGV > 25 mL and pH < 2.5.1 Older age and the presence of diabetes mellitus (DM) are perceived to be associated with slow gastric emptying due to age or diabetes related autonomic nerve dysfunction. A previous study showed younger age and male gender to correlate with higher RGV assessed endoscopically. All patients were fasting for 2 and 6 hours after fluid and solid intake, respectively, but results have not been replicated by any other studies with the same methodology, especially with PEG-containing bowel preparation.2 Objective: We examined RGV and pH in patients undergoing combined esophagogastroduodenoscopy (EGD) and colonoscopy after split-dose bowel preparation with PEG. Methods: In a single-center study, 892 consecutive patients undergoing combined EGD and colonoscopy were prospectively recruited between Dec 2013 and May 2016. RGV was recorded immediately after stomach intubation. Patient demographics, diabetic status including treatment, and runway time were recorded. Mean ± standard deviation (SD) was recorded for all data and Spearman–Rho was used to detect correlation between non-parametric data. Univariate and multivariate logistic regression analysis were used to detect variables associated with higher RGV. P-value <0.05 was considered to be statistically significant. Results: Mean age ± SD was 55 ± 16 (range 18–94) years, with 61% female. pH could not be estimated in 158 patients with RGV < 1 mL. Mean RGV, pH, runway time were 17.7 ± 19.2 mL, 2.6 ± 1.5, 5.8 ± 2.0 hours, respectively. There was a negative and significant correlation between RGV and age (r = −0.253, P < 0.001). Males had a higher RGV than females (20.9 ± 1.4 vs. 16.2 ± 0.8, P = 0.002). pH was inversely correlated with RGV (r = −0.190, P < 0.001). Ninety-eight (11%) patients were diabetic with 27 patients on insulin therapy (27.5%) and 70 patients using oral hypoglycemic therapy (71.4%). There were no differences in mean RGV ± SD between diabetics and non-diabetics (17.4 ± 0.86 vs. 18.8 ± 2.4, P = 0.48). Similarly, RGV did not differ between insulin dependent diabetics and patients on oral hypoglycemics. “Younger age” and “male gender” were the only significant predictors of higher RGV on multivariate analysis (P < 0.001) (Table 1). Conclusion: Older age and DM did not correlate with higher RGV in this large cohort of patients following the use of PEG-containing bowel preparation. In fact, young age and male gender are the only two factors that correlate with high-volume acidic gastric content that could result in chemical pneumonitis should aspiration occur, which is in agreement with previously published work (2). Table 1 Univariate and multivariate logistic regression analysis of predictors of higher RGV Univariate β (95% CI) P Age (per 10 years) −2.23 (−3.10 to −1.36) <0.001 Male sex 4.66 (1.72–7.60) 0.002 Diabetes 1.42 (−2.54–5.39) 0.482 Runway time −1.21 (−0.73–0.49) 0.699 Multivariate Age (per 10 years) −2.53 (−3.40 to −1.66) <0.001 Male sex 6.09 (3.17–9.01) <0.001 References 1Roberts RB, Shirley MA. Reducing the risk of acid aspiration during cesarean section. Anesth Analg 1974; 53: 859– 868. 2Phillips S, Liang SS, Formaz-Preston A, Stewart PA. High-risk residual gastric content in fasted patients undergoing gastrointestinal endoscopy: a prospective cohort study of prevalence and predictors. Anaesth Intensive Care 2015; 43: 728– 733. Residual gastric volume (RGV) after split-dose bowel preparation using polyethylene glycol (PEG) in obese patients A Alghamry1,2, ST Yerkovich1,2, H Moattar1, A Wan1,2, A Vandeleur1, Endoscopy Nurses Collaborative (ENC)1, J Thomas1,2, J Croese1,3, T Rahman1,3 and R Hodgson1,2 1Centre for Service and Quality Improvement, Department of Gastroenterology & Hepatology, The Prince Charles Hospital, Brisbane; 2School of Medicine, University of Queensland, Brisbane; 3College of Medicine and Dentistry, Division of Tropical Health and Medicine, James Cook University, Cairns, Australia Background: Although previous clinical trials reported similar RGV and pH in obese compared to non-obese patients, aspiration related