Background: Colonoscopy has a miss rate for adenomas that may partly relate to poor visualization of the colonic surface. Dynamic position changes during colonoscope withdrawal can improve luminal distension.
Objective: To assess whether position changes also improve adenoma and polyp detection.
Design: Randomized crossover clinical trial.
Setting: Academic endoscopy unit.
Patients: This study involved 130 patients who presented for routine colonoscopy.
Intervention: Examination either entirely in the left lateral position followed by position changes (cecum to hepatic flexure, left lateral; transverse colon, supine; splenic flexure and descending colon, right lateral) or vice versa. After both examinations, polyps were removed for histopathology.
Main outcome measurements: Proportion of patients with ≥1 polyp or adenoma detected between the hepatic flexure and the sigmoid-descending colon junction. Luminal distension was measured on a scale of 1 to 5: 1, total collapse; 5, fully distended.
Results: At least 1 adenoma was detected in 34% of patients in colon areas in which the patient position differed from left lateral (transverse colon, splenic flexure, descending colon) compared with 23% examined with the patient in the left lateral position alone (P = .01). At least 1 polyp was detected in 52% of patients with position changes versus 34% of patients examined in the left lateral position alone (P < .001). Adenoma and polyp detection were positively correlated with an improved distension score (correlation coefficient, 0.12; P < .001). Adenomas were detected in 16% of colon areas with adequate distension scores (4 and 5) compared with 7% of those with borderline or nondiagnostic scores (1-3; P < .001).
Limitations: Single-operator study.
Conclusion: Dynamic position changes during colonoscope withdrawal significantly improved polyp and adenoma detection. (
Clinical trial registration number: NCT00234650).
Copyright © 2011 American Society for Gastrointestinal Endoscopy. Published by Mosby, Inc. All rights reserved.