AGA Abstracts
Medical, Bedford, MA). Patients were included if they underwent endoscopy to screen for BE and had no prior BE diagnosis. Results: Of the 565 patients in this registry, 84 (15%) met the inclusion criteria. The average age was 57.5 years (range 22-85) and 48 patients (57%) were female. Nearly half (48%) had visible salmon-colored mucosa suggestive of Barrett's esophagus, while another 11% had an irregular Z-line. Erosive esophagitis was appreciated in 21% of the cases. In 15 (30%) of the patients with salmon-colored mucosa or irregular z-line present, who also had random biopsies, findings on VLE aided in targeting additional forceps biopsies. Four of these 15 patients (27%) had more advanced disease on VLE-targeted biopsies, including cases with low grade dysplasia and intramucosal carcinoma missed with random biopsies. VLE-targeted biopsies confirmed the grade of pathology seen on random biopsies in an additional 11 patients (73%). VLE showed no findings suspicious for BE or dysplasia in 33 procedures (39%), including 8 cases where salmon-colored mucosa was present. In the 17 cases where random tissue sampling was performed despite these results, 16 (94%) were negative for BE. Conclusion: Suspicious findings on VLE during endoscopic screening increase the likelihood of making a new diagnosis of BE or associated dysplasia, even when no visual abnormalities are present on white light endoscopy. Additional cases of metaplasia, dysplasia and even carcinoma are detected, which can lead to significant changes in patient management. Negative VLE scans also serve a clear benefit in these cases, with a high negative predictive value (94%). Further improvements in targeting tissue for sampling, such as a laser marking system, will only increase the yield of VLE-guided biopsies performed while screening patients for BE. Patient Demographics
Sa1265 Guideline Adherence Evaluation through Barrett's Esophagus Reporting Shelley Shi, Jitin Makker, Conklin Jeffrey, Eric Esrailian, Daniel Hommes, Christine Yu Background Routine endoscopic surveillance is recommended for patients with Barrett's Esophagus (BE) to detect early progression to esophageal adenocarcinoma (EAC), a cancer with a survival rate of 13%. Adherence to endoscopic surveillance guidelines has wide variability, even within the same institution. Reasons for such variability are not clear. AIM To assess adherence to BE surveillance guidelines and identify variations in reporting patterns at an academic tertiary care referral center. Methods This study was done in the Division of Digestive Diseases at the University of California Los Angeles as part of the GI Quality Program. Cases of intestinal metaplasia were identified through natural language processing of a pathology database. Those with missing reports, gastric intestinal metaplasia or no endoscopic suspicion for BE on retrospective chart review were excluded. Data elements included (1) Prague criteria measurements of circumference (C), maximum (M), and gastric folds (G). Procedures were categorized as using Complete Prague if both C and M could be calculated and Incomplete Prague if any other combination or single value of C, M and G were reported. (2) Seattle protocol biopsies by quantity of 4 or greater every 2 centimeters for non-dysplastic BE and technique of four-quadrant biopsy (3) Appropriateness of surveillance intervals defined as 1 and 3-5 years for new diagnoses and surveillance cases, respectively. Results Of 838 cases identified through natural language processing, 428 were excluded and 410 procedures from a total of 238 patients were reviewed. Measurement of the BE segment occurred in 97.9% (334/341) of procedures, but only 27.2% (91/334) of those cases reported Complete Prague. As compared to 19.7% (23/117) of new diagnoses, Complete Prague was used in 36.0% (68/189) of surveillance procedures (p=0.004). Biopsies were taken in 98.2% (335/341) of procedures. Seattle Protocol by quantity was used in 72.5% (148/204) of procedures and by technique was used in 41.6% (47/113). Surveillance endoscopy intervals were documented in 40.9% (167/408) of procedure reports and were correct in 84.0% (21/25) new diagnoses and 77.3% (34/44) surveillance procedures. Conclusion Our data suggests that endoscopists recognize that BE measurement and biopsies are important, but adherence rates to systematic methods and standardization of reporting need improvement. Further studies identifying reasons for low adherence are warranted and results could serve as areas for quality improvement projects.
AGA Abstracts
X : 10052$CH01 03-28-16 00:57:27 PDFd : 10052B : e
Sa1266 Social Class Distribution of Patients With Barrett's Esophagus and Esophageal Adenocarcinoma Santanu Bhattacharjee, Christine Caygill, Piers Gatenby, Anthony Watson Introduction - Many Western Nations, including the UK, are experiencing rapidly increasing incidence of esophageal adenocarcinoma (EAC) over the last three to four decades. Barrett's esophagus (BE) is the only known precursor condition for EAC. Surveillance of BE provides an opportunity to diagnose the disease early, but identifying ‘at risk' populations would appear to be a logical step in targeting surveillance to those who are most likely to benefit. Anecdotally, the increased incidence of EAC appears to have been accompanied by an increase in affected social class I and II, but there has not been a study to confirm this. Aim - To identify a trend in diagnosis of BE and EAC in a cohort of patients based on their occupation and social class. Methods - Data of patients from two centres in the UK, who have consented and registered with the United Kingdom Barrett's Oesophagus Registry (UKBOR), were used for the purpose of this study. These centres have an active BE surveillance programme running for at least 3 decades. Data from 2896 patients formed the basis of the study. Data on patient's occupation were obtained from hospital records, or from death certificates where available. The most widely used measure of social class in the UK is the ‘Registrar General's Classification' which assigns a social class category against individual occupation. Social class based on occupation has five classes, I through to V, with class III further subdivided into III N (non-manual) and III m (manual). Social class I are the professionals and higher managers, II and III N are junior managers and supervisors and those in III M and IV are manual occupations and V are unskilled workers. A Chi-squared goodness of fit test was used to test for significance of the association of social class in both BE and EAC, with results less than 0.05 taken as significant. Results - Table 1 shows the observed cases of BE and EAC in each of the social classes. 1751 patients with BE and 126 with EAC had information on occupation. Of the BE group, 1715 had usable information on occupation (Table 1). The observed distribution of BE across social class does not seem to be very different from that expected. However, the observed number of EAC (51) is higher than expected (36.4); O/E 1.4 and in social class III M, whereas in social class V, the observed EAC (4) is lower than expected (11.4); O/E 0.4 Conclusion - Our study shows that more EAC is observed in BE patients in social class III M than expected whereas fewer are observed in social class V than expected.
S-262
Page 262