concerns always exist when obese patients are undergoing esophagogastroduodenoscopy (EGD) and colonoscopy under deep sedation.1, 2 Prior studies suggest that pulmonary aspiration risk increases with RGV > 25 mL and pH < 2.5.3 Objective: We examined RGV and pH in obese patients undergoing combined EGD and colonoscopy after split-dose bowel preparation with PEG compared to normal body weight patients. Methods: In a single-center study, 202 consecutive patients undergoing combined EGD and colonoscopy were prospectively recruited between May 2015 and May 2016. RGV was recorded immediately after stomach intubation and pH was measured using pH indicator strips. Patient demographics, BMI, and runway time were recorded. Mean ± standard deviation (SD) was recorded for all data and regression analysis used to detect differences between BMI groups. Results: Mean age ± SD was 54.2 ± 16.8 years, with 60% female. pH could not be estimated in 40 patients with RGV < 1 mL. Mean RGV, pH, runway time, and BMI were 16.9 ± 18 ml, 2.4 ± 1.4, 5.5 ± 2.0 hours, and 28.2 ± 6.5, respectively. Stratified by BMI, there was no difference in RGV between normal weight and all obesity classes (P = 0.49). Similarly, pH did not differ between any of the groups (P = 0.99) (Table 1). There were no reported cases of aspiration pneumonia from routine phone follow up. Conclusion: Obese patients have similar RGV and pH to their normal weight counterparts when using PEG-based split-dose bowel preparation with no features to suggest high risk gastric content. Table 1 Demographics and clinical data Normal weight BMI 18.5–24.9 (n: 69) Overweight BMI 25–29.9 (n: 71) Class 1 obesity BMI 30–34.9 (n: 37) Class 2 obesity BMI 35–39.9 (n: 16) Class 3 obesity BMI ≥ 40 (n: 9) Demographics Age* 48.7 ± 17.5 58.5 ± 15.6 55.6 ± 18.7 57.6 ± 11.4 52.5 ± 9.5 BMI* 22 ± 1.7 27.4 ± 1.4 32.6 ± 1.4 36.7 ± 1.3 47.8 ± 5.7 Sex (Female %) 65% 50% 67% 56% 66% Runway time (hours)* 5.6 ± 2.2 5.4 ± 1.4 5.5 ± 2.9 5.2 ± 1.7 5.4 ± 1.2 Characteristics of gastric aspirate Residual gastric volume (mL)* 16.8 ± 18 15.9 ± 17 15.1 ± 15 22 ± 15 24.4 ± 33 Gastric pH* 2.4 ± 1.4 2.4 ± 1.3 2.4 ± 1.4 2.7 ± 1.7 2.6 ± 2 Patients with gastric volume >25 mL 25% 21% 24% 31% 33% Patients with gastric pH < 2.5 12% 28% 32% 12% 22% Patients with gastric volume > 25 mL and pH < 2.5 0% 0% 0% 0% 0% Asterisk symbol (*) = values are mean ± SD; remainder are percentage. References 1Juvin P, Fevre G, Merouche M, Vallot T, Desmonts JM. Gastric residue is not more copious in obese patients. Anesth Analg 2001; 93: 1621– 1622, table of contents. 2Harter RL, Kelly WB, Kramer MG, Perez CE, Dzwonczyk RR. A comparison of the volume and pH of gastric contents of obese and lean surgical patients. Anesth Analg 1998; 86: 147– 152. 3Roberts RB, Shirley MA. Reducing the risk of acid aspiration during cesarean section. Anesth Analg 1974; 53: 859– 868. Non adherence to NHMRC post polypectomy surveillance intervals in an Australian tertiary institution B Baraty, H Wang, D Zhou, P Wu, M Szczesniak, P Craig and L Choo Department of Gastroenterology and Hepatology, St. George Hospital, Sydney, Australia Introduction: Adherence to the National Health and Medical Research Council of Australia (NHMRC) guidelines for post polypectomy surveillance colonoscopies (PPSC) is unclear. Premature deviations from guidelines may increase risks to patients and costs to the healthcare system. Aims: To determine rates of adherence to NHMRC guideline for PPSC To identify factors associated with premature PPSC To determine if premature PPSC infers additional benefit in detection of advanced colonic neoplasia. Methods: An audit of all PPSC performed at our two tertiary public hospitals in 2013 was performed using electronic and paper-based medical records. For each identified PPSC, data from the prior colonoscopy including histology reports were collected. Premature deviations from recommended intervals were determined using NHMRC guidelines. Multivariate analysis was performed to determine demographic and endoscopic factors associated with the incidence of premature