Keywords
cataract - training - residents - residency - stress - heart rate - steps
Stress level monitoring, while only sparingly used within surgical training, has previously
been used in other professions such as aviation, military, and athletics that also
require high-intensity training.[1]
[2]
[3]
[4]
[5]
[6]
[7] While a certain degree of stress can assist in task performance, psychological evidence
from professionals working within a variety of high-stress fields such as aviation
and military suggests that excessive levels can have a detrimental effect on performance.[8]
[9] Stress has been shown to potentially impair technical skills, vigilance, memory,
and other cognitive processes.[8]
[10] These effects are of great significance when considering how they may negatively
affect surgeon performance and thereby have an effect on patient safety.
Stress during surgical training can result from learning new techniques and mastering
the nuances of surgery, conducting complex procedures, and working under a time crunch.[11]
[12] While the degree of participation of a resident surgeon may not always be well-defined,
cataract surgery is one area where the resident must frequently have a significant
lead role as required by the Accreditation Council for Graduate Medical Education
(ACGME). The ACGME requires each resident to perform a minimum of 86 cataract surgeries
as the primary surgeon, with many programs performing a higher number.[13]
[14]
[15]
[16]
[17] When performing cataract surgery as the primary surgeon, the resident sits at the
main scope, while the attending sits at the side scope. This positioning ensures that
the resident performs the surgery with limited, if any, direct surgical input from
the attending. The attending provides guidance and usually provides surgical intervention
only when deemed necessary.
Although managing stress has a clear role in optimizing surgical outcomes, few studies
have closely evaluated intraoperative stress in relation to events in the surgery
itself.[1]
[2]
[3] Cataract surgery offers a unique and exciting opportunity to study intraoperative
stress for several reasons: these surgeries are typically of shorter duration, can
be divided into fixed steps of surgery, are more frequently recorded given the standard
use of microscopes, and have a high minimum number required to meet ACGME requirements.[17]
[18] As such, it is feasible to monitor events during the surgery and simultaneously
track their effects on the trainee stress levels over many cases. It is important
to make note that stress may not always have a detrimental effect on surgical performances.
Certain studies have shown that moderate levels of stress could lead to improvements
in performance and selective attention.[19]
[20] While the delineation between productive stress levels and detrimental stress levels
requires further study, stress certainly plays a role in the operating room, and a
better understanding of the resident experience could be useful in improving surgical
training. For instance, measuring stress levels over the course of training may provide
insight on the minimum number of cases necessary before a resident begins to feel
comfortable with each step of cataract surgery and may even help determine the delineation
between productive and detrimental levels of stress. Since surgical complications
during cataract surgery often occur without advance notice, monitoring stress levels
could provide cues for intraoperative intervention by the attending physician. Improving
the resident experience would not only facilitate better teaching but could also pave
the way for improved surgical outcomes in resident-performed cataract surgeries.
This study sought to monitor resident stress levels during surgery in relation to
different steps of the cataract surgery: incisions, continuous curvilinear capsulorrhexis,
hydrodissection, nucleus disassembly, quadrant removal, cortical cleanup, intraocular
lens (IOL) insertion, and closure. Stress levels were quantified using heart rate
(HR), which has been previously used as a reliable marker for stress.[1]
[2]
Design and Methods
The resident HR is measured through a Beets BLU chest-strapped Bluetooth device ([Fig. 1]). The Bluetooth device is linked to an iPod application specifically designed for
this study to simultaneously collect HR and audio during surgery. The iPod application
creates timestamps for the audio and HR recordings so that they can be properly aligned
with various steps during surgery. A SHURE MV88 iOS digital stereo condenser microphone
was placed below the microscope oculars for audio recording and connected through
a USB cable to an Apple iPod Touch placed on the arm of the machine, as shown in [Fig. 1]. The USB cable is kept firmly against the body of the microscope using adhesive
tape. Furthermore, video recording is obtained for each procedure. The audio and video
components are synchronized using the audiovisual cue of “incision” as the paracentesis
incision is made, which is synchronized with the video of the paracentesis incision.
