Introduction
Use of colonoscopy as a diagnostic and therapeutic tool is likely to rise. This is
mainly a result of implementation and expansion of colorectal cancer screening projects,
targeting ever younger patients [1]. Endoscopic procedures are associated with embarrassment, pain, and discomfort [2]. This proves an important barrier to undergoing colonoscopy and may subsequently
make patients less willing to be subjected to repeat surveillance colonoscopies [3]
[4].
Indeed, a relevant proportion of patients (18 % to 29 %) experience anxiety due to
concerns related to preparation for, execution of, and anticipation of the result
of colonoscopy [5]. Sedation to relieve anxiety is the method of choice used to mitigate the discomfort
patients experience during colonoscopy [6]. However, drug-induced sedation comes with adverse effects related to suppression
of pulmonary and circulatory function [7]
[8]. There is a higher post-procedural risk of pneumonia in elderly patients [9]. Deep sedation even puts patients at increased risk for procedure-related complication
of perforation [10]. Also, monitoring patients during and after sedation is both logistically demanding
and costly [11].
Therefore, several studies have examined non-pharmacological interventions to reduce
anxiety and pain during endoscopy [12]
[13]
[14]
[15]
[16]
[17]
[18]. These studies used a mix of visual [13]
[15]
[18]
[19] or auditory stimuli [12]
[14]
[18] and found that while true efficacy is not fully established, combined visual and
auditory distraction is better at reducing discomfort than auditory distraction alone
[18].
Virtual reality (VR) integrates computer-generated visual and auditory signals to
recreate an illusionary perception of the actual physical world [20]
[21]. The distraction that comes with immersive VR induces an analgesic effect and has
been used as an adjunct to control pain and anxiety during operative procedures [22]
[23]. VR technology has become more affordable and portable, adding to its immersive
qualities [24]. VR reduces pain during burn wound debridement [25]
[26]
[27], and discomfort during dental procedures [28].
A questionnaire study found that up to 25 % patients are willing to undergo colonoscopy
with VR glasses instead of sedation. Key patient motive is the reduction in sedative
use, which allows patients to drive their cars home themselves afterwards [29].
But still unknown is patient acceptance (e. g. feasibility) of performing colonoscopy
on patients actually wearing VR glasses. Wearing VR glasses could potentially be disadvantageous
to the colonoscopy procedure, as it could obstruct communication with patients. Conversely,
the procedure itself might compromise the VR effect, as positional changes of the
patient are sometimes necessary. Therefore, we set out in this pilot study to investigate
use of VR distraction during colonoscopy. The primary aim was to assess patient acceptance
of wearing VR glasses while undergoing colonoscopy. We were also interested in whether
VR reduces discomfort, pain, and anxiety and increases satisfaction in patients compared
to the standard practice.
Patients and methods
This experiment was designed as a pilot study to evaluate patient acceptance and practical
feasibility. A control group was designed to allow evaluation of procedural and patient-related
outcomes. The sample size was set at 12 subjects per group. This computation was based
on a rule of thumb for pilot studies [30]. Ethical permission from the Radboudumc Ethics Committee was obtained prior to commencement
of the study (number 2016–2750). The trial was registered with the Dutch trial Registry
(NTR6175).
Patients
We screened patients who were already scheduled for outpatient colonoscopy. Inclusion
criteria of the study were adult age and any elective indication of colonoscopy. Exclusion
criteria of the study were visual and/or auditory impairments, dementia, limited Dutch
language skills, and a diagnosis of balance disorders or epilepsy.
After evaluation of the above criteria, informed consent was obtained from all participants
and they were allocated to the VR (intervention) or non-VR (control) group. Allocation
was based on the day the colonoscopy was planned. Participants were informed about
the group to which they were allocated on the day of the procedure.
