Page 38 - Journal of Special Operations Medicine - Summer 2016

Time to Stop Bleeding Figure 5 Results of pressure applied by model of tourniquet.
Results of the pairwise comparison of mean time to stop
bleeding by model of tourniquet indicated no significant
difference among two pneumatic models (TPT2 and
EMT) from the C-A-T, the control model (p > .0528
for all three), whereas all three means were significantly
different than that of the TPT3, which had the slowest
mean time at 50 seconds (p < .0065 for all three) (Figure
4). Pairwise difference in means ranged from 1 second
for the TPT2 and EMT pairing (p = .9953) to 20 seconds
for the TPT3 and C-A-T pairing (p < .0001). Similarly,
ANOVA results were two-tiered, with the TPT3 alone in
the slow tier and the three other models in the fast tier.
Figure 4 Results of time to stop bleeding by model of Results of pressure applied by model of tourniquet were that the EMT
tourniquet. had the lowest mean pressure. In pairwise comparison, there were
significant differences among pneumatic models for the TPT3-EMT
and TPT2-EMT pairings, and the amount of pressure applied with
the C-A-T was significantly different than all three pneumatic mod-
els. Analysis of variance results were three-tiered, with the C-A-T and
TPT3 in the high tier, the TPT3 and TPT2 in the middle tier, and the
EMT in the low tier. C-A-T, Combat Application Tourniquet; EMT,
Emergency and Military Tourniquet; TPT2, 2-in. Tactical Pneumatic
Tourniquet; TPT3, 3-in. Tactical Pneumatic Tourniquet.

significant in their difference (adjusted p ≥ .05). If groups
are statistically significant in their difference they will be
in different tiers, but the same group, which cannot be
different from itself, may be in more than one tier—for
example, TPT3 is in more than one tier. A model, such
as TPT3, may be in two tiers if its mean is between those
The TPT3 had the slowest mean time to stop bleeding. In pairwise of two other groups, such as C-A-T and TPT2, and both
comparison, there was no significant difference among mean times for
the TPT2, EMT, and C-A-T, whereas all these three means were signif- differences, e.g., TPT3–CAT and TPT3–TPT2, are not
icantly different than that of the TPT3. Similarly, analysis of variance statistically significant (Table 1).
indicated that the TPT3 was alone in the slow tier and the three other
models were in the fast tier. The vertical box plots depict the 25th
percentile as the box bottom, 75th percentile as the box top, 5th per- Blood Loss
centile as the down bar, 95th percentile as the up bar, the dashed line Pairwise comparison results for mean blood loss volume
as the mean, and the solid line as the median. Data points beyond bars recorded by model of tourniquet differed significantly
are dots. C-A-T, Combat Application Tourniquet; EMT, Emergency
and Military Tourniquet; TPT2, 2-in. Tactical Pneumatic Tourniquet; among all models except the TPT2 and EMT pairing
TPT3, 3-in. Tactical Pneumatic Tourniquet. (7mL; p = .0658) (Figure 6). The greatest difference in
means was between the TPT3 and C-A-T pairing; the for-
Mean Pressure Achieved mer was 176mL more (p < .0001). ANOVA results were
Results of the pairwise comparison of mean pressure two-tiered, with the TPT3 alone in the high tier (mean,
achieved by model of tourniquet were significantly differ- 348mL) and all others in the low tier, where means ranged
ent among the pneumatic models for the TPT3 and EMT from 172mL for the C-A-T to 239mL for the EMT.
pairing and the TPT2 and EMT pairing (p < .0044 for
both), and the C-A-T was significantly different than all Results by User
three pneumatic models (p < .0071 for all three) (Figure The difference of mean time to stop bleeding by user
5). Pairwise difference in means ranged from 6mmHg for was not statistically significant (p = .9886), with the less
the TPT3 and TPT2 pairing (p = .1153) to 42mmHg for experienced user 8 seconds slower than the more expe-
the EMT and C-A-T pairing (p < .0001). ANOVA results rienced user. The difference of mean pressure by user
were three-tiered, with the C-A-T and TPT3 in the high was statistically significant (p = .0210) (Figure 7), with
tier (mean, 202mmHg and 190mmHg, respectively), the the more experienced user achieving a mean pressure
TPT3 and TPT2 in the middle tier (mean, 183mmHg), 13mmHg greater than that of the less experienced user.
and EMT in the low tier (mean, 160 mmHg). Treatment The difference of mean blood loss by user was not sta-
groups, namely models of tourniquet, within the ANOVA tistically significant (p = .8287), with mean blood loss
model are tiered based on statistical significance be- 6mL more when the less experienced user applied the
tween means; groups in the same tier are not statistically tourniquets.

24 Journal of Special Operations Medicine Volume 16, Edition 2/Summer 2016