Introduction ERCP is a technically demanding procedure, with significant risk of complications. Cannulation success is widely regarded as a key performance indicator of high- quality ERCP practice (1). Multiple factors affecting the complexity of a complete ERCP procedure have been suggested as a way of adding relevance to success rates and complication incidence. We propose a novel concept where complexity stratification is used to predict outcome based on ampullary characteristics.
Methods 200 ERCPs performed on a virgin ampulla were prospectively recorded. We classified ampulla as non-prominent, prominent or distorted by tumour. Cannulation method and number of ampullary contacts were recorded. Cannulation success and incidence of complications were the primary outcome measures.
Results The most common indication was biliary duct stones (54%). Ampullae were classified as non-prominent in 107 cases, prominent in 78 and involving tumour in 15.
Overall deep cannulation was successful in 189 cases (94%), with significant variation between ampullary groups. Non-prominent and prominent ampullae were more likely to be cannulated successfully, (95.3%, 94.9% respectively), than those involving tumour (80%) (p = 0.004).
Fewest contacts prior to cannulation were made on the non-prominent ampullae and most on those involving tumour (p < 0.001). Needleknife assisted cannulation was used most frequently on ampullae involving tumours and least often on non-prominent ampullae (p = 0.044).
The presence of a peri-ampullary diverticulum or a covering mucosal fold, did not reduce cannulation success.
15 patients had complications (7.5%) - pancreatitis (10), perforation (3), infection (1) and bleeding (1). 12 complications occurred in the non-prominent group, with increased incidence of perforation, pancreatitis and bleeding (p = 0.04). Complications were more likely in younger patients (p = 0.03). Complication rate was not affected by patient gender, cannulation outcome, number of ampullary contacts or trainee involvement. Neither presence of diverticulae nor covering folds increased complication rate.
Conclusion The assessment of ampullary characteristics may prove to be a novel means of predicting cannulation difficulty and anticipating risk of complication. Non-prominent ampullae appear to be easier to cannulate, with fewer ampullary contacts and less use of needleknife fistulotomy, but complication rates appear highest. Statistical significance is limited by the sample size and low incidence of cannulation failure and complications, so further study is required. These findings may have implication for case selection in ERCP training, and may add validity to key outcome quality indicators in practice.
Reference 1 ASGE. Quality Indicators for ERCP. 2015.
Disclosure of Interest None Declared