Intervention
The hardware we used to generate VR distraction was the Samsung Gear VR (Consumer
Edition–SM-R322, combined with Galaxy S7). This is an inexpensive ($ 172) off-the-shelf
wide field-of-view, three-dimensional VR headset that projects video and rendered
graphics into two independent lenses. The current model is the size of a small pair
of ski goggles, with a combined weight of 470 g, and is positioned on the head with
elastic straps. The video content that was visualized on the VR hardware consisted
of several short clips (with a total length 19 minutes, 59 seconds) of moving 360-degree
cameras featuring tropical islands and forests in the Caribbean (supported by VR firm
Visyon, Eindhoven, The Netherlands). The VR content had not audio, to allow optimal
communication with the patient. The authors considered the chosen VR content to be
of a relaxing and not overly thrilling character, generating an adequate level of
distraction for all participants.
Study design
At T1, all participants filled out a baseline form on a tablet with information about
demography, prior experience undergoing colonoscopy, prior VR experience, and a validated
general health questionnaire (RAND-36) [31]. On T2, the day of colonoscopy, about 15 minutes before the procedure, all patients
received a second form that included validated questionnaires on anxiety (STAI) and
pain (NRS) [32]
[33]. Patients in the intervention group also tested the VR glasses before colonoscopy.
During colonoscopy, T3, one researcher (NK) observed the patient’s well-being and
positioning together with several procedural aspects e. g., time to cecul intubation
and time of total procedure. All patients received conscious sedation with midazolam
and/or alfentanyl according to the standard of care, with the dose increased at physician
discretion. After colonoscopy, patients completed a set of questionnaires at T4, including
questions about anxiety (STAI), pain (NRS), net promoter score, and willingness-to-return
questions. A short qualitative interview was held with the patients in the intervention
group to explore their experiences with VR glasses.
Measures
Primary outcome
The main outcome was patient acceptance of wearing VR glasses during the procedure.
That included adequate positioning of the VR glasses during the entire procedure,
even during patient repositioning. In addition, we recorded cecal intubation rate,
cecum and total procedure time as well as administered sedatives and analgesics.
Secondary outcomes
Patient comfort
Patient comfort was measured using a five-point Gloucester Comfort scale: 1, comfortable
and 5, severe discomfort [34].
Patient pain
An 11-point numeric rating scale (NRS) was used to measure pain of the patient before
and during the procedure: 0, no pain and 10, highest imaginable pain [32].
Patient anxiety
The State Trait Anxiety Inventory (STAI) was used to measure patient anxiety before
and after the procedure. The 20-item STAI is widely used with scores ranging from
20 (absence of anxiety) to 80 (high anxiety) [33].
Patient satisfaction
The general health of the participants was measured using the RAND 36 questionnaire
[31]. Net Promoter Score (NPS) [35] and an 11-point scale of willingness to return: 0, no willing at all and 10, definitely
willing, were used to measure participant satisfaction with the procedure.
Statistical analysis
Statistical analysis was performed using SPSS version 22 (International Business Machines
Corporation, Armonk, New York, United States). Mann Whitney U-tests were used to test
whether the median scores for, i. e., age, pain, dose of medication, duration of the
procedure, anxiety, satisfaction, NPS, and willingness to return, of the VR (intervention)
and non-VR (control) group were comparable to each other. Fisher’s Exact tests were
used to test categorical data. P ≤ 0.05 was considered statistically significant.
Results
In total 24 patients entered the trial ([Fig. 1]). Patients were recruited at the endoscopy outpatient clinic. There were 55 eligible
patients scheduled for colonoscopy within a 4-week timeframe and we invited 38 consecutive
patients, 24 of whom accepted our invitation. Informed consent was obtained from all
patients. After allocation, two patients in the VR (intervention) group and three
patients in the non-VR (control) group were excluded (three patients cancelled the
scheduled appointment, one was admitted to the hospital, and in one patient there
was a technical problem with the endoscopy equipment). As a result, 19 patients were
included in the final analysis, 10 in the VR group and nine in the non-VR group. All
patients in the intervention group used the VR glasses during the whole procedure
([Fig. 2]). No adverse events associated with VR distraction in combination with medication
were observed. One endoscopist performed all the procedures (FV) except one in the
VR group (BvH). FV had > 5 years of experience, BvH > 3 years.
