Article Text


Endoscopy II
Does colonoscopists' performance vary between bowel screening and patients with symptoms – a comparison of colonoscopy completion rates
  1. N Nagrath *1,
  2. P Phull2
  1. 1University of Aberdeen Medical School, Aberdeen, UK
  2. 2Gastrointestinal & Liver Service, Aberdeen Royal Infirmary, Aberdeen, UK


Introduction Colonoscopy completion rates (CCR) are an important quality assurance parameter for individual endoscopists and for the national Bowel Screening Programme (BSP). The standard is set at an unadjusted completion rate of 90%. However, individuals undergoing colonoscopy for bowel screening and patients with symptoms represent different.

Aim To determine if colonoscopy completion rates varied between screening and patients with symptoms, patient groups and their characteristics may affect colonoscopy performance.

Methods Retrospective audit of completion rates for patients undergoing colonoscopy at our institution during the year 2008. Data was retrieved from an endoscopy database (Unisoft) in addition to the local BSP database. Colonoscopy completion rates were calculated for those colonoscopists participating in the BSP, and compared between BSP and symptomatic groups. Information on factors that may affect completion of colonoscopy such as poor bowel preparation, presence of diverticular disease, polyps, tumour and strictures were recorded.

Results During 12-month period of the audit, a total of 3582 colonoscopies were performed; of these 508 (14.2%) were undertaken for the BSP. Five colonoscopists performed 488 (95.9%) of these procedures. The remaining 20 colonoscopies were performed by 4 colonoscopists; these individuals were excluded from further analysis due to the small number of procedures performed. The 5 main BSP colonoscopists also performed 568 colonoscopies in patients with symptoms. Colonoscopy completion rates are shown in the table 1:

Table 1


CCR was 90% or higher for all colonoscopists in BSP procedures. However, for 2 of the 5 colonoscopists the CCR was below 90% in procedures undertaken in patients with symptoms; in the case of one colonoscopist, this difference was statistically different.

Data on the case-mix revealed that the mean age of patients was higher in the BSP group for all colonoscopists. However, there were no consistent findings in the gender distribution, poor bowel preparation, presence of diverticular disease, polyps, tumour or strictures mix that could explain the lower CCRs for 2 colonoscopists in the patients with symptoms.

Conclusion Colonoscopists' performance, as assessed by the CCR, demonstrated some variation between screening and patients with symptoms. This audit demonstrates that in addition to monitoring overall CCR as a performance indicator, it is the important to also look at CCRs separately in BSP and patients with symptoms.

Statistics from


  • Competing interests None.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.