Introduction
High quality colonoscopy cannot be performed without adequate bowel preparation [1 ]. Inadequate bowel preparation is associated with lower lesion detection rates and
is a major predictor of failed cecal intubation. Moreover, the need for repeat procedures
increases healthcare costs. Importantly, patients with poor bowel preparation are
less likely to be satisfied about their colonoscopy and have a lower willingness to
repeat colonoscopy [2 ]. This is detrimental to the benefit of colorectal cancer (CRC) screening, as lower
participation rates significantly decrease colonoscopy screening efficiency [3 ].
Efforts have been made to increase both bowel preparation efficiency and tolerability.
Meta-analyses have demonstrated a higher tolerability of lower volume bowel preparation
fluids [4 ]. A 2 L polyethylene glycol (PEG) solution with added ascorbate (2L PEG+Asc) has
been shown to be noninferior to the “gold standard” of 4 L of PEG, and is associated
with a higher willingness to repeat [1 ]. Recently, even lower volume bowel preparations have been introduced, such as 1L
PEG+Asc with added sodium sulfate.
Nevertheless, the impact of bowel preparation on patients’ working lives and health-related
quality of life (HRQoL) may be significant, potentially leading to postponement and
sometimes to patients refraining from undergoing colonoscopy [2 ]
[5 ]. Despite its negative correlation with both bowel preparation efficacy and colonoscopy
uptake, little is known about the effect of bowel preparation on patient-reported
outcomes, and the differences between various volumes of laxative. In a prospective
cohort of 1100 patients, Collatuzzo et al. reported a correlation between patient-related
factors and symptoms caused by bowel preparation [6 ]. Fuccio et al. demonstrated that bowel preparation and colonoscopy significantly
impacted work, with up to one-third of patients reporting work absenteeism or reduced
performance, particularly when using same-day full-dose regimens or experiencing procedure-related
symptoms [7 ]. They did not however include the 1L PEG+Asc regimen, and the studies were not randomized,
which may lead to selection bias of patients choosing the laxative of their choice.
Additionally, unemployed patients were excluded, while a large proportion of the patient
population undergoing colonoscopy also consists of patients who are retired or not
working owing to illness. Other studies on the impact of bowel preparation on patient-reported
outcomes and work productivity are scarce and have mainly involved larger volume preparations
(2 L and higher).
Insight into the impact on symptoms experienced, work, and HRQoL may help patients
and caregivers in selecting a bowel preparation regimen and could reduce the barriers
experienced to undergoing colonoscopy. Therefore, this study aimed to compare low
volume and intermediate volume bowel preparation in terms of their bowel cleansing
efficacy, tolerability, and impacts on work and HRQoL.
Methods
Study design
We performed an open-label, noninferiority randomized controlled trial in four Dutch
hospitals. We included adult patients referred for diagnostic, screening, or surveillance
colonoscopies. Exclusion criteria included: therapeutic procedures, inpatient status,
emergency colonoscopy, inflammatory bowel disease, American Society of Anesthesiologists
(ASA) score ≥4, (partial) colectomy, inability to provide informed consent, inability
to complete Dutch questionnaires via email (owing to language barrier, lack of email,
visual impairment, or dementia), common contraindications for bowel preparation or
its ingredients, and indications for intensified bowel preparation as assessed by
the treating physician.
The study was approved by the Radboud University Medical Center Medical Ethics Committee
(Medisch Ethische Toetsingscommissie Oost-Nederland, registration number NL79014.091).
It was conducted in accordance with the Declaration of Helsinki and Good Clinical
Practice guidelines. All patients provided written informed consent. An independent
monitor reviewed the study data and informed consent process, and two research team
members independently verified all study data.
Randomization
After informed consent had been obtained, eligible patients were randomized 1
:
1 using a block design with variable block sizes (4, 6, 8) via a secure web-based
system (Castor Electronic Data Capture, Amsterdam, The Netherlands). Randomization
was stratified only by study site and prior experience of bowel preparation. The randomization
sequence was blinded to study team members, but patients and healthcare providers
were not blinded to the allocation to minimize interference with routine medical care.
Blinding of patients was not possible owing to the differing volumes. We included
additional questions on patient experiences and satisfaction in the questionnaires
to assess the impact of potential modifying factors without disrupting standard care.
