Introduction: The goals of colonoscopy are changing over time and it is important to regularly determine if endoscopists are achieving key performance indicators.
Methods: Data on key performance indicators were recorded independently by nursing staff for all colonoscopies performed during a 10-year period. The results were discussed at regular meetings and feedback given to endoscopists.
Results: Audit data was recorded for 67,570 procedures. The key performance indicators (time to caecum, withdrawal time, adjusted caecal intubation rate and polyp detection rate) all improved over the audit period (p<0.0001 for trend). For each endoscopist the mean withdrawal time was highly variable ranging from 3.1 mins (95%CI 3.0; 3.1) to 11.2 mins (11.0; 11.3). For each endoscopist mean polyp detection rate varied from 29% (CI 26, 31%) to 69% (CI 68, 70%). There was a significant correlation between mean withdrawal time and mean polyp detection rate for each endoscopist (r=0.42; p=0.03). The polyp detection rate improved from 29% in 1999 to 49% in 2010. The proportion of procedures with more than 2 polyps increased from 22% in 2001 to 33% in 2010. There was a significant association of patient discomfort with time to caecum and also to level of consciousness, p<0.0001. There was a significant decrease in the proportion with significant discomfort over the audit period, p<0.0001.
Conclusions: Colonoscopy audit as a routine process with data collection by endoscopy nurses over several years may be able to improve key performance indicators by the process of regular feedback to endoscopists. Audit should be encouraged as a routine process rather than simply as a research tool for a limited period.