The Utility of Timed Segment Withdrawal During Screening Colonoscopy for Increasing Adenoma Detection Rate: A Pilot Randomized Controlled Clinical Trial

Introduction: Adenoma detection during colonoscopy is an important quality measure in colon cancer screening. There are conflicting findings regarding the usefulness of segmental withdrawal and increasing withdrawal time on adenoma detection rate (ADR). We conducted a pilot prospective, factorial randomized trial to assess the impact of these protocols on polyp detection rate (PDR) and ADR. Our hypothesis was that these will increase detection rates. Methods: Study participants were recruited from patients scheduled for a screening or surveillance colonoscopy from January 2014 to May 2015. Patients were initially randomized to either segmental or standard nonsegmental protocol. Further, subjects of each group were randomized to either 6 min or 8 min protocol. This design leads to 4 groups: A) at least 3 min dedicated to the right side of the colon, minimum 6 min total withdrawal time, B) at least 4 min right side, min 8 min total withdrawal time, C) 6 min total withdrawal time regardless of time spent in any segment, D} 8 min total withdrawal time. Randomization was done using sealed opaque envelopes. Outcome variables included: PDR, ADR, advanced adenoma detection rate (AADR), number of polyps and number of adenomas per subject. Generalized estimating equations with binomial distribution were used to assess the unadjusted and adjusted effect of different protocols on detection rates. Results: There were 326 patients included in the study: mean age 58, 64.7% female, 92% Hispanic. There was no difference between groups in gender, ethnicity, indication, preparation quality, and prior abdominal surgery. Results for PDR, ADR, AADR for the 4 groups were: A} 60.5%, 40.7%, 4.9%, B) 54.4%, 39.2%, 5.1%, C} 45.7%, 34.6%, 7.4%, D} 54.1%, 38.9%, 8.2%. The average number of adenomas per patient were A} 0.73, B} 0.87. C) 0.80, D} 0.66. Although not significantly different, use of segmental protocol or increased time protocol showed a trend for increased ADR or PDR compared with the alternatives. Group A showed a trend towards superiority for PDR (p=0.05) compared with group C. Conclusion: In our study, we found a trend for segmental withdrawal to be superior compared with nonsegmental withdrawal for the 6 minutes protocol. Further studies to see if this trend is significant should be done and our study provides data with which to estimate sample size for the conduct of future studies to improve ADR.