Keywords
Endoscopy Small Bowel - Capsule endoscopy - Quality and logistical aspects - Preparation
- Endoscopy Lower GI Tract
Introduction
Colon capsule endoscopy (CCE) was introduced in 2006 as a new modality for imaging
the colonic mucosa [1]. It is sedation-free and may be able to reduce colonoscopy needs and minimize interaction
with health care personnel [2]. For a long time, completion rates and the bowel preparation regimen for CCE have
not been comparable to that for colonoscopy [3]. However, evidence is accumulating to suggest increased diagnostic accuracy and
convenience for CCE as compared with optical colonoscopy [3]
[4]. This underscores the need for more studies exploring how to improve completion
rates for CCE to meet established standards for colonoscopy.
Small bowel capsule endoscopy (SBCE) was introduced 6 years earlier than CCE as a
new imaging technique for observing the small bowel mucosa and today it is a well-established
and accepted procedure in clinical practice [5]
[6]. Prior studies have shown that accelerating SBCE transit time may increase the likelihood
of a complete investigation [7]. However, the most optimal bowel preparation regimen for SBCE is yet to be determined
[6].
The completion rate for capsule endoscopy is a limitation of both SBCE and CCE [3]
[7]. Over the past decade, there has been significant improvement in bowel preparation
regimens for both CCE and SBCE [5]
[8]. As of now, the bowel preparation regimen for CCE is already more extensive than
that for colonoscopy, and therefore, additional preparations should be introduced
with caution. Some studies have investigated the ability to accelerate transit time
by use of sham feeding to stimulate motor and sensory activity [9]
[10]. Using chewing gum as a substitute for sham feeding has shown good results in accelerating
motility throughout the gastrointestinal tract [9]
[11]
[12]
[13].
This systematic review with meta-analysis aimed to assess the efficacy of chewing
gum as a booster in bowel preparation regimens for patients undergoing either SBCE
or CCE. The effectiveness was examined by comparing gastric transit time (GTT), small
bowel transit time (SBTT), colon transit time (CTT), bowel cleanliness rate, and pooled
estimates of complete investigations.
Methods
This systematic review was prepared according to PRISMA guidelines and registered
in PROPSPERO (PROSPERO ID: CRD42022385214) without any further additions to the protocol
after submission [14].
Literature search
A systematic literature search was performed in four electronic databases: PubMed,
Embase, Cochrane, and Web of Science. The search string was initially developed for
all four databases in three predefined search areas: investigation, comparator and
outcome. The words for investigation were identifying studies on SBCE and/or CCE.
Comparator words were used to identify but also limit the number of search hits in
search of studies with a bowel preparation regimen involving chewing gum. Words for
outcome were to further limit the number of search hits to studies with reported completion
rates, effect on transit time and/or diagnostic yield. Both within and across the
three search areas, search strings with relevant search terms were created and combined
using Boolean expressions. The Boolean expression “OR” was used within search areas,
and Boolean expression “AND” was used across search areas.
In addition, indexed search terms from the database thesauruses and free text search
terms with truncation were included in the search strategy. The search strategy was
then revised and edited by adding applicable MeSH terms, in collaboration with a research
librarian from the University of Southern Denmark. This allowed the search strategy
to be as comprehensive as possible to search multiple terms. The last search was performed
on November 8, 2023. The search strategy is shown in Appendix A as Table S1, along with search strings in Table S2.
Reference screening
Endnote version X9 was used to process papers [15]. Two reviewers completed the entire screening process and data extraction to reduce
risk of bias and validate eah other’s results (S.S.J. and U.D.) [16]. After excluding duplicates, titles and abstracts of all remaining citations were
independently assessed for inclusion by the same two authors. If a discrepancy occurred
and agreement could not be reached, the reference was evaluated by a third reviewer
(T.B-M.). The third reviewer would decide whether the reference was eligible for inclusion.
Following abstract and title screening, the same two authors independently assessed
the relevant full-text manuscripts. No third opinions were necessary. Finally, one
reviewer (S.S.J.) examined the references of each included study to identify any additional
references of possible relevance that met the inclusion criteria. However, further
studies were not retrieved.
Eligibility criteria
We included studies of patients who underwent SBCE or CCE for any indication. The
bowel preparation regimen had to involve chewing gum and be compared with a standard
bowel preparation regimen without chewing gum. Included studies had to be randomized
controlled trials (RCTs), cross-sectional studies, or cohort studies. Studies not
reporting the completion rate between groups were excluded. Eligible studies for inclusion
had to be published in English, Danish, German, French, or Spanish; however, no non-English
studies were retrieved. Case reports, conference abstracts, and reviews were excluded.