Study procedures
Bowel preparation and colonoscopy
Patients received either 1L PEG+Asc (Pleinvue, Norgine) or 2L PEG+Asc (Moviprep, Norgine)
[8 ] in an overnight split-dose regimen, each dose followed by 0.5 L clear liquids (additional
fluids were allowed to reflect real-life conditions and improve adherence). The last
dose needed to be completed 2 hours before travel to the hospital. Patients were also
instructed to follow a low residue diet starting 2 days before colonoscopy, followed
by a liquid diet upon starting to take the laxatives. This was based on local clinical
practice, although this does deviate from the European Society of Gastrointestinal
Endoscopy (ESGE) bowel preparation guideline, which advises 1 day of low residue diet
[1 ]. Instructions were provided by specialist nurses, supplemented with a detailed leaflet
and patient diary to record the volume of additional fluids consumed.
Colonoscopies were performed as part of routine medical care, with quality control
ensured under the Dutch national CRC screening program [8 ], by experienced endoscopists, with the patients under conscious sedation or propofol.
Bowel preparation quality was assessed using the Boston Bowel Preparation Scale (BBPS)
after flushing, in line with routine clinical practice, with adequate preparation
defined as a minimum score of 2 per segment [1 ]. High quality bowel preparation was defined as a BBPS of 3–3-3 (maximum score).
Questionnaires
Participants completed online questionnaires at two timepoints: before starting preparation
(baseline) and within 1 week after colonoscopy. The first questionnaire collected
baseline sociodemographic and lifestyle characteristics, clinical details including
bowel symptoms, prior bowel preparation experience, risk factors for inadequate preparation
based on the existing literature [9 ], and HRQoL data. The second questionnaire assessed patients’ experiences with the
bowel preparation and colonoscopy, impact on work productivity and costs, and HRQoL.
To assess work productivity impact, we used the validated Institute for Medical Technology
Assessment Productivity Cost Questionnaire (IPCQ). Unlike the commonly used Work Productivity
and Activity Impairment questionnaire, the IPCQ also considers the impact of unpaid
labor [10 ]. HRQoL was assessed using the EuroQol Group five dimensions five levels (EQ-5D-5L).
Previous studies have indicated that the endoscopic procedure minimally affects HRQoL
outcomes, with bowel preparation being the primary factor influencing HRQoL [11 ]
[12 ].
The second questionnaire also included the Mayo Florida Bowel Preparation Tolerability
Questionnaire (MBTQ), validated in outpatient colonoscopy settings in the USA [13 ]. For our study, the MBTQ was translated using a forward–backward translation process
by a professional translator. In addition to the MBTQ, we assessed patient satisfaction
using the Dutch Gastrointestinal Endoscopy Satisfaction Questionnaire (D-GESQ) [14 ]. Finally, we recorded self-reported post-procedural adverse events (AEs) with the
second questionnaire, and supplemented this with information from hospital records.
Outcomes and objectives
Our primary outcome was the proportion of patients with adequate bowel preparation,
defined as a segmental BBPS score of ≥2. The aim was to demonstrate noninferiority
of 1L PEG+Asc vs. 2L PEG+Asc. Secondary outcomes included tolerability (assessed via
the MBTQ), HRQoL (measured by EQ-5D-5L), and work productivity (evaluated using the
IPCQ) to provide an exploratory comprehensive assessment of patient experience before
and after bowel preparation.
Statistical analysis
Baseline characteristics were summarized using mean and SD, median and interquartile
range (IQR), or proportions, as appropriate. Comparisons were made using Student’s
t test, Mann–Whitney U test, Wilcoxon’s rank test, chi-squared test, or Fisher’s exact test.
Primary outcome analysis was conducted on both an intention-to-treat (ITT) and per-protocol
basis. The difference between the two groups is presented with a 95%CI. The noninferiority
margin was set to 5%. If noninferiority was confirmed, superiority was tested using
Fisher’s exact test. Questionnaire outcomes were analyzed per their manuals. The EQ-index
of the EQ-5D-5L was compared with the general population [15 ]. A minimal clinically important change was defined as 0.5 SD of the change [16 ]. Costs of absenteeism, presenteeism, and unpaid labor were calculated using Dutch
government-set reference costs (Zorginstituut Nederland), with hourly costs of €34.75
for paid work and €14 for unpaid labor.
