BACKGROUND/RATIONALE:
Although there are several approaches to colorectal cancer screening, colonoscopy is the only test that allows for identification and removal of polyps from the entire colon.4 Case-control studies reveal that polypectomy is associated with a 50-90% reduction in colorectal cancer mortality vs. no polypectomy.(5, 6)

Successful colonoscopy requires that patients effectively evacuate their bowels through proper use of purgatives and adherence to strict dietary restrictions.(7-9) But as most recipients of colonoscopy can attest, preparing for the procedure is often more demanding than the procedure itself. Patients must sustain a clear liquid diet for 1-2 days, consume a purgative which induces large volume diarrhea, and often experience abdominal discomfort, bloating, and nausea. For many patients the preparatory experience is difficulty, unpleasant, and disruptive of daily routines. Thus, it is not surprising that inadequate bowel preparation is common; up to one-quarter of patients presenting for colonoscopy have an inadequate preparation, although rates vary considerably among patient populations and centers.(7, 8)

Inadequate bowel preparation is important because it is linked to poor outcomes and high costs. Rex and colleagues found that patients with inadequate preparation had 45% fewer polyps detected and 5% more incomplete or aborted procedures compared to those with adequate preparation.(10) The authors further demonstrated that patients with inadequate preparation cost 22% more on average - an incremental expense largely driven by repeated procedures. In short, compared to adequately prepared patients, those with inadequate bowel preparation have more incomplete examinations, fewer polyps found, more repeat colonoscopies, and higher cost of care.

Although preparation quality is partly determined by the prescribed purgative and the timing of administration (e.g. single dose vs. split dose), purgatives generally require that patients follow similar dietary guidelines. The effectiveness of a bowel preparation is closely linked with patient compliance with both the pharmacy and dietary instructions. The reliance on patient compliance may explain why randomized trials reveal only minimal differences in efficacy between competing single-dose purgatives.(11) Although there have been extensive attempts to study competing preparations,(11, 12) with recent emphasis on the use of split-dose preparations,(1-3) there has been relatively less attention paid to understanding non-pharmacological factors that may optimize bowel preparation quality. It is likely that even the most effective bowel regimens can be further enhanced through efforts to maximize patient compliance during the preparatory period.

In this grant, we developed and validated a novel educational booklet aimed at addressing patient knowledge, attitude, and belief barriers to colonoscopy preparation. We hypothesized that patients receiving the booklet in a randomized controlled trial of single-dose purgatives would achieve better bowel preparation quality vs. controls independent of the specific purgative prescribed.

References Cited
1.Rex DK. Dosing considerations in the use of sodium phosphate bowel preparations for colonoscopy. Ann Pharmacother 2007;41:1466-75.
2.Park JS, Sohn CI, Hwang SJ, Choi HS, Park JH, Kim HJ, Park DI, Cho YK, Jeon WK, Kim BI. Quality and effect of single dose versus split dose of polyethylene glycol bowel preparation for early-morning colonoscopy. Endoscopy 2007;39:616-9.
3.Unger RZ, Amstutz SP, Seo da H, Huffman M, Rex DK. Willingness to undergo split-dose bowel preparation for colonoscopy and compliance with split-dose instructions. Dig Dis Sci;55:2030-4.
4.Winawer S, Fletcher R, Rex D, Bond J, Burt R, Ferrucci J, Ganiats T, Levin T, Woolf S, Johnson D, Kirk L, Litin S, Simmang C. Colorectal cancer screening and surveillance: clinical guidelines and rationale-Update based on new evidence. Gastroenterology 2003;124:544-60.
5.Winawer SJ, Zauber AG, Ho MN, O'Brien MJ, Gottlieb LS, Sternberg SS, Waye JD, Schapiro M, Bond JH, Panish JF, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med 1993;329:1977-81.
6.Muller AD, Sonnenberg A. Protection by endoscopy against death from colorectal cancer. A case-control study among veterans. Arch Intern Med 1995;155:1741-8.
7.Harewood GC, Sharma VK, de Garmo P. Impact of colonoscopy preparation quality on detection of suspected colonic neoplasia. Gastrointest Endosc 2003;58:76-9.
8.Froehlich F, Wietlisbach V, Gonvers JJ, Burnand B, Vader JP. Impact of colonic cleansing on quality and diagnostic yield of colonoscopy: the European Panel of Appropriateness of Gastrointestinal Endoscopy European multicenter study. Gastrointest Endosc 2005;61:378-84.
9.Ness RM, Manam R, Hoen H, Chalasani N. Predictors of inadequate bowel preparation for colonoscopy. Am J Gastroenterol 2001;96:1797-802.
10.Rex DK, Imperiale TF, Latinovich DR, Bratcher LL. Impact of bowel preparation on efficiency and cost of colonoscopy. Am J Gastroenterol 2002;97:1696-700.
11.Belsey J, Epstein O, Heresbach D. Systematic review: oral bowel preparation for colonoscopy. Aliment Pharmacol Ther 2007;25:373-84.
12.Tan JJ, Tjandra JJ. Which is the optimal bowel preparation for colonoscopy - a meta-analysis. Colorectal Dis 2006;8:247-58.