Data handling
Two authors (S.S.J and U.D.) performed data extraction from the included studies separately
and discussed any discrepancies until a consensus was reached. [Table 1] shows an overview of the data extracted. Simultaneously, both authors assessed the
quality of the included studies using the MINORS index [17].
Table 1 Data extracted for descriptive purposes and statistical analyses.
|
Number
|
Description
|
|
I
|
Number of individuals included in the study
|
|
II
|
Number of individuals with complete small bowel or colon capsule endoscopy
|
|
III
|
Descriptive data: first author, publication year, data origin (country), year of data
collection, study type, single- or multicenter study, indications for either small
bowel or colon capsule endoscopy, type of capsule, reported bowel/procedure preparation
medicine (incl. boosters and contrast agents), type of bowel cleansing rate scale,
transit time; total, gastric, small bowel and colon, participant characteristics,
sex, and age distribution
|
Completion rate was defined as the proportion of investigations with complete transit.
Complete transit was defined as visualization of the cecum for small bowel endoscopy
and as visualization of the hemorrhoidal plexus or excretion within battery lifetime,
along with acceptable bowel preparation for CCE. Completion rates were calculated
as the number of complete examinations from the total number of procedures in the
included studies. Additional descriptive data were retrieved; however, they were only
used for descriptive reasons and not for subgroup analysis as initially intended due
to the limited number of references.
Statistical analysis
All extracted data ([Table 1]) were collected in Excel spreadsheets, whereafter they were imported to Stata 17
for further statistical analyses. The proportions of complete SBCE and CCE, including
their 95% confidence intervals (CIs), were calculated for each study included in the
analyses. These estimates were then pooled using Freeman-Tukey double arcsine transformation
in random effects models using the Metaprop command [18]. Furthermore, as a sensitivity analysis, we repeated the calculation of these pooled
estimates using fixed effects models. To evaluate the reliability of the results,
the degree of heterogeneity was estimated by I2 statistics for each pooled estimate. Furthermore, Egger’s test was used to investigate
potential small study effects and publication bias in each subgroup illustrated by
funnel plots [19]. Stata 17 was used to conduct the analyses [18]
[20].
Results
The final literature search resulted in 90 articles. Of those, 55 duplicates were
removed and 30 articles were excluded after preliminary title and abstract screening.
This left five articles for full-text screening. One of these references was not retrievable
from any of the searched databases. Four studies were included after a full-text review,
as shown in [Fig. 1]
[21]
[22]
[23]
[24]. No further articles were identified from the reference list screening of the included
studies for full-text reading.
Fig. 1 Flow chart of literature search and study inclusion.
Characteristics of all four studies are presented in [Table 2]. Three studies investigated SBCE [21]
[22]
[23] and one evaluated CCE [24]. In the included studies, a total of 267 individuals received a bowel preparation
regimen with chewing gum and 270 individuals received a bowel preparation regimen
without chewing gum. The proportion of male individuals ranged from 44% to 66% and
the median age ranged from 47 to 58 years between studies. All four studies were prospective,
single-center RCTs conducted between 2008 and 2021 in Greece, Canada, China, and Denmark.
In one study, SBCE was performed with the small bowel first-generation PillCam (Given
Imaging Inc., USA) [21], in two studies, SBCE was performed with the small bowel second-generation PillCam
(Given Imaging Inc., USA and Medtronic, USA) [22]
[23], and in one study, CCE was performed with the second-generation colon PillCam2 (Given
Imaging Inc., USA) [24]. In all studies, all individuals received dietary instructions before and during
the capsule endoscopy procedure. Bowel preparation regimens are listed in [Table 3]. Only one study reported adverse events among the chewing gum group, with three
cases of capsule retention in the chewing gum group and six capsule retentions in
the control group [23].
Table 2 Overview of study characteristics.
|
Publication, year
|
Country
|
Study type
|
Endoscopy type
|
Adverse events in chewing gum group
|
No. of cases/controls
|
MINORS score*
|
|
*Score from 0–24 for comparative studies.
|
|
Apostolopoulos, 2008 [21]
|
Greece
|
Prospective, single-center
|
SBCE, PillCamSB1 (Given Imaging, Yoqneam, Israel)
|
0
|
47/46
|
22/24
|
|
Ou, 2014 [22]
|
Canada
|
Prospective, single-center
|
SBCE, PillCamSB2 (Given Imaging, Yoqneam, Israel)
|
0
|
60/62
|
24/24
|
|
Huang, 2021 [23]
|
China
|
Prospective, single-center
|
SBCE, PillCamSB2 (Medtronic, Minnesota, America)
|
Capsule retention (n=3)
|
103/102
|
24/24
|
|
Buijs, 2018 [24]
|
Denmark
|
Prospective, single-center
|
CCE, PillCam2, (Given Imaging, Israel)
|
0
|
57/60
|
22/24
|
Table 3 Bowel preparation regimens and administration of chewing gum in the included studies.
