Factors That Influence Colonoscopic Technique and Adenoma Detection Rates At An Academic Gastroenterology Unit

Abstract

Introduction: Colonoscopy is the dominant screening exam for colorectal cancer. Recent publications have highlighted the significant variation in colonoscopic withdrawal times and adenoma detection rates among gastroenterologists. The purpose of this study was to evaluate the quality of colonoscopy at a tertiary academic institution with a focus on factors influencing withdrawal times and adenoma detection rates. Methods: This is a review of the procedural data and pathologic results of 550 consecutive screening colonoscopies in average risks patients (mean [ ± SD] age, 57 ± 7.6, 44 % male) completed between 10/2/06 to 3/15/07 by 10 board certified gastroenterologists. The colonoscopies were all morning procedures completed by staff attending endoscopists only. Endoscopist's experience, cecal intubation time, withdrawal time, and time of day the procedure started were examined to study their effect on adenoma detection rates. Results: Two hundred twenty five (41%) patients had at least one polyp found. One hundred twenty one patients (22 %) had at least one adenoma detected. Advanced lesions (adenomas ≥10 mm in size or with advanced pathology) were found in 25 (4.5%) patients. Colon cancer was detected in 3 (0.5%) patients. Per individual gastroenterologist, the adenoma detection rates (number of adenomas detected per patient screened) ranged widely from .09 to .82 with a mean of .46 for the group. The mean withdrawal time was 7.0 minutes for the group and ranged from 3.4 to 9.6 minutes. There was a significant positive relationship between the number of adenomas detected and the withdrawal time, with more adenomas detected the longer the colonoscopic withdrawal time (P < .01). There was a significant inverse relationship between cecal intubation time and withdrawal time, with shorter withdrawal times associated with longer times taken to reach the cecum (P < .01). Endoscopist's years of experience and hour of day procedure performed did not have a significant influence on withdrawal times, cecal intubation times or adenoma detection rates. Conclusions: 1) Significant variation in gastroenterologists' abilities to detect adenomas during screening colonoscopies exists. 2) Colonoscopic withdrawal time is the most important factor associated with increased adenoma detection rates. 3) It is paramount that future research and endoscopic training focus on increasing adenoma detection rates as well as decreasing interexaminer variation.