Although this study was a randomized controlled trial and not designed for prediction
modeling, we performed exploratory multivariable logistic regression to identify factors
potentially associated with willingness to repeat the bowel preparation regimen. These
analyses were conducted to better understand patient-reported experiences, and not
for causal inference or predictive accuracy. The intervention group was included as
a covariate in the model, and other variables were selected based on the literature
and outcome events. After linearity and multicollinearity had been assessed, variables
with a P value <0.2 on univariable analysis were included in the multivariable analysis, using
backward propagation for model selection. Odds ratios (ORs) and 95%CIs were reported.
Missing questionnaire data were handled through multiple imputation after confirming
data were missing at random, using predictive mean matching with 50 iterations to
create 50 imputed datasets.
All analyses were performed using SPSS statistics (IBM, version 25.0), with a two-sided
significance level of 5%. Secondary outcomes were assessed in an exploratory fashion,
therefore multiple-testing corrections were not applied.
Sample size calculation
Assuming a 95% adequate preparation rate in both groups and a noninferiority margin
of 5%, a sample of 470 patients (235 per group) was calculated as being required to
demonstrate noninferiority using one-sided testing, with an alpha of 5% and 80% power.
Results
Baseline characteristics
From May 2022 to February 2023, we enrolled 509 patients across four centers. After
42 patients had been excluded because of consent withdrawal or other exclusion criteria,
238 were randomized to the 1L PEG+Asc group and 229 to the 2L PEG+Asc group. The ITT
population included 467 patients. In the per-protocol analysis, 35 patients were excluded
owing to noncompliance or logistical treatment arm switches, resulting in 432 patients
([Fig. 1 ]).
Fig. 1 Flow chart of patient inclusions and exclusions. IBD, inflammatory bowel disease;
PEG+Asc, polyethylene glycol with added ascorbate.
The baseline characteristics of the two groups were well balanced ([Table 1 ]). The questionnaire response rate was 87.8% (n = 410) and was not significantly
different between groups (P = 0.32). Half of the study population (49.3%) had previous experience with bowel
preparation, with 2L PEG+Asc being the most common prior regimen (24.9% and 29.8%).
As for pre-existent risk factors for inadequate bowel preparation, most patients had
a fair physical activity level, and 88.2% and 89.9% did not have any risk factors
for poor bowel preparation. The background of the patients was predominantly Dutch
(96.2% and 96.6%), with intermediate to high levels of education; around half were
employed (51.7% and 54.0%).
Table 1 Baseline characteristics of the patients in the two groups that received either 1
L or 2 L of polyethylene glycol (PEG) and ascorbate (Asc) in the intention-to-treat
(ITT) analysis.
1L PEG+Asc (n = 238)1
2L PEG+Asc (n = 229)2
ASA, American Society of Anesthesiologists; BMI, body mass index; CVA, cerebrovascular
accident; IQR, interquartile range; TCA, tricyclic antidepressant.
