ERCP outcomes: Defining the operators, experience, and environments

Research output: Contribution to journalArticlepeer-review

40 Scopus citations


A potpourri of associations has been identified for the common ERCP outcomes of success/failure, overall complication rates, and the specific complications of pancreatitis, hemorrhage, and cholangitis. Generally consistent results demonstrate improved success rates for endoscopists with cumulative experience of 150 to 200 procedures or case volumes of more than 40 to 50 sphincterotomies per year. Whether these 2 indicators of experience track together is not known. Similarly, prospective evaluation of fellows in training suggests that over 200 procedures are required to reach an acceptable level of competence in the performance of essential maneuvers. Global or specific complications have been less consistently associated with experience, although most studies suggest an association, and the major prospective studies concur in this regard. Pancreatitis appears to be less associated with pure measures of experience than with various indices of difficulty. Difficulty, which also correlates with other negative outcomes, probably reflects a blend of patient and endoscopist elements. The practice environment appears to influence outcomes predominantly through case mix and volumes. From the little that is known about training environments it appears that the case mix and the responsible instructor influence outcome more than the participation of a fellow. The influence of the operator, his or her experience, and the environment on outcomes beyond success rates and complications has not been studied extensively.

Original languageEnglish (US)
Pages (from-to)953-958
Number of pages6
JournalGastrointestinal endoscopy
Issue number7
StatePublished - Jun 2002

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Gastroenterology


Dive into the research topics of 'ERCP outcomes: Defining the operators, experience, and environments'. Together they form a unique fingerprint.

Cite this