Digital Cholangioscopic Interpretation: When North Meets the South.

Area of Science:

• Gastroenterology
• Endoscopy
• Biliary Tract Diseases

Background:

• Digital single-operator cholangioscopy (DSOC) offers high-definition imaging of the biliary tree.
• Two criteria sets, Monaco and Carlos Robles-Medranda (CRM), evaluate indeterminate biliary strictures.
• Assessing interrater agreement (IA) for these criteria is crucial for clinical application.

Purpose of the Study:

• To evaluate the interrater agreement (IA) of DSOC interpretation for indeterminate biliary strictures.
• To compare the IA of the Monaco criteria versus the CRM criteria.
• To assess the impact of these criteria on diagnostic accuracy.

Main Methods:

• 15 interventional endoscopists scored 40 DSOC videos using Monaco and CRM criteria.
• Scoring included specific features for each criteria set.
• Intraclass correlation (ICC) analysis determined IA for criteria and final diagnosis.

Main Results:

• IA ranged from poor to excellent for both Monaco (0.1-0.76) and CRM (0.1-0.62) criteria.
• Monaco criteria showed excellent IA for lesion and papillary projections; CRM criteria had good IA for villous, flat, and honeycomb patterns.
• Diagnostic IA was good for Monaco (0.65) and fair for CRM (0.58), with overall accuracy of 61% and 57%, respectively.

Conclusions:

• Interrater agreement and accuracy for DSOC interpretation of biliary strictures are similar between Monaco and CRM criteria.
• Some criteria within both sets exhibit poor IA, impacting overall diagnostic accuracy.
• Formal training and criteria refinement are necessary to enhance diagnostic performance.