1 Missing data, n = 14 (5.9%).
2 Missing data, n = 22 (9.6%).
Response rate, n (%)
223 (93.7)
206 (90.0)
218 (91.6)
204 (89.1)
212 (89.1)
198 (86.5)
Median time between questionnaires in days (IQR)
8 (4–13)
9 (5–14)
Age, median (IQR), years
63 (55–71)
64 (55–70)
Sex, female, n (%)
101 (42.6)
98 (42.8)
ASA score, median (IQR)
2 (1–2)
2 (1–2)
BMI, median (IQR)
25.9 (23.6–29.7)
26.3 (23.4–29.4)
Prior experience with bowel preparation, n (%)
113 (49.3)
113 (47.5)
Number of prior colonoscopies, median (IQR)
2 (1–3)
2 (1–3)
Prior laxative used, n (%)
71 (29.8)
57 (24.9)
7 (2.9)
14 (6.1)
3 (1.3)
3 (1.3)
2 (0.8)
4 (1.7)
9 (3.8)
7 (3.1)
3 (1.3)
2 (0.9)
34 (14.3)
34 (14.8)
Willingness-to-repeat prior laxative, n (%)
65 (91.5)
49 (86.0)
6 (85.7)
13 (92.9)
3 (100.0)
3 (100.0)
2 (100.0)
4 (100.0)
7 (77.8)
6 (85.7)
2 (66.7)
2 (100.0)
28 (82.4)
26 (76.5)
Colonoscopy indication, n (%)
75 (31.5)
89 (38.9)
77 (32.4)
71 (31.9)
84 (35.3)
Smoking, n (%)
26 (11.1)
25 (11.2)
113 (48.3)
107 (48.0)
94 (40.2)
91 (40.8)
Migration background, n (%)
217 (96.9)
200 (96.2)
3 (1.3)
5 (2.4)
2 (0.9)
0
0
2 (0.9)
2 (0.9)
1 (0.5)
Education level, n (%)
3 (1.3)
0
4 (1.8)
3 (1.4)
126 (58.0)
119 (57.5)
89 (39.8)
82 (39.6)
2 (0.9)
3 (1.4)
Marital status, n (%)
21 (9.4)
20 (9.6)
149 (66.5)
150 (72.1)
28 (12.5)
25 (12.0)
10 (4.5)
4 (1.9)
14 (6.3)
7 (3.4)
2 (0.9)
2 (1.0)
Physical activity level, n (%)
0
2 (1.0)
9 (4.0)
3 (1.4)
118 (52.7)
100 (48.3)
43 (19.2)
33 (15.9)
54 (24.1)
69 (33.3)
Paid job, n (%)
121 (54.0)
107 (51.7)
103 (46.0)
100 (48.3)
Work activity per week, mean (SD)
31.8 (11.7)
32.6 (11.4)
4.2 (1.1)
4.3 (1.1)
Occupational level, n (%)
1 (0.4)
2 (1.0)
90 (40.2)
79 (38.2)
19 (8.5)
19 (9.2)
17 (7.6)
6 (2.9)
2 (0.9)
1 (0.5)
7 (3.1)
7 (3.4)
82 (36.6)
85 (41.1)
6 (2.7)
8 (3.9)
Risk factors for poor bowel preparation, n (%)
214 (89.9)
202 (88.2)
28 (11.8)
26 (11.4)
27 (11.3)
27 (11.8)
22 (9.2)
18 (7.9)
3 (1.3)
1 (0.4)
3 (1.3)
0 (0)
1 (0.4)
1 (0.4)
4 (1.7)
4 (1.7)
191 (80.3)
189 (79.9)
6 (2.5)
3 (1.3)
3 (1.2)
2 (0.9)
Bristol stool scale score, n (%)
5 (2.2)
4 (1.9)
14 (6.3)
8 (3.9)
39 (17.4)
39 (18.8)
91 (40.6)
87 (42.0)
31 (13.8)
29 (14.0)
39 (17.4)
36 (17.4)
5 (2.2)
4 (1.9)
4 (3–5)
4 (4–5)
History of inadequate bowel preparation, n (%)
7 (6.5)
6 (6.1)
Bowel cleansing efficacy
In the 1L PEG+Asc group, 96.1% (95%CI 92.6% to 98.0%) and 96.4% (95%CI 92.9% to 98.3%)
of colonoscopies were adequately prepared in the ITT and per-protocol analyses, respectively,
compared with 96.4% (95%CI 93.0% to 98.3%) and 96.6% (95%CI 92.9% to 98.5%), respectively,
in the 2L PEG+Asc group (intention to treat, P = 0.84; per protocol, P = 0.92) ([Fig. 2 ]; Table 2 ). The between-group difference was −0.4 percentage points (95%CI −4.0 to 3.3 percentage
points) for the ITT analysis and −0.2 percentage points (−3.9 to 3.6 percentage points)
for the per-protocol analysis. This met our predefined 5% noninferiority margin.
Fig. 2 Proportion of adequately cleansed patients using 1L or 2L polyethylene glycol (PEG)
bowel preparation (95%CI). The difference falls within our predefined noninferiority
margin of 5% (∆), establishing noninferiority.
The median BBPS scores were significantly higher in the 1L PEG+Asc group for the right
(P = 0.01) and transverse colon (P = 0.004), but not significantly different in the left colon. High quality cleansing
(BBPS 3–3-3) was more common in the 1L PEG+Asc group, with 72.7% (95%CI 67.0% to 78.0%)
vs. 63.8% (95%CI 57.5% to 70.0%) in the 2L PEG+Asc group for the ITT population (P = 0.04). Because of the high rates of adequate cleansing, in-depth analysis of the
reasons of inadequate bowel preparation was not performed. Other colonoscopy quality
parameters are detailed in [Table 2 ].
