Keywords Quality and logistical aspects - Training - Quality management - Endoscopy Lower GI
Tract - Endoscopic resection (polypectomy, ESD, EMRc, ...)
Introduction
Prevalence of musculoskeletal disorders (MSD) among endoscopists is estimated to range
from 29% to 89%, leading to health issues for practitioners and negatively impacting
procedure safety and performance during procedures [1 ]
[2 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ].
Factors associated with MSDs among endoscopists include a high number of procedures
(over 20/week), prolonged procedure times (over 16 hours/week), the age of the endoscopist
(>40 years vs ≤ 40 years), and female sex [3 ]
[5 ]
[6 ]
[7 ]
[9 ]
[12 ]
[13 ]. Awkward postures and invariability of the precise gesture are biomechanical risk
factors of MSD.
Evolution of interventional digestive endoscopy, especially with introduction of endoscopic
submucosal dissection (ESD), has led to increased procedure complexity characterized
by an increase in procedure duration, greater precision of the task, and maintenance
of prolonged static posture. The spread of ESD techniques in endoscopic procedures
may elevate risk of MSDs due to longer procedure time [14 ]
[15 ]
[16 ]
[17 ].
Recommendations have been proposed to enhance practitioner comfort, akin to those
utilized by digestive surgeons [18 ]
[19 ]. These include utilizing a height-adjustable examination table, positioning the
screen at eye level, and using a seat during procedures.
Most of the studies on endoscopist MSDs are based on surveys, which introduce biases
when quantifying the role of endoscopic procedures in development of MSDs. Recently,
Shergill et al. demonstrated that during colonoscopy, left wrist extensor muscle activity
exceeded established thresholds with the greatest risk of constraints occurring during
colonic insertion [20 ].
Still, there are limited data on biomechanical stresses encountered by endoscopists
during many specific procedures (endoscopic retrograde cholangiopancreatography, endoscopic
mucosal resection) and particularly ESD. Quantitative assessment of muscular and postural
strains associated with ESD procedure activity and tools is crucial to evaluate them
against risk thresholds and to prevent MSD occurrence.
The goal of this study was to assess, for the first time, biomechanical risk factors
such as muscular, kinematic angulation and awkward posture strains during ESD and
to compare them with strains of diagnostic colonoscopy.
Methods
Population
Interventional endoscopists from three French university hospital (Amiens, Lyon, Lille)
were recruited to participate in 2024. The endoscopists provided informed consent.
The study was approved, according to national guidelines, by the CNIL committee (Comité
Consultatif sur le Traitement de l’Information en matière de Recherche dans le domaine
de la Sante) - record number NCT06549894. Each gastroenterologist completed a survey
collecting demographic information such as sex, age, body mass index, years of endoscopy
practice. Type of colonoscope used, procedure duration, use of a height-adjustable
examination table, and positioning of the screen at eye level were noted. Participants
performed one diagnostic colonoscopy and one ESD after placement of biomechanical
measurement devices. To increase accuracy and because we suspected that gesture and
therefore biomechanical strain were different, the ESD procedure was split in two
specific sub-tasks: the initial phase of incision to access to the submucosal space
(ESD-1) and the second phase of submucosal dissection (ESD-2).
Outcome measures
Perceived physical effort or discomfort
To assess physical effort or discomfort during the different endoscopy procedures,
physicians were asked to complete a Borg scale [21 ]. The 10-point rating scale was used (0 = no effort, 10 = maximal effort).
Forearm muscle activity
A wireless transmission module T-Sens was used to measure bilateral muscle surface
electromyographic (EMG) activity of the common flexor and extensors carpi radialis.
Bipolar surface electrodes were placed on the skin after abrasion and cleaning with
alcohol (EMG Triode Electrode, Thought Technology Ltd) using standard anatomic locations
([Fig. 1 ] and [Fig. 2 ]). Root mean square value of EMG was directly collected at a sampling rate of 128
Hz and was calculated and smoothed with a time window of 300 ms.
Fig. 1 Schematic representation of electromyographic and angular sensor positions Yellow
stars: deltoid EMG sensor; red star: flexor carpi radialis; green star: extensor radi
carpialis; orange circle: angular sensor for neck, trunk, shoulder, elbow and wrist.
Fig. 2 Schematic representation of flexor and extensor carpi radialis.
Three maximum voluntary contractions (MVCs) for each muscle were elicited for 3 to
5 seconds by maximally resisting appropriate joint motions [22 ]. The signal for each muscle was then normalized to %MVC.
