Keywords
exoscope - plastic surgery - ergonomics - postural exposure - inertial measurement
unit
Musculoskeletal disorder (MSD) prevalence is high among surgeons,[1]
[2] resulting in lost productivity and decreased quality of surgical care.[3] In addition, the frequency of MSD within the surgical profession may also deter
medical students from pursuing surgical specialties.[4] Studies have shown that surgeons' necks overall are the most at-risk body area intraoperatively.[5]
[6] Specifically, the use of surgical equipment such as loupes,[7]
[8] headlights,[9] and operating microscopes[10] may be associated with risk factors (e.g., higher neck postural exposure with the
use of loupes, awkward neck postural exposure from microscope use, and cervical spine
loading when using loupes and headlights).
Plastic surgeons and residents may be at high risk of musculoskeletal problems because
of chronic work-related neck pain and discomfort.[11]
[12]
[13]
[14] This may be true especially during autologous breast reconstruction using deep inferior
epigastric perforator (DIEP) flap surgery, involving harvesting the fat, skin, and
blood vessels from the abdomen and reattaching it using microsurgery to the breast
area after it is dissected to reconstruct the breast. Each DIEP flap surgery can be
thought of in three main phases: (1) the abdominal flap harvest, (2) chest vessel
dissection, and (3) anastomosis in which the harvested flap is reattached to the chest
using microsurgery. The DIEP flap presents several ergonomic risks to surgeons, such
as its long average duration, the use of loupes and headlights, and the use of a surgical
microscope. Attempts to address the ergonomic issues associated with operating microscopes
have included employing different imaging methods such as stereoscopic video displays
during microsurgery. Video displays, when compared with traditional microscopes, have
been shown to reduce neck angles but increased the shoulder flexion.[15]
Adopting exoscopes in place of microscopes may significantly improve surgeon postures
and ergonomics, in addition to offering high-resolution three-dimensional visualization
and increased field-of-view adjustability.[16]
[17] Studies have shown additional ergonomic advantages with exoscope use.[18]
[19]
[20]
[21] Potential advantages with intraoperative exoscope use include decreased neck flexion
and fatigue, which may hinder surgeon performance.[22] However, the utility of an exoscope as a potential ergonomic intervention to reduce
the risk of MSDs among plastic surgeons (e.g., DIEP flap cases) is not yet well known.
The current study evaluated the ergonomics (e.g., surgeons' intraoperative postural
exposure and workload) of DIEP flap surgery using a 3D exoscope compared with those
using loupes, headlights, and microscopes for the various phases of the case.
Methods
Participants
Two surgeons who regularly perform DIEP flap autologous breast reconstruction surgeries
at a quaternary care academic medical center (United States) participated during their
regularly scheduled dual-surgeon DIEP flap cases. One surgeon's role was to work on
the flap harvest for most of the surgeries (surgeon at abdomen), while the other surgeon's
role was to work on the chest dissection and anastomosis for most of the surgeries
(surgeon at chest). The surgeons were novices with exoscopes but had performed cadaveric
practice sessions prior to the study. This study was approved by the medical center's
institutional review board.
Apparatus
The DIEP flap surgery was performed as either a baseline or exoscope (ORBEYE, Olympus
Corporation, Tokyo, Japan) surgical case. In baseline surgical cases, the surgeons
used loupes with a headlight during abdominal flap harvest and chest dissection ([Fig. 1A]) and used a microscope (MM51, Mitaka USA, Denver, CO) for the microsurgical anastomosis
([Fig. 1B]). In the exoscope surgical case, the surgeons used the exoscope ([Fig. 1C]) (ORBEYE, Olympus Corporation, Tokyo, Japan) during the main phases of the surgery:
abdominal flap harvest, chest dissection, and anastomosis. The exoscope features a
camera mounted on a semirobotic arm that is controlled by the surgeon either by directly
manipulating controls on the device or remotely by using a foot pedal. Images captured
on the exoscope were displayed on a 4K-3D monitor, and the surgeons and their team
wore 3D glasses to view the surgical field projection ([Fig. 1C]). The exoscope also had optical and digital zoom capabilities of up to 26× for visualization
of small anatomical structures in detail, which was crucial for the microsurgical
anastomosis. For this study, two exoscope units and four monitors were used simultaneously;
one exoscope with two monitors (one each on the left and right side of the patient)
was used at the chest for chest dissection and anastomosis and the second exoscope
(ORBEYE, Olympus Corporation, Tokyo, Japan) also with two monitors was used at the
abdomen for flap harvest. Each exoscope was used by an attending surgeon and a resident.