Table 2 Comparison of colonoscopy and bowel preparation outcomes for the patients in the two
groups that received either 1 L or 2 L of polyethylene glycol (PEG) and ascorbate
(Asc).
1L PEG+Asc1
2L PEG+Asc2
P value
Asc, ascorbate; BBPS, Boston Bowel Preparation Scale; IQR, interquartile range; PEG,
polyethylene glycol.
1 Missing from questionnaires, n = 17 (7.1%).
2 Missing from questionnaires, n = 24 (10.5%).
3 Adequate bowel preparation was defined as a minimal segmental BBPS score of ≥2.
4 Complete BBPS could not be assessed owing to incomplete colonoscopy, n = 8.
5 Complete BBPS could not be assessed owing to incomplete colonoscopy, n = 4.
6 High quality bowel preparation was defined as a BBPS score of 3 in all segments.
Intention-to-treat analysis
n = 238
n = 229
Adequate cleansing3 , n (%) [95%CI]
221 (96.1)4
[92.6 to 98.0]
217 (96.4)5
[93.0 to 98.3]
0.84
Difference in proportions, % (95%CI)
−0.4 (−4.0 to 3.3)
High quality cleansing6 , n (%)
[95%CI]
173 (72.7)
[67.0 to 78.0]
146 (63.8)
[57.5 to 70.0]
0.04
Per-protocol analysis
n = 220
n = 203
Adequate cleansing3 , n (%)
[95%CI]
212 (96.4)
[92.9 to 98.3]
196 (96.6)
[92.9 to 98.5]
0.92
Difference in proportions, % (95%CI)
−0.2 (−3.9 to 3.6)
High quality cleansing6 , n (%)
[95%CI]
166 (73.5)
[67.7 to 78.8
131 (63.6)
[57.0 to 70.2
0.03
BBPS score, median (IQR)
3 (3–3)
3 (2–3)
0.01
3 (3–3)
3 (3–3)
0.004
3 (3–3)
3 (3–3)
0.08
Adequate cleansing stratified per center, n (%)
(P = 0.41)
(P = 0.34)
36 (92.3)
32 (91.4)
51 (98.1)
51 (96.2)
55 (98.2)
55 (98.2)
79 (95.2)
79 (97.5)
Experience complying with low residue diet, n (%)
0.25
50 (22.6)
56 (27.3)
122 (55.2)
101 (49.3)
37 (16.7)
42 (21.0)
11 (5.0)
5 (2.4)
1 (0.5)
0 (0)
Experience complying with liquid diet, n (%)
0.75
39 (17.6)
40 (19.5)
103 (46.6)
97 (47.3)
58 (26.2)
48 (23.4)
16 (7.2)
18 (8.8)
5 (2.3)
2 (1.0)
Additional liquids taken, median (IQR), L
3 (2–4)
4 (2.4–5.1)
<0.001
≥75% compliance with laxative, n (%)
218 (99.5)
202 (98.5)
0.36
Physical activity level during bowel preparation, n (%)
0.81
85 (38.5)
85 (41.5)
131 (59.3)
116 (56.6)
5 (2.3)
4 (2.0)
Cecal intubation, n (%)
219 (92.8)
219 (96.1)
0.13
Reason for failed cecal intubation, n (%)
0.10
3 (17.6)
0 (0)
5 (29.4)
0 (0)
5 (29.4)
6 (66.7)
4 (23.5)
3 (33.3)
Gloucester comfort scale score, median (IQR)
2 (1–2)
2 (1–2)
0.60
Sedation, n (%)
0.18
16 (6.8)
16 (7.0)
207 (87.7)
207 (90.8)
13 (5.5)
5 (2.2)
Adenoma detection rate, %
48.7
51.5
0.40
Adenomas per colonoscopy, median (IQR)
0 (0–1)
0 (0–1)
0.64
Serrated polyp detection rate, %
17.6
16.2
0.67
Colorectal cancer, n (%)
10 (4.2)
4 (1.8)
0.12
Polyp detection rate, %
55.9
62.0
0.18
Polyps per colonoscopy, median (IQR)
1 (0–2)
1 (0–2)
0.55
Withdrawal time, median (IQR), minutes
12 (8–17)
13 (9–19)
0.047
Impact on work and related costs
No clinically relevant differences in the proportions of absenteeism or presenteeism
were present between the groups ([Fig. 3 ]; Table 1s , see online-only Supplementary material). Among working patients, the mean baseline
absenteeism was 16.2%, decreasing to 7.9% after the colonoscopy. Additionally, 12.3%
of patients in both groups reported reduced productivity (presenteeism) in the preceding
4 weeks, with median associated costs of €1390.