The video is then divided down into the stages of cataracts as follows: incisions,
continuous curvilinear capsulorhexis, hydrodissection, nucleus disassembly, quadrant
removal, cortical cleanup, IOL insertion, and closure. Each step then has its corresponding
audio, video, and HR data available for analysis.
Fig. 1 Beets BLU chest-strapped Bluetooth devices (left) and the setup showing placement
of microphone and Ipod device on the microscope (right).
Emory ophthalmology residents who operated at Grady Memorial Hospital during the course
of the study from April 2017 to February 2018 were eligible to participate in the
study. All participants provided written informed consent after being informed of
the purpose of the study and the associated risks and benefits. No personal health
information was collected from the patients. The research protocol and informed consent
were approved by the Emory University Institutional Review Board.
Statistical Methods
HR data were collected continuously during the surgery. For each surgery, the minimum,
maximum, and mean HRs were calculated for each stage. For each of these measures,
a repeated measure analysis of the outcome was performed with a means model, calculating
the adjusted means and confidence intervals at each stage of surgery, over the course
of the entire surgery. A compound–symmetric variance–covariance form was assumed and
used to control for the correlation that inherently exists in the data since the HR
measures were taken on the same resident, both over time in a single surgery and across
time in multiple surgeries. All missing data were assumed to be missing at random,
and t-tests were used to compare differences over time. The mean HR data were considered
the primary model for analysis. The minimum and maximum HRs were also examined as
secondary outcomes, each in a separate mixed means model. For all three outcomes,
unadjusted means and confidence intervals were also calculated for comparison purposes.
These unadjusted measures did not control for the correlation across time or within
each resident. An α of 0.05 was used for all statistical tests, and SAS 9.4 (SAS Institute Inc.) was
used to complete the analyses.
Results
Thirteen residents, seven postgraduate year (PGY) 3 and six PGY-4, were enrolled in
the study. A total of 549 surgeries were recorded over a 10-month period starting
on April 4, 2017, and ending on February 28, 2018, during which all cataract surgeries
performed at Grady Memorial Hospital were recorded. Out of the 549 surgeries recorded,
427 (77.8% of the total surgeries) yielded viable data. Reasons for data dropout included
forgetting to press record on the audio or video recording devices, improperly placed
HR chest straps, physical movement such as standing up or walking, and malfunctioning
of iPod application software.
The mean, minimum, and maximum HR values were analyzed for each step of surgery described
earlier. There was a statistically significant difference between the mean HRs for
the stages (p < 0.0001). Quadrant removal had the highest adjusted mean HR of 90.3 (95% confidence
interval [CI]: 82.9, 97.7) beats per minute followed by nucleus disassembly at 89.1
(82, 97.7) beats per minute and incisions at 88.5 (80.7, 96.2) beats per minute ([Fig. 2]). The average maximum HR showed a similar trend with quadrant removal yielding the
highest value at 97.3 (89.5, 105.3) beats per minute followed by incisions at 97 (88.4,
105.5) beats per minute and nucleus disassembly at 96.8 (89.2–104.5) beats per minute
([Fig. 3]). The average minimum HR data still has quadrant removal highest at 82.1 (75.2,
89) beats per minute, whereas nucleus disassembly at 81.4 (74, 88.8) beats per minute
and hydrodissection at 81 (74.1, 88) followed ([Fig. 4]). [Table 1] demonstrates the comprehensive findings for all the steps.