OBJECTIVE(S):
RESEARCH HYPOTHESES AND SPECIFIC AIMS

We hypothesized that compared to usual practice, a novel educational booklet aimed at improving preparation for colonoscopy in patients undergoing colorectal cancer screening would lead to: improved bowel preparation quality, improved detection of colon polyps, improved completion rates of colonoscopy and less attributable resource utilization.

METHODS:
Study Overview

This was a 3-phase study conducted in patients referred for colonoscopy at the West Los Angeles VA Medical (WLAVA) Center and the Durham VA (DVA). In Phase I we established the content validity of an educational booklet for colonoscopy preparation. In Phase II we conducted a randomized controlled trial to compare bowel preparation quality between outpatients receiving the booklet vs. no booklet in the West Los Angeles VA. In Phase III we re-tested the booklet in the Durham VA and again in the West Los Angeles VA. We obtained approval from the West Los Angeles VA Institutional Review Board for each phase of the study. The Phase II trial was registered with Clinicaltrials.gov (# NCT00975247).

Phase I: Booklet Development

Content Development

We conducted cognitive interviews to identify patient knowledge, attitude, and belief deficits hypothesized to drive inadequate colonoscopy preparation in a sample of 15 patients presenting for outpatient colonoscopy at the West Los Angeles VA Gastrointestinal Procedures Unit. We prepared a standard introductory script for the interviews developed in concert with the research team and a psychometrician. Following the introduction, each interview began with an open-ended probe employing the "think aloud" technique of cognitive interviewing.(13, 14) The interviewer then focused the respondent with a series of directed scripted probes.(13) The interviews assessed patient perceptions regarding facilitators and barriers to colonoscopy, their experiences during preparation, and their knowledge, attitudes, and beliefs regarding bowel preparation and colorectal cancer screening. We abstracted each interview onto a semi-structured form that included open fields to comment on dominant trends. This process identified major domains and associated minor domains of knowledge and belief barriers to effective colonoscopy preparation.

We repeated these methods in separate interviews with providers. We interviewed 10 gastroenterologists and 5 experienced ancillary staff members, including 3 endoscopy nurses, 2 endoscopy technicians, and 1 gastroenterology physician assistant. For these interviews we prepared a provider-specific introductory script for the interviews, along with a list of "think aloud" probes.

Booklet Construction

We developed a draft version of an educational booklet for patients preparing for colonoscopy. The booklet was designed to meet 4 criteria: (1) address the knowledge and belief barriers identified in the Phase I cognitive interviews; (2) feature high quality visual elements to assist patients in preparing for colonoscopy; (3) ensure that frequently asked questions are explicitly addressed; and (4) ensure the language did not exceed a 6th grade level using the Simple Measure of Gobbledygook (SMOG) readability formula,(15) as calculated by the SMOG readability calculator.(16)

The booklet was based on the principles of Rosenstock's Health Belief Model,(17) which describes the factors contributing to a patient's decision of whether to follow recommended health behaviors - in this instance proper preparations for colonoscopy. The model posits that adherence to recommended behaviors results from several factors, including perceived risk of underlying disease (e.g. "I may already have colon cancer"), perceived severity of the disease (e.g. "If I have cancer I may not live"), perceived benefits of following the recommended behaviors (e.g. "If I follow the instructions the doctor is more likely to find polyps or cancer"), and perceived barriers to following recommended behaviors (e.g. "The diet is just too hard to follow.").

