Trial Information

Does Endoscope Position Detecting Unit Facilitate Caecal Intubation? A Randomized Controlled Trial

Hypothesis UPD guided colonoscopy improves cecal intubation rate.

Methods

Trial design This is a prospective randomized controlled trial comparing the cecal
intubation rate of colonoscopy being performed with and without UPD guidance by endoscopists
of different levels of experience. The trial will be conducted in the Endoscopy Unit of a
university affiliated hospital.

Patients who will have elective colonoscopy in the Endoscopy Unit during the study period,
and meet the following inclusion criteria are recruited. The colonoscopy will be done by 6
dedicated endoscopists with different levels of experience in performing colonoscopy.

Interventions

UPD-3 and the specific colonoscope with an in-built magnetic imaging facility (Olympus
Optical Co, Ltd, Japan) will be used in all examinations. A mobile sensor will be attached
to the assistant's hand, so that the hand position in relation to the colonoscope could be
shown in the imager screen.

Patient controlled sedation (PCS) will be given to all patients. It is delivered by means
of Graseby 3300 PCA pump (Graseby Medical Ltd), using a mixture of propofol (200mg in 20ml)
and alfentanil (0.5mg in 1ml) in a 25-ml syringe. The drugs will be delivered in response
to pressure on a hand-held button controlled by the patient. Each bolus (0.5ml) delivers
4.8mg propofol and 12microgram alfentanil. No loading dose was used and the lock-out time of
the PCA machine was set at zero. All patients will be instructed in the use of the hand-held
button before colonoscopy. Nasal oxygen (2l/min) will be given to all patients, and oxygen
saturation and blood pressure will be continuously monitored throughout the procedure.

The procedure will be randomized to either with the colonoscopist viewing the imager screen
or not.

UPD guided group: Both the colonoscopist and assistant will be viewing the imager screen
during the whole procedure. Information about colonoscopy looping will interpreted and
documented by the colonoscopist. Anterior-position view is the standard reference. When
abdominal compression is required, the assistant nurse should have the hand held probe
attached, such that abdominal compression is guided by the image.

Conventional group (non- UPD guided): The imager screen will be turned to the other side
such that both the colonoscopist and assistant is blind to UPD images. When the
colonoscopist encountered looping as defined by paradoxical movement of the tip or
disproportionate advancement in endoscopic view to the shaft insertion, the colonoscopist is
requested to estimate the type of loop formed. The images will be captured for later
interpretation.

The primary outcome is cecal intubation rate. Cecal intubation is defined as visualization
of appendix orifice and ileocaecal valve, procedure started by a trainee and taken over by
supervisor before caecal intubation is regarded as incomplete.

Sample size

The cecal intubation rate could be increased by at least 11% for trainee in a randomized
trial by Shah [14] and 16% for trainee and experienced endoscopist in another trial [15].
Since the overall cecal intubation rate in our Endoscopy unit without UPD guide is at least
90% already, the investigators expect a modest benefit from UPD guidance. Assume a
difference of 5% in caecal intubation rate (from 90% in conventional group to 95% in the UPD
guided group) with a power of 80% and a type I error of 5%, 2 sided analysis, 435 patients
will be needed in each group. Thus a sample size of 870 patients is required.

Randomization

Informed consent will be obtained before colonoscopy in the Endoscopy Unit, if the patient
fulfils the inclusion criteria. Block randomization will be done on a 1:1 basis, to ensure
equal number of UPD guided or conventional colonoscopies are done by each endoscopist.

A sealed non-opaque envelope (produced according to the computer generated random sequence)
will be opened to determine the appropriate limb of entry before scope insertion.

It is not possible to blind the study.

Statistical analysis

The data will be analyzed according to the intention-to-treat principle. The chi-squared
test (or Fisher's exact test when appropriate), Student's t-test, and Mann-Whitney U test
will be used to compare categorical, parametric, and non-parametric data, respectively. A P
value <0.05 is considered statistically significant.

Planned interim analysis will be conducted after recruiting 50% of patients.

Data collection Data will be collected by the in-charged endoscopist and assistant after
each procedure in a standardized data sheet.

Accrual rate It is expected that about 20 cases per week will be recruited for
randomization. The study could be finished in around 1 year time.