Page 54 - JSOM Fall 2022

outer surface and the adhesive coating on its inner surface and “elevated” perfusion pressures as presented in Table 3.
physically interact in a way that allows the wrap to be re­ Ease­of­use scores across all trials were similar for both tour­
peatedly adhered to and de­bonded from itself. During appli­ niquet types and are reported in Table 4.
cation, the ST is tightly wound around a limb two or more
times in an application process similar to the placement of an TABLE 4 Tourniquet Ease-of-Use Scores
Esmarch tourniquet. Next, the ST is tightened to the desired Scored Response to Statement,
level of limb compression by application of additional wraps “Tourniquet was easy to apply.” ST CAT
while under user­applied, pull tension. Once hands­free hem­ 4 ­ Strongly agree 17 (71%) 19 (79%)
orrhage control is achieved, any remaining unused length of 3 ­ Somewhat agree 6 (25%) 3 (13%)
the adhesive self­secured ST can be left intact or removed if de­ 2 ­ Neutral 1 (4%) 2 (8%)
sired. Post­application, the ST is adjusted by peeling back the 1 ­ Somewhat disagree 0 0
free end from the limb and unwinding a portion of the wrap 0 ­ Strongly disagree 0 0
before rewinding it around the limb under modified tension Pooled self­reported user scores from each tourniquet trial.
until the desired compression is achieved.
Tourniquet Completion Pressure
Statistical Analysis As shown in Table 3, the mean tourniquet pressure immedi­
The mean, standard deviation, and standard error for each per­ ately following completed application was found to be > 40%
formance metric as well as the calibration curve for pressure lower for the ST than for the CAT under both “normal” and
transducer readings were calculated using Microsoft Excel for “elevated” arterial perfusion pressures with statistical signif­
Microsoft 365 (Microsoft Corporation, Redmond, WA, USA). icance (p < .05). It should be noted that a positioning error
Additional statistical computations were performed using the which compromised the collection of pressure readings from
permuter R package (version 0.1.0, https://github.com/statlab/ both cuffs was discovered and corrected by an independent
permuter) and the R language and environment for statistical member of the laboratory support staff prior to the eighth
computing and graphics (version 4.0.5, www.r­project.org). For tourniquet trial conducted at “normal” perfusion pump pres­
each tourniquet performance metric, evidence for the null hy­ sures. As a result, tourniquet pressure measurements obtained
pothesis of no difference in means between the two devices was during the first four CAT trials and the first three ST trials
assessed against an alternative hypothesis of (non­directional) under “normal” perfusion parameters were excluded from the
inequality. A difference­in­means test statistic (two­sample data analysis described herein and presented in Table 3. The
t­statistic with pooled variance) was used to examine perfor­ pressure sensor issue did not impact or result in the exclusion
mance metrics between the two tourniquets, with a permuta­ of any other experimental data points collected during the first
tion test used to generate a null distribution by shuffling the seven tourniquet trials.
tourniquet identities 10,000 times while stratifying over tourni­
quet users. This resulted in an exact test, using the permutation
distributions to derive the reported p­values. Significance of all Discussion
p­values was assessed against the standard cutoff of p = .05. Tourniquet Completion Pressures
The most significant finding of this laboratory study was that
the 10.2 cm wide, minimally elastic, adhesive wrap–based ST
Results
can provide rapid and effective hemorrhage control perfor­
Time to Achieve Arterial Occlusion mance at tourniquet application completion pressures sub­
Success rates for achieving arterial occlusion were identical stantially lower than the 3.8­cm­wide, windlass­based CAT.
for the CAT and ST tourniquets at both the “normal” and While other studies of the ST have not been published to­date,
“elevated” arterial perfusion pressure regimens used. As sum­ previous studies of the CAT conducted on healthy, adult hu­
marized in Table 3, the mean time to achieve arterial occlusion man volunteers and which used similar pressure monitoring
was similar between the two tourniquet types for both sets of systems reported mid­thigh, tourniquet completion pressures
perfusion parameters. that corroborate the CAT completion pressures measured in
this cadaver study. 16,33 As a result, our research finding has
Time to Complete Tourniquet Application significant patient safety implications since a leading concern
The mean time required to complete tourniquet application surrounding the use of tourniquets is the risk of permanent
was nearly identical between the CAT and ST under “normal” nerve and soft tissue damage resulting from high, sustained

TABLE 3 Tourniquet Performance
ST CAT ST CAT
146 ± 29mmHg Perfusion 471 ± 3mmHg Perfusion
Tourniquet (“Normal” Pressure) (“Elevated” Pressure)
Success, n (%) 20 (100) 20 (100) 4 (100) 4 (100)
Time to achieve occlusion, s
Mean ± SE 25 ± 2 22 ± 2 p = .23 24 ± 7 24 ± 7 p = .94
Median, Minimum, Maximum 22, 15, 51 20, 13, 37 23, 10, 40 21, 12, 42
Time to complete application, s
Mean ± SE 27 ± 2 26 ± 2 p = .56 25 ± 7 25 ± 7 p = .91
Median, Minimum, Maximum 23, 16, 52 22, 16, 42 24, 11, 41 21, 13, 43
Pressure at completion, mmHg
Mean ± SE 110 ± 20 210 ± 30 p = .009 190 ± 50 340 ± 30 p = .03
Median, Minimum, Maximum 77, 31, 315 188, 44, 420 207, 51, 309 367, 257, 383
SE = standard error

52 | JSOM Volume 22, Edition 3 / Fall 2022