Fig. 1 Baseline surgical procedure using (A) loupes and headlights and (B) microscope and (C) exoscope. (Reproduced with permission of Mayo Foundation for Medical Education and
Research. All rights reserved.).
Data Collection Instrumentation
The data collection instrumentation consisted of inertial measurement units (IMUs;
objective assessment) and an electronic survey (subjective assessment). IMU sensors
(Opal, APDM, Inc., Portland, OR) were used to collect surgeons' postural exposure.
Three IMUs were affixed to the surgeon's scrub top at their upper back and right/left
shoulders to capture their intraoperative back and shoulders angles. The fourth IMU
sensor was placed in an elastic headband worn by the surgeon with the sensor positioned
at the back of their head to capture their intraoperative neck angle.[7] IMUs recorded accelerometer, gyroscope, magnetometer, and time (in UTC) data at
a sampling rate of 128 Hz.
After each surgical procedure, the participants received an electronic survey that
evaluated their intraoperative workload, including mental demand, physical demand,
temporal demand, performance, effort, and frustration based on NASA-TLX,[23] distractions based on SURG-TLX,[24] degree of difficulty (difficulty) based on Global Operative Assessment of Laparoscopic
Skills (GOALS),[25] and fatigue based on Borg CR-10 Scale[26] (all ranged from “0 = min” to “20 = max”). The surgeons were asked to complete the
survey within 24 hours after the end of each surgical procedure.
Experimental Procedure
Prior to the start of each surgical case, study personnel were present in the operating
room (OR) to place the IMUs on the participants and calibrate them following a method
developed and explained in previous studies.[8]
[27] The start and end times of each phase of the surgery as well as the surgeon who
performed the phase were recorded for both sides of the patient. The surgeons then
completed the workload survey after each procedure.
Data Processing
Joint angles were calculated from the IMU data using a custom program in MATLAB (MathWorks,
Natick, MA).[27] Joint angles were defined as deviation of a body segment from the vertical axis.
The average joint angle across each surgical phase was calculated for each body segment.
Additionally, the joint angles for each body part during the surgery was divided to
four risk areas[7] using a modified Rapid Upper Limb Assessment (RULA)[28] where larger joint angle deviations from neutral were allocated to higher risk scores.[8] Postural risk scores on a scale of 1 (lowest) to 4 (highest) were then calculated
for each body segment and each phase of the surgery based on a methodology described
by Norasi et al.[8]
Statistical Analysis
Statistical analysis was performed in SPSS (IBM, Armonk, NY). Paired t-tests were used to compare the joint angles of the participants between the DIEPs
performed on the left and right sides of the patient. Risk scores associated with
each body segment were compared between the baseline and the exoscope using a t-test after checking data normality. To determine any changes in the frequency of
participants adjusting the exoscope using the camera controls, the percentage of each
phase that the right arm elevation angle was greater than 45 degrees was calculated
and any trends over time was identified using a least-squares linear regression. The
normality of the skin-to-skin duration of surgical procedures was also checked and
t-test was used to evaluate the difference in this dependent variable between baseline
and exoscope surgical procedures. Furthermore, operating time, flap harvest, chest
dissection, and anastomosis time were compared between baseline and exoscope. The
Kruskal–Wallis test was used for chest dissection completion time because of violation
of the normality assumption.
Additionally, the Kruskal–Wallis test was employed to investigate the effects of surgical
procedure (baseline and exoscope) on the nine discrete subscales of the survey for
each surgeon individually. All tests had a significance level of 0.05.
Results
Objective Evaluation Using IMUs (Postural Risk Scores)
A total of 23 bilateral DIEP flap surgeries (10 baseline and 13 exoscope) were included
in this study.
A paired t-test comparing joint angles observed during the left and right DIEPs showed that
there were no significant differences in postures for any of the phases (abdominal
flap harvest, chest dissection, anastomosis); therefore, we did not distinguish laterality
in any further analyses.
Joint angles for each body segment, neck, torso, and right and left upper arms, during
each phase of the DIEP were compared between the baseline and the exoscope cases ([Table 1]).
