Inter-institutional variations regarding Barrett's esophagus diagnosis.

backgroundBarrett's esophagus (BE) is a known precursor for development of esophageal adenocarcinoma and surveillance of affected patients is necessary when cancer progression risk is considered to be high. However, the accuracy of BE diagnosis may not be homogenous among institutions with endoscopy units. We investigated inter-institutional variability by examining the accuracy of endoscopic diagnosis of BE at 4 different hospitals.

methodsThe accuracy of BE diagnosis at the 4 hospitals was retrospectively reviewed by 6 expert endoscopists, who independently reviewed endoscopic images of approximately 500 consecutive patients examined at each hospital without information regarding the diagnosis by the on-site endoscopists. When the expert reviewers made different diagnosis, a final diagnosis was made by consensus. That was then compared with the diagnosis of the attending endoscopists at each hospital and their concordance was calculated separately for each endoscopy unit. In addition, the relationship between diagnostic accuracy and endoscopic experience was assessed.

resultsThe prevalence of BE diagnosis by the on-site endoscopists was not homogenous and varied widely (17.2-96.8%). In 1 hospital, over-diagnosis was the cause of dissimilarity, while under-diagnosis was the cause in two hospitals. Diagnostic accuracy by the attending endoscopists in all 4 hospitals ranged from 44.6 to 83.1% (P < 0.05). There was no significant association between diagnostic accuracy and endoscopic experience or board licensing status of the on-site endoscopists.

conclusionDiagnostic accuracy for BE was not homogenous among 4 hospitals, and problems related to over- and under-diagnosis should be considered.