BackgroundIn the early 2000s we introduced a prioritization model for referrals based on involvement of primary care physicians (PCPs) and specialists. AimsAssess the application of that model of prioritisation, comparing gastroscopies performed 8 years apart, with respect to priority level, appropriateness and relevant endoscopic findings (REFs). MethodsThe studies included 247 and 354 out-patients, who had undergone gastroscopy in 2006 and in 2014, respectively. To reduce interspecialists variability, both studies were performed by the same specialist as investigator. ResultsIn both years, most patients were assigned low-priority referral by PCPs (78.6% and 75.1% respectively). The agreement PCPs versus specialist on referral priority was moderate in 2006 (0.60, Landis–Koch scale 0.41–0.60) and high in 2014 (0.81, Landis–Koch scale 0.81–1.00). In both years we observed a similar rate of inappropriateness: 27.5% and 27.1%, respectively. Due to multiple logistic regression, the odds ratio (OR) for REF increased when: (i) very high-priority referral versus nopriority referral was indicated (8.813 OR, p = 0.0012), (ii) referral followed the guidelines (9.29 OR, p<0.0001), and (iii) agreement of priority occurred (1.911 OR, p = 0.0308). ConclusionsOur findings highlighted that the issues of low-priority referrals should be addressed in order to discontinue gastroscopy overusing and reduce related operational costs.