Introduction
Since its introduction into clinical practice in the late 1960 s, colonoscopy has
now become a standard procedure [1]
[2]. After the initial report of successful ileal intubation during colonoscopy, [3] ileoscopy is often recommended by some in unselected patients undergoing colonoscopy
[4] or at least in those patients presenting with diarrhea or suspected Inflammatory
Bowel Disease (IBD) or intestinal tuberculosis (ITB) [5]
[6]
[7]
[8]
[9]
[10]. In addition, identification of the ileocecal (IC) valve and ileal mucosa documents
completeness of full length colonoscopy [11]
[12].
However, various difficulties are encountered during ileoscopy. The IC valve may be
difficult to locate when it is hidden behind large semilunar folds or if excess air
is insufflated with stretching of the lumen or if the valve is papillary in form and
flush with the wall [1]
[13]. In various studies, the success of ileoscopy, without any special maneuvers, varies
from 77.8 % to 81.3 %, with a time required to insert the scope into the ileum of,
on average, 3.4 min (range: 30 s to 10 min) [14]
[15]
[16]. Despite this, with various endoscopic maneuvers, it is possible to successfully
intubate the terminal ileum in a significant number of patients [16].
During colonoscopy, the endoscopist employs various maneuvers, including changing
the patient’s posture, to achieve complete colonoscopic examination [1]
[13]
[16]. If one position fails, another is attempted. Posture change has also been reported
to increase the success rate of ileal intubation, with the supine posture often bringing
the IC valve to a 6 – 7 o’clock position ([Fig. 1]) [16]
[17]. However, there has been no randomized trial which has shown that a particular posture
of the patient increases the success rate of ileoscopy.
Fig. 1 Upper panel shows the “down to left” position taken by the IC valve when shifting
the patient’s posture from left-lateral decubitus to supine. Lower panel shows the
four types of IC valve.
The present study was performed to determine the impact of the patient’s posture (left
lateral vs supine position) on the success rate of ileal intubation. In addition,
we also assessed the yield (clinically relevant positive findings) of ileal intubation
and the predictors of successful terminal ileal intubation.
Patients and Methods
Study Design
All outpatients (> 12 years old) referred for colonoscopy between June and December
2010 were eligible for entry into the study provided they did not have the following
conditions: 1) acute fulminant colitis; 2) acute intestinal obstruction; 3) suspected
intestinal perforation; 4) peritonitis; 5) pregnancy; 6) severe cardio-respiratory
disease (ASA grade > II); 7) decompensated liver disease; 8) recent pelvic or colonic
surgery (in the last 6 months); 9) large aortic or iliac artery aneurysm; 10) Human
Immunodeficiency Virus (HIV) infection; and 11) were non-consenting.
Colonoscopy was performed in the Endoscopy Suite of the School of Digestive and Liver
Diseases (SDLD), IPGME & R, Kolkata, a tertiary referral-cum-teaching hospital with
a videocolonoscope (EC-201 WL, Fujinon) by endoscopists with different skill levels,
including third-year gastroenterology trainees and departmental consultants, under
conscious sedation with titrated intravenous midazolam (dose: 0.05 – 0.1 mg/kg) while
the patients were monitored by continuous pulse oxymetry (Draco-oxy Biptronics Plethysmogram).
Randomization
In this randomized, open-label study, patients were randomized to undergo ileal intubation,
either in the left-lateral decubitus (LL) or supine (S) posture, after the cecum had
been successfully intubated by a 1:1 randomization (equal proportions) by the sealed
envelope method. An independent statistician generated random numbers on a computer
using SPSS 13.0 for Windows (SPSS Inc, Chicago, Ill). He prepared hundreds of sealed
envelopes containing random numbers for allocation, which had equal odd – even proportions.
An odd number allocated the patient to the LL posture group and an even number to
the S posture group. An assistant nurse opened the envelopes just after completion
of cecal intubation at colonoscopy.
Sample-size calculation
This was done with the free software G*Power 3.0.10 (Universität Kiel, Germany). Based
on the literature, the reported success of ileoscopy without any maneuver was taken
to be ~ 80 %. [14]
[15]
[16] Expecting at least a 15 % increase in success at ileal intubation after changing
the patient’s posture to supine, the sample size required using Fisher’s exact 2-sided
test to detect a statistically significant difference (α-error probability = 0.05)
was 216 (108 in each group) with 90 % power (β-error = 0.1).
