Introduction
Use of colonoscopy for colorectal cancer (CRC) screening has been increasing [1]
[2]. Furthermore, colonoscopy is the diagnostic procedure to be performed when there
is an abnormality detected by any other screening modality because of the opportunity
to perform biopsies and remove polyps seen during the examination. However, colonoscopy
requires full bowel preparation using oral laxative agents. The need for a complete
bowel preparation process with dietary restriction and laxative ingestion can be a
major barrier to undergoing colonoscopy for many patients. Lack of tolerability of
the bowel preparation laxatives reduces the willingness to undergo colonoscopy and
portends poor clinical outcomes including poor bowel preparation, missed colorectal
lesions, and reduction in willingness to undergo future CRC screening.
Reduction in volume of laxatives to consume and flavoring of the laxatives have been
undertaken with inconsistent results [3]
[4]
[5]
[6]
[7]
[8]. Furthermore, low-volume laxatives tend to be “non-preferred brands” for some third-party
payers and they are generally associated with higher co-pays. Therefore, the 4 L polyethylene
glycol (PEG), solution remains the major colonoscopy laxative used in many practices,
especially among underserved populations with low socioeconomic status. Tolerability
of this preparation is essential and it is the target of this study.
We hypothesized that patients would have a better experience if they tasted the bowel
preparation laxative with and without flavoring and then decided how they wanted to
drink the rest of the laxative since taste preferences vary widely from person to
person. In this clinical trial, we investigated the effect of patterns of flavoring
of laxatives on bowel preparation experience of patients undergoing outpatient colonoscopy
and their clinical outcomes with adequate bowel preparation and subsequent adenoma
detection.
Patients and methods
Study population
We recruited competent, non-institutionalized adult men and women who were scheduled
for colonoscopy from gastroenterology clinics at Howard University from February 11,
2014 to September 10, 2014. The patients provided information on their demographic
characteristics and lifestyle factors. All participants gave written informed consent
and gave permission for their colonoscopy and pathology reports to be reviewed. The
project was approved by the Howard University Institutional Review Board (IRB-13-MED-60).
This was a pilot/feasibility project in preparation for a larger study to improve
bowel preparation process and experience for outpatient colonoscopy. The investigators
determined that approximately 100 participants in each arm would yield sufficient
baseline data to guide subsequent studies.
Inclusion and exclusion criteria
We included adult patients who were scheduled for colonoscopy and gave informed consent.
We excluded patients with a personal history of CRC and those at high risk for CRC
as determined by history of familial adenomatous polyposis syndrome (FAP), family
history of hereditary non-polyposis colorectal cancer syndrome (HNPCC), and personal
history of inflammatory bowel disease (ulcerative colitis or Crohn’s disease). Furthermore,
we also excluded participants who had undergone bowel resection regardless of the
indication for the surgery.
Randomization assignment
This was a single-blind (endoscopist), parallel design, three-arm pilot clinical trial.
A total of 313 patients were randomly assigned to one of three groups using a computer-generated
simple randomization sequence with instructions placed in sealed envelopes (concealed
assignment). Group 1 patients (n = 101) were issued a gallon of PEG and were instructed
to drink it without flavoring (unflavored group). Group 2 patients (n = 113) were
issued PEG with flavoring packs (citrus or lemon per patient’s preference) and were
instructed to flavor the entire gallon and drink. Group 3 patients (n = 99) were issued
PEG with flavoring packs but were instructed to taste a cup of PEG without flavoring
and another cup with flavoring and decide for themselves how they wanted to drink
the rest (choice group). A clear liquid diet the day before colonoscopy and split-dose
bowel preparation were recommended for all patients. Of those randomized, those who
completed the study by undergoing colonoscopy were group 1 (n = 84), group 2 (n = 90)
and group 3 (n = 82) and constituted the primary analysis.
Outcome assessment
On the day of their colonoscopy, patients completed outcome forms detailing their
bowel preparation experience (subjective assessment) using a rating scoring scale
from 1 to 10 where (1 = unbearable; 5 = neutral; 10 = pleasant). Information gathered
included their ratings of the bowel preparation laxative's taste, ease of consumption,
presence or absence of nausea, actual vomiting, whether they consumed the entire recommended
laxative, and their overall ratings of the bowel preparation experience. They were
also asked whether they would like to have the same bowel preparation laxative if
they needed colonoscopy in the future. Furthermore, they were asked whether they felt
that their bowel was adequately clean for their colonoscopy (subjective assessment
by patients).