|
Publication
|
Bowel preparation regimen
|
Administration of chewing gum
|
Bowel preparation evaluation
|
|
Apostolopoulos, 2008 [21]
|
Day -5:
Interrupt use of medication that could limit mucosal visualization
Day -1:
All day – clear liquids
Evening – 45 mL of sodium phosphate (Fleet Phospho-Soda, Botania, Greece) with water
Exam day:
Fasting overnight, at least 8 hours prior to capsule endoscopy examination
|
1 piece of sugarless gum to be chewed for 30 minutes every 2 hours (maximum of four
pieces total)
|
Bowel preparation evaluation scale proposed by the 2005 International Conference on
Capsule Endoscopy
|
|
Ou, 2014 [22]
|
Day -5:
Interrupt use of oral iron supplementation
Day -1:
Breakfast – soft diet
Lunch -> rest of day – clear fluid diet
3.00 PM – 2L polyethylene glycole electrolyte solution (PEG3350e)
Exam day:
Take nothing per mouth 2 hours before the scheduled appointment
|
1 piece of sugarless gum to be chewed for 20 minutes every 2 hours (maximum of four
pieces total)
|
Not reported
|
|
Huang, 2021 [23]
|
Day -1:
For dinner – clear liquids
Exam day:
Fasting overnight (at least 8 hours prior to undergoing SBCE)
4.00–5.00 – two sachets of polyethylene glycol electrolyte powder (59 g polyethylene
glycol 4,000, 5.68 g sodium sulfate, 1.68 g sodium bicarbonate, 1.46 g sodium chloride,
and 0.74 g potassium chloride per sachets) dissolved in 2 L of water (within 2 hours)
|
1 piece of sugarless gum to be chewed for 15 minutes every 30 minutes in the first
hour (maximum of two pieces total)
|
Poor bowel preparation if < 75% of the mucosa was visualized
|
|
Buijs, 2018 [24]
|
Day -2:
All day – 2 L water in addition to normal intake
2 times daily 1000 mg magnesium oxide (oral)
Day -1:
All day – clear liquids only
Until 16.00 – white pasta with oil, only
17.00–19.00 – 1 L Moviprep followed by
1.5 L water
Exam day:
6.00–7.30 – 1 L Moviprep followed by at least 1.5L water
7.30–9.30 – No food/liquid ingestion
±10.00 – Ingestion of CCE with 20 mg domperidone (oral)
After alarm – 0.75 L Moviprep followed by at least 0.6 L water
+ 3 hours – 0.25 L Moviprep followed by 0.2 L water
+ 2 hours – Bisacodyl enema
|
2 pieces of sugarless gum to be chewed for 30 minutes when capsule left stomach
|
Leighton-Rex scale
|
Three studies used polyethylene glycol (PEG) solutions for bowel preparation [22]
[23]
[24] and one study used sodium phosphate [21]. The CCE study was the only study in which a split-dose regimen was used and a booster
solution was combined with prokinetics after capsule ingestion [24]. The booster was a PEG solution along with bisacodyl as the prokinetic agent. Among
the studies, chewing gum dosage varied from two to a maximum of four pieces to be
chewed after capsule ingestion at varying intervals. An overview of chewing gum administration
is shown in [Table 3].
Completion rate
The pooled estimated completion rate for capsule endoscopy with the chewing gum regimen
was not significantly higher at 85% (95% CI 69%–96%) compared with 78% (95% CI 60%–91%)
for the non-chewing gum regimen as shown in [Fig. 2]. Test of heterogeneity resulted in 89.41% (P <0.001) in the chewing gum groups and 89.72% (P <0.001) in the control groups. Sensitivity analyses using fixed effects models resulted
in similar pooled estimates of 87% (95% CI 83%–91%) compared with 80% (95% CI 75%–85%)
in the chewing gum and control groups, respectively.
Fig. 2 Forest plot of estimated completion rates of capsule endoscopy with and without chewing
gum in bowel preparation regimen. ES, estimated completion rate.