Table 1
Adjusted and unadjusted means for mean, maximum, and minimum HRs along with the corresponding
CIs
|
Adjusted mean for mean HR (95% CI)
|
Adjusted mean for max HR (95% CI)
|
Adjusted mean for min HR (95% CI)
|
Unadjusted mean for mean HR (95% CI)
|
Unadjusted mean for max HR (95% CI)
|
Unadjusted mean for min HR (95% CI)
|
|
Incisions
|
88.5 (80.7, 96.2)
|
97 (88.4, 105.5)
|
79.3 (72.1, 86.6)
|
94.6 (93, 96.1)
|
102.6 (100.9, 104.4)
|
86.4 (85, 87.8)
|
|
CCC
|
87.6 (80.2, 95)
|
95.4 (87.5, 103.3)
|
79.9 (72.3, 87.4)
|
93.7 (92.2, 95.3)
|
101.1 (99.3, 102.8)
|
86.9 (85.4, 88.4)
|
|
Hydrodissection
|
86.9 (80.1, 93.8)
|
93.4 (85.8, 100.9)
|
81 (74.1, 88)
|
93 (91.6, 94.5)
|
99 (97.4, 100.6)
|
88.1 (86.6, 89.5)
|
|
Nucleus disassembly
|
89.1 (82, 97.7)
|
96.8 (89.2, 104.5)
|
81.4 (74, 88.8)
|
95.2 (93.8, 96.7)
|
102.5 (100.8, 104.1)
|
88.4 (86.9, 89.9)
|
|
Quadrant removal
|
90.3 (82.9, 97.7)
|
97.3 (89.5, 105.3)
|
82.1 (75.2, 89)
|
96.4 (94.9, 97.9)
|
103 (101.4, 104.6)
|
89.2 (87.7, 90.6)
|
|
Cortical cleanup
|
87.1 (80.2, 94)
|
94 (86.8, 101.2)
|
80.3 (73.3, 87.2)
|
93.2 (91.8, 94.6)
|
99.6 (98.1, 101.1)
|
87.3 (86, 88.6)
|
|
IOL insertion
|
86.9 (79.9, 93.8)
|
95 (87.5, 102.6)
|
79.2 (72.4, 85.9)
|
93 (91.6, 94.4)
|
100.7 (99.2, 102.2)
|
86.3 (85, 87.5)
|
|
Closure
|
86.2 (79.4, 93.1)
|
93.5 (86.1, 100.8)
|
79.2 (72.4, 86)
|
92.3 (91, 93.6)
|
99.1 (97.6, 100.5)
|
86.2 (85, 87.5)
|
Abbreviations: CCC, continuous curvilinear capsulorhexis; CI, confidence interval;
HR, heart rate; IOL, intraocular lens.
Fig. 2 Average values for mean heart rate during each step of cataract surgery. CCC, continuous
curvilinear capsulorhexis; IOL, intraocular lens.
Fig. 3 Average values for minimum heart rate during each step of cataract surgery. CCC,
continuous curvilinear capsulorhexis; IOL, intraocular lens.
Fig. 4 Average values for maximum heart rate during each step of cataract surgery. CCC,
continuous curvilinear capsulorhexis; IOL, intraocular lens.
Discussion
Measuring a resident's stress is difficult. Real-time questionnaires, which are often
used to assess stress perception, are not feasible during surgical procedures.[12]
[16] Cortisol monitoring, which is another method of measuring acute stress, is also
not practical given the need to monitor stress continuously throughout the procedures.[3] What is more practical, however, is measuring the resident's HR during a procedure.
