Visible Lesion Detection in Barrett Esophagus Remains Low Among Endoscopists

stomach, esophagus, gi system
Researchers assessed the visible lesion detection rate in Barrett esophagus among community and academic endoscopists.

High-definition white light endoscopy (HD-WLE) and narrow band imaging (NBI) may not meet the quality indicator threshold for visible lesion detection in Barrett esophagus (BE), according to study findings in Gastrointestinal Endoscopy.

Researchers assessed the rate at which endoscopists could correctly identify visible lesions using HD-WLE and NBI in patients with BE.

Of 50 endoscopists enrolled in the video-based prospective cohort study, researchers included 44 (22 academic endoscopists and 22 community endoscopists) from 5 academic centers: Northwestern University, University of Colorado, Washington University, University of Kansas, and University of California at Los Angeles. Of the study participants, 57.1% reported using NBI consistently, 76.2% had been practicing for at least 5 years, and 57.1% performed at least 5 endoscopies to screen for BE monthly.

For the video survey, endoscopists viewed 25 video clips (13 clips with visual lesions and 12 clips without) of patients who underwent endoscopic therapy. Study participants were also asked to identify visible lesions in the Barrett segment, to describe if the lesion was identified using HD-WLE or NBI, and to recommend a treatment strategy.

To establish a gold standard of visible lesion identification, 5 experts in BE from each of the academic institutions reviewed the 25 videos and completed the same video assessment as the 44 endoscopists, which included noting anatomical landmarks, visible lesion detection, and recommended treatment.

Researchers used Fleiss’ Kappa analysis to compare the 44 endoscopists’ answers to the gold standard. Correlation and intraclass correlation coefficients were calculated. Correlation was measured as: less than 0.2 (poor), 0.21 to 0.40 (fair), 0.41 to 0.60 (moderate), 0.61 to 0.80 (good), 0.81 to 1.00 (very good).

The primary outcome was visible lesion identification rate using HD-WLE and NBI.

Endoscopists correctly identified the top of gastric folds (57.7%) and top intestinal metaplasia (73.9%). When expanding the range of correct identification to within 1 and 2 cm, identification improved to 70.3% and 97.4% for the top of gastric folds and 87.6% and 99.5% for the top of intestinal metaplasia, respectively.

Study participants identified 72% of visible lesions using HD-WLE and 69% using NBI. While type of endoscopist practice showed no difference when using HD-WLE, it was significant with NBI, with community endoscopists showing higher rates of lesion detection compared with academic endoscopists (odds ratio, 2.07; CI, 1.2-3.5; P <.008).

Regardless of type of endoscopist, the detection rate of visual lesions did not meet quality indicator thresholds of 90%. For visible lesion detection, 18% and 5% of participants who used HD-WLE and NBI, respectively, reached the threshold.

Study limitations include the survey study design; response and reporting bias, which may affect generalizability; and the possibility that the BE experts may have missed visible lesions.

“Regardless of practice type or endoscopist experience the overall rate of visible lesion detection remains low,” the study authors wrote. “These findings highlight a significant need for further education and training to improve VL [visible lesion] detection during BE surveillance.”

Disclosure: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.


Beveridge CA, Mittal C, Muthusamy R, et. al. Identification of visible lesions during surveillance endoscopy for barrett’s esophagus: A video-based survey study. Gastrointest Endosc. Published online August 22, 2022. doi:10.1016/j.gie.2022.08.024