Fig. 1 Study flowchart.
Fig. 2 Samsung Gear VR shown on a patient during colonoscopy (with permission).
Baseline characteristics
No significant differences were observed in baseline characteristics of the two groups,
i. e., gender (55.6 % women in the control group, versus 60 % women in the intervention
group), age (median, 64, versus 65 years), level of education, RAND-36, previous colonoscopy,
and prior experience with VR ([Table 1]).
Table 1
Baseline characteristics.
|
Control (non-VR)
(n = 9)
|
Intervention (VR)
(n = 10)
|
P value
|
|
Age (years)[1]
|
64 [47.5; 67.5]
|
65 [62, 67]
|
0.414[2]
|
|
Gender (male:female)
|
4:5
|
4:6
|
1.000[3]
|
|
RAND-36
|
|
Physical functioning[1]
|
90 [70, 100]
|
82.5 [72.5; 95]
|
0.549[2]
|
|
Role limitations due to physical health[1]
|
87.5 [68.75; 100]
|
68.75 [50; 84.38]
|
0.156[2]
|
|
Role limitations due to emotional problems[1]
|
100 [53.13; 81.25]
|
83.33 [47.92; 100]
|
0.133[2]
|
|
Energy/ fatique[1]
|
75 [72.5; 90]
|
59.38 [48.44; 81.25]
|
0.497[2]
|
|
Emotional well-being[1]
|
85 [75, 100]
|
75 [50; 81.25]
|
0.113[2]
|
|
Social functioning[1]
|
100 [73.47; 94.9]
|
81.25 [62.5; 100]
|
0.113[2]
|
|
Pain[1]
|
89.79 [73.47; 67.5]
|
72.45 [67.35; 100]
|
0.497[2]
|
|
General health[1]
|
55 [35; 67.5]
|
57.5 [52.5; 66.25]
|
0.497[2]
|
|
Health change[1]
|
50 [25; 62.5]
|
37.5 [25, 50]
|
0.549[2]
|
|
Number of previous colonoscopy[1]
|
2.5 [1.75; 5]
|
2 [1.25; 3.75]
|
0.515[2]
|
|
Level of education
|
|
|
0.733[4]
|
|
Primary school
|
0 (0)
|
0 (0)
|
|
|
Lower vocational education
|
0 (0)
|
10 (1)
|
|
|
Lower general secondary school
|
0 (0)
|
0 (0)
|
|
|
Intermediate general secondary school
|
11.1 (1)
|
10 (1)
|
|
|
Intermediate vocational education
|
22.2 (2)
|
20 (2)
|
|
|
Upper general secondary school
|
22.2 (2)
|
10 (1)
|
|
|
Higher vocational education
|
33.3 (3)
|
50 (5)
|
|
|
University
|
11.1 (1)
|
0 (0)
|
|
|
Prior experience with VR (yes) % (n)
|
22.2 (2)
|
30 (3)
|
1.000[3]
|
VR, virtual reality
1 Variables are denoted as median (interquartile range).
2 Mann-Whitney U test.
3 Fisher’s Exact test.
4 Chi-square test.
Procedure characteristics
There were no differences in procedural characteristics. The time to reach the cecum
(median 10.48 minutes in the control group, versus 6.83 minutes in the intervention
group), time to complete procedure (median 21.20 minutes in the control group, versus
22.60 minutes in the intervention group), and completed colonoscopies (100 % in the
control group, versus 90 % in the intervention group) were comparable in the two groups
([Table 2]).