Procedure
Colonoscopy and preparation
Patients were advised to take clear liquid on the night prior to colonoscopy. Bowel
preparation was done by taking two tablets of Bisacodyl (Tablet Dalculax 10 mg) at
bedtime on the night before colonoscopy and a standard poly-ethylene glycol (PEG)
preparation (containing 118.0 g PEG, 2.93 g sodium chloride, 1.48 g potassium chloride,
3.37 g sodium bicarbonate, 11.3 g anhydrous bicarbonate; ColopegTM, J.B. Chemicals & Pharmaceuticals Ltd, India) in the early morning on the day of
colonoscopy. Adequacy of bowel preparation was assessed by the validated Boston Bowel
Preparation Scale (BBPS or “bee-bops”) [18].
Colonoscopy was started in the LL posture. The endoscopist was allowed to apply all
maneuvers to successfully intubate the cecum, which was confirmed by standard landmarks,
i. e., appendicular orifice, IC valve and transillumination [11]. The shape ([Fig. 1]) and position of the IC valve were recorded.
Ileoscopy and randomization
After confirmation of cecal intubation and with the scope shortened and de-looped,
patients were randomized to either the LL or S position if they satisfied the predefined
inclusion criteria. Then terminal ileal intubation was attempted. Partial suction
and scope maneuvers for ileal intubation were allowed [13]. Use of anti-peristalsis agents, biopsy forceps (as an “anchor” to facilitate IC
valve intubation), and intubation in the retroflexed position were not allowed. If
the ileum was not intubated within 6 min, [19] it was considered a failure and the study end point was reached. Further attempts
at ileoscopy were at the discretion of the endoscopist.
Outcome measures
The primary outcome was whether ileal intubation was achieved or not within the stipulated
6 min. The other outcomes that were recorded were: 1) time taken to intubate the ileum;
2) depth of ileal intubation; and 3) presence or absence of ileal pathology.
Successful ileal intubation was defined as visualization of the ileal mucosa or villi
(confirmed by digital photography). [20] Time to ileal intubation was calculated from the first attempts, after randomization,
until visualization of the ileal mucosa, and was limited by pain or time ≤ 6 min.[19] After ileal intubation, ileal exploration was performed with minimal air insufflation
and was limited by pain and/or length of the colonoscope. Depth of ileoscopy was calculated
at the time of withdrawal of the scope, from the furthest point reached in the ileum
to the IC valve. Any abnormal finding was recorded and biopsies were taken, if deemed
necessary, by the performing endoscopist.
Patient tolerability of ileoscopy was rated by a three-point score: 0, no pain; 1,
pain not limiting the procedure; 2, pain limiting the procedure [19]. All adverse events were recorded including hypoxia, tachy- and/or bradycardia.
Ethical clearance and statistical analysis
Each patient’s data were meticulously recorded by one of the investigators (MA) in
a structured proforma. Written informed consent about the study protocol was taken
from each patient before colonoscopy and the study was approved by the Institutional
Ethics Committee (IEC).
All statistical analyses were performed on an intention-to-treat basis. Mean, median,
standard deviation (SD), standard error of the mean (SEM), range and proportions were
calculated, as appropriate. Categorical variables were analyzed using the χ2 test with the Yates correction or Fisher’s exact test, as appropriate. Continuous
variables were compared using the Student’s t-test. Multivariate analysis, to identify the predictors of successful ileal intubation,
was done by binary logistic regression analysis using variables found to be significant
on univariate analysis (P < 0.05). Statistical analysis was done using SPSSTM (Version 13 for Windows) software. A P-value of < 0.05 was considered to be statistically significant.
Results
Between June 1 and December 31, 2010, of the 653 patients referred for colonoscopy,
623 (95.4 %) had successful cecal intubation. Of the 653 patients, 320 (49 %) were
eligible for inclusion ([Fig. 2]). Of these, 217 (67.8 %; 150 males), were randomized, 106 and 111 undergoing ileoscopy
in the left-lateral decubitus and supine posture, respectively ([Fig. 2]: CONSORT Diagram).
Fig. 2 Flowchart of the study population.
