Abstract
Background Assessment of mucosal visualization during esophagogastroduodenoscopy (EGD) can be improved with a standardized scoring system. To address this need, we created the Toronto Upper Gastrointestinal Cleaning Score (TUGCS).
Methods We developed the TUGCS using Delphi methodology, whereby an international group of endoscopy experts iteratively rated their agreement with proposed TUGCS items and anchors on a 5-point Likert scale. After each Delphi round, we analyzed responses and refined the TUGCS using an 80% agreement threshold for consensus. We used the intraclass correlation coefficient (ICC) to assess inter-rater and test-retest reliability. We assessed internal consistency with Cronbach's alpha and item-total and inter-item correlations with Pearson's correlation coefficient. We compared TUGCS ratings with an independent endoscopist's global rating of mucosal visualization using Spearman's rho.
Results We achieved consensus with 14 invited participants after three Delphi rounds. Inter-rater reliability was high at 0.79 (95%CI 0.64-0.88). Test-retest reliability was excellent at 0.83 (95%CI 0.77-0.87). Cronbach's alpha was 0.81, item-total correlation range was 0.52-0.70, and inter-item correlation range was 0.38-0.74. There was a positive correlation between TUGCS ratings and a global rating of visualization (r=0.41, P=0.002). TUGCS ratings for EGDs with global ratings of excellent were significantly higher than those for EGDs with global ratings of fair (P=0.01).
Conclusion The TUGCS had strong evidence of validity in the clinical setting. The international group of assessors, broad variety of EGD indications, and minimal assessor training improves the potential for dissemination.