Colon Cleansing in Patients with Predictors of a Poor Bowel Preparation: An Analysis of the See Clear II Study

Michael Epstein, MD, Nav K Grandhi, MD, Phillip O Katz, MD, Douglas K Rex, MD

Digestive Disorders Associates, Gastroenterology Research Consultants of Greater Cincinnati, Albert Einstein Healthcare Network, Indiana University School of Medicine

Objectives: Poor bowel cleansing may increase colonoscopy procedure time and decrease the detection rate for precancerous lesions. Previous reports have established that specific demographic and clinical predictors are associated with inadequate bowel preparation. We conducted a post hoc analysis of the SEE CLEAR II study to investigate the effectiveness of 2 colon cleansing products in patients with known predictors of a poor bowel preparation.

Methods: SEE CLEAR II was a phase 3, randomized, multicenter, assessor-blinded study that investigated the efficacy of day-before administration of P/MC (Prepopik™), a nonphosphate, orange-flavored, dual-action, low-volume preparation containing sodium picosulfate, magnesium oxide, and citric acid (n = 294) compared with conventional dosing of 2L polyethylene glycol solution and two 5-mg bisacodyl tablets (2L PEG + 2 bisacodyl tablets, HalfLytely® and Bisacodyl Tablets Bowel Prep Kit) (n = 300) in patients preparing for colonoscopy. To be included in the post hoc analysis, patients must have had ≥1 predictor of a poor bowel preparation (predictor) including male sex, history of constipation or stroke, no history of gastrointestinal polyps, concomitant tricyclin antidepressant use, or incompletion of bowel preparation. Aronchick and Ottawa scale scores were analyzed for each predictor.

Results: Greater than 50% of patients in each treatment arm of the SEE CLEAR II study had ≥1 predictor. A total of 321 patients were included in the analysis; of these, 152 patients received P/MC and 169 patients received 2L PEG + 2 bisacodyl tablets. Demographics and baseline characteristics were similar between the treatment arms. Predictor groups that had >10 patients in at least 1 of the treatment arms are summarized below. In patients with ≥1 predictor, no significant difference was observed in Aronchick scores between the treatment arms (P > .05). However, in male patients, Aronchick scores were significantly better for P/MC compared with 2L PEG + 2 bisacodyl tablets (P = .0121); excellent = 48.5% vs 30.4%, good = 32.0% vs 44.6%, fair = 19.4% vs 21.4%, and inadequate = 0% vs 3.6%. Furthermore, males receiving 2L PEG + 2 bisacodyl tablets had Aronchick scores that were significantly worse compared with females in the same treatment arm (P = .0138); excellent = 30.4% vs 46.3%, good = 44.6% vs 36.2%, fair = 21.4% vs 17.0%, and inadequate = 3.6% vs 0.5%. Conversely, no significant difference by sex was noted in the P/MC treatment arm (P > .05). No significant differences were evident in Aronchick scores between the treatment arms in patients with a history of constipation (P > .05) or without a history of gastrointestinal polyps (P > .05). Similar patterns in Ottawa scores were observed between P/MC and 2L PEG + 2 bisacodyl tablets for the mid and rectosigmoid segments of the colon.

Conclusion: In patients with ≥1 predictor of a poor preparation, including a history of constipation or no history of gastrointestinal polyps, day-before dosing of P/MC provided similar cleansing scores compared with 2L PEG + 2 bisacodyl tablets. P/MC may offer greater benefit than 2L PEG + 2 bisacodyl tablets in males preparing for colonoscopy.

Session: Poster Presentation

Program Number: P032

« Return to SAGES 2013 abstract archive