Influence of reviewers' clinical backgrounds on interpretation of confocal laser endomicroscopy findings.

BACKGROUND AND STUDY AIMS: Substantial differences in endoscopic strategy for gastric cancer exist between Western and Eastern countries, owing to clinicoepidemiological diversity, including differences in the prevalence of gastric cancer. This international multicenter study involved German and Japanese institutions and aimed to evaluate the influence of reviewers' clinical backgrounds on interpreting probe-based confocal laser endomicroscopy (pCLE) findings for diagnosis of superficial gastric lesions. PATIENTS AND METHODS: 39 reviewers answered questionnaires about their clinical background and then reviewed 30 sets of white light endoscopy (WLE) and pCLE video clips via an online questionnaire. For each set of clips, reviewers were asked to classify lesions as neoplastic or non-neoplastic. RESULTS of video reviews were compared with the final histopathological diagnosis for each lesion. The accuracy of diagnosis based on WLE + pCLE was compared with that based on WLE alone for each aspect of clinical background.
RESULTS: The overall accuracy of diagnosis based on WLE + pCLE was higher than that based on WLE alone (73.93 % vs. 65.64 %, P = .0002). Outcomes of expert gastroenterologists were better than those of pathologists (P = .038 for WLE, P = .002 for WLE + pCLE) and outcomes of reviewers at Japanese institutions were better than those of reviewers at German institutions (P = .001 for WLE, P < .001 for WLE + pCLE).
CONCLUSIONS: Reviewers from Japanese institutions and expert gastroenterologists performed well in the pCLE interpretation. Substantial experience in conventional endoscopy is important for interpreting pCLE images for the diagnosis of gastric cancer.University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR; www.umin.ac.jp/ctr/index.htm) number UMIN 000013437). © Georg Thieme Verlag KG Stuttgart · New York.