Pilot Testing for Comprehensiveness, Comprehensibility, and Helpfulness

We performed a pilot study to establish patient receipt, understanding, and acceptance of the educational booklet in 60 consecutive patients scheduled for outpatient colonoscopy at the West Los Angeles VA Medical Center. Patients with dementia or other forms of cognitive impairment were excluded. All patients received the booklet by mail 1 week prior to their scheduled colonoscopy. We interviewed patients in person on the morning of their procedure and transcribed responses onto a data collection form. We asked patients about their experiences with the booklet, including whether they received and read the booklet, whether they believed the booklet was clear and understandable, and whether they found the booklet interesting, informative, and helpful. Patients also completed closed-ended scoring sheets to rate each section of the draft booklet across 3 domains: (1) clarity and understandability, (2) interest, and (3) helpfulness. Assessments were rendered on a 5 point Likert scale (e.g. 1=Not Helpful; 5=Extremely Helpful). We calculated mean scores for each section and planned to remove any sections that achieved a mean score of 3 (Somewhat Helpful) or less. Based on qualitative and quantitative feedback, coupled with reviews from our consulting physicians and ancillary staff, we iteratively revised the booklet to its final form.

Phase II: WLAVA Randomized Controlled Study

Study Procedures

In Phase II we prospectively randomized patients scheduled for outpatient colonoscopy in a 1:1 ratio to either receive usual care instructions or receive the booklet by mail 1 week prior to their scheduled colonoscopy. Patients were included if they had been scheduled for an outpatient, non-urgent, screening, surveillance, or diagnostic colonoscopy, were over 18 years of age, and did not have dementia or other forms of cognitive impairment. Patients in both arms received standard pharmacy directions for bowel preparation consisting of written dietary and purgative instructions. The study did not specify which purgatives to prescribe. Physicians selected between one of 3 preparations according to usual institutional practices, including: (1) sodium phosphate (Fleet Phosphosoda, C.B. Fleet); (2) magnesium citrate, or (3) 2 liter oral lavage of polyethylene glycol (MoviPrep, Salix Pharmaceuticals). Patients were instructed to take their regimens the night prior to colonoscopy; split-dosing was not employed in the unit during the course of this study. During the study period there were 8 gastroenterology fellows and 11 faculty members who performed colonoscopies in the unit. Allocation was determined by a random number generator. All patients, physicians, nurses, and technicians were blinded to group allocations.

Study Outcome Measures

The primary outcome was bowel preparation quality at the time of colonoscopy. Endoscopists photographed and labeled representative segments of the ascending, transverse, and descending colon, and research coordinators abstracted data on the volume of water infused and fluid aspirated (measured in cubic centimeters) during the procedure. These data were uploaded into an electronic interface, programmed specifically for this study, which displayed the images and fluid volumes on a standardized template. In order to ensure blinding of the rater, the template did not reveal any patient identification and the program accessed subjects in random order. Two blinded rater employed the standardized Ottawa scoring system(18) using a point-and-click template function to measure the preparation quality in each bowel segment based on photodocumentation and data regarding amount of aspiration required during the procedure. The Ottawa scale yields a composite score reflecting overall fluid (0=small; 1=moderate; 2=large) and cleanliness in the right, mid, and rectosigmoid colon, each rated between 0 and 4 (0="excellent;" 1="good;" 2="fair;" 3="poor;" 4="inadequate"). The total score is calculated by adding the 3 segmental scores and the overall fluid score (range=0-14; lower is better).