Fig. 3 Bar charts comparing the 1L and 2L PEG+Asc groups for: a the impact on work and work productivity measured after bowel preparation and colonoscopy
(rates of absenteeism and presenteeism calculated in the working subpopulation; rates
of unpaid labor loss, caregiver help, and caregiver absenteeism in the entire study
population); b the costs of absenteeism, presenteeism, unpaid labor loss, and median self-reported
costs in euros.
Unpaid labor loss rates remained stable before and after the procedure in both groups
(9.0% before vs. 8.5% after colonoscopy), with mean costs of €1012 and €1065, respectively.
Caregiver involvement was common, with 82.5% of patients requiring assistance, resulting
in caregiver absenteeism in 30.6% of cases.
The out-of-pocket costs of patients were modest, with median out-of-pocket costs of
€15 (IQR €8.5–49.5), and no significant difference between the study groups (P = 0.59). No significant differences between the study arms were observed in any categories.
Tolerability and impact on quality of life
Tolerability
Most respondents drank ≥75% of the bowel preparation fluid (99.5% in the 1L PEG+Asc
group and 98.5% in the 2L PEG+Asc group; P = 0.18) (Table 2s ). Tolerability was rated as “easy” by 51.6% in the 1L PEG+Asc group and 57.1% in
the 2L PEG+Asc group (P = 0.22). Willingness to repeat taking the bowel preparation was higher with 1L PEG+Asc
(59.8% vs. 48.3%; P = 0.04). Among patients with previous experience of bowel preparation, more patients
in the 1L PEG+Asc group found the tolerability better than their previous experience
(26.9% vs. 9.3%; P < 0.001). The total symptom scores from bowel preparation were not significantly
different between the groups, but more patients in the 1L PEG+Asc group reported a
moderate or severe bad taste and nausea/vomiting compared with the 2L group (Table 2s ).
On univariable regression analysis, married or living together (OR 1.86, 95%CI 1.12
to 3.09), higher endoscopy satisfaction scores (OR 1.07, 95%CI 1.04 to 1.10), higher
score on the visual analog scale in the HRQoL questionnaire (EQ-VAS; OR 1.03, 95%CI
1.02 to 1.05), and EQ-index (OR 1.04, 95%CI 1.02 to 1.06) were significantly associated
with a higher willingness to repeat the bowel preparation. In contrast, 2L PEG+Asc
vs. 1L PEG+Asc (OR 0.62, 95%CI 0.43 to 0.90), female sex (OR 0.49, 95%CI 0.34 to 0.71),
fair tolerability (OR 0.18, 95%CI 0.11 to 0.29) and difficult tolerability compared
with good tolerability (OR 0.03, 95%CI 0.01 to 0.07) were significantly associated
with a lower willingness to repeat the bowel preparation.
On multivariable analysis, lower tolerability (fair, OR 0.23, 95%CI 0.13 to 0.40;
difficult, OR 0.05, 95%CI 0.02 to 0.14) and higher symptom score from bowel preparation
(OR 0.85, 95%CI 0.78 to 0.93) had a negative impact on willingness to repeat the bowel
preparation, while patients with a higher endoscopy satisfaction score (OR 1.04, 95%CI
1.01 to 1.08), and patients who received 1L PEG+Asc (OR 2.61, 95%CI 1.59 to 4.27))
compared with 2L PEG+Asc (OR 0.40) had a higher willingness to repeat the bowel preparation
(P < 0.001) (Table 3s ).
