Research Article: A randomized clinical prospective trial comparing split-dose picosulfate/ magnesium citrate and polyethylene glycol for colonoscopy preparation

Date Published: March 28, 2019

Publisher: Public Library of Science

Author(s): Alaa Rostom, Catherine Dube, Kirles Bishay, Lilia Antonova, Steven J. Heitman, Robert Hilsden, John Green.


Colonoscopy remains the gold standard for the investigation of abnormalities within the colon. However, its success is highly dependent on the quality of bowel preparation. The objective of this study was to compare the bowel preparation efficacy of picosulfate/magnesium citrate (PMC) vs polyethylene glycol (PEG) in a one-day vs two-day split dose regimen.

A prospective, randomized, controlled trial was conducted at the Forzani & MacPhail Colon Cancer Screening Centre in Calgary, Canada. 171 colonoscopy outpatients were randomized to split-dose PMC or PEG lavage as well as into one-day split or two-day split regimens in blocks of eight. Bowel preparation quality was recorded in a blinded manner by the endoscopist using the Ottawa Bowel Preparation Scale (OBPS) prior to washing or suctioning. The scale results were analyzed using a two-factor analysis of variance.

141 patients received complete colonoscopies (PMC-71; PEG-70). PEG was found to be superior to PMC (mean OBPS: 4.14 ± 2.64 vs 5.11 ± 3.44, p = 0.019), when adjusted for administration regimen, leading to significantly more adequate bowel preparations (79.7% vs 59.7%, p = 0.007). A two-day split dose was superior to a one-day split dose regimen (mean OBPS: 3.68± 2.82 vs 5.69 ± 3.06, p<0.001). Two-day split dosing also resulted in a better right colon cleanliness score (right bowel OBPS 1.27±0.11 vs 2.10±0.12 for one-day split, P<0.001). Optimal bowel preparation was achieved with the use of PEG lavage when administered in a two-day split dose regimen. This trial is registered with under identifier NCT01415687.

Partial Text

Colonoscopy is the gold standard for the investigation of abnormalities within the colon and is an integral part of all colorectal cancer screening programs. The ability of colonoscopy to detect high-risk lesions is greatly dependent on the quality of pre-colonoscopy bowel preparation. Poor bowel preparation results in longer procedures, need for repeat colonoscopy and missed lesions [1] and has been observed in as many as 25% of cases.[1–3] One level of consideration in selecting an agent relates to the physico-chemical properties of the agent itself, including solution tonicity, safety profile, as well as the volume required to produce adequate bowel preparation. Patient factors, such as comorbidity burden and cost of various preps, also represent important considerations in selecting an agent. Lastly, a decision should be based on published evidence of the agent’s efficacy as a colonic lavage solution. Based on these factors, PEG and PMC have emerged as two of the most commonly used agents clinically.

This trial was approved by the University of Calgary institutional ethics board on April 21, 2011 and registered with under identifier NCT01415687. A small delay in registration was caused by an administrative error. No changes in the protocol occurred during this period. This study has no other related past or ongoing trials.

In total, between May 2011 and December 2011, 171 patients were enrolled in the trial (Fig 1). Of these, 141 underwent a complete colonoscopy. Colonoscopy could not be completed for reasons other than poor bowel preparation in 6 patients and 22 patients either withdrew from the trial or had scheduling conflicts for colonoscopy post-randomization. The distribution of patients that did not complete the study was similar between treatment groups, as shown in Fig 1. No adverse events occurred during the trial that resulted in discontinuation. A total of 70 patients received PEG and 71 received PMC. Table 2 shows the baseline characteristics of the two groups. PMC and PEG groups were well matched in terms of both age (57.6[95% CI:55.8,59.4] vs 56.4[95%CI:55.0,57.8]), P = 0.395) and sex (percent female: 47.8% vs 51.4%, P = 0.539). In the PEG group, 34(48%) patients received a one-day split regimen and 36(52%) patients received a two-day split regimen. In the PMC group, 33(46%) patients received a one-day split regimen and 38(54%) received a two-day split regimen. Patient compliance to their assigned regimen exceeded 95% in all groups, due to the personalized nurse-led patient training provided at the screening center prior to colonoscopy preparation.

In this trial, we compared the bowel-cleansing effectiveness of PEG and PMC, as well as that of day-before versus two-day preparation regimen. Our results demonstrate that PEG is superior to PMC for lavage of the bowel, after adjustment for one-day vs two-day split dosing. Previous findings indicate that the difference of nearly 1 point on the OBPS produced by PEG administration is likely to be clinically significant.[10,16] Furthermore, we found that a two-day preparation regimen produces superior bowel cleansing, as compared to a one-day preparation regimen. The difference produced by preparation timing was 2 points on the OBPS. Our results indicate that a change to a two-day split dose regimen can have a greater impact on colon cleansing than the choice of preparation agent. Optimal OBPS was obtained when PEG was utilized in a two-day preparation regimen. Segment scores did not differ between the preparation agents. However, when assessing the right colon, there was a strong effect in favor of two-day bowel preparation.

The results of this trial demonstrate that bowel preparation quality can be optimized through the use of a two-day preparation regiment and the administration of a PEG based lavage. This suggests that even for early morning procedures, a second dose of preparation should be administered 5–6 hours prior to the procedure. It is imperative that patients are educated with regard to the need for good preparation, in order to optimize colonoscopy outcomes and reduce the need for repeat procedures.




Leave a Reply

Your email address will not be published.