During colonoscopy, the endoscopists who were all blinded to the randomization assignment
of the patients rated the bowel preparation of the patients (objective assessment
by endoscopists) using Aronchick scale selected in a drop-down menu format in their
endoscopy reports. The bowel preparation was scored as poor, fair, good, very good
or excellent as a function of how clearly the colonic mucosa was seen and the percentage
examined clearly. Bowel preparation of poor and fair were considered as to be inadequate
while good, very good and excellent designations were regarded as adequate bowel preparation.
Findings of colorectal neoplasia were also recorded from the endoscopy and pathology
reports.
Statistical analyses
We compared baseline characteristics of the participants by randomization groups.
We used chi squared test to compare the reported bowel preparation challenges of patients
such as experience of nausea and vomiting by randomization arm. We compared subjective
ratings of the patients’ bowel preparation experience in terms of ease of drinking
and taste of the laxatives using analysis of variance (ANOVA). We calculated and compared
the percentage of patients with adequate bowel preparation as graded by the endoscopist
in the colonoscopy reports. Adequate bowel preparation was defined as a description
of good, very good or excellent on the Aronchick scale while inadequate bowel preparation
was defined as fair or poor rating. The reference group for our analyses was the unflavored
arm (group 1). We compared those issued with flavoring packs (group 2 and group 3)
with group 1. We used log binomial models to compare the adequacy of bowel preparations
and the presence of colorectal adenoma by randomization assignment. We calculated
relative risks (RR) and 95 % confidence intervals (95 % CI). P < 0.05 was regarded as statistically significant. We used Stata statistical software
version 14.2 for all our analyses.
Results
Participant characteristics
A total of 313 participants were randomized (group 1 = 101; group 2 = 113 and group
3 = 99). Of
them, 57 did not show up for their procedures or cancelled their colonoscopy appointments.
Therefore, 256 (81.8 %) participants showed up for their procedures, underwent colonoscopy
and
were included in the final analysis (group 1 = 84; group 2 = 90 and group 3 = 82).
CONSORT
diagram ([Fig. 1]) shows the flow of participants through the study.
The characteristics of the subjects in the three arms were comparable ([Table 1]).
Fig. 1 CONSORT diagram of flow of participants through the study. From: Moher D, Liberati
A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic
Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097
Table 1
Characteristics of study participants.
Characteristics
|
Unflavored group (n = 84)
|
Flavored group (n = 90)
|
Choice group (n = 82)
|
Mean age in years (SD)
|
57.0 (10.0)
|
56.6 (8.5)
|
58.0 (9.9)
|
Sex, n (%)
|
|
37 (44.0)
|
41 (45.6)
|
44 (53.7)
|
|
47 (56.0)
|
49 (54.4)
|
38 (46.3)
|
Black, n (%)
|
|
10 (12.4)
|
10 (11.8)
|
2 (2.7)
|
|
71 (87.6)
|
75 (88.2)
|
72 (97.3)
|
Married, n (%)
|
|
59 (76.6)
|
61 (70.1)
|
48 (64.0)
|
|
18 (23.4)
|
26 (29.9)
|
27 (36.0)
|
Highest education, n (%)
|
|
47 (56.0)
|
52 (59.1)
|
49 (60.5)
|
|
37 (44.0)
|
36 (40.9)
|
32 (39.5)
|
Yearly household income, n (%)
|
|
19 (23.5)
|
20 (25.0)
|
26 (33.8)
|
|
62 (76.5)
|
60 (75.0)
|
51 (66.2)
|
History of smoking, n (%)
|
|
41 (49.4)
|
49 (55.7)
|
31 (38.8)
|
|
42 (50.6)
|
39 (44.3)
|
49 (61.2)
|
Indication for colonoscopy, n (%)
|
|
45 (54.2)
|
46 (52.9)
|
42 (53.2)
|
|
38 (45.8)
|
41 (47.1)
|
37 (46.8)
|
Laxative drinking pattern of choice group participants (group 3)
Participants in group 3 (choice group) were issued flavor packets and were asked to
taste the laxative flavored and unflavored and decide how they want to drink the rest
of the laxative. Of 75 participants in this group who answered questions about how
they eventually drank the laxative after tasting the flavored and unflavored options,
13 (17.3 %) participants drank the rest of the laxatives without flavoring while 53
(70.7 %) participants drank the rest with flavoring. The remaining nine participants
(12 %) drank the rest of the laxative by flavoring some cups before drinking and drinking
some cups of laxative without flavoring. However, we did not ascertain the exact percentage
flavoring that they did.
Bowel preparation experience
There were no differences in ratings of the bowel preparation process across the randomization
groups in terms of nausea, vomiting, drinking the entire laxative, taste of the laxatives,
ease of drinking the laxatives, and overall experience with the bowel preparation
process. However, a higher percentage of those in the choice arm (group 3) reported
that they would be willing to take the same laxative if they needed to undergo repeat
colonoscopy in the future ([Table 2]).