Head, neck, back and upper-limb joint positions and movements
11 Inertial Measurement Units (IMU) T-Sens (T-Sens Motion, TEA Ergo, France) were
used to record position of the different joints and to record 3 degrees of freedom
of the upper-limb movements during endoscopy procedure ([Fig. 1 ]). T-Sens wireless motion sensors integrate a 3-axis accelerometer, a 3-axis gyroscope,
and a 3-axis magnetometer (IMU characteristics: precision head/pitch, roll: 2°/0.5°;
gyroscope: ± 2000°/s; accelerometer ±16 g; magnetometer ± 2.5 Gauss). They were attached
to the hands, forearms, arms, head, trunk, low back and thighs. Their acquisition
frequency was 32 Hz. Using the manufacturer's software, the IMUs were automatically
calibrated to similar parameters landmark. At the end of the acquisitions, the software
automatically generates processed data on the selected joints. The data is then streamed
into CAPTIV (TEA Ergo, France), a software that enables accurate displays and analysis
of human body motion angles on joints, angular speed and acceleration.
Biomechanical parameters
Muscle activity
Muscle activity was expressed as %MVC. High muscle strain of muscle groups was defined
as any contraction of the muscles more than 10% of the maximum voluntary contraction
which allows to determine the time spent over this threshold [23 ].
IMU measures
For neck, trunk, and wrist postures, angles were reported as flexion/extension in
the sagittal plane. For the shoulders and elbow, angulation was reported with 0° being
at the neutral position of the arm. Posture angles were measured during a static endoscopic
task. Static posture is defined as joint angular velocity lower than 1°/s [24 ]
[25 ]. Percent time spent in demanding postures was calculated. Such postures were defined
as angles outside the range of the recommended limits [26 ]
[27 ]
[28 ]: >10° neck flexion or extension, >20° trunk flexion, >45° shoulder elevation, <60°
or >100° elbow flexion (forearm position), >0° wrist flexion or >30° wrist extension.
Time in demanding postures was compared between the different operating procedures.
Statistical analysis
Due to the low number of participants, non-parametric statistic tests were performed
and data were expressed with the median (min-max). All metrics were compared between
three tasks: colonoscopy, the ESD first phase (ESD-1): injection and incision of the
mucosa and the second phase of ESD (ESD-2): dissection, using non-parametric, paired
Wilcoxon test. Effect size between endoscopic procedures was calculated using r coefficient.
Effect size r was calculated as Z statistic divided by square root of the sample size
(N). Interpretation of effect size was: small (r = 0.1), medium (r = 0.3), large (r
≥ 0.5).
Results
Population
Six endoscopists were included in this study. The majority were men (83.3%) with a
median age of 40 years (37.8–40) and practice of endoscopy for 11 years (8–32) ([Table 1 ]).
Table 1 Gastroenterologist characteristics.
N = 6
IQR, interquartile range.
Men, n (%)
5 (83.3)
Age, years, median (IQR)
40 (37.8–40)
Body mass Index, kg/cm2 ) median (IQR)
21.9 (21.1–23.6)
Years of endoscopy practice, median (IQR)
10.8 (6.4–11.8)
Right-handed, n (%)
5 (83.3)
Musculoskeletal disorders, n (%)
6 (100)
3 (50)
1 (16.7)
2 (33.3)
3 (50)
5 (83.3)
2 (33.3)
1 (16.7)
0 (0)
Hours spent in endoscopy/week, median (IQR)
24.3 (22.9–26.3)
Data were fully available for 12 procedures (6 colonoscopies, 6 ESD). Median time
for a colonoscopy was 9 minutes (7–64) and 43 minutes (14–98) for ESD (P = 0.01). ESD-1 lasted 19 minutes (18–44) and ESD-2 lasted 24 minutes (9–77) (P = 0.35).
Perceived physical effort
Perception of physical effort was significantly greater for ESD than for colonoscopy
(median Borg of 3.5 [3.0–4.8] vs 2 [.2–2.0]; P = 0.03).
Muscle activation
[Table 2 ] presents levels of muscle activation during the procedure. Levels of muscle activation
during ESD-1 and -2 were not significantly different from muscle activation measured
during colonoscopy. Only time spent >10% MVC for right extensor carpi radialis during
ESD-1 procedure was significantly higher (+15%) than during colonoscopy ([Fig. 3 ]). No difference was observed with ESD-2. The effect size was r = 0.90, which demonstrates
a huge effect.
Table 2 Muscle activation levels (%MVC) (median, min-max) during different endoscopic procedures.
Colonoscopy
ESD-1
ESD-2
P value
ESD-1, endoscopic submucosal dissection phase 1; ESD-2, endoscopic submucosal dissection
phase 2; %MVC, % of maximal voluntary contraction; NS, nonsignificant.
