Endoscopist can be more powerful than age and male gender in predicting adenoma detection at colonoscopy

Am J Gastroenterol. 2007 Apr;102(4):856-61. doi: 10.1111/j.1572-0241.2006.01054.x. Epub 2007 Jan 11.


Background and aims: Both advancing age and male gender are known predictors of adenomas and large adenomas at colonoscopy. However, the importance of endoscopist compared with both age and gender as predictors of adenomas is not known. In this study, we assessed the adenoma detection rates of nine endoscopists performing colonoscopy and the effects of endoscopist on adenoma detection compared with the established predictors of advancing age and male gender.

Methods: Consecutive colonoscopy reports by nine attending gastroenterologists at Indiana University Hospital between January 1999 and January 2004 were obtained. Patients with indications for inflammatory bowel disease (IBD), inherited colon cancer syndromes, or who were referred for polypectomy were excluded.

Results: There were 10,034 included colonoscopies. The overall mean age (yr) +/- SD for the entire cohort was 56.0 +/- 13.5, range 14-93 yr. Multivariate analysis indicated that both advancing age and male gender were powerful predictors of adenomas, both for the overall cohort and in those >or=50 of age (P<or= 0.0001). The prevalence of adenomas increased from 14.6% at age 30 to 35.2% at age 70. Men had more adenomas than women (24.4%vs 16.6%, P<or= 0.0001). Among patients >or=50 yr of age, the range of detection of at least one adenoma per colonoscopy by nine colonoscopists was 15.5-41.1%, at least two adenomas was 4.9-20.0%, at least three adenomas was 0.8-10.8%, and at least one adenoma >or=1.0 cm was 1.7-6.2%, and the range of adenomas detected per colonoscopy was 0.21-0.86. Differences between endoscopists were significant (P < 0.001).

Conclusion: Who performs the procedure can be more important than age and gender in predicting adenomas at colonoscopy. Our results underscore the need to measure adenoma detection rates in the continuous quality improvement process and to report them to endoscopists. Additional data are needed to determine whether corrective measures can reduce variation in adenoma detection rates.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adenoma / diagnosis*
  • Age Factors
  • Chi-Square Distribution
  • Clinical Competence*
  • Colonoscopy / standards*
  • Colorectal Neoplasms / diagnosis*
  • Humans
  • Logistic Models
  • Male
  • Predictive Value of Tests
  • Sex Factors