There is little doubt that acute stress leads to physiological effects on the cardiovascular
system. Among the numerous cardiovascular parameters investigated, HR increases have
been consistently reproducible in response to acutely stressful events.[21]
[22]
[23] While the degree of HR reactivity may depend on several factors unique to each individual,
the presence of this reactivity has been uniform across several studies.[21]
[22]
[23] Since HR increases are related to increased stress, HR measurements are an excellent
surrogate in assessing the resident's stress.[1]
[2]
[3]
[24] While HRs have been used previously to study stress during surgeries, as done by
Becker et al and Arora et al for general, orthopaedic, and cardiac surgeries, previous
studies have not measured stress continuously in real time as a function of actual
events during the operation, as we have done here.[1]
[3]
The primary criticism for using HR as a measure for stress levels has been that physical
exertion or movement also can affect HR.[3] In this study, however, that effect is minimal considering that cataract surgery
is performed while in a stable, seated position. Any significant movement by the resident,
such as standing, was noted and accounted for by omission in the final analysis. A
limitation of the study was that while 13 PGY-3 and PGY-4 residents took part in the
study, the six PGY-4 residents, who operated most frequently, make up a large portion
of the dataset. Statistically, this was controlled for with mixed means models to
ensure that the data were not inappropriately skewed by the large proportion of the
total surgeries performed by certain residents. Furthermore, we did not note consumption
of stimulants such as caffeine and cardiovascular modulators such as β-blockers, which
could potentially have had an effect on intraoperative HR. While it is possible that
these agents may have affected all the steps of surgery evenly, it is also quite possible
that depending on the particular agent and time of administration, they may have affected
the different steps unevenly. As such, data collection in future studies should record
use of stimulants and cardiovascular modulators. Another limitation is that the study
did not have a control group such as a group of experienced attendings operating with
minimal stress. Having such a group in future studies would provide context and help
more thoroughly interpret resident stress levels.
Based on our results, quadrant removal, nucleus disassembly, and incisions represent,
in that order, the three most stressful steps of cataract surgery for the resident
surgeon. Our results were not surprising as these particular steps are often when
surgical complications such as posterior capsular rupture may occur (quadrant removal
and nucleus disassembly) or, in the case of incisions, when the resident may be experiencing
the most nervousness before settling into the surgery. With this knowledge, it follows
that these particular steps may require increased attention as the resident surgeon
prepares for the challenges of cataract surgery. This preparation may come in the
form of supplementary effort in a controlled environment such as the wet laboratory
or additional attending guidance during the cataract surgery itself.
Additionally, this study paves the way for further research using intraoperative HR
monitoring to improve resident training and surgical outcomes. While the cataract
surgeries performed during this period did not result in enough complications to draw
conclusions regarding the effect of HR fluctuations on complication rate, a more longitudinal
examination may shed light on this relationship. Multiple studies have commented on
the resident surgeon cataract learning curve using measures such as phacoemulsification
times, complication rates, postoperative visual acuity, and completion rate.[13]
[25] The method of intraoperative HR monitoring we describe here could provide another
objective way to track the resident learning curve. Such a study could help us better
understand the number of cases required for residents to feel more comfortable and
confident performing cataract surgery. In turn, this knowledge could contribute to
the important discussion of determining the ideal number of resident-performed cataract
surgeries during training and ultimately establishing the minimum number of cases
to demonstrate competence in phacoemulsification.
In addition to the resident learning curve, intraoperative HR monitoring could also
be a useful aid in better understanding the attending learning curve. Puri et al found
significant differences between a novice and an experienced attending's complication
rates when they did a retrospective analysis of resident-performed cataract surgeries.[26] Since the complication rates were higher with the novice attending, they concluded
that surgical programs should aim to reinforce areas for improvement to create top
surgical educators.[26] Intraoperative stress monitoring could provide an objective manner in which to identify
areas where more junior attendings may feel less confident. Perhaps, these areas are
not true weaknesses in the attending's technical skills but rather a reflection of
the resident's anxiety affecting the attending surgeon's confidence in surgical decision-making.
Simultaneous intraoperative HR monitoring of both the attending and resident during
surgery could help elucidate the relationship between attending and resident stress
levels.
Conclusion
Overall, our data suggest that certain steps of cataract surgery are significantly
more stressful for resident surgeons. Surgical training programs should devote additional
time and resources in preparing residents for these steps. Whether it requires additional
practice in the wet laboratory or more attending guidance in the operating room, an
emphasis on these steps of cataract surgery will improve the resident experience,
surgical training, and ultimately patient outcomes. Looking toward the future, the
scope of intraoperative monitoring can be widely expanded to shed further light on
the longitudinal stress levels of residents throughout the training program as well
as the resident and attending experience and interaction during resident-performed
cataract surgery.