Table 2
Comparison of bowel preparation experience by randomization arm.
Experience
|
Unflavored group (n = 84)
|
Flavored group (n = 90)
|
Choice group (n = 82)
|
P value
|
Had nausea (yes)
|
23.5 %
|
26.9 %
|
30.3 %
|
0.63
|
Vomited (yes)
|
3.7 %
|
11.5 %
|
10.5 %
|
0.16
|
Drank entire laxative solution (yes)
|
88.9 %
|
91.0 %
|
89.5 %
|
0.90
|
Thinks colon is adequately clean for colonoscopy (yes)
|
96.3 %
|
97.4 %
|
98.7 %
|
0.64
|
Want the same laxative next time (yes)
|
72.5 %
|
81.3 %
|
88.9 %
|
0.04
|
Mean rating of taste of laxative (SD)[1]
|
6.8 (2.8)
|
7.7 (2.5)
|
7.3 (2.6)
|
0.67
|
Mean rating of ease of consuming laxative (SD)
|
7.5 (2.4)
|
8.2 (2.2)
|
7.8 (2.4)
|
0.53
|
Mean rating of overall experience (SD)
|
7.9 (2.1)
|
8.1 (2.1)
|
7.9 (2.5)
|
0.18
|
1 Using a rating scoring scale from 1 to 10 where (1 = unbearable; 5 = neutral; 10 = pleasant).
Adequacy of bowel preparation and adenoma detection
Overall, when asked if they felt that their colon was adequately clean for colonoscopy,
97.4 % of all patients affirmed that their colon was clean without any difference
by randomization arms: 96.3 %, 97.4 %, 98.7 % for groups 1, 2, and 3, respectively
(P = 0.64). This contrasts with an actual overall 81.6 % adequacy of bowel preparation
from endoscopic assessment. Adequacy of bowel preparation was highest at 89.3 % among
group 1 patients who drank the unflavored laxative when compared to 80 % adequacy
among group 2 patients who flavored all the laxatives and 75.6 % among group 3 patients
who tasted flavored and unflavored laxatives before making their choice about how
they want to drink the rest ([Table 3]). This represented a 15 % reduced risk of adequate bowel preparation with the choice
group 3 when compared to those who drank the entire laxative without flavoring. Adenomatous
polyps were detected in 40.5 % of patients in group 1, 23.3 % in group 2 and 39.0 %
in group 3 representing a 42 % statistically significant reduced risk of adenoma detection
among those who flavored all the laxative as compared to group 1 patients who drank
unflavored laxative ([Table 3]). Adenoma detection was comparable among group 1 and group 3 subjects.
Table 3
Comparison of satisfaction with bowel laxative, adequacy of bowel preparation, and
adenoma detection by patterns of flavoring of PEG.
Randomization group
|
% willing to have to have same laxative again
|
RR (95 % CI)
|
% with good to excellent bowel preparation
|
RR (95 % CI)
|
% with adenoma
|
RR (95 % CI)
|
Unflavored
|
72.5
|
Reference
|
89.3
|
Reference
|
40.5
|
Reference
|
Flavored
|
81.3
|
1.12 (0.94–1.33)
|
80.0
|
0.90 (0.79–1.02)
|
23.3
|
0.58 (0.37–0.91)
|
Choice
|
88.9
|
1.23 (1.05–1.44)
|
75.6
|
0.85 (0.73–0.98)
|
39.0
|
0.96 (0.66–1.40)
|
Discussion
It is well established that the bowel preparation process remains a major barrier
to uptake of colonoscopy both as a screening modality and as a diagnostic tool in
patient care. Efforts to improve patient bowel preparation experience can potentially
increase uptake of colonoscopy among the population. In this study, we examined whether
the pattern of flavoring can improve patient experience with attendant improvement
in actual bowel preparation and adenoma detection at colonoscopy. Our study revealed
that flavoring the bowel preparation laxative did not improve the overall experience
for patients, but those who tasted unflavored and flavored solution and then decided
how they would take the bowel laxatives (choice group 3) were more willing to have
the same laxative if they needed to undergo colonoscopy in the future. Those who drank
the unflavored laxatives were more dissatisfied and were the least likely to want
to have the same laxative in the future if they needed to undergo colonoscopy. However,
they had the highest adequacy of bowel preparation and subsequent highest percentage
of adenoma detection. This suggests a minimum justification effect in the bowel preparation
process.