Table 1
The comparison of neck, torso, and right and left upper arm average deviation angles
between baseline and exoscope cases for three phases of the surgery (abdominal flap
harvest, chest dissection, and anastomosis)
|
Baseline, mean (SD) [degrees]
|
Exoscope, mean (SD) [degrees]
|
p-Value
|
|
Abdominal flap harvest
|
|
Neck
|
30.3 (6.8)
|
14.4 (2.4)
|
<0.001
|
|
Torso
|
17.9 (6.7)
|
12.5 (3.9)
|
0.008
|
|
Right upper arm
|
15.0 (4.9)
|
22.0 (11.3)
|
0.009
|
|
Left upper arm
|
16.7 (3.8)
|
21.3 (15.5)
|
NSS
|
|
Chest dissection
|
|
Neck
|
28.0 (8.9)
|
15.5 (4.2)
|
<0.001
|
|
Torso
|
19.8 (6.7)
|
14.3 (5.0)
|
0.006
|
|
Right upper arm
|
21.8 (8.5)
|
25.8 (10.7)
|
NSS
|
|
Left upper arm
|
20.8 (5.4)
|
25.4 (9.7)
|
NSS
|
|
Anastomosis
|
|
Neck
|
12.5 (5.1)
|
17.5 (5.3)
|
0.009
|
|
Torso
|
13.0 (4.3)
|
13.4 (3.5)
|
NSS
|
|
Right upper arm
|
23.1 (9.9)
|
31.9 (9.5)
|
0.013
|
|
Left upper arm
|
19.0 (5.9)
|
29.9 (5.7)
|
<0.001
|
Abbreviations: NSS, not statistically significant; SD, standard deviation.
There were significant decreases in the surgeons' neck angle and torso angle for exoscope
cases compared with the baseline during the abdominal flap harvest and the chest dissection
phases; however, there were modest but statistically significant increases in the
neck angles for exoscope cases compared with baseline during the anastomosis (all
p-values < 0.05). Additionally, the exoscope cases were associated with increased upper
arm elevation angles during the abdominal flap harvest (only right upper arm) and
the anastomosis (both upper arms) (all p-values < 0.05). Although the right arm elevation angles during abdominal flap harvest
were higher in the exoscope, the percentage of time spent in the most extreme right
arm elevation angles (>45 degrees) gradually decreased as more cases were performed.
Although not statistically significant, there was a negative slope in the right arm
angles as more cases were performed (slope = −1.4 degrees/case, p = 0.054). Furthermore, the right arm angles during the anastomosis using the exoscope
significantly decreased as more cases were performed. A linear regression showed a
statistically significant negative slope with the more cases being performed (slope=
−4.1 degrees/case, p < 0.001), suggesting that the surgeons reached up less often to position or adjust
the exoscope camera over time.
Postural risk scores for each body segment during each section of the DIEP were compared
between the baseline and the exoscope DIEP cases as well. The percentage of time that
the participants spent in each risk score by body segment and phase of the DIEP is
illustrated in [Fig. 2]. Average risk scores by body segment and phase are illustrated in [Fig. 3] and [Fig. 4].
Fig. 2 Percentage of time spent in each risk category for each body segment compared between
the baseline and exoscope cases during the (A) abdominal flap harvest, (B) chest dissection, and (C) anastomosis.
Fig. 3 Mean (SD) risk scores for the baseline and exoscope during the (A) abdominal flap harvest, (B) chest dissection, and (C) anastomosis.
Fig. 4 Mean (±SD) of the nine subscales of the survey for the (A) surgeon at abdomen and (B) surgeon at chest. Asterisks represent statistical differences (p < 0.05).
The results of the statistical analyses for the postural risk scores ([Table 2]) were consistent with the results of average angles ([Table 1]). Additionally, the percentage of time spent in a high-risk neck angle (postural
risk score 3 or 4) decreased from 82% to 21% during the abdominal flap harvest and
from 63% to 22% during the chest dissection but increased from 21.7% to 32.2% during
the anastomosis. Torso angle also showed a similar decrease in high-risk postures
(postural risk score of 3 or 4) when using the exoscope, from 35% to 13% during the
abdominal flap harvest and from 42% to 23% during the chest dissection phase ([Fig. 2]).