Because of high heterogeneity, a sensitivity analysis excluding the CCE study was
conducted to explore the potential impact on results. [Fig. 3] shows that heterogeneity was reduced in both the chewing gum group (62.60%, P=0.07) and the control group (71.76%, P=0.03). The pooled estimated completion rate for the chewing gum regimen (91%, 95%
CI 83%–97%) was not significantly different from that for the standard regimen (85%,
95% CI 74%–93%) when limiting the analysis to SBCE studies. Sensitivity analyses using
fixed effects models resulted in similar pooled estimates of 92% (95% CI 88%–95%)
compared with 86% (95% CI 81%–90%) in the chewing gum group and control group respectively.
Fig. 3 Forest plot of estimated completion rates of small bowel capsule endoscopy with and
without chewing gum in bowel preparation regimen. ES, estimated completion rate.
Bowel preparation
Only two studies reported an adequate bowel cleanliness rate; therefore, no pooled
estimate was calculated [23]
[24]. One study reported an adequate bowel preparation rate of 100% in both the chewing
gum and control groups; however, it was not stated how bowel cleanliness was evaluated
[23]. Another study reported a 48% cleanliness rate in the chewing gum group and 52%
in the control group based on the Leighton-Rex scale [24].
Transit time
Pooled GTT, SBTT, and CTT estimates were not calculated because only one study reported
variance information [21]. Median GTT ranged from 18.38 to 29.0 minutes and median SBTT ranged from 229.43
to 318.5 minutes in the chewing gum regimen groups. One study found a mean GTT of
40.81 SD ± 30.28 in the intervention group, and 56.41 SD ±42.77, as shown in [Table 4]
[21]
[22]
[23]. In the control groups, the median GTT ranged from 19.43 to 42.5 minutes, and the
median SBTT ranged from 232.52 to 287.0 minutes with standard bowel preparation regimens.
Lastly, a study found a mean SBTT of 229.05 SD ±75.99 in the intervention group and
266.69 SD±69.88 in the control group, as shown in [Table 4]
[21]
[22]
[23]. Only one study reported a significant impact for chewing gum in reducing both GTT
and SBTT [21]. In addition, another study established a considerable GTT reduction but not for
SBTT [23]. A third study found no significant influence on either GTT or SBTT [22]. The study on CCE did not find any difference between the intervention group and
the control group regarding CTT and total transit time [24].
Table 4 Overview of transit times in included small bowel capsule endoscopy studies.
|
Study
|
GTT
|
SBTT
|
|
|
Gum
|
Control
|
Gum
|
Control
|
|
Apostolopoulos, 2008 [21] mean ± SD
|
40.81±30.28
|
56.41±42.77
|
229.05±75.99
|
266.69±68.88
|
|
Ou, 2014 [22]median
|
18.38
|
19.43
|
229.43
|
232.52
|
|
Huang, 2021 [23]median
|
29.0
|
42.5
|
318.5
|
287
|
Publication bias and small study effects
Funnel plots are presented in [Fig. 4]. The Egger’s tests were significant for both pooled estimates of completion rate
in capsule endoscopy in the chewing gum group (P <0.001) and control group (P <0.001), ([Fig. 4], plots A and B). However, when the CCE study was removed and the analysis repeated,
Egger’s tests were no longer significant for either chewing gum groups (P=0.075) or control groups (P=0.051) ([Fig. 4], plots C and D).
Fig. 4 Funnel plots visualizing the symmetry of the effect sizes for completion rate in small
bowel and colon capsule endoscopy with and without chewing gum in the bowel preparation
regimen. P values from Egger’s tests provided for each plot. Plot A+B includes both small bowel
capsule endoscopy and colon capsule endoscopy. Plot C+D includes only small bowel
capsule endoscopy.
Discussion
This review evaluated the effect of adding chewing gum to the bowel preparation regimen
for capsule endoscopy. The primary finding was a non-significant improvement in the
completion rate of 91% in SBCE patients who chewed gum during the procedure compared
with a completion rate of 85% in control groups as shown in [Fig. 3]. When including the CCE study by Buijs et al. [24], the conclusion did not change. However, the completion rate for both intervention
groups and control groups decreased. Including the CCE study by Buijs et al. [24] also introduced substantial heterogeneity, as shown in [Fig. 2]. The CCE study is the only study to date to explore the influence of chewing gum
as a booster for CCE completion rate.
The included studies also evaluated transit times in capsule endoscopy within control
and intervention groups. However, pooled analysis was not performed due to unavailable
variance information. Even without the pooled analysis, we did see a tendency for
chewing gum to positively affect transit times in the small bowel studies. GTT was
significantly decreased by chewing gum, as shown in Apostolopoulos et al [21].