At baseline ([Table 1]), the two groups were evenly matched with respect to demography, degree of colonic
preparation, type of IC valve, indication for attempted colonoscopy, and the endoscopist
performing the procedure. Endoscopist 1 (JD) was senior with > 15 years’ experience
in colonoscopy doing > 70 ileoscopies per year. Endoscopist 2 (SB) and 3 (RS) had
> 3 years’ experience in colonoscopy while endoscopist 4 (MA) had > 2 years’ experience
in colonoscopy. Endoscopist 2, 3 and 4 had already done > 100 ileoscopies, more than
twice the number regarded as necessary to achieve competence [19].
Table 1
Baseline characteristics of the two groups.
|
Left lateral decubitus
(n = 106)
|
Supine
(n = 111)
|
P-value
|
|
Age (mean ± SEM)
|
40.7 ± 1.5
|
39.8 ± 1.4
|
0.7
|
|
Male (%)
|
69
|
69
|
1.0
|
|
R-colon preparation (BBPS) [%]
|
|
|
0.5
|
|
Sub-optimal (score = 1)
|
23
|
22
|
|
|
Average (score = 2)
|
50
|
44
|
|
|
Good (score = 3)
|
27
|
34
|
|
|
IC valve type (%)
|
|
|
0.3
|
|
Thin-lipped (n = 70)
|
39
|
30
|
|
|
Single-bulged (n = 42)
|
19
|
21
|
|
|
Double-bulged (n = 66)
|
32
|
31
|
|
|
Volcanic (n = 31)
|
10
|
18
|
|
|
Indication (%)
|
|
|
0.3
|
|
Bleeding P/R
|
37
|
26
|
|
|
Diarrhea
|
10
|
14
|
|
|
Anemia
|
6
|
14
|
|
|
SAIO
|
12
|
10
|
|
|
FBD
|
12
|
13
|
|
|
Endoscopist (%)
|
|
|
0.7
|
|
No 1
|
13
|
9
|
|
|
No 2
|
27
|
24
|
|
|
No 3
|
14
|
15
|
|
|
No 4
|
46
|
52
|
|
P/R, per rectal; SAIO, subacute intestinal obstruction; FBD, functional bowel disorder
AU: Please check that FBD has been defined correctly.
Overall, the primary end point, i. e., successful ileal intubation, was achieved in
145 of 217 (66.8 %) patients, significantly more often in the supine (74.8 %) versus
the left-lateral decubitus (58.5 %) posture (P = 0.014). The predominant reason for failure in 69 patients was inability to intubate
the ileum in the stipulated predetermined time of ≤ 6 min (n = 55, 80 %) alone or
combined with poor right colon preparation (n = 9, 13 %).
Among the secondary outcomes ([Table 2]), the median time taken for ileal intubation was 74.0 s (range: 2 to 349 s) overall,
with a longer time required in LL posture versus S posture, which did not achieve
statistical significance. The mean (± SEM) depth of ileal intubation was 8.4 ± 0.3 cm
overall, with no difference between the two groups. No major adverse event developed
in any of the patients in the trial. Among the minor adverse events, there was no
difference between the two groups with respect to pain, hypoxia and bradycardia. Only
asymptomatic tachycardia was significantly more common in the supine posture ( [Table 2]).
Table 2
Secondary outcome measures in the two groups.
|
Total
|
Left lateral decubitus
(n = 106)
|
Supine
(n = 111)
|
P-value
|
|
Successful ileal intubation, n (%)
|
145 (66.8)
|
62 (58.5)
|
83 (74.8)
|
0.014
|
|
Depth of ileal intubation
(cm; mean ± SEM)
|
8.4 ± 0.3
|
8.6 ± 0.4
|
8.2 ± 0.3
|
0.5
|
|
Time to ileal intubation
(s; median, range)
|
74.0
(2 – 349)
|
84.5
(2 – 310)
|
71.0
(4 – 349)
|
0.3
|
|
Patient-perceived pain (%)
|
|
|
|
0.3
|
|
Nil
|
93
|
90
|
95
|
|
|
Present, not limiting
|
7
|
10
|
5
|
|
|
Adverse events (%)
|
|
|
|
|
|
Hypoxia
|
2
|
2
|
2
|
1.0
|
|
Tachycardia
|
16
|
9
|
22
|
0.015
|
|
Bradycardia
|
0.5
|
1.0
|
0
|
0.5
|
Clinically relevant findings were recorded in 13 (9 %) patients in whom successful
ileoscopy was performed: a small bowel source of bleeding was confirmed in 2; ulcers
± nodularity ± strictures were seen in 8 patients (Tuberculosis 5; Crohn’s 1; Idiopathic
2); terminal ileal nodularity in 1 and polyp in another 1 patient.