Health-related quality of life
In the EQ-5D-5L, patients scored a median baseline EQ-index of 0.89 (1L PEG+Asc) and
0.92 (2L PEG+Asc), both improving to 1.0 after colonoscopy (P < 0.001). Both groups showed a slight decrease in EQ-VAS after the procedure, reflecting
a modest drop below the minimum clinically important difference in perceived health
status. The similarity in score change between the arms suggests that the choice of
bowel preparation did not differentially impact overall perceived health (Table 4s ). No differences were observed between the study arms in the subdomains ([Fig. 4 ]; Table 4s ), although pain and anxiety in the 1L PEG+Asc group and mobility in the 2L PEG+Asc
group significantly decreased below the minimum clinically important difference.
Fig. 4 Violin plots of health-related quality of life outcomes measured by the EQ-index and
EQ-VAS score in the 1L and 2L PEG+Asc groups. Higher scores indicate higher quality
of life. A wider plot indicates more frequent occurrence of that score. Scores were
not significantly different between the treatment arms.
Endoscopy satisfaction
In general, patients were satisfied about the endoscopic care with a mean score of
85.1 (SD 8.3) (Table 5s ). Patients were especially satisfied about the endoscopist’s skills (mean 92.4 [SD
8.8]) and overall hospital impression (mean 90.5 [SD 11.5]). Scores were not significantly
different between the treatment groups.
Adverse events
Sixteen AEs were reported, equally split between the 1L PEG+Asc and 2L PEG+Asc groups.
In the 1L PEG+Asc group, all AEs were related to the colonoscopy. In the 2L PEG+Asc
group, five AEs were colonoscopy related, while two involved bowel preparation (both
atrial fibrillation) – these cases resolved within 1 day, but the colonoscopies were
postponed and alternative laxatives were used. One serious AE, a head injury 30 days
post-procedure, was unrelated to the colonoscopy or bowel preparation.
Discussion
While the tolerability of bowel preparation is negatively associated with both bowel
preparation efficacy and colonoscopy participation, little is known about the impact
on patient-reported outcomes of bowel preparations with presumed higher tolerability
owing to their lower volumes. Our findings demonstrate that the low volume preparation
(1L PEG+Asc) was noninferior to the intermediate volume preparation (2L PEG+Asc) in
achieving adequate bowel cleansing (96.1% vs. 96.4%), had higher high quality cleansing
rates (72.7% vs. 63.8%), and was associated with greater patient willingness to repeat
the preparation (59.8% vs. 48.3%). Patient-reported outcomes, including HRQoL and
productivity loss, were not significantly different between the treatment arms.
Adequate efficacy is a prerequisite for research on patient-reported outcomes of bowel
preparation. The noninferiority of 1L PEG+Asc to 2L PEG+Asc in our study is in line
with previous studies [17 ]
[18 ]
[19 ]. Our rate of 96% adequately cleansed colonoscopies is at the high end of the spectrum.
This could be due to the exclusion of patients at risk for inadequate bowel preparation
from our study; however, this patient group has also been excluded from other studies
[19 ]
[20 ]
[21 ]. The high compliance in both groups could also be a contributing factor. Furthermore,
we included a low residue diet and split-dose protocol, combined with enhanced instructions,
all known to improve efficacy and adherence [1 ].
Our results demonstrated that tolerability is likely a critical factor in bowel preparation
because it directly impacts patients’ compliance and their willingness to repeat,
with patients who found the preparation “difficult” being 94.7% less likely (OR 0.05)
to be willing to repeat the preparation compared with those who found it “easy.” Repici
et al. demonstrated that 1L PEG+Asc had significantly higher compliance compared with
4L PEG, resulting in more effective bowel cleansing [19 ]. Similarly, Bednarska et al. concluded that patients were more willing to repeat
a regimen of 1L PEG+Asc than 2L PEG+Asc [18 ], consistent with our findings (59.8% vs. 48.3%; P = 0.04). Despite a more frequently reported bad taste (P = 0.007) and nausea/vomiting (P = 0.006) in the 1L PEG+Asc group, overall tolerability rates were similar between
the groups. Patients who had prior experience with taking bowel preparation found
the 1L PEG+Asc regimen significantly more tolerable than the bowel preparation they
had taken previously, as also reported by Bednarska et al. [18 ]. This may well suggest that volume is an important factor in negatively affecting
patient experience, even more than taste or side effects. Therefore, the superior
experience of 1L PEG+Asc contributes to the efficacy of the bowel preparation. Furthermore,
management of discomfort and anxiety by effective communication likely also plays
an important role in improving compliance and overall patient experience [22 ].
