Is size, morphology, site, and access scoring system consistent between endoscopists? Interobserver and intraobserver...

This study aimed to evaluate interobserver and intraobserver agreement in Size/Morphology/Site/Access (SMSA) scoring among practicing endoscopists with varying levels of experience.

A total of 102 fully independent endoscopists participated in the study. Ten short video clips of colonic polyps of varying size and complexity were recorded using Olympus 290 colonoscopes and included white light, near focus, narrow band imaging, and chromoendoscopy where applicable. These videos were embedded in an online questionnaire. Endoscopists were asked to assign SMSA scores based on three variables-size, morphology, and access-with the site provided for standardization. A subgroup of five participants repeated the assessment after 2 to 3 weeks to evaluate intraobserver consistency. Data were analyzed using Fleiss' kappa via SPSS (v26), and Kappa interpretation followed the Landis and Koch classification.

Overall interobserver agreement for SMSA level across all participants was fair (κ = 0.346). Among individual parameters, morphology had the highest agreement (κ = 0.505, moderate), followed by access (κ = 0.408, moderate) and size (κ = 0.241, fair). Subgroup analysis of experienced endoscopists (> 1000 lifetime colonoscopies) yielded slightly improved kappa values, with morphology still demonstrating the highest consistency. Intraobserver agreement showed moderate to almost perfect reliability for size (κ = 0.444 to 1.000) and moderate to substantial agreement for SMSA level (κ = 0.429 to 0.846).

Morphology was the most consistently scored parameter. Although the SMSA tool remains valuable, efforts such as standardized training and simplification of variable categories may be warranted to improve interobserver consistency and enhance clinical-utility.