Table 2
The comparison of neck, torso, and right and left upper arm risk scores between baseline
and exoscope DIEP cases for three phases of the surgery (abdominal flap harvest, chest
dissection, and anastomosis) with a minimum of 1 and a maximum of 4
|
Baseline, mean (SD)
|
Exoscope, mean (SD)
|
p-Value
|
|
Abdominal flap harvest
|
|
Neck
|
2.78 (0.20)
|
1.80 (0.20)
|
<0.001
|
|
Torso
|
2.06 (0.35)
|
1.67 (0.31)
|
0.001
|
|
Right upper arm
|
1.24 (0.19)
|
1.55 (0.45)
|
0.004
|
|
Left upper arm
|
1.32 (0.15)
|
1.49 (0.55)
|
NSS
|
|
Chest dissection
|
|
Neck
|
2.56 (0.38)
|
1.95 (0.34)
|
<0.001
|
|
Torso
|
2.22 (0.30)
|
1.83 (0.40)
|
0.002
|
|
Right upper arm
|
1.52 (0.34)
|
1.70 (0.44)
|
NSS
|
|
Left upper arm
|
1.51 (0.23)
|
1.72 (0.42)
|
NSS
|
|
Anastomosis
|
|
Neck
|
1.69 (0.44)
|
2.12 (0.40)
|
0.005
|
|
Torso
|
1.77 (0.38)
|
1.75 (0.25)
|
NSS
|
|
Right upper arm
|
1.57 (0.41)
|
2.01 (0.33)
|
0.002
|
|
Left upper arm
|
1.39 (0.30)
|
1.91 (0.21)
|
<0.001
|
Abbreviations: NSS, not statistically significant; SD, standard deviation.
Also, comparing the skin-to-skin duration of baseline and exoscope surgical procedures
using a t-test showed that the skin-to-skin duration of exoscope surgical procedures was significantly
longer than baseline surgical procedures (p = 0.0273). The mean (standard deviation (SD)) of skin-to-skin surgical duration was
552.1 (70.3) minutes for the baseline cases, compared with 635.5 (92.5) minutes for
the exoscope cases. However, the results of the statistical analyses did not show
a significant difference between baseline and exoscope for the three surgical phases.
The mean (SD) duration for abdominal flap harvest were 143.25 (42.34) and 140.08 (53.70)
minutes for baseline and exoscope, respectively. The values for chest dissection were
90.80 (37.39) and 70.95 (43.08) minutes for baseline and exoscope, respectively. Finally,
the mean (SD) anastomosis duration were 78.41 (17.43) and 79.44 (18.40) minutes for
baseline and exoscope, respectively. The longer duration of the exoscope cases may
have been due to the additional time needed for the exoscope equipment setup during
surgery. It should also be noted that since two phases of the surgery are often being
performed simultaneously, a shorter or longer duration of any individual surgical
phase may not be reflected in the skin-to-skin duration.
Subjective Evaluation (Surveys)
[Fig. 4] illustrates the mean (±SD) of the nine questions in the survey by surgeon. The Kruskal–Wallis
test revealed that for the surgeon harvesting the abdominal flap, the mental demand
(p = 0.0346), temporal demand (p = 0.0321), and effort (p = 0.0414) were all higher during exoscope cases compared with baseline cases. The
surgeon's self-perceived performance for exoscope cases was lower than for baseline
cases (p = 0.0499). Furthermore, for the surgeon performing the chest dissection and anastomosis,
the physical demand (p = 0.0255) and fatigue (p = 0.0188) were lower during exoscope cases compared with baseline cases. There was
no statistically significant difference in other workload subscales between exoscope
and baseline surgical procedures for the two surgeons.
Discussion
The exoscope was effective in improving postures and reducing the ergonomic risk of
the neck and torso compared with baseline (i.e., loupes, headlights) during the abdominal
flap harvest and the chest dissection phases. Exoscope use was associated with significantly
less neck flexion and torso flexion angles during abdominal flap harvest and chest
dissection in comparison to the use of loupes and headlights. During the abdominal
flap harvest phase, the neck and torso risk scores in the exoscope cases were significantly
reduced compared with the baseline. Using a threshold average RULA value of 2.5 or
higher to determine high-risk postures,[29] the neck angles during the baseline abdominal flap harvest and chest dissection
can be categorized as posing a high ergonomic risk to participants. Use of the exoscope
improved postures and reduced the ergonomic risks during these procedures such that
the high-risk neck angles observed during the baseline abdominal flap harvest and
chest dissection phases were no longer observed during the exoscope cases. The torso
risk score associated with the baseline abdominal flap harvest and chest dissection
cases using loupes and headlights were not considered to be high-risk using the RULA;
however, they were further reduced in the exoscope cases.
