Fifteen years later: colonoscopic retroflexion revisited.

OBJECTIVE: Retroflex views of the rectal vault are included in the teaching of colonoscopic technique but are not pervasive in clinical practice. The utility of adding a retroflex maneuver at the end of colonoscopy has yet to be determined. The aim of this study was to evaluate the additional benefit of a retroflex view of the rectal vault at the completion of colonoscopic examination.
METHODS: A prospective study of consecutive colonoscopies performed by a single physician was conducted. The rectal vault was first visually inspected upon withdrawal of the colonoscope. The endoscope was then readvanced into the rectum and retroflexed to view the vault. Endoscopic findings on both views were recorded along with demographic patient information. The six groups of findings sought on the two views were: retained stool, abnormal hemorrhoids, erosions/ulcerations, polyps, masses, and normal examinations. A determination on whether retroflex views influenced patient diagnosis was recorded by the endoscopist.
RESULTS: There were 453 patients enrolled: 182 (40.2%) male and 271 (59.8%) female, consisting of 216 African-Americans, 232 Caucasians, and five Asians. The retroflex maneuver was performed successfully in 445 of 453 patients. In all but nine cases, the retroflex view did not produce additional information. The nine findings included three inflammatory pseudopolyps, five hyperplastic polyps, and one case of erosions/ulcerations.
CONCLUSIONS: In the majority of cases, retroflexing the endoscope does not produce additional information compared with the thorough examination in straight view. The retroflex view may be of benefit if there is suspicion of pathology upon insertion or withdrawal of the colonoscope.