PPSC. Results: A total of 461 PPSC (191 by gastroenterologists; 270 by surgeons) were identified. With 6-month flexibility around the NHMRC guidelines for PPSC, a total premature surveillance rate of 76% was identified. PPSC was performed at a median of 44 (IQR 37–54) months, 40 (IQR 31–43) months, and 14 (IQR 11–21) months compared to recommended 120, 60, and 36 month intervals, respectively. Multivariate analysis identified that in patients less than 65 years old, those with poor bowel preparation at index colonoscopy, and polyp features including diminutive size, hyperplastic histology, and low total polyp counts at index colonoscopy were associated with incidence of subsequent premature PPSC (Table 1). Neither the level of training nor the specialty of the doctor arranging PPSC was associated with premature surveillance. Premature PPSC had a lower detection rate of advanced colonic neoplasms (>10 mm, villous histology, or high grade dysplasia) when compared to PPSC performed at recommended intervals (5% vs. 22% respectively, P < 0.001). No cancer was detected in either group. Conclusion: In our cohort, adherence to NHMRC guidelines is poor, with 76% of PPSC performed significantly earlier than the recommended intervals. We identified that at the index colonoscopy, the younger patients (<65 years old), poor bowel preparation, and low risk polyp features (size less than 10 mm, less than 2 polyps, and hyperplastic histology) were associated with premature PPSC. Premature PPSC did not convey any additional benefit in detection of advanced colonic polyps. Ongoing study involving anonymous surveys are underway to elucidate factors that may explain these findings. Table 1 Factors associated with premature surveillance colonoscopy Odds Ratio P 95% confidence interval Older Age (<65) 1.03 0.016 1.00–1.06 Male 0.87 0.616 0.52–1.48 Poor bowel prep 2.16 0.036 1.05–4.45 <10 mm polyp 7.40 0.000 3.73–14.69 <2 diminutive polyps 5.36 0.000 2.68–10.72 Hyperplastic polyps 4.75 0.011 1.43–15.75 Surgeon arranged surveillance 1.14 0.786 0.43–3.02 Trainee arranged surveillance 1.65 0.187 0.78–3.48 Description of disease activity in patients with ulcerative colitis by endoscopists is suboptimal M Chew1,2, N Parthasarathy1 and M Garg1,2 1Department of Gastroenterology and Hepatology, Eastern Health, Melbourne, Victoria, Australia; 2Eastern Health Clinical School, Monash University, Melbourne, Victoria, Australia Background: Accurate and comprehensive description of endoscopic findings – including the use of endoscopic scores such as the Mayo endoscopic score – in patients with ulcerative colitis (UC) is important in the determination of appropriate therapy. This project aimed to retrospectively assess adequacy and accuracy of description of endoscopic findings amongst patients with UC at a tertiary metropolitan health service. Methods: Endoscopy reports from patients with UC who underwent colonoscopy or flexible sigmoidoscopy within the previous 12 months were identified and assessed for use of descriptors of inflammation, namely, edema, loss of vascular pattern, granularity, erythema, friability, ulceration, stricturing, and pseudopolyps. The reporting of endoscopist global assessment of severity of inflammation (normal, mild, moderate, severe) and endoscopic scoring systems (including Mayo endoscopic score) were assessed for consistency with descriptors of inflammation and for concordance with photographs and histological assessment of severity when available. Results: Eighty-one endoscopy reports from 76 patients (mean age 49 [range 17–90] y, 37 [49%] females) were identified. Sixty reports comprised colonoscopy findings, and 21 flexible sigmoidoscopy findings. Thirty were performed by gastroenterologists with a sub-specialty interest in inflammatory bowel disease (IBD), 32 by other gastroenterologists, 15 by gastroenterology trainees, 3 by colorectal surgeons and 1 by a general surgeon. Fifty-two reports were from patients with extensive colitis, 24 from patients with left-sided colitis, and 5 from patients with proctitis. The endoscopy was reported