Trial design and setting
This prospective, single-blinded, randomized control trial (clinicaltrials.gov NCT01978509) was approved by the Institutional Review Board (IRB) and adheres to Consolidated Standards for Reporting Trial (CONSORT) guidelines for reporting clinical trials. It was conducted as a pilot study at Mayo Clinic Arizona between September 2013 and March 2019.
Eligible subjects included hospitalized patients 18 years or older who were able to provide consent and in whom colonoscopy was deemed medically necessary while hospitalized. Patients who were unable to give consent, were pregnant or breastfeeding, had renal impairment, ileus, ascites, toxic megacolon, evidence of gastrointestinal obstruction, or presence of an allergy to study drug were excluded. Risks and benefits were explained to all subjects and written informed consents were obtained.
After informed consent was obtained, patients were randomly assigned to the high-volume solution polyethylene glycol – PEG (GoLYTELY®), PEG+ascorbic acid medium-volume solution (MoviPrep®) or sodium picosulfate low-volume solution (Prepopik®), see table 1 for full list of ingredients. All doses were prescribed and administered as split dose, with half of the required preparation being administered the night before the procedure starting at six in the evening and the other half being administered the morning of the procedure starting at three in the morning. All patients were required to complete the liquid purgative two hours prior to their procedure. These are further described in Table 1. All subjects received a clear liquid diet the day before the procedure.
Physicians performing the endoscopy were blinded to what type of bowel preparation each patient received. Fellow physicians within the gastroenterology department on service at the hospital enrolled participants. The allocation ratio was 1:1:1 for the intervention. Randomization was carried out using a computer-generated random numbers model and performed by a nurse practitioner who then placed bowel preparation orders without informing the inpatient gastroenterology service or endoscopist(s) performing the colonoscopy. Additionally, the patients were told not to speculate or inform their nursing staff, physicians, or performing endoscopist(s) if they were aware which bowel preparation regimen they consumed. Both fellow physicians and faculty physicians completing the procedures were involved in the creation of the BBPS for formal reports and blinded to the preparation the patient received.
Colonoscopy procedures were performed by the inpatient gastroenterology hospital service, which included attending faculty and fellows (under direct supervision of an attending), using Olympus Exera II 180 series colonoscopes in 4 subjects and Olympus 190-series (CF-HQ190AL and PCF-H190L) colonoscopes in the remaining cases (Olympus Corp., Tokyo, Japan). The success of cecal intubation was established by visualization of anatomic landmarks (appendiceal orifice and ileocecal valve). Procedures were performed under conscious sedation (IV fentanyl, IV midazolam) in 21 subjects while 4 patients underwent deep sedation (IV fentanyl, IV midazolam, and IV propofol) with assistance of anesthesia providers.
Therapeutic interventions, such as biopsies, polypectomies, clip placement, argon plasma coagulation, or other electrocoagulation modality were performed as indicated. During the colonoscopy, quality metrics (BPSS, and cecal intubation) were obtained. No procedures were required to be repeated due to inadequate preparation. Withdrawal time was not a quality metric tracked in this study due to inpatient procedures being performed for diagnostic purposes, not screening or surveillance.
To determine quality of bowel preparation among three different volume solutions, we used the Boston Bowel Preparation Scale (BBPS) in the three colon segments (right, transverse and left colon) along with total score.
After colonoscopy, all subjects received a questionnaire about their experiences with the colonoscopy preparation (Table 2), and cecal intubation rate was obtained. The questionnaire given to patients included questions on percentage of bowel preparation completion, perceived unpleasant taste, nausea and vomiting.
This was a pilot study, in which a modest sample size was achieved. Number of patients enrolled in this study was determined by willingness for participation. The trial ended due to a difficulty of recruiting participants, partly due to faculty and fellow bias that low volume colon preparation would lead to poor bowel preparation and need for repeat procedures.
Continuous variables are described with their mean and standard deviation while categorical variables are described by count and percentage. The Chi-Square test was used for demographics, indication for colonoscopy, diabetic status, history of constipation, purgative type, and if patient completed bowel preparation. The three groups were compared in terms of BMI, quality of bowel preparation (using Boston Bowel Preparation Scale), and side effects (unpleasant taste, nausea, vomiting, abdominal pain scored in a five-point scale) by the one-way analysis of variance (one-way ANOVA) test. All hypothesis tests were two-sided with P < 0.05 considered statistically significant. Analyses were performed using SAS 9.4 (SAS Institute, Inc.; Cary, NC).