BACKGROUND: Validated, accepted grading tools for preprocedural complexity assessment in ERCP are lacking. We therefore created a grading system for ERCP based on the classification used by the American Society for Gastrointestinal Endoscopy (ASGE).
METHODS: Data on ERCP adverse events (AE) and success were collected in a multicenter, prospective uncontrolled study. Multiple logistic regressions were applied to success and AEs in accordance with the ASGE classification. Each procedure suggested by ASGE was tested against different outcomes. Results were used to create a score and were evaluated in a control cohort.
RESULTS: 16,327 ERCPs were documented in 27 centers. Analysis of ASGE categorization (10,904 cases) showed that this model fails to adequately predict parameters of complexity; only for cardiopulmonary AEs and perforation was no significant variance evident. Depending on the specific clinical circumstances, probability of success of the intervention sometimes varied significantly in risk, implying a twofold score, one part for probability of success and one for risk. A split score with three levels each was designed and tested in a validation cohort (5,423 procedures). Achieving therapeutic targets / post-ERCP pancreatitis could be correctly predicted in 87.0%/95.3%.
CONCLUSIONS: Grading ERCP success and AEs have to be considered independently. Onefold grading systems appear incomplete and unable to provide an adequate classification of severity. SASE (Success and Adverse Event Score in Endoscopic Retrograde Cholangiopancreatography) was created to incorporate these findings. Showing high predictive value, this score could be a potent tool for planning ERCP and training in endoscopy.