This is the only systematic review exploring chewing gum influence on completion rate
and the only review in the current literature to compare transit times for both SBCE
and CCE. Therefore, no other reviews were available for results comparison. However,
a letter to the editor from 2013 describing 415 SBCE patients, of whom 207 had chewing
gum administered as part of the bowel preparation regimen, described an equivalent
pooled analysis of chewing gum impact on completion rate and transit time [25]. This meta-analysis comprised three studies, of which two are also included in this
review [21]
[22]. The third study was non-retrievable for this review, as noted in [Fig. 1] and the results paragraph. As our meta-analysis supports, those authors did not
find an increased completion rate by adding chewing gum as a booster, but they did
conclude an overall significant decrease in SBTT. However, the result was reported
as non-applicable in clinical practice due to a lack of solid evidence. If the number
of SBCE studies in this review had been larger and variance information had been provided
for each study, transit times could have been pooled and compared to advance evidence.
The three studies about SBCE all indicated that chewing gum had an accelerating effect
on transit times. It is possible that in the future, chewing gum should be used for
patients with low motility to speed up transit times. As previously mentioned, accelerating
transit times in SBCE is associated with higher completion, underscoring the advantages
of further exploration of the effect of chewing gum in that regard [7].
In the four studies included, administration of chewing gum varied. Apostolopoulos
et al. and Ou et al. administered sugarless gum for as long as 8 hours, whereas Huang
et al. and Buijs et al. administered sugarless chewing gum for as long as 4 hours
[21]
[22]
[23]
[24]. The chewing gum administered for 8 hours could stimulate motor and sensory activity
in both the stomach and small intestine; however, the chewing gum administered in
Huang et al. and Buijs et al. stimulated only motor and sensory activity in the stomach.
This might have influenced completion rates in the studies. However, the results do
not indicate that administration of chewing gum for only 4 hours by Huang et al. produced
a lower completion rate than was achieved in the two other SBCE studies [23].
With both SBCE and CCE, the bowel preparation regimen is essential to obtain a complete
investigation. However, bowel preparation for CCE is more extensive than for colonoscopy
because it is impossible to intervene in other ways than by administering boosters
before or after capsule ingestion. This has resulted in numerous studies searching
for the optimal composition of the bowel preparation regimen to increase acceptable
bowel cleanliness and excretion rates [3]. In addition, add-on boosters such as castor oil, prucalopride, and a sulfate-based
solution have influenced completion rates, suggesting that they have potential [26]
[27]
[28].
This review was strengthened by all included studies being prospective RCTs. Moreover,
all included studies were quality assessed and scored high evaluations from the MINORS
index. Limitations of this review include the low number of included studies (n=4)
and that all four studies were single-center reports. Therefore, the power of the
analysis is limited, and the insignificant differences found between the pooled estimates
may, in fact, be the result of type II errors. Furthermore, all studies were single-blinded
and might have been candidates for hidden bias. This makes it challenging to support
changes in clinical practice due to the need for external validity. However, because
there might be a benefit from providing chewing gum during capsule transit, and the
risks and costs associated are minuscule, it seems that it could be a reasonable add-on
in patients with suspected decreased peristalsis. The data were too heterogeneous
to include Buijs et al. in the pooled completion rate analysis because it was the
only study to represent CCE [24]. This was also based on the fact that Egger’s test demonstrated significant P values for all included studies when Buijs et al. was part of the test, but when
excluded, P values were non-significant, as shown in [Fig. 4]. This also suggests that this review was limited by small study effects and publication
bias when Buijs et al. was part of the pooled completion rate estimates. Heterogeneity
was still moderately high without Buijs et al., but this is to be expected from pooled
analyses with few studies.
In terms of future research, more prospective and multicenter RCTs are warranted to
disprove this review's findings. In addition, future issues that need addressing are
the timing of booster administration and patient tolerability, regardless of which
capsule endoscopy type or prokinetic agent is being evaluated. These might be critical
factors in implementing changes in clinical practice, although evidence supporting
this has yet to be established. Some of the studies included in this meta-analysis
indicate that chewing gum might have a relevant advantage in accelerating transit
times, which could be relevant to further investigation.
Conclusions
This meta-analysis did not find a significant difference in completion rate among
patients receiving chewing gum as a booster in capsule endoscopy. With chewing gum’s
safety profile in consideration, more capsule endoscopy studies evaluating its efficacy
can be performed with a low risk of causing complications. More CCE studies, in particular,
are essential for proper evaluation of the actual impact of chewing gum as a booster
in the bowel preparation regimen, and those data would enhance empirical evidence
for analyses to be repeated. The findings about transit times are more optimistic,
and it may be clinically relevant to consider use of chewing gum to decrease transit
time, if further studies come to the same conclusion.