While the exoscope was associated with an increase in joint angles of the right upper
arm during the abdominal flap harvest, as well as the neck and left and right upper
arms during the anastomosis, these increases did not present a high degree of ergonomic
risk to surgeons. Elevated arm angles observed during abdominal flap harvest could
be due to the overhead exoscope visual field, lighting, and zooming adjustments made
by the surgeon using their right hand. It may also reflect the ability of the exoscope
to allow the surgeon to sit during surgery, something that the traditional equipment
of loupes, headlights, and microscope did not afford. On the other hand, the ability
to sit while either dissecting the abdominal flap and chest vessels or performing
the anastomosis may be beneficial to reducing lower body fatigue and discomfort associated
with performing surgery.
Inclusion of a survey in the study allowed exploration of surgeons' perceived cognitive
and physical workload. The results for the “surgeon at abdomen,” who mostly performed
the abdominal flap harvest task, showed that her mental demand, temporal demand, and
effort were all higher during the exoscope cases, and her performance was lower during
exoscope cases compared with the baseline. The results for the “surgeon at chest,”
who mostly performed the chest dissection and anastomosis tasks, revealed that her
physical demand and fatigue were significantly lower during exoscope cases compared
with baseline. These findings indicate that there may be other workload drivers in
addition to the difference in postural exposure (e.g., neck angle, torso angle) between
exoscope and baseline cases that should be considered. Such factors may include (1)
the effects of learning curve for both the surgeon[30]
[31] and surgical team, (2) smaller focal area during the use of exoscope compared with
loupes, (3) longer skin-to-skin duration of exoscope cases (83 minutes longer on average)
compared with baseline (although the phase-by-phase comparison did not show any significant
difference in duration between baseline and exoscope), (4) managing multiple monitors
due to setup and integration of two exoscope units within the OR, and (5) the need
to adjust the exoscope cameras during the surgery. These potential underlying causes
may have negatively affected workload during exoscope surgical procedures.
While this study focused on the ergonomics of the two primary surgeons, experimenters
and OR staff noted additional findings based on researcher's observation. Setting
up two ORBEYE exoscope units presented a challenge due to the amount of equipment
involved. Additionally, surgical staff and residents expressed space concerns due
to the two exoscope units and four screens present, even though the 4K-3D screens
provide the residents and medical students with a high-quality view of the surgical
field without being obstructed by the surgical team in real time.
There are limitations to this study that should be considered when generalizing these
findings. Data were collected from two primary surgeons at one hospital during a single,
complex surgery (DIEP flap). Similar studies with larger samples of surgeons with
additional surgical expertise will help verify these findings. While this study focused
solely on the primary attending surgeons, it should be acknowledged that the exoscope
may have also affected the postural exposure and workload of other surgical team members,
such as the residents and surgical assistant. These factors should be noted in comparing
the findings for the exoscope and baseline surgical procedures. It should be noted
that due to the long durations of procedures, our evaluation using IMU sensors was
confined to joint angles as the deviation of the body segment from the vertical axis
running inferosuperior to the body. However, the IMU sensors allowed us to capture
surgeons' postural exposure in the restricted and challenging environment of the OR
during long DIEP flap surgeries. Lastly, there could be learning taking place while
using the exoscope. While this was not the focus of this study, future work should
focus on exploring the learning curve associated with using the exoscope.
Conclusion
The findings of this study showed that the intraoperative exoscope significantly reduces
the postural exposure and the corresponding ergonomic risk at the neck and torso during
the abdominal flap harvest and chest dissection when compared with baseline DIEP flap
cases where loupes and headlights were used. Conversely, the right arm during the
abdominal flap and the neck, right arm, and left arm during the anastomosis were associated
with higher joint angles when using the exoscope. Subjective measures showed that
the “surgeon at abdomen” experienced greater mental demand, temporal demand, and effort
as well as lower performance when using the exoscope. The “surgeon at chest” experienced
lower physical demand and lower fatigue when using the exoscope. The exoscope improved
postures at the neck and torso during the abdominal flap harvest and chest dissection
phases using the objective postural data from the IMUs; however, the subjective data
showed mixed results and may reflect the participants' personal preferences and characteristics.
