Background: Although 140 colonoscopies is the recommended minimal requirement for gastroenterology fellows, it is unclear whether this minimum is a surrogate for competence.
Objective: To assess whether 140 colonoscopies is an adequate threshold to determine > or =90% colonoscopy performance independence.
Design: Retrospective analysis on a database constructed for quality control/improvement.
Setting: Gastroenterology fellowship training program at a veterans hospital.
Patients: Consecutive patients who underwent colonoscopy primarily for symptoms, previous polyps, or family history of cancer (a minority were performed for screening only) from April 2007 to September 2008. This study involved 11 gastroenterology fellows who performed 770 colonoscopies during 18 individual month-long rotations.
Intervention: Assessment of various procedure-related parameters.
Main outcome measurements: Determining when > or =90% independence in colonoscopy performance was reached.
Results: Total colonoscopy time, time to cecal intubation, withdrawal time, and independent completion rates all significantly improved when first and third years of training were compared (P < .001 for all comparisons). The adenoma detection rate did not change between years of training. Independent completion was achieved in > or =90% of cases for all fellows after 500 colonoscopies, whereas no fellow reached a > or =90% independent colonoscopy completion rate after 140 colonoscopies.
Limitations: Number of participants, single center.
Conclusions: Becoming a competent colonoscopist requires repeated practice. Our study suggests that, although there is variability between a trainee's ability to become colonoscopy independent, 500 colonoscopies are likely required to ensure reliable (> or =90%) independent completion rates. Competency requires more than a single parameter.