Our secondary outcome was bowel preparation quality as measured by the principal endoscopist for each procedure using a standard global 6-point Likert scale (1=inadequate, 2=poor; 3=adequate; 4=fair; 5=good; and 6=excellent). The scores were entered into an electronic form immediately after the procedure using a drop-down menu interface (endoPRO operating system, Pentax Medical). In a previous audit of 932 patients receiving colonoscopy in our institution, we found a linear and statistically significant relationship between bowel preparation Likert scores and polyp yield, supporting the construct validity of the scale in our population.(19) Patients with at least a "good" preparation quality (i.e. quality>5) had, on average, a 0.5 more polyps found than patients with lower preparation quality using the Likert scale (1.6 vs. 1.1 polyps; p=0.0006).(19)

Sample Size Calculation

Although both the Ottawa and the Likert scales are widely used measures of bowel preparation quality, there are no data measuring the minimally clinically important difference on these scales. Therefore, we powered our study to achieve a half standard deviation difference in mean scale scores between the groups - i.e. an effect size of 0.5. This effect size is considered to be clinically meaningful using the rules of Cohen.(20) Assuming a two-tailed 5% significance level with a power of 80%, we required a minimum of 60 subjects per group to demonstrate an effect size of 0.5 standard deviation in Ottawa and Likert scale scores.

Analyses

We calculated descriptive statistics for patients in each group. We then performed bivariate analyses to compare mean Ottawa scores between groups using a t-test, and compared Likert scale scores between groups with a non-parametric rank sum test. We compared the proportion achieving at least a "good" preparation (Likert scale score >5) between groups using chi-squared, and calculated the "number needed to send" in order to achieve one additional "good" preparation with the booklet vs. no booklet. We then performed multivariable logistic regression analysis to predict dichotomous Likert scores while controlling for other predictors of bowel preparation quality, including age, sex, race, and body mass index.(9) We report the odds ratio and 95% confidence intervals of booklet receipt on achieving a "good" or better preparation. We adopted a P-value of <0.05 as evidence for statistical significance, and calculated 95% confidence intervals around point estimates. All analyses were performed using Stata for Windows version 8.2 (StataCorp, College Station, TX).

Phase III: WLAVA vs. DVA Study

For this study, we reproduced all of the same analyses and methods outlined in Phase II, above. The goal of Phase III was to conduce a second validation study in a different VA, and to compare the results with a second cohort at the WLAVA exposed to the booklet. This allowed us not only to test the generalizability of the booklet to another VA center, but also allowed us to measure results between centers. Notably, the WLAVA uses "single-dose" bowel preparations, whereas the DVA uses "split does" preparations. The difference is that split does allows patients to take half their dose the evening before, and the rest the morning of their procedure. Data now reveals that split-dosing is superior to single-dosing. In Phase III, we further tested the booklet in a multi-center RCT to evaluate its additive effect on both single and split-dose preps, and also tested its impact on adenoma detection rate (ADR) and cecal intubation rate (CIR).
We recruited patients from 2 geographically distinct VA Centers: WLAVA and DVA. The prep instructions were standardized for single-dose magnesium citrate or polyethylene glycol (PEG) at WLAVA or split-dose PEG at DVA. We measured "suboptimal" prep as any score less than "good" on a 6-point Likert scale or a total score of <8 on the Ottawa scale. We measured CIR and ADR based on chart and path review in concert with Dr. Deborah Fisher in the Gastroenterology Department at DVA. We performed bivariate analysis with t-tests and X2 to first compare outcomes across all patients, and then stratified by site.