Because of the multiple instructions patients have to adhere to during bowel preparation,
and possible discomfort and anxiety [2 ], we hypothesized that these could also have an impact on HRQoL. We found comparable
HRQoL scores to the general Dutch population, with no clinically relevant difference
in either group. Andronis et al. reported that patients undergoing colonoscopy experienced
a temporary decrease in HRQoL owing to bowel preparation discomfort [23 ]. In contrast, Niv et al. observed that, while patients had some discomfort during
bowel preparation, the overall HRQoL scores remained relatively stable after colonoscopy,
in line with our results [11 ]. As patients report some discomfort during bowel preparation, traditional HRQoL
measures are unlikely to be able to fully capture the impact on patients [24 ]; however, given the short duration of the bowel preparation process, any long-term
impact on HRQoL by bowel preparation is likely negligible.
Bowel preparation can have a significant impact on patients’ work as they need to
take time off to complete the bowel preparation and undergo the colonoscopy. Fuccio
et al. reported in a prospective cohort (n = 1137) that absenteeism or presenteeism
owing to bowel preparation or symptoms after colonoscopy was 30.5% in patients undergoing
colonoscopy [7 ]. Our study’s lower rate of both absenteeism (7.9%) and presenteeism (12.3%) may
be due to the use of a split-dose regimen and a difference in the laxatives used (4L
and 2L vs. our study’s 1L and 2L). Other studies have observed absenteeism in 20%–32%
of patients owing to bowel preparation, with an association found between a higher
symptom burden and absenteeism [25 ]
[26 ]. Additionally, our study observed that the loss of unpaid labor was 8.5% after colonoscopy,
which has been poorly investigated so far. Both direct (hospital) and indirect nonmedical
costs are needed for a clearer understanding of the total expenses associated with
bowel preparation [27 ]
[28 ]. Our findings may therefore offer a perspective on these costs from both patient
and societal viewpoints.
Ongoing interest needs to be paid to bowel preparation tolerability to ensure adequate
preparation but also to minimize the possible negative impact on patients. Possible
variations include a split-dose regimen, diet liberations such as low residue diets
with shorter durations [29 ]
[30 ], and also developing other low volume laxatives. This will likely reduce the barriers
experienced to colonoscopy and CRC screening [31 ]. Furthermore, our results on HRQoL and patient costs, using validated instruments,
may inform healthcare policies and allow policies to be based on informed decisions
that improve CRC screening participation, while minimizing impact on patients.
Our study has several strengths. In contrast to previous studies, our multicenter
randomized design reduces the risk of bias. Additionally, we used validated instruments
to increase generalizability and comparability. To minimize possible remaining bias,
we used multiple imputations to compensate for missing information, although the response
rate was already nearly 90%. The use of clinical and socioeconomic variables enabled
us to correct for potential confounders and provided more optimal information on the
impact of bowel preparation on patients and relatives.
We also acknowledge some limitations. The 2-day low residue diet was not in line with
ESGE recommendations for bowel preparation and might potentially have influenced our
results, but it is unlikely it would have introduced differences between the study
groups. Additionally, as our secondary analyses were only performed exploratorily,
external confirmation is needed. Our study sample lacked patients of non-Dutch ethnicity
and those with lower education levels. This may limit generalizability to patients
with other ethnicities or lower socioeconomic levels, while these patients are known
to be at slightly higher risk for inadequate bowel preparation [9 ]
[32 ]. Furthermore, because of the mixed colonoscopy indications, the adenoma detection
rate (ADR) presented in this study may limit comparison with fecal immunochemical
test (FIT)-only cohorts, but the inclusion of various colonoscopy indications improves
generalizability to a wider patient group. Lastly, although we are confident that
relevant post-procedural AEs were captured through the questionnaires and hospital
records, smaller events outside the hospital system may have been missed.
In conclusion, low volume 1L PEG+Asc bowel preparation is an effective and tolerable
alternative to an intermediate volume 2L PEG+Asc preparation, with high patient satisfaction
and minimal impact on quality of life and productivity. Prioritizing tolerability
of bowel preparation is essential to increase adequate bowel cleansing and reduce
barriers to colonoscopy.