On univariate analysis, in addition to supine posture, successful ileal intubation
was more frequent in younger patients (mean ± SEM: 38.0 ± 1.2 vs 44.7 ± 1.9 years;
P < 0.01), in those with better right-colon preparation (BBPS average and/or good vs suboptimal:
72 % vs 48 %; P < 0.01) and in those who did not have a thin-lipped IC valve (other types vs thin-lipped:
79 % vs 43 %; P < 0.0001). Sex, endoscopist and other factors were not significant in univariate analysis.
On multivariate analysis, age (OR 0.97; 95 % CI: 0.95 – 0.99), supine posture (OR
2.2; 95 % CI: 1.1 – 4.3), average and/or good BBPS of right colon (OR 4.1; 95 % CI:
1.9 – 8.9) and IC valve type that was not thin-lipped (OR 5.1; 95 % CI: 2.6 – 9.8)
were independent predictors of successful ileal intubation ([Table 3]).
Table 3
Univariate and multivariate analysis of the factors associated with successful ileal
intubation.
|
Univariate
|
Multivariate
|
|
Success
(n = 145)
|
Failure
(n = 72)
|
P-value
|
OR
(95 % CI)
|
P-value
|
|
Age (mean ± SEM)
|
38.0 ± 1.2
|
44.7 ± 1.9
|
0.002
|
0.97
(0.95 – 0.99)
|
0.02
|
|
Male (%)
|
71
|
65
|
0.4
|
|
|
|
R-colon preparation (BBPS) [%]
|
|
|
0.007
|
|
|
|
Sub-optimal (score = 1)
|
16
|
35
|
|
1
|
|
|
Average/Good (score = 2 /3)
|
84
|
65
|
|
4.1
(1.9 – 8.9)
|
< 0.0001
|
|
Patient posture (%)
|
|
|
0.014
|
|
0.02
|
|
Left lateral
|
43
|
61
|
|
1
|
|
|
Supine
|
57
|
39
|
|
2.2
(1.1 – 4.3)
|
|
|
IC valve type (%)
|
|
|
< 0.0001
|
|
< 0.0001
|
|
Thin-lipped
|
22
|
58
|
|
1
|
|
|
Others
|
78
|
42
|
|
5.0
(2.6 – 9.8)
|
|
|
Indication (%)
|
|
|
0.6
|
|
|
|
Bleeding P/R
|
28
|
37
|
|
|
|
|
Diarrhea
|
13
|
10
|
|
|
|
|
Anemia
|
10
|
10
|
|
|
|
|
SAIO
|
12
|
8
|
|
|
|
|
FBD
|
12
|
12
|
|
|
|
|
Endoscopist (%)
|
|
|
0.2
|
|
|
|
No 1
|
9
|
15
|
|
|
|
|
No 2
|
26
|
24
|
|
|
|
|
No 3
|
17
|
8
|
|
|
|
|
No 4
|
48
|
53
|
|
|
|
When analyzing the independent predictors of successful ileal intubation separately
in the LL and S postures, by multivariate analysis, age, degree of right-colon preparation
and type of IC-valve were significant independent predictors of ileal intubation in
the LL posture, whereas the latter two alone were significant independent predictors
of ileal intubation in the S posture (data not shown).
Discussion
This prospective randomized trial demonstrated that, within a predetermined stipulated
time frame, ileal intubation is significantly more successful in the supine than in
the left-lateral decubitus posture. Moreover, in addition to supine posture, younger
age, better right colon preparation and type of IC valve (non-thin-lipped) are independent
predictors of successful ileal intubation. The major strengths of our study are its
prospective randomized nature, sample-size generation, blinded allocation and analysis,
using predefined stringent criteria for success and failure and use of a well-validated
bowel preparation score.