as being normal in 16 reports. Of the remaining 65 reports, a median of 2 (range 0–8) descriptors were used – most commonly ulceration (n = 48), erythema (n = 46) and granularity (n = 43). Thirty-three (50.7% of 65 non-normal) reports mentioned an endoscopic scoring system – the Mayo endoscopic score in all. Fifty (76.9%) reports mentioned global assessment of severity. Internal consistency between the Mayo endoscopic score and descriptors was present in 51 (78.5%) reports. Adequate quality photographs were present in 70 reports, including 64 that had active disease and 28 that reported a Mayo score of 1 to 3. Concordance between the reported Mayo endoscopic score and photographs was present in 20 (71%) of these 28 reports. Concordance between global endoscopic assessment of severity and histological severity was present in 36 of 62 (58.1%) reports. A change in therapy was undertaken following 34 endoscopies, with escalation in 28 patients and de-escalation in 6 patients. Amongst these 34 patients in whom therapy was changed, the Mayo endoscopic score was reported and adequate quality photographs were present in 16 patients. Lack of concordance between the Mayo endoscopic score and photographs was present in only 1 (6%) of these 16 reports. Conclusion: Endoscopic findings in UC were described to an insufficient extent by clinicians. Accepted descriptors of inflammation in UC were reported in three-quarters of patients with active disease, the Mayo endoscopic score was only reported in half of patients with active disease and lack of consistency between this score and descriptors was present in one quarter. Electronic reporting systems incorporating mandatory fields, as well as education in description of findings, may result in improvement and consistency of reporting and lead to optimal therapeutic decisions in real-world practice for patients with UC. Outcomes of non-pedunculated high grade-dysplastic polyps in the English Bowel Cancer Screening Programme C Dailami, R Willert and HY Lee Department of Gastroenterology, Central Manchester University Hospitals NHS Foundation Trust Background and Aims: Colonic polyps over 20 mm, sessile or flat, with villous histology have a higher malignant potential.1 High grade dysplasia (HGD) and piecemeal EMR (pEMR) may also confer a higher risk of subsequent cancer.2 This study assessed the outcomes of resected, non-pedunculated HGD polyps ≤19 mm, and ≥20 mm removed en-bloc or by pEMR within the English Bowel Cancer Screening Programme (BCSP). Methods: All colonoscopies undertaken in the BCSP (2008–2015) with HGD polyps found were retrospectively reviewed. The primary outcome assessed incidence of cancer development. Secondary outcomes assessed polyp size and location. Descriptive statistical analyses and chi-squared tests were used to analyze the data using SPSS Version 23. Results: There were 2245 flat or sessile HGD polyps, in 2159 patients, identified (1510 males, mean age 66.8 years). Mean endoscopic diameter was 23.25 mm (range 2–100 mm). Small HGD polyps were defined as ≤19 mm (n = 1026 polyps), large polyps ≥20 mm (n = 1213); 6 polyps had no size recorded. En-bloc resection occurred in 553 small polyps (53.9%); 698 large polyps underwent pEMR (57.5%). Cancers recorded <62 days from index colonoscopy were classified as incident cancers (n = 627). Excluding these and polyps with missing information (n = 113) gave 1505 non-malignant HGD polyps (669 small, 836 large). The incidence of malignant progression irrespective of polyp size was 3.7% (n = 56), where most (84%, n = 47) were left sided cancers. Seventeen cancers arose from polyps ≤19 mm (30.3%). Thirty-nine cancers arose from polyps ≥20 mm (69.7%), of which 20 (51.3%) were previously resected piecemeal; 1 (2.6%) resected en bloc; 18 (46.2%) unspecified. Polyps removed by pEMR and ≥20 mm were associated with a higher incidence of malignant progression (P < 0.05). Conclusion: The majority of HGD polyps removed at colonoscopy, irrespective of resection technique, did not subsequently progress to cancer. However, the incidence of malignant progression was greater in polyps ≥20 mm and following pEMR. Large HGD polyps therefore require close surveillance post pEMR. References 1Knabe M, Pohl J, Gerges C, Ell C, Neuhaus H, Schumacher B. Standardized long-term follow-up after endoscopic resection of large nonpedunculated colorectal lesions: a prospective two-center study. Am J Gastro 2014; 109: 183– 189. 