Minimum justification is the phenomenon in which individuals rationalize overcoming
some inconvenience to achieve small benefit at the individual level. Individuals use
internal motivation to justify behavior. Drinking unpalatable unflavored PEG is an
unpleasant activity in order to undergo an unpleasant procedure (colonoscopy) to prevent
a rare event (colon cancer). Yet patients understand the need for this unpleasant
activity and use their internal resources to ensure compliance, suggesting that people
give themselves a positive reason for why they do what they generally do not want
to do. Our study suggests that internal motivation of patients rather than merely
more palatable taste of the laxative may play more of a role in terms of achieving
adequate bowel preparation. Afterall, achieving adequate bowel preparation involves
more than drinking laxatives because there are additional dietary modification and
fluid ingestion requirements.
We are not aware of any study that has evaluated patterns of flavoring of PEG laxative
to assess patient bowel preparation experience, actual bowel preparation and yield
of adenoma at colonoscopy for direct comparison with our study. Nevertheless, our
findings are comparable to the report of Hayes et al. [3] of their randomized control trial that involved 130 patients. The authors reported
that flavoring the PEG solution did not make any difference in bowel cleansing with
adequate bowel preparation occurring among 75 % of participants who received flavored
laxative and 76 % among those who received unflavored laxative.
In another study which evaluated taste and volume of polyethylene glycol laxative
use among patients scheduled for elective colonoscopy in the Netherlands, Szojda et
al. [4] compared 3-L sulphate-free PEG (SF-PEG) and a 4-L PEG solution and evaluated patient
acceptability and tolerability of the laxatives and effectiveness of the bowel preparation
process among 102 patients who underwent colonoscopy. They reported that there were
no differences in compliance, taste, tolerance and adequacy of bowel preparation.
However, a statistically significant percentage of those who received the lower-volume
solution expressed willingness to have the same laxative in the future.
For our study, we hypothesized that patients would have a better experience if they
tasted the bowel preparation laxative with and without flavoring and then decided
how they wanted to drink the rest of the laxative because taste preferences vary widely
from person to person. Participants randomized to this choice arm had a more favorable
opinion of the laxative in terms of taking it the next time if they needed colonoscopy
in the future, but there was no difference in their tolerability of the bowel preparation
process when compared to those who drank unflavored PEG, who were less likely to want
the same laxative in the future. In contrast to our hypothesis, those in the choice
group actually had lower adequacy of bowel preparation for their colonoscopy. The
findings of Szojda et al. [4] and our study suggest that internal motivation to follow bowel preparation instructions
probably has more impact on the adequacy of bowel preparation than the volume or taste
of PEG. However, other studies suggested that 2-L PEG solution with ascorbic acid
was noninferior to the 4-L PEG in bowel cleansing but was better in taste and compliance
[5]
[6]
[7]. Therefore, there may be residual differences that could not be captured in our
study. Furthermore, lower-volume laxative with better-tasting solutions may increase
the willingness of patients to undergo colonoscopy in the future.
Another important finding of our study is the fact that participants grossly overestimated
the adequacy of their bowel preparation for their colonoscopy. Subjective estimation
by patients of their bowel preparation adequacy was 96.3 %, 97.4 % and 98.7 % for
groups 1, 2 and 3, respectively, whereas the objective bowel preparation adequacy
from endoscopic examination was 89.3 %, 80.0 % and 75.6 % for groups 1, 2 and 3, respectively.
In fact, among six patients who opined that their bowel preparation mighty be less
than optimal, four patients (66.7 %) actually had adequate bowel preparation. This
suggest that endoscopists should not rely on the positive estimation of bowel preparation
adequacy from their patients.
A major strength of our study was that our randomized clinical trial was conducted
among a predominantly underserved inner city minority populations who are frequently
under-represented in biomedical research. Yet, our findings are comparable to those
obtained among majority populations in the United States and studies conducted in
other parts of the world. This suggests that improving the bowel preparation experience
of patients is a global necessity to reduce the burden of colorectal diseases including
colorectal cancer in the world. Our study is limited by the relatively small number
of participants. Therefore, it is possible that clinically relevant effects may be
seen in larger studies. Nonetheless, the number of included patients in our study
is larger than many previous studies on this important topic.
Conclusion
In conclusion, the patterns of flavoring of PEG laxative were not associated with
improved tolerability of bowel preparation process for colonoscopy but those who tasted
unflavored and flavored PEG and then decided how they wanted to drink the laxative
were more willing to have the same laxative in the future. However, those who drank
unflavored PEG were less satisfied but had better clinical outcomes in terms of adequacy
of bowel preparation and subsequent adenoma detection, suggesting minimum justification
effect in the bowel preparation process. The implication of this is that patients
should be encouraged to use their internal motivation to overcome the lack of palatability
they may experience in the course of undergoing the bowel preparation process for
their colonoscopy.