Messmann, Helmut

INTRODUCTION: The performance of a high quality esophagogastroduodenoscopy (EGD) is dependent on the mucosal cleanliness. Recently, the Polprep: Effective Assessment of Cleanliness in EGD (PEACE) scale was created to assess the degree of mucosal cleanliness during EGD. The aim of this study was to validate this scoring system in a cohort of international endoscopists. METHODS: In total, 39 EGD videos, with different degrees of mucosal cleanliness were retrieved from a previously conducted prospective trial. All experts rated the cleanliness of the mucosa on each video using the PEACE scale. To evaluate agreement of all scores (0–3), intraclass correlation coefficient 2.1 was used. The agreement on adequate (scores 2 and 3) and inadequate (scores 0 and 1) cleanliness was assessed using kappa values. RESULTS: Videos evaluating esophagus, stomach, and duodenum cleanliness were reviewed by 16 endoscopists. The PEACE scores demonstrated good agreement (intraclass correlation coefficient 0.82, 95% CI 0.75–0.89), especially for esophagus (0.84; 95% CI 0.71–0.95) and stomach (0.81; 95% CI 0.69–0.91), while agreement was moderate for the duodenum (0.69; 95% CI 0.51–0.87). The agreement was similar between Eastern (0.86; 95% CI 0.79–0.92) and Western experts (0.80; 95% CI 0.72–0.88). Similarly, agreement regarding adequate cleanliness was comparable between Eastern (0.70; 95% CI 0.55–0.85) and Western (0.74; 95% CI 0.64–0.84) endoscopists being overall 0.75 (95% CI 0.65–0.85). DISCUSSION: The PEACE scoring system is a simple and reliable scale to assess the cleanliness during EGD. The score is now validated among international experts with high concordance, justifying its use in clinical practice.

Background Piecemeal endoscopic mucosal resection (EMR) is an acceptable technique for T1a esophageal adenocarcinoma, but en bloc R0 excision is advocated for T1b disease as it may offer a potential cure and mitigate recurrence. Thus, distinguishing between T1a and T1b disease is imperative under current treatment paradigms. We investigated whether expert Barrett’s endoscopists could make this distinction based on optical evaluation. Methods Endoscopic images of histologically confirmed high grade dysplasia (HGD), T1a, and T1b disease (20 sets for each) were compiled from consecutive patients at a single institution. Each set contained four images including an overview, a close-up in high definition white light, a near-focus magnification image, and a narrow-band image. Experts predicted the histology for each set. Results 19 experts from 8 countries (Australia, USA, Italy, Netherlands, Germany, Canada, Belgium, and Portugal) participated. The majority had been practicing for > 20 years, with a median (interquartile range) annual case volume of 50 (18–75) for Barrett’s EMR and 25 (10–45) for Barrett’s endoscopic submucosal dissection. Esophageal adenocarcinoma (T1a/b) could be distinguished from HGD with a pooled sensitivity of 89.1 % (95 %CI 84.7–93.4). T1b adenocarcinoma could be predicted with a pooled sensitivity of 43.8 % (95 %CI 29.9–57.7). Fleiss’ kappa was 0.421 (95 %CI 0.399–0.442; P < 0.001), indicating fair-to-moderate agreement. Conclusions Expert Barrett’s endoscopists could reliably differentiate T1a/T1b esophageal adenocarcinoma from HGD. Despite fair-to-moderate agreement for T staging, T1b disease could not be reliably distinguished from T1a disease. This may impact clinical decision making and selection of endoscopic techniques.

ObjectivesThere have been significant advances in the management of large (≥20 mm) laterally spreading tumors (LSTs) or nonpedunculated colorectal polyps; however, there is a lack of clear consensus on the management of these lesions with significant geographic variability especially between Eastern and Western paradigms. We aimed to provide an international consensus to better guide management and attempt to homogenize practices.MethodsTwo experts in interventional endoscopy spearheaded an evidence‐based Delphi study on behalf of the World Endoscopy Organization Colorectal Cancer Screening Committee. A steering committee comprising six members devised 51 statements, and 43 experts from 18 countries on six continents participated in a three‐round voting process. The Grading of Recommendations, Assessment, Development and Evaluations tool was used to assess evidence quality and recommendation strength. Consensus was defined as ≥80% agreement (strongly agree or agree) on a 5‐point Likert scale.ResultsForty‐two statements reached consensus after three rounds of voting. Recommendations included: three statements on training and competency; 10 statements on preresection evaluation, including optical diagnosis, classification, and staging of LSTs; 14 statements on endoscopic resection indications and technique, including statements on en bloc and piecemeal resection decision‐making; seven statements on postresection evaluation; and eight statements on postresection care.ConclusionsAn international expert consensus based on the current available evidence has been developed to guide the evaluation, resection, and follow‐up of LSTs. This may provide guiding principles for the global management of these lesions and standardize current practices.

Abstract Background This study evaluated the effect of an artificial intelligence (AI)-based clinical decision support system on the performance and diagnostic confidence of endoscopists in their assessment of Barrett’s esophagus (BE). Methods 96 standardized endoscopy videos were assessed by 22 endoscopists with varying degrees of BE experience from 12 centers. Assessment was randomized into two video sets: group A (review first without AI and second with AI) and group B (review first with AI and second without AI). Endoscopists were required to evaluate each video for the presence of Barrett’s esophagus-related neoplasia (BERN) and then decide on a spot for a targeted biopsy. After the second assessment, they were allowed to change their clinical decision and confidence level. Results AI had a stand-alone sensitivity, specificity, and accuracy of 92.2%, 68.9%, and 81.3%, respectively. Without AI, BE experts had an overall sensitivity, specificity, and accuracy of 83.3%, 58.1%, and 71.5%, respectively. With AI, BE nonexperts showed a significant improvement in sensitivity and specificity when videos were assessed a second time with AI (sensitivity 69.8% [95%CI 65.2%–74.2%] to 78.0% [95%CI 74.0%–82.0%]; specificity 67.3% [95%CI 62.5%–72.2%] to 72.7% [95%CI 68.2%–77.3%]). In addition, the diagnostic confidence of BE nonexperts improved significantly with AI. Conclusion BE nonexperts benefitted significantly from additional AI. BE experts and nonexperts remained significantly below the stand-alone performance of AI, suggesting that there may be other factors influencing endoscopists’ decisions to follow or discard AI advice.