Right flexor carpi radialis
5.5
3.8
3.7
NS
Left flexor carpi radialis
8.9
9.0
10.3
NS
Right extensor carpi radialis
7.9
10.0
7.3
NS
Left extensor carpi radialis
11.8
19.6
18.4
NS
Fig. 3 Mean time spent >10% of maximum voluntary contractions for right and left extensor
and flexor carpi radialis during colonoscopy and ESD.
Value of time spent >10% MVC measured for the left extensor carpi radialis for all
the operating procedures was higher than 50% of time. These values were increased
during the ESD-1 and ESD-2 procedures vs colonoscopy (+ 67% and +59%, respectively)
without reaching statistical difference.
Posture and angulation during the endoscopic procedures
[Table 3 ] presents posture angles measures, and the statistical difference of percent time
spent in demanding posture in the different conditions.
Table 3 Percent time spent in demanding posture (median (min–max)) (as defined in Methods)
during different endoscopic procedures.
Colonoscopy
ESD-1
ESD-2
Comparison
P value
r (effect size)
ESD-1, endoscopic submucosal dissection phase 1; ESD-2, endoscopic submucosal dissection
phase 2; NS, nonsignificant.
Neck flexion
27.0
22.2
37.9
Co vs ESD-2
P = 0.046P = 0.028
0.81
Trunk flexion
0.09
8.1
5.8
NS
Right shoulder flexion
0.9
1.1
2.2
NS
Left shoulder flexion
0.04
0.1
0.2
NS
Right elbow
80.6
66.7
68.9
Colonoscopy vs ESD-1
P = 0.028
0.89
Left elbow
19.3
47.0
83
Colonoscopy vs ESD-2
P = 0.046
0.81
Right wrist (extension)
41.2
40.9
70.3
NS
Left wrist (extension)
43.1
54.2
50.6
Colonoscopy vs ESD-1
P = 0.028
0.89
Neck demanding flexion posture is significantly increased during ESD-2 vs the other
operating procedures (P = 0.046). For trunk and shoulder, time spent in demanding postures seemed relatively
low and was not different whatever the endoscopic procedures. For the other joints,
time spent in angles outside of the recommended range was from 19.3% to 83%.
Independently on the endoscopic procedure, right elbow (forearm position) angle was
lower than 60° and left elbow angle was higher than 100°. Time in demanding posture
for right elbow was significantly higher during colonoscopy than during ESD-1. For
left elbow, the constraint was higher during ESD-2 than during colonoscopy (P = 0.046).
For all three endoscopic procedures, left and right wrist position were in extension
demanding posture and lasted more than 40% of time. The most important percentage
of time spent in a demanding posture was for the right wrist during the ESD-2 procedure,
although the differences with the other endoscopic procedures were not significant.
In contrast, percentage of time spent in a demanding posture for the left wrist was
significantly higher during ESD-1 than during the other two endoscopic procedures.
Discussion
In this prospective pilot study, we showed that endoscopy is associated with biomechanical
stresses whatever the typology of technique for diagnostic or therapeutic means.
First, this study revealed that perception of physical effort using the Borg scale
was significantly greater for ESD compared with diagnostic colonoscopy even if the
level of exertion perceived during colonoscopy and ESD remain moderate for highly
trained physician. We further analyzed biomechanical stresses by measuring forearm
muscle activity and joint strain using IMU sensors. We have pointed out that except
for the right elbow, most of the biomechanical parameters show that the level of strain
is even greater in ESD than in colonoscopy. Indeed, the strain is considerable for
all the forearm muscles, with more than 25% of the time spent higher than 10% of the
MVC regardless of type of endoscopy. Right extensor carpi radialis muscle activity
was more important during the ESD-1 procedure than during classic colonoscopy, which
introduces risk of even more musculoskeletal injuries, such as epicondylitis.
The greatest strain seems to relate to the left extensor carpi radialis extensors.
Shergill et al. showed that the left wrist extensor muscle is intensively used during
endoscopy, potentially contributing to risk of MSDs [20 ]. Our study confirms these results because relative time spent higher than 10% MVC
is particularly important, more than 50% for colonoscopy and even more than 80% during
ESD-1 and -2 procedures. This could be particularly worrying for the left upper limb
because the position of the left forearm is in the risk zone (i.e. with elbow angle
higher than 100°) during ESD-1 and mostly during ESD-2 procedures, the biomechanical
strain increases at elbow level. The percent time of left wrist in demanding posture
is also important during the ESDs procedures, more than 50% of time, and significantly
higher during ESD-1 when compared with the two other procedures. To our knowledge,
this is the first study to specifically suggest that ESD procedures may be associated
with a higher risk of MSDs compared with standard endoscopy. Indeed, most published
studies rely on self-reported data, which introduces bias. For example, in 2022, Sturm
et al. analyzed responses from 151 gastroenterologists who completed an online questionnaire
[1 ]. In their study, the average number of endoscopic procedures was 86 per week, including
only four EMRs and fewer than one ESD per week. Due to the relatively low volume of
ESD and EMR procedures, their analysis did not identify any particular type of endoscopic
procedure as being more strongly associated with MSDs, despite the clear differences
in technical complexity between, for instance, a diagnostic EGD and an interventional
ERCP. In contrast, our study specifically focused on ESD, a technically demanding
procedure, and highlights its potentially greater biomechanical impact.
