Position changes improve visibility during colonoscope withdrawal: a randomized, blinded, crossover trial

Gastrointest Endosc. 2007 Feb;65(2):263-9. doi: 10.1016/j.gie.2006.04.039. Epub 2006 Dec 4.


Background: Adequate distension is essential to maximize neoplasia detection during colonoscope withdrawal. Position changes may improve distension but are often not performed in routine practice.

Objective: To assess whether routine position changes improve luminal distension during colonoscope withdrawal.

Design: Randomized, blinded, crossover trial.

Setting: Single tertiary-referral center, United Kingdom.

Patients: Fourteen patients attending for routine colonoscopy.

Interventions: Videotaped, back-to-back examination of colon proximal to rectosigmoid junction in left lateral position only, then with position changes: left lateral for the cecum to the hepatic flexure, supine for the transverse colon, and right lateral for the splenic flexure and the descending colon, or vice versa.

Main outcome measurements: Luminal distension as scored by a blinded video reviewer. Luminal distension was scored on a scale of 1 to 5 for each colonic area: 1, complete collapse; 5, widely distended to limit of view. A score of 2 or less was considered inadequate for diagnosis.

Results: Scores for the 2 examinations from the blinded video reviewer were significantly higher in the transverse, the splenic flexure, and the descending colon, P = .02, .002, and <.001, respectively. Without position changes, 6 of 14 of patients (43%) would have had a nondiagnostic distension score (1 or 2) in at least 1 colonic area, P = .03.

Limitations: Nonvalidated scoring system for luminal distension, however, good agreement between endoscopist and blinded reviewer, weighted kappa 0.53, 95% confidence interval 0.38-0.69.

Conclusions: Position change, a cost-neutral intervention, during colonoscope withdrawal improved luminal distension between hepatic flexure and sigmoid-descending junction and has the potential to reduce adenoma and early cancer miss rates.

Publication types

  • Randomized Controlled Trial

MeSH terms

  • Colon
  • Colonoscopy / methods*
  • Cross-Over Studies
  • Humans
  • Posture*
  • Single-Blind Method
  • Videotape Recording