Table 2
Procedure characteristics.
|
Control (non VR)
(n = 9)
|
Intervention (VR)
(n = 10)
|
P value
|
|
Dose midazolam, in mg[1]
|
2.5 [2.5; 3]
min 2; max 3.75
|
2.5 [2.38; 3]
min 2.5; max 3
|
0.842[2]
|
|
Dose alfentanyl, in mg[1]
|
0.25 [0.25; 0.50]
min 0.25; max 0.50
|
0.25 [0.25; 0.5]
min 0.25; max 0.50
|
0.278[2]
|
|
Completed colonoscopies % (n)
|
100 (9)
|
90 (9)
|
1.000[3]
|
|
Patient acceptance of VR glasses % (n)
|
n/a
|
100 (10)
|
n/a
|
|
Time to reach the cecum, in minutes[1]
|
10.48 [8.65; 13.80]
min 6.10; max 19.00
|
6.83 [5.75; 10.77]
min 2.66; max 11.92
|
0.094[2]
|
|
Time to complete procedure, in minutes[1]
|
21.20 [19.72; 35.15]
min 19.18; max 44.07
|
22.60 [16.25; 25.45]
min 9.95; max 26.43
|
0.340[2]
|
VR, virtual reality
1 Variables are denoted as median (interquartile range).
2 Mann-Whitney U test.
3 Fisher’s Exact test.
Similarly, both groups were comparable in terms of initial intravenous bolus of sedatives
and analgesics, i. e., dose of midazolam (median, 2.5 mg in both groups), dose of
alfentanyl (median, 0.25 mg in both groups).
Patient pain, comfort and anxiety
Results of pain scores, patient comfort, and anxiety scores are summarized in [Table 3]. Median pain score before (0 in both groups) and during (3 in both groups) the procedure
was similar in both groups. The Gloucester Comfort scale did not reveal significant
differences in patient comfort between the two groups (4 patients [44 %] in the control
group were rated comfortable, versus 4 patients [40 %] in the intervention group).
No significant difference was observed in median anxiety score prior to the procedure
(49 in the control group, versus 48.5 in the intervention group). Median baseline
anxiety score (trait) was similar in the intervention group and the control group
(29 in the control group, versus 35 in the intervention group). Median anxiety score
increased after the procedure (50 and 50).
Table 3
Pain, patient comfort and anxiety results.
|
Control (non-VR)
(n = 9)
|
Intervention (VR)
(n = 10)
|
P value
|
|
Pain score (pre-procedure)[1]
|
0 [0, 3]
|
0 [0; 1.75]
|
0.968[2]
|
|
Pain score (during procedure)[1]
|
3 [1]
[4]
|
3 [1.5; 5.5]
|
0.661[2]
|
|
Gloucester comfort scale % (n)
|
|
|
0.699[3]
|
|
Comfortable
|
44.4 (4)
|
40 (4)
|
|
|
Minimal
|
44.4 (4)
|
30 (3)
|
|
|
Mild
|
11.1 (1)
|
20 (2)
|
|
|
Moderate
|
0 (0)
|
0 (0)
|
|
|
Severe
|
0 (0)
|
10 (1)
|
|
|
STATE (Pre-procedure)[1]
|
49 [48, 50]
|
48.5 [45.75; 50.25]
|
0.497[2]
|
|
TRAIT[1]
|
29 [21; 36.5]
|
35 [28; 41.5]
|
0.156[2]
|
|
STATE (Post-procedure)[1]
|
50 [48; 52.5]
|
50 [47.75; 51.25]
|
0.549[2]
|
VR, virtual reality
1 Variables are denoted as median (interquartile range).
2 Mann-Whitney U test.
3 Chi-square test.
Patient satisfaction
No differences were observed between the two groups in patient satisfaction. All patients
scored high satisfaction rates in the scales that were used (median score was of 9
out of 10 in both groups). Results on patient satisfaction are summarized in [Table 4].
Table 4
Satisfaction results.
|
Control (non-VR)
(n = 9)
|
Intervention (VR)
(n = 10)
|
P value
|
|
Patient satisfaction[1]
|
9 [8]
[10]
|
9 [6.5; 10]
|
0.905[2]
|
|
NPS[1]
|
9 [8]
[10]
|
9 [7.75; 10]
|
0.905[2]
|
|
Willingness to return[1]
|
9 [7.5; 10]
|
9 [6.75; 10]
|
0.720[2]
|
VR, virtual reality.
