Page 91 - JSOM Winter 2022

of variance (ANOVA) with post hoc localizing pairwise com- FIGURE 2 Time to application.
parisons and correction for multiple comparisons. Binomial
outcomes (survival, hemostasis, rebleed) were assessed using
2
χ tests. Application attempts, which could not be assumed to be
from a Gaussian distribution, were assessed with the Kruskal-
Wallis test. Repeated-measures (HR, RR) between-group dif fer-
ences over time were assessed using repeated-measures ANOVA.
Statistical tests were performed using Prism 8 (GraphPad Soft-
ware, https://www.graphpad.com/scientific-software/prism/);
12
sample size calculations were performed using G*Power. All
differences were considered statistically significant at the p < (A) (B)
.05 threshold.
Time to application for iTClamp with XSTAT (IT+XS), iTClamp with
Results Combat Gauze (IT+CG), iTClamp alone (IT), and Combat Gauze
with pressure dressing (CG) including 3 min of pressure held for CG
Swine in each treatment group were statistically similar in (A) and excluding 3 min of pressure held for CG (B) Symbols reflect
body weight, and at baseline HR and RR; mean and standard individual data points; horizontal midpoint line represents the mean
deviations for each measurement are shown in Table 2. and error bars indicate standard deviation. (Horizontal lines represent
pairwise comparisons, p < .0001).
TABLE 2 Baseline Values the iTClamp does not cover a wound but is intended to seal
it, any bleeding from the wound was considered a failure and
Baseline Values IT+XS IT+CG IT CG was a binary outcome.
Weight (kg) 40.62 43.91 39.33 41.03
(4.34) (5.44) (5.52) (3.07) Failure to achieve initial hemostasis as defined above triggered
Heart rate 80.5 96.13 93.44 93.57 the removal of the iTClamp or pressure dressing and reapplica-
(bpm) (12.31) (20.77) (17.59) (9.43) tion. This process was repeated until hemostasis was achieved.
Respiratory rate 29.13 24.13 29.56 26.00 CG required one application attempt for each animal, while IT
(breaths/min) (6.18) (5.11) (10.94) (7.14) required a second attempt for one animal. IT+CG required a
Weights were recorded prior to instrumentation on the day of surgery. second attempt for two animals and IT+XS required a second
Heart rate and respiratory rate baselines were recorded after anesthe- attempt for two animals and a third attempt for one animal.
sia and instrumentation and prior to injury. Values are mean (standard
deviation).
Rebleed
Any evidence of external blood loss after the initial hemostatic
Application Time period was considered a rebleed episode. Occurrence of re-
2
After a 45-second free bleed, each treatment strategy was ap- bleed was statistically similar across groups, χ (df = 3) = 0.64,
plied. Application time was measured from the time the se- p = .88. IT+XS had two rebleeds, while other groups had one
lected hemostatic agent was picked up until the iTClamp was rebleed each (Table 2).
locked or the pressure dressing was fully wrapped. Mean time
to apply was significantly slower for CG [mean (SD) = 315 sec Blood Loss
(18.9)], compared to IT+CG, [mean (SD) = 70.3 sec (22.0); Percent blood loss trended the highest for CG treated swine
p < .0001], IT+XS (mean (SD) = 26 sec (8); p < .0001], and [mean (SD) = 18.4% (8)], followed by IT+XS [mean (SD) =
IT [mean (SD) = 12 sec (3); p < .0001] (Figure 2A). IT+CG 16.1% (14.6)] and IT [mean (SD) = 13.4% (10.2)] (Figure 3).
was significantly slower than IT+XS (p < .0001) and IT IT+CG had the lowest average blood loss [mean (SD) = 9.5%
(p < .0001). IT+XS and IT were not significantly different (p = (4.5)], which trended lower than CG (p = .09).
.51). Even if the 3-minute (180 sec) pressure application time
was excluded from the CG treatment time, CG was still sig-
nificantly slower [mean (SD) = 135.1 sec (18.9)] than IT or
IT+XS and IT+CG (each p < .0001) (Figure 2B). These findings
support our hypothesis that addition of a mechanical closure FIGURE 3 Total blood loss.
mechanism speeds the time of application. Symbols reflect individual data
points, horizontal midpoint
Hemostasis and Application Attempts line represents the mean, and
Ultimately, hemostasis was obtained in all the subjects. How- error bars indicate standard
deviation.
ever, there were differences in the rate of initial hemostasis
achieved on the first attempt. We defined initial hemostasis
as the cessation of blood loss outside of the wound during
the 30-second interval immediately following completion of
hemostatic application. CG was able to reliably achieve this Heart Rate
in 100% of cases. Conversely, all the iTClamp treatments had Subjects’ vital signs were monitored as a measure of physi-
some failed attempts, with success on the first attempt 88% ologic response to hemorrhage to assist in determination of
for IT applications, 75% for IT+CG applications, and 63% the significance of blood loss. Figure 4A shows that treatment
for IT+XS applications. Chi square revealed no statistically groups were not significantly different in the average of three
significant differences between groups in initial hemostasis, χ baseline heart rate (HR) measures (p = .19), though IT+XS
2
(degrees of freedom [df] = 3) = 3.77, p = .29 (Table 2). Because trended lowest. Treatment groups did not significantly differ
iTClamp-Mediated Wound Closure | 89