References Cited
9.Ness RM, Manam R, Hoen H, Chalasani N. Predictors of inadequate bowel preparation for colonoscopy. Am J Gastroenterol 2001;96:1797-802.
10.Rex DK, Imperiale TF, Latinovich DR, Bratcher LL. Impact of bowel preparation on efficiency and cost of colonoscopy. Am J Gastroenterol 2002;97:1696-700.
11.Belsey J, Epstein O, Heresbach D. Systematic review: oral bowel preparation for colonoscopy. Aliment Pharmacol Ther 2007;25:373-84.
12.Tan JJ, Tjandra JJ. Which is the optimal bowel preparation for colonoscopy - a meta-analysis. Colorectal Dis 2006;8:247-58.
13.Ericsson A SH. Protocol Analysis: Verbal Reports as Data (2nd edition). MIT Press, 1993.
14.Willis G. Cognitive Interviewing and Questionnaire Design: A Training Manual (working paper #7). National Center for Health Statistics, March, 1994.
15.McLaughlin G. SMOG grading: a new readability formula. J Reading
1969 12:639-46.
16.McLaughlin G. SMOG: Simple Measure of Gobbledygook. Volume 2009.
17.Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the Health Belief Model. Health Educ Q 1988;15:175-83.
18.Rostom A, Jolicoeur E. Validation of a new scale for the assessment of bowel preparation quality. Gastrointest Endosc 2004;59:482-6.
19.Spiegel B CM, Shekelle P, Aberbrook M, Cohen H Adherence to Colonoscopy Quality Indicators in a Large Regional VA Healthcare System. Veteran Administration Health Services Research and Development (VA HSR&D) National Meeting. Baltimore, MD, 2007.
20.Cohen J. Statistical Power Analysis for the Behavioral Science. Academic Press, 1960.
21.Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol 1997;32:920-4.
22.Chen LA, Santos S, Jandorf L, Christie J, Castillo A, Winkel G, Itzkowitz S. A program to enhance completion of screening colonoscopy among urban minorities. Clin Gastroenterol Hepatol 2008;6:443-50.
23.Wolf MS, Fitzner KA, Powell EF, McCaffrey KR, Pickard AS, McKoy JM, Lindenberg J, Schumock GT, Carson KR, Ferreira MR, Dolan NC, Bennett CL. Costs and cost effectiveness of a health care provider-directed intervention to promote colorectal cancer screening among Veterans. J Clin Oncol 2005;23:8877-83.
24.Denberg TD, Coombes JM, Byers TE, Marcus AC, Feinberg LE, Steiner JF, Ahnen DJ. Effect of a mailed brochure on appointment-keeping for screening colonoscopy: a randomized trial. Ann Intern Med 2006;145:895-900.


FINDINGS/RESULTS:
Phase I Results
Booklet Construction

Based on feedback from cognitive interviews we created an initial draft of the educational booklet, entitled Getting Ready for Your Colonoscopy, with the subtitle One and Done: Let's to This Once and Do it Right! The booklet includes a variety of sections based on the needs expressed by patients, including:
-An overview that emphasizes the importance of patient participation to ensure a successful procedure, provides information about risk and consequences of colon cancer, and highlights the risk reduction afforded by screening colonoscopy.
-Use of a visual analogy, based on feedback from provider interviews, to highlight the importance of proper bowel preparation.
-Daily preparatory instructions including the sequence of steps to follow on each day preceding the colonoscopy.
-Pictures of allowable and prohibited foods, including large, clear, color photographs of foods that can and cannot be consumed during preparation for colonoscopy.
-Description of clear liquids, including instructions on how to distinguish a "clear liquid" from other liquids. The booklet introduces the "newsprint test," suggested by nurses in our interviews, that a clear liquid is defined by the ability to read newsprint through the liquid.
-A "guide to stool effluent" based on feedback from patients that most did not know how to interpret their stool effluent - i.e. how to know when they "were done." We created a simple visual color scale for interpreting graded effluent color and transparency; the scale is analogous to the Bristol Scale for objectively judging stool form as a clinical diagnostic aid,21 yet in this instance the scale is designed for interpreting stool output to judge preparation quality.
-Answers to frequently asked questions identified by patients during content development.
-An interactive colonoscopy checklist of sequential steps to follow while preparing for colonoscopy.
In the content validation cohort all sections of the draft booklet achieved helpfulness scores of 4 (quite a bit helpful) or higher for each assessment; no sections were dropped.

Phase II Results
We recruited 220 and 216 patients in the control and booklet groups, respectively. There were no significant differences in characteristics between groups. Seventy-eight percent of evaluable patients randomized to the intervention arm reported receiving the booklet. In intention-to-treat primary analysis including all evaluable patients, blindly rated Ottawa scores were superior in patients receiving the booklet vs. controls (4.4+2.3 vs. 5.1+2.9 [lower score=higher quality]; difference=0.7; 95% CI for difference=0.06 to 1.25; p=0.03). When excluding subjects who never received the booklet intervention (and therefore could not benefit from its content in the first place), patients receiving the booklet had significantly improved bowel preparation quality vs. controls in per protocol analysis (4.2+2.2 vs. 5.1+2.9; difference=0.9; 95% CI=0.27 to 1.53; p=0.005). Mean bowel preparation Likert scores were higher in patients receiving the booklet vs. controls (4.2+1.1 vs. 3.8+1.3; rank sum p=0.006). Preparation was rated as "good" or better in 76% of patients who received the booklet compared to 46% of controls (difference=30%; 95% CI=20% to 40%; p<0.00001). Therefore, for every 3.3 booklets sent out there was 1 additional "good" or better prep vs. controls. In multivariable regression analysis adjusting for type of purgative received, age, sex, race, and body mass index, booklet receipt increased the odds of a "good" prep by 3.7 times (95% CI=2.3-5.8).
From Phase III, we concluded that provision of a novel educational booklet considerably improves preparation quality in patients receiving single-dose purgatives. The effect of the booklet on split-dose purgatives remains untested and was evaluated next in Phase III.