NPS, net promoter score.
1 Variables are denoted as median [interquartile range].
2 Mann-Whitney U test.
Qualitative comments
The majority of patients (9/10) rated use of VR glasses as positive. Four patients
indicated that they preferred to select the VR content themselves. Two patients complained
about the quality of the movie and one patient indicated that the resolution of the
VR movie was too low. The physician who performed the colonoscopy was able to communicate
with all patients in the intervention group and did not experience any limitations
with use of VR.
Discussion
Our pilot study shows that it feasible to use VR distraction during colonoscopy as
we observed complete patient acceptance of the device during all procedures. Procedural
time was not longer as a result of our intervention.
Comfort, pain, anxiety and patient satisfaction were not affected by VR, but patients
reported a positive distracting effect of the VR glasses.
This pilot study indicates that there are no obstacles to investigating VR glasses
further in a larger sample of patients. Important to the design of subsequent trials
from the endoscopist perspective is that use of VR glasses did not interfere with
the completion colonoscopy.
Various studies have found that visual and/or auditory distraction during endoscopic
procedures reduces pain and improves satisfaction as a result [13]
[15]
[18]. In this pilot we were not able to identify these advantages for VR. This is similar
to outcomes in two trials of VR in burn wound victims in which the authors suggested
resolving this issue by developing a better-customized VR system instead of off-the-shelf
VR sets [36]
[37].
Indeed, patients reported that the effect of the VR distraction was less immersive
probably because of the content shown. Other studies have found that content is relevant
to the level of distraction [12]
[15]. Low pixel resolution of the VR content influenced the experience of at least one
participant and previous studies hhave shown that low-resolution videos reduce the
quality of the experience [38].
The literature on VR for patients in endoscopy is scarce. A retrospective study of
190 patients found that VR allowed unsedated transnasal gastroscopy in children and
young adults. In this study, VR-assisted transnasal gastroscopy was safe and cost-effective
for staging of eosinophilic esophagitis [39]. The argument has been made that VR makes it possible to avoid sedation for colonoscopy,
which fuels patient experience [40]. Therefore it is probable that in selected patients, VR during colonoscopy will
be the preferred option [29].
Strengths and limitations
Our study was performed in a real-life setting and a representative sample of patients,
which add to its external validity. By using Samsung Gear VR to provide distraction,
we chose a widely available and relatively inexpensive VR device, enhancing the generalizability
of the results.
Our study also comes with limitations. First, the small sample size does not allow
robust statements on clinically relevant endpoints like reducing anxiety or pain or
improving patient satisfaction. Also, recent literature suggests that our sample size
computation carries the risk of overestimation of the required sample size when designing
a main trial to confirm our results [41]. Second, the physicians who performed the procedure were not blinded, because the
patients in the control group did not wear VR glasses. Although we did not observe
a difference in administration of drugs in the control and intervention groups, this
could have affected the choice and dose of sedatives. The ideal set-up is a direct
comparison of sedation vs VR, instead of VR combined with sedation as done in our
pilot.
We used patient-reported measures for pain and comfort after patients were recovering
from sedative administration. The post-amnesia effect of midazolam might have had
some effect, but the Gloucester scale rated by the nursing staff revealed no differences
between groups.
Because of the low costs of the VR device, VR distraction may easily be deployed in
colonoscopy. There are several technical shortcomings such as low resolution, orientation,
and content, which if resolved may improve the distractive effect and help ensure
enhanced patient comfort and satisfaction.
To achieve a maximal immersive effect, VR content must be developed that provides
specific targeted distraction for colonoscopy, such as relaxing colors, relaxing music,
and properly selected visualizations.
Conclusion
In summary, patients accepted VR distraction undergoing colonoscopy, without compromising
the technical success of the procedure. Future studies are justified to evaluate the
possible substitution of sedation with VR. Patients reported that the VR experience
was pleasant and distracting, facilitating recruitment for these trials.