In view of these results, the left forearm, therefore, seems particularly exposed
to biomechanical risk factors during ESD and particularly during its second phase,
which is at least as biomechanically stressful as colonoscopy. Moreover, the position
of the neck is also more concerning during ESD-2 sub-task.
Awareness of this topic is important not only for short-term physical comfort, but
also can significantly impact the longevity of a provider’s career. Our results provide
evidence that muscular constraints may lead to conditions such as epicondylitis, epitrochleitis,
carpal tunnel syndrome, tendonitis, and tenosynovitis of the hand and fingers.
The question now is how to prevent this stress. The American Society of Gastroenterology
(AGA) has issued good practice recommendations. Working in an environment with a temperature
of 19°C, soft lighting, and adjustable screens and tables is advised [18 ]. However, the key factor leading to disorders is repetition and duration of tasks.
Further strategies to mitigate these risks need to be developed and implemented. First,
reducing the length of procedures could help to minimize risk of developing MSDs.
While shortening the duration of colonoscopies may not be feasible, because a minimum
time is required to ensure quality, progress has been made in reducing the time required
for ESD procedures. This reduction is largely due to practitioner experience and adoption
of traction strategies, leading to a significant decrease in procedure time in recent
years [14 ]
[16 ]
[29 ]. Teamwork with other physicians to split the longer procedure is also a way. Second,
the advent of robot-assisted surgeries has demonstrated potential in reducing postoperative
discomfort and muscle strain in the upper extremities, particularly on the dominant
side of the surgeon [30 ]
[31 ]. However, this advancement comes with its own set of challenges, including increased
static neck positioning and subjective back stiffness [32 ]. This highlights the importance of continuous development in ergonomics and physical
and occupational therapy interventions aimed at reducing MSD and enhancing surgeon
longevity. In addition, use of exoskeletons has also emerged as a promising option,
with studies indicating a reduction in fatigue without negatively impacting surgical
procedures [33 ]
[34 ]
[35 ]. In that regard, Shergill et al. (2021) have evaluated an endoscope assist device
(antigravity support arm) during a simulated task and shown a small reduction of 1.6%
MVC in the left extensor carpi radialis muscle activity.
Integration and development of such innovative tools and techniques could pave the
way for more sustainable practices in interventional endoscopy and surgery, ultimately
improving the well-being and efficiency of healthcare professionals. In light of our
results, it seems urgent to consider procedure time and frequency to mitigate risk
of developing MSDs of the left arm. This risk appears to be even higher for doctors
performing ESD. If robotization strategies are not feasible now, the question remains
how to better organize operational planning and increase the number of trained doctors
to help reduce incidence of MSDs.
We are aware that our study has some bias. First, the number of subjects evaluated
was low and the endoscopist population was homogeneous, which prevented a meaningful
analysis of the impact of endoscopic experience on biomechanical measurements. Second,
our study did not evaluate psychological stress generated by the procedure, but it
can be assumed that ESD increases this type of stress due to the higher risk of perforation
and bleeding complications compared with colonoscopy, which in turn could increase
risk of MSDs. Further studies must be conducted to identify, assess, and prevent psychosocial
risk factors related to endoscopy activity and notably ESD procedures. Finally, we
did not monitor all the muscles, we focused on the extensor and flexor carpi radialis
for EMG measurement. To minimize potential bias related to fatigue or cumulative workload,
recordings were conducted in the morning during the first colonoscopy or ESD procedure
of the day. The choice of these muscles is due to their continuous use during endoscopic
procedures: pressing valves and switches, rotating the wheels, holding the endoscope,
and applying forces/torques to the insertion tube.
Conclusions
To conclude, although ESD offers significant benefits in treating complex gastrointestinal
lesions, it also introduces substantial biomechanical strain, increasing long-term
risk of MSDs among ESD practitioners.
It is, therefore, essential to develop prevention programs, such as education and
ergonomics training, to benefit the health of endoscopists and quality and safety
of care. In addition, development and integration of new devices are essential to
combat these issues and enhance the well-being of healthcare professionals.
Bibliographical Record