There have been very few studies about the technical feasibility of terminal ileoscopy
during colonoscopy [14]
[16]
[17] with most of the studies predominantly focusing on the yield of ileoscopy and/or
ileal biopsies [5]
[6]
[7]
[8]
[9]
[10]
[21]
[22]
[23]
[24]. Most endoscopists intubate the terminal ileum in the left-lateral decubitus posture
with the ileocecal valve in the 6 – 7 o’clock position, using downward deflection
and anticlockwise torque while withdrawing from the cecal cone [16]. Prospective studies have shown that the ileum is successfully intubated in this
position in 65 – 78 % of patients, depending on the training level of endoscopists
[16]
[19]
[25]. Shifting the posture to supine, however, achieves an ileoscopy completion rate
of > 95 %. [16]
[19] It is against this background that our trial conclusively demonstrates that the
supine posture is technically the easier posture to intubate the terminal ileum, with
a higher success rate.
While our analysis was ongoing, another randomized trial was published that showed
that the prone posture is equivalent to the left-lateral decubitus posture with regard
to the frequency of ileal intubation achieved, albeit within a shorter time [26]. However, that trial did not stratify or analyze the confounding factors such as
endoscopist, bowel preparation and IC valve morphology [26].
In addition to posture, we also found that age, degree of right-colon preparation
and morphology of IC valve are independent predictors of ileal intubation success.
Younger age was associated with increased success at ileal intubation overall, and
especially in left lateral decubitus posture. Age was not an independent predictor
of ileal intubation success in the supine posture. The quality of right-colon preparation
was an independent predictor of success overall, as well as in both postures independently.
This is to be expected as good bowel preparation is an important determinant of technical
success in colonoscopy and in any colonoscopic procedure.
In compliance with a previous study, the morphology of the IC valve was an independent
predictor of ileal intubation success. [19] The thin-lipped type of valve is difficult to intubate as it is less distensible
and often flush with the semicircular fold of the cecum.
One drawback in our study was the lower rate of successful ileal intubation overall
(~ 67 %). Although this was well within the range reported in the literature, the
major reason (~ 80 %) for failure was the inability to intubate the ileum within the
stipulated time frame of ≤ 6 min. We had stipulated a cut-off time of 6 min, in compliance
with a previous study, and in addition, this is double the average time required to
intubate the ileum according to the literature [19]. Thus, in a prospective controlled trial setting with a time-pressure, the rate
of success might be lower for complex visuo-spatio-psychomotor procedures such as
colonoscopy and ileoscopy [27]. Moreover, our population was predominantly male and had a higher prevalence of
the difficult thin-lipped type of IC-valve vis-à-vis, the Italian population reported
in the literature [20]. This could also have contributed to the lower rate of success.
Reviewing the literature, we found two studies that prospectively assessed the technical
success of ileoscopy during colonoscopy. In the first study, overall success was achieved
in 79 %, with a longer time (mean: 3.4 min; range: 30 s to 10 min) and no difference
between endoscopists having different levels of training [14]. In the second study, ileoscopy, performed by a single colonoscopist, was achieved
without any intervention in 77.8 %, which increased to ~ 97 % when changing the posture
to supine and/or use of an anti-peristalsis agent (Hyoscine-n-butyl bromide) in one-fifth
of patients, respectively [16]. The median time taken for intubation was 55 s (range: 2 – 1140 s) [16], not much different from our rates. Moreover, we did not use any anti-peristalsis
agent which is proven to facilitate ease of ileal intubation [28].
One of the biggest impediments to the technical success of colonoscopy is the result
of “looping” in the tortuous and often redundant areas of the colon, especially the
sigmoid and transverse colon [29]
[30]
[31]. We speculate that the same should hold for ileoscopy, as it is difficult to intubate
the terminal ileum if the colonoscope is not de-looped. Older age is one of the risk
factors for increased looping, which is reflected in our study by the decreased success
at ileoscopy in older age [31]. Posture change is one of the maneuvers that successfully de-loops the colon, especially
in the sigmoid and transverse colon [30]. We speculate that this might be the reason for the increased success of ileoscopy
in the supine vis-à-vis the left lateral decubitus posture.
In conclusion, this randomized controlled trial proves that ileal intubation is more
successful in supine than left-lateral decubitus posture. Age, bowel preparation and
type of IC valve also determine successful ileoscopy.
