Page 70 - JSOM Summer 2018

FIGURE 7 The Advanced Brain Monitoring B-Alert X10 Cap used FIGURE 8 Average percent reduction in blood flow rate by day.
to measure electroencephalographic signals and heart rate variability.

one of four classifications of stress: (1) high vigilance, (2) low
vigilance, (3) relaxed wakefulness, and (4) sleepy. HRV (also
known as R-to-R variability) dynamically monitored the vari- what level of blood flow reduction meets the requirement of an
ance of sympathetic and parasympathetic nervous systems by adequate packing for this type of injury. Therefore, the findings
measuring the intervals between each R wave, thus indicating are provided as a measure of effectiveness rather than a measure
levels of stress and arousal. of quality. In addition, packing effectiveness improved as the
number of attempts of performing the procedure increased.
Self-reported perceived workload information was collected via
survey in addition to the response variable from physiological Surgical procedure was the strongest predictor of performance
measures and direct observations. After execution of each surgi- with “worst” performance being correlated with the femur
cal procedure, all participants completed the TLX (a variation of and pelvic procedures. The correlation between surgical team
the National Aeronautics and Space Administration TLX), tai- and procedure was also significant, demonstrating the differ-
lored to collect subjective workload data from surgical person- ences in skill levels for various participant team combinations.
nel, including subscales of mental demands, physical demands,
temporal demands, own performance, effort, and frustration. Physiologic Workload
Within the EEG-based cognitive workload, no statistically
The self-reported MASQ has become an accepted tool in ac- significant main effect was found for motion condition (p =
ademia for assessing symptoms relevant to motion sickness .002, no motion condition versus the mean of the two motion
and to sopite syndrome. Immediately after each procedure, conditions; Figure 9).
9
testers administered the MASQ to the surgical teams to as-
sess perceived symptoms such as disorientation, nausea, and FIGURE 9 Electroencephalography-determined workload.
lightheadedness.

Analysis Methods
Multivariate analysis of variance tests were used to evaluate
continuous outcome measures (i.e., EEG and HRV); nonpara-
metric methods (i.e., ordinal logistic regression, generalized
linear model with binomial distribution) were used for per-
formance and error severity scores. The metrics collected for
each of the hypotheses addresses aspects of the physiological,
performance, and subjective report results. For each individ-
ual and procedure (across both trials), EEG-based workload,
HRV-based arousal, and TLX-based perceived workload for
the no-motion case were compared with the mean of the two
motion cases in a repeated-measures design, conducted as a
multivariate analysis of variance.
To determine what effect EEG workload may have had on
surgical errors, a bivariate correlation was conducted between
Results
mean EEG-measured surgeon workload and mean errors per
Surgical Performance procedure (Figure 10). Except for one team that consistently
Although some teams performed better in motion than others, committed errors throughout the entire experiment (the first
motion did not significantly affect overall performance. This team to perform), no significant correlation was found be-
conclusion is based on an analysis of the quantitative results as tween workload and errors when this team’s data were re-
measured by the surgical SME. The total average blood flow moved from the data set.
reduction of all procedures was approximately 29% (Figure 8).
With the elimination of one group that was consistently below Within the HRV data, no significant effects were found due to
this average, the new average flow reduction becomes approxi- motion based on the high- to low-frequency ratio (r = −0.18,
mately 35%. In either case, because of the use of a noncoagulat- p = .767), which is an indicator of the balance between the
ing blood simulant, there are no data or standard that suggest sympathetic and parasympathetic nervous systems (Figure 11).

68 | JSOM Volume 18, Edition 2/Summer 2018