The Impact of Diet Liberalization on Bowel Preparation for Colonoscopy

Abstract

Purpose: Dietary restrictions pose a barrier to compliance with colonoscopy. We evaluated the effect of diet liberalization on the adequacy of bowel preparation. Methods: Single-center, prospective, randomized, investigator-blinded study comparing clear liquid diet (CLD) to low-residue diet (LRD) in patients undergoing outpatient colonoscopy. A LRD instruction sheet was provided to and reviewed with patients randomized to LRD. LRD was permitted up to 1 PM day prior to colonoscopy followed by clears up to 2.5 hours before colonoscopy. The CLD group was permitted clears day prior and up to 2.5 hours before colonoscopy. All patients took split dose (PM/AM) 2 L polyethylene glycol-electrolyte solution purgative (MoviPrep®). Preparation was scored using the Boston Bowel Preparation Scale (BBPS). Primary endpoint was the dichotomous index of preparation quality (adequate [BBPS > 5] vs. inadequate [BBPS ≤ 5]) for the whole colon. A one-sided Wald test for non-inferiority with a pre-specified margin of -13.5% was used. Secondary endpoints were: BBPS scores for whole colon and individual segments, purgative completion, polyp and adenoma detection, adverse events, satisfaction, and quality-of-life measures. This study was approved by Jefferson's IRB and registered at ClinicalTrials.gov. Results: One hundred eighty-eight patients were randomized, 48 withdrew prior to colonoscopy, and 140 were analyzed (72 CLD, 68 LRD). The groups were well-matched except the CLD group had more men (47.2% vs. 38.2%) and fewer with a colonoscopy indication of “symptoms” (13.9% vs. 23.9%). Preparation was adequate in 68 (94.4%) of the CLD group vs. 60 (88.2%) of the LRD group (risk difference -6.2; non-inferiority p=0.06, two-sided 95% CI -15.5%, 3.1%). Overall preparation completion was similar, but 79.1% of LRD group vs. 91% of CLD group completed >90% of the AM dose (p=0.09). There were no significant differences between the groups in BBPS scores for the whole (p=0.22), right (p=0.33), transverse (p=0.93), and left (p=0.08) colon. Polyp and adenoma detection (percent of patients with >1) was not significantly different between the study groups. However, the CLD group had a significantly greater number of polyps (IRR 1.43; 95% CI 1.04-1.96; p=0.03) and adenomas (median adenomas/patient, CLD vs. LRD: 1.0 vs. 0.53; IRR 1.84; 95% CI 1.16-2.91; p=0.01). Satisfaction was significantly greater with LRD vs. CLD (1.99 vs. 3.04; p=0.01). There were no significant differences in adverse events or productivity between the study groups. Conclusion: Using the endpoint of adequate bowel cleansing, a low-residue diet was not non-inferior to clear liquids. Total adenoma detection was superior with clear liquids, and satisfaction greater with a lowresidue diet. General adoption of a low-residue diet may result in more preparation failures. Disclosure - Dr. James Walter - None; Dr. Apurva Patel - None; Dr. Becky Matro - None; Dr. Leo Katz - None; Dr. Ramalinga Kedika - None; Rachael Grosso - None; Alexis Sharpe - None; Dr. Louis Broad - None; Dr. Daniel Quirk - None; Dr. Scott Keith - None; Dr. Constatine Daskalakis - None; Dr. Marianne Ritchie - None; Dr. Linda Greenbaum - None; Dr. Patricia Kozuch - None; Jocelyn Andrel Sendecki - None; Dr. David Kastenberg - Consultant: Salix Pharmaceuticals, Inc.