Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Volumetric laser endomicroscopy in Barrett’s esophagus: interobserver agreement for interpretation of Barrett’s esophagus and associated neoplasia among high-frequency users
Background and Aims Targeting neoplasia in Barrett’s esophagus (BE) is challenging. Volumetric laser endomicroscopy (VLE) is a new imaging technique that allows for real time cross-sectional microstructure imaging that can detect BE neoplasia. The interobserver agreement among users in practice is unknown. Methods Eight high-volume users of VLE from different academic centers in the United States evaluated 120 stored VLE images blinded to the endoscopic and clinical findings. There were 30 images for each tissue type: gastric cardia, esophageal squamous mucosa, nonneoplastic BE, and neoplastic BE. Each image with BE had corresponding histology confirming the tissue diagnosis. Each normal esophagus and gastric cardia had matching endoscopic images confirming normal mucosa. These were considered the criterion standard. Respondents were asked to classify the images into 1 of each category. Agreement among the users was measured. Results The overall agreement among users was almost perfect (kappa = 0.81; 95% confidence interval [CI], 0.79-0.83). For esophageal squamous and gastric cardia, the agreement was almost perfect (kappa 0.95 and 0.86, respectively [95% CI, 0.92-0.98 and 0.83-0.89]). For nonneoplastic BE and neoplastic BE, the agreement was strong (kappa 0.66 [95% CI, 0.63-0.69] and 0.79 [95% CI, 0.75-0.82], respectively). When compared with the criterion standard, the median accuracy for identifying normal squamous mucosa, normal gastric mucosa, nonneoplastic BE, neoplastic BE, and all tissue types was 99% (95% CI, 98%-100%), 97% (95% CI, 95%-99%), 93% (95% CI, 88%-98%), 95% (95% CI, 91%-99%), and 96% (95% CI, 94%-99%), respectively. Conclusions VLE has a high level of agreement and accuracy among high-volume users.