Consecutive patients with a solid pancreatic lesion were included. Endosonographer was submitted to a basic pancreatic cythpatology training programme at two institutitions. The patients were randomly allocated to the ROSE group - in which the number of needle passes required to obtain a sample suitable for cytopathologic categorization was established by the endosonographer's on-site evaluation - or to the non-ROSE group - in whom adequacy of the specimen was evaluated macroscopically and up to five needle passes could be performed, assuring sample adequacy. The gold standard was the final cytopathologist's diagnosis. The number of needle passes, procedure duration, specimen adequacy, diagnostic yield and adverse events rates were compared between groups and the performance measures of ROSE by endosonographer were determined.
