Effect of simply recording colonoscopy withdrawal time on polyp and adenoma detection rates.

Background: According to national recommendations, colonoscopy withdrawal time (WT) on negative screening examinations should average more than 6 minutes because this time is associated with a higher rate of polyp detection. Attempts have been made to increase the WT; however, simply knowing that a quality measure, such as the WT, is being monitored, by itself, may improve the quality of an examination.
Objective: To measure changes in the polyp detection rate associated with recording the WT.
Design: Retrospective.
Setting: Single tertiary care center.
Patients: Patients undergoing colonoscopy within 5 months immediately before (group A) and after (group B) initiation of WT recording. Colonoscopies were excluded if procedure times were incomplete, missing, or nonsensical.
Main Outcome Measurements: Polyp detection, compared by using the chi(2) test and logistic regression multivariate analyses; pathology manually reviewed from a sample of 200 consecutive polyp cases (100 per group).
Results: The average WT in group B was 14.5 minutes (11.0 minutes in negative screening examinations). In group A, polyps were detected in 530 (37.7%) of 1405 colonoscopies compared with 571 (41.2%) of 1387 colonoscopies in group B (difference 3.5%; 95% CI, -0.2% to 7.1%), a nonsignificant 9.3% relative increase. Longer procedure time, age, sex, and indication were significant predictors; monitoring the WT was not. Nonsignificantly, more polyp examinations in group B found all nonadenomas compared with group A (36% vs 27%; P = .17), and polyps were nonsignificantly smaller (P = .30).
Limitations: Retrospective database data, pathology performed only on a subsample.
Conclusion: WT recording was associated with a nonsignificant increase in polyp detection, but this was likely attributable to a slight increase in the detection of (smaller) nonadenomatous polyps.
(Copyright 2010 American Society for Gastrointestinal Endoscopy. Published by Mosby, Inc. All rights reserved.)