Introduction: GAGES (Global Assessment of Gastrointestinal Endoscopic Skills) was created to provide an objective measure of endoscopic skills to better assess training curricula. Pilot data have shown GAGES to meet high standards of reliability and validity. The aim of this multicenter study was to demonstrate the practicality and external validity of this instrument, while providing further evidence for reliability and validity. Methods: GAGES upper endoscopy (UE) and colonoscopy (C) are 5-point Likert rating scales developed by expert endoscopists. For UE, domains assessed were esophageal intubation, scope navigation, maintenance of a clear field, instrumentation (when biopsy, injection or polypectomy were performed), and overall quality of the examination; for C, these were scope navigation, strategies for scope advancement, clear field, instrumentation (when performed), and overall quality. GAGES was completed by the attending (A), a trained observer (O, in some cases), and in self-assessment (S) during clinical procedures to establish reliability (intraclass correlation coefficient, ICC) and internal consistency (Cronbach’s alpha). Instrumentation was evaluated when possible and correlated with total scores. Construct and external validity were examined by comparing scores between novice (NOV) and experienced (EXP) endoscopists in 9 internationally dispersed academic institutions (Student’s t-test). Correlations were calculated to compare GAGES-UE and C assessments in participants who had performed both. Results: 114 evaluations (68 NOV, 46 EXP) were completed (65 UE, 49 C); 36 participants performed both. Interrater reliability was excellent (A vs O) with an ICC of 0.92 (95% CI 0.82-0.97) for UE and 0.97 (0.92-0.99) for C. Interrater reliability between S and A assessments was also very good for UE (0.76; 0.64-0.85) and C (0.93; 0.88-0.96). Internal consistency of rating items was 0.89 for UE, and 0.95 for C. There were significant differences (mean (SD), maximum score=20) between NOV and EXP endoscopists for UE 15.8 (±3.7) vs 18.8 (±1.5) (p
Session: Podium Presentation
Program Number: S111
« Return to SAGES 2009 abstract archive