Phase III Results
We analyzed data from 235 and 148 colonoscopies in the control and booklet arms, respectively (95% male, 46% white, mean age 61+11). There was no difference in age, BMI, or black vs. white race between groups. Across sites, 36% vs. 44% of booklet vs. controls had sub-optimal prep, respectively (p=0.04). When limited to DVA, where all patients received split-prep, there was no difference in sub-optimal prep between booklet vs. controls (12% vs. 8%; p=0.27). Across sites, the CIR was 93% vs. 87% in booklet vs. controls (p=0.039); but this effect was absent in patients with split-prep at DVA (98.1% vs. 97.5%; p=0.70). There was no difference in ADR between booklet vs. controls (28.9% vs. 28.5%; p=0.92), regardless of site or prep type.
From Phase III, we concluded that provision of the educational booklet improves bowel prep quality and CIR in patients receiving single-dose, but not split-dose preps. Split-dose was so highly effective that there was little need for ancillary education - the booklet was not additive. In contrast, the booklet was beneficial in less effective single-dose preps. In neither case did the booklet improve ADR.

IMPACT:
The impact of our study has been far-reaching. Locally, we now distribute the booklet to every patient scheduled for colonoscopy at the WLAVA. This is an exciting result of our research, and an example of directly implementing findings from an HSR&D Merit study. Patients report positive experiences with the booklet, and our staff has noticed improvements in bowel preparation among those using the booklet.
UCLA Medical Center also obtained a copy of our booklet and converted it for their own needs. They also now administer the booklet to their patients, and report excellent results.
Furthermore, we have been contacted by innumerable organizations both nationally and internationally since the publication of our Phase I and II results. Our booklet is now being used in centers across the world, and has even been translated into Turkish, among other languages.
Finally, a publishing company (SLACK Incorporated) company learned about our booklet, obtained it from us under the public domain accessibility, and converted it into a commercial product. Our group has no involvement in the production of the booklet or profits in any way, but we are gratified that the booklet now enjoys even wider attention as a result of SLACK's involvement.

---

External Links for this Project

NIH Reporter

Grant Number: I01HX000143-01
Link: https://reporter.nih.gov/project-details/7748638

---

PUBLICATIONS:

Journal Articles

1. Kaneshiro M, Ho A, Chan M, Cohen H, Spiegel BM. Colonoscopy yields fewer polyps as the day progresses despite using social influence theory to reverse the trend. Gastrointestinal endoscopy. 2010 Dec 1; 72(6):1233-40. [view]
2. Spiegel BM. Does time of day affect polyp detection rates from colonoscopy? Gastrointestinal endoscopy. 2011 Mar 1; 73(3):476-9. [view]
3. Spiegel B. H. pylori: should we still be looking for it? What should we be doing about it? A case-based approach. Digestive endoscopy : official journal of the Japan Gastroenterological Endoscopy Society. 2012 May 1; 24(3):208. [view]
4. Khanna D, Krishnan E, Dewitt EM, Khanna PP, Spiegel B, Hays RD. The future of measuring patient-reported outcomes in rheumatology: Patient-Reported Outcomes Measurement Information System (PROMIS). Arthritis care & research. 2011 Nov 1; 63 Suppl 11:S486-90. [view]
5. Saliminejad M, Bemanian S, Ho A, Spiegel B, Laine L. The yield and cost of colonoscopy in patients with metastatic cancer of unknown primary. Alimentary pharmacology & therapeutics. 2013 Sep 1; 38(6):628-33. [view]

DRA: Health Systems, Cancer
DRE: Treatment - Comparative Effectiveness
Keywords: Cancer, Education (patient), Satisfaction (patient)
MeSH Terms: none