This study tests the impact of narrow band imaging (NBI) on endoscopists' accurate detection of visible lesions and dysplasia in patients with Barrett's esophagus, as well as the effect of NBI on the choice of primary treatment modality among endoscopists performing endoscopic eradication therapy (EET).

Barrett's esophagus (BE), is a condition whereby normal esophageal squamous epithelium is replaced by metaplastic columnar epithelium, predisposing patients to esophageal adenocarcinoma (EAC). It is estimated that about 5.6% of adults in the United States have BE with risk factors including: long standing gastroesophageal reflux disease, tobacco use, male gender, central obesity, and age over 50 years. EAC is believed to progress in a step-wise pattern with the following order of non-dysplastic BE, low-grade dysplasia (LGD), and high-grade dysplasia (HGD). Each carries a risk of progression to EAC, differing by degree of dysplasia: 0.2-0.5%, 0.7%, and 7% per year, respectively. Given this association, it is common practice to perform endoscopic surveillance with biopsies in patients with BE. Endoscopic surveillance has been shown to detect EAC at earlier stages and improve survival in asymptomatic presentations. As dysplasia in BE may not always be seen as a distinct lesion, surveillance programs entail use of the Seattle Protocol, a systematic four-quadrant biopsy technique obtained at 1 to 2 cm increments. Current guidelines recommend the use of high-definition white light endoscopy (HD-WLE) as it is superior over standard-definition in regards to improved targeted detection of dysplasia.

Advanced endoscopic imaging techniques have been proposed to improve dysplasia detection with preference for electronic chromoendoscopy, specifically narrow band imaging (NBI), as it does not require dye sprays. NBI has been shown to be more accurate in detecting intestinal metaplasia and HGD. HGD is more often detected in areas with subtle mucosal and vascular abnormalities, which may be more difficult to see on HD-WLE alone. However, subtle lesions may go undetected, as NBI is not routinely used in the community with a recent survey showing only about a third of practicing gastroenterologists use advanced endoscopic imaging. The widespread use of NBI has been potentially limited by a perceived complexity of interpretation and lack of standardization. Recently, Sharma et al introduced the BING criteria - a standardized classification system to detect dysplasia and EAC with NBI. While a few studies have demonstrated no significant difference in detection of dysplasia or neoplasia between HD-WLE and NBI, they have had some limitations. The studies occurred prior to the BING classification system, and participants were limited to a few expert tertiary medical centers.

The current standard of care for visible lesions identified by HD-WLE (nodules, ulcers, erosions, or plaques) is endoscopic mucosal resection (EMR). Endoscopic recognition and appropriate resection of visible lesions is essential for optimal patient outcomes. Staging EMR is critical as it allows for histopathological "upgrading" or "downgrading" of dysplasia and ultimately is the best tool for identifying and treating early EAC. Despite the importance of EMR for BE-AN, survey data suggests it is underutilized in practice with 39% of academic endoscopists and 13% of community-based endoscopists performing EMR. While many endoscopists utilize NBI to assist in identification of visible lesions, the resection of areas deemed "abnormal" by NBI alone is not widely accepted. Moreover, endoscopists at community hospitals detect neoplastic lesions at significantly lower rates than at BE expert centers.

Given these data, routine use of NBI prior to EET could significantly impact treatment decisions among all endoscopists with highly accurate rates of dysplasia detection. It's been shown that NBI increases the accuracy and positive predictive value of predicting histology than if HD-WLE is used alone. This study is limited by the use of still-images, which does not accurately reproduce live images seen during endoscopy. Nevertheless, the current standard of using HD-WLE for identification of visible lesions likely underestimates the presence of dysplastic areas in patients undergoing Endoscopic Eradication Therapy (EET) for BE-AN. We hypothesize that the routine use of narrow band imaging (NBI) for identification of visible lesions will improve dysplasia detection and have a significant effect on the choice of primary treatment modality among endoscopists performing EET. To this end, we propose a video-based study to evaluate the impact of NBI on choice of treatment modality during EET.

Inclusion Criteria:

• Endoscopists familiar with EET.

Exclusion Criteria:

• Endoscopists not familiar with EET or non-endoscopists.
• Special populations will not be included in this study.

Endoscopist sub-groups include academic endoscopists performing greater than 50 EET procedures per year, community endoscopists performing greater than 50 EET procedures per year, 25 academic gastroenterologists performing greater than 500 upper endoscopies per year not including EET, and 25 community gastroenterologists performing greater than 500 upper endoscopies per year not including EET.