2Facciorusso A, Di Maso M, Serviddio G, Vendermiale G, Spada, C, Costamagna et al. Factors associated with recurrence of advanced colorectal adenoma after endoscopic resection. Clin Gastro Hep 2016 (in press). Do visible vessels within the post endoscopic mucosal resection (EMR) defect predict post EMR bleeding? L Desomer1, DJ Tate1, F Bahin1, B Holt1, K Byth2 and MJ Bourke1,3 1Department of Gastroenterology and Hepatology, Westmead Hospital, Sydney, NSW, Australia; 2Research and Education Network, Westmead Hospital and Sydney University, Sydney, NSW, Australia; 3Westmead Clinical School, University of Sydney, Sydney, NSW, Australia Introduction: Clinically significant post EMR bleeding (CSPEB) is the most frequent serious complication after wide-field endoscopic mucosal resection (EMR) of laterally spreading lesions ≥ 20 mm (LSLs). 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. Methods: A prospective study of LSLs ≥ 20 mm referred for EMR at a single tertiary referral center was performed. Data collection included patient and lesion characteristics, defect features including submucosal vessels, submucosal hemorrhage, fibrosis, fat and exposed muscle, and the rate of 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 were taken forward for binomial logistic regression modelling. Results: Over 60 months, to April 2016, 576 lesions (51.5% located proximal to the transverse colon) in 576 patients (mean age 66.8 years, 50.7% male) were eligible for analysis. The frequency of CSPEB was 35/576 (6.1%). Defect features and statistical analysis are outlined in Table 1. No features of the PED, including number of visible vessels (median 3, IQR 0.25–7.75, P = 0.308), diameter of largest vessel (14/165 > 0.5 mm (8.5%), P = 0.853), and herniation of vessels (12/146 (8.2%), P = 0.620), were significantly associated with CSPEB at univariate analysis and were therefore not used in the multivariate model. However, vessels were more often seen in the left colon (76% vs 63.9%, P = 0.002), and these were significantly larger (19.6% vs 9.2% ≥ 1 mm, P = 0.008), more numerous (median 4 (IQR 2–6) vs 3 (IQR 1.25–4.1) vessels, P < 0.001) and showed herniation more often (32.6% vs 21.7%, P = 0.005). On multivariate analysis CSPEB was associated with use of aspirin within 7 days (RR 3.31, 95% CI 1.5–7.6, P = 0.01), size ≥ 40 mm (RR 3.54, 95% CI 1.6–7.8, P = 0.001), and right colon location (RR 2.508, 95% CI 1.177–5.344, P = 0.017). Conclusion: Number, size or presence of herniation of vessels within the PED does not predict CSPEB. Vessel number, size, and herniation is significantly greater in the left colon; however, the bleeding rate is less. Other features within the PED also do not predict CSPEB, including submucosal hemorrhage and intraprocedural bleeding. The visible vessels within the PED should not be considered a therapeutic target for prevention of CSPEB and should not be used for risk stratification of CSPEB. Lesion location and size are the dominant risk factors for CSPEB. Table 1 Defect and procedural features in patients with clinically significant post EMR bleeding (CSPEB) as compared with patients with no CSPEB Lesions (N = 576) Univariate P Multivariate P Defect features Presence of visible vessels (%) 402 (69.8) .314 / °Presence of arteries 34 (5.9) .713 °Presence of veins 394 (68.4) .141 Number of visible vessels (median, IQR) 3 (0.25–7.75) .187 / Herniation of vessels (%) 12/146 (32.6) .620 / Estimated diameter of largest vessel (in comparison to snare wire) (%) .446 / ° ≤1 mm 21 (60.0) ° >1 mm 6 (17.1) Intraprocedural bleeding (%) 182 (31.6) .429 / Submucosal hemorrhage 58 (10.1) .787 / Procedural Aspirin use within 7 days (%) 53 (9.2) .009 .005 Lesion size ≥ 40 mm (%) 26 (4.5) .001 .002 Lesion location in the right colon (%) 23 (7.8) .051 .017 Reference 1Burgess NG, Metz AJ, Williams SJ et al. Risk factors for intraprocedural and clinically significant delayed bleeding after wide-field endoscopic mucosal resection of large colonic lesions. Clin Gastroenterol Hepatol 2014 Apr; 12 (4): 651– 661. Defining the optimal management of large colorectal laterally spreading lesions: a cost effectiveness analysis comparing wide-field endoscopic mucosal resection and endoscopic submucosal dissection FF Bahin1,2, SJ Heitman1,4, KN Rasouli1, H Mahajan1,3, D McLeod1,3, EYT Lee1, SJ Williams1 and MJ Bourke1,2 1Department of Gastroenterology and Hepatology, Westmead Hospital; 2Westmead Clinical School, University of Sydney, Sydney, New South Wales, Australia; 3Department of Anatomical Pathology, Westmead Hospital, Sydney, New South Wales, Australia; 4Department of Medicine, University of Calgary, Calgary, Alberta, Canada Introduction and Aims: Endoscopic submucosal dissection (ESD) of large sessile and laterally spreading colorectal lesions (LSL) is advocated as a superior treatment to wide-field endoscopic mucosal resection (WF-EMR) due to reduced adenoma recurrence and potential cure of low risk submucosal invasive cancer (LR-SMIC). The cost-effectiveness of ESD for treatment of LSL has not been determined. The aim was to perform an economic evaluation comparing potential ESD strategies to standard WF-EMR for the treatment of colorectal LSL. Methods: A hypothetical decision tree model was applied to a prospective observational cohort of WF-EMR for LSL ≥ 20 mm at a tertiary referral center. An incremental cost-effectiveness analysis was performed comparing the following strategies: WF-EMR, universal ESD, and selective ESD for lesions suspicious for SMIC (WF-EMR for remainder). Lesion assessment for suspicion of SMIC was based on the presence of non-granular topography, Paris 0-IIc morphology, Kudo pit pattern V, and non-lifting. LR-SMIC was defined as well-differentiated, vertical submucosal involvement <1000 microns and absence of lymphovascular invasion or tumor budding. Inflation adjusted cost estimates were derived from the Australian Refined Diagnosis Related Group codes (Version 7.0). Effectiveness data and procedural risks were supplemented with systematic review literature. Outcomes included costs, the number and types of surgeries performed and incremental cost per surgery avoided. Results: Over 88 months, 1317 patients with 1375 lesions (mean size 37.1 mm, 52.4% right colon) were enrolled (Table 1). The prevalence of SMIC and LR-SMIC was 8.3% and 3.7%, respectively. Lesion assessment for SMIC had a sensitivity and specificity of 21.1% and 98.9%, respectively. Selective ESD cost slightly less and prevented 18 surgeries compared to WF-EMR, but only 28 ESDs would be performed per 1000 patients treated. Universal ESD would prevent an additional 27 surgeries compared to selective ESD, but at an incremental cost of $132,759 per surgery avoided. Expanding the criteria for ESD within the selective ESD strategy to mirror contemporary Japanese guidelines resulted in improved performance of endoscopic lesion assessment for SMIC (sensitivity and specificity of 65.1% and 67.6%, respectively). Selective ESD by Japanese criteria would prevent 34 surgeries through 351 ESD procedures performed at an incremental cost of $29,701 compared to WF-EMR. Conclusions: Whilst a selective ESD strategy was cost-effective compared to WF-EMR, it is predominantly a WF-EMR strategy with ESD applicable to a small proportion of LR-SMIC in an enriched tertiary center referral population. Expanded ESD criteria result in additional surgeries avoided, but most ESD procedures are completed on lesions without SMIC. A universal ESD strategy cannot be justified. Due to the low absolute number deriving a clinically meaningful benefit even from selective ESD, its introduction requires careful consideration given a significant learning curve, lack of training infrastructure, and high costs. WF-EMR remains a relatively cost-effe