Variation in polyp size estimation among endoscopists and impact on surveillance intervals

Gastrointest Endosc. 2014 Oct;80(4):652-659. doi: 10.1016/j.gie.2014.01.053. Epub 2014 Mar 27.

Abstract

Background: Accurate estimation of polyp size is important because it is used to determine the surveillance interval after polypectomy.

Objective: To evaluate the variation and accuracy in polyp size estimation among endoscopists and the impact on surveillance intervals after polypectomy.

Design: Web-based survey.

Participants: A total of 873 members of the American Society for Gastrointestinal Endoscopy.

Interventions: Participants watched video recordings of 4 polypectomies and were asked to estimate the polyp sizes.

Main outcome measurements: Proportion of participants with polyp size estimates within 20% of the correct measurement and the frequency of incorrect surveillance intervals based on inaccurate size estimates.

Results: Polyp size estimates were within 20% of the correct value for 1362 (48%) of 2812 estimates (range 39%-59% for the 4 polyps). Polyp size was overestimated by >20% in 889 estimates (32%, range 15%-49%) and underestimated by >20% in 561 (20%, range 4%-46%) estimates. Incorrect surveillance intervals because of overestimation or underestimation occurred in 272 (10%) of the 2812 estimates (range 5%-14%). Participants in a private practice setting overestimated the size of 3 or of all 4 polyps by >20% more often than participants in an academic setting (difference = 7%; 95% confidence interval, 1%-11%).

Limitations: Survey design with the use of video clips.

Conclusion: Substantial overestimation and underestimation of polyp size occurs with visual estimation leading to incorrect surveillance intervals in 10% of cases. Our findings support routine use of measurement tools to improve polyp size estimates.

MeSH terms

  • Clinical Competence*
  • Colonic Polyps / diagnosis
  • Colonic Polyps / pathology*
  • Colonoscopy / standards*
  • Colonoscopy / trends
  • Confidence Intervals
  • Female
  • Health Care Surveys
  • Humans
  • Internet
  • Male
  • Sensitivity and Specificity
  • Severity of Illness Index
  • Societies, Medical
  • United States
  • Video Recording*