Page 33 - JSOM Winter 2017

.5518, respectively). User 4 put the tourniquet atop the wound Table 1 Pairwise Comparison of Group Means in Time to
a
and not 2–3 inches above; this test was with device U4-2 (user Determination of Bleeding Control b
4’s second device) and the user was wearing leather gloves. Mean Difference
Glove Groups Compared (seconds) p Value
Trial Status Cold gloves and mittens: bare hands 25.85 <.0001
For users, gloves, and devices, 90% of the tests (162 of 180) Cold gloves and mittens: 19.80 .0002
ended with a satisfactory result. Interuser, interglove, and in- examination gloves
terdevice differences were not significant (p = .8732, .1727, Cold gloves and mittens: glove liners 18.40 .0004
and .6776, respectively). Mittens: bare hands 17.90 .0006
Cold gloves and mittens: flight gloves 16.95 .0011
Time to Determination of Bleeding Control Cold gloves: bare hands 16.55 .0015
The mean time to determination of bleeding control was 34 Leather gloves: bare hands 13.70 .0082
± 17.4 seconds (median, 31 seconds; minimum, 11 seconds;
maximum, 111 seconds; range, 100 seconds). The 10-fold Cold gloves and mittens: glove liners 13.50 .0092
and leather gloves
range in times indicated that performance varied broadly, with Glove liners and leather gloves:
long times causing the wide breadth in times. bare hands 12.35 .0169
Cold gloves and mittens:
For users, results could be separated into three levels. User 1, leather gloves 12.15 .0188
the most experienced, was alone in the fast level; user 4, the Mittens: examination gloves 11.85 .0219
least experienced, was alone in the slow level; users 2 and 3, Cold gloves: examination gloves 10.50 .0418
at 33 and 34 seconds, respectively, were of intermediate expe-
rience with midlevel results. The analysis showed that 62% Mittens: glove liners 10.45 .0428
of the variance of times to determination of bleeding control Cold gloves and mittens: cold gloves 9.30 .0711
could be attributed to the users. Users showed no or mild Cold gloves: glove liners 9.10 .0773
learning when time in regression was checked by use number Mittens: flight gloves 9.00 .0806
(R < .2662 for all four users). Flight gloves: bare hands 8.90 .0840
2
Cold gloves and mittens: mittens 7.95 .1224
For gloves, time to bleeding control was fastest with bare hands Cold gloves: flight gloves 7.65 .1370
and examination gloves (the two glove groups which were thin- Leather gloves: examination gloves 7.65 .1370
nest), and slowest with cold gloves layered under mittens—the Glove liners: bare hands 7.45 .1475
thickest-glove group. Each level followed the pattern, which Glove liners and leather gloves:
was generally scalable, of time to bleeding control being slower examination gloves 6.30 .2203
with the thicker-glove group. Among 36 pairwise comparisons Leather gloves: glove liners 6.25 .2240
of difference between group means, 13 were significant (p <
.0428, all 13 pairs; Table 1). In regression of comparisons, the Examination gloves: bare hands 6.05 .2391
difference in glove-thickness by group was moderately associ- Mittens: glove liners and 5.55 .2800
leather gloves
ated with the difference in mean time (time difference = 4.3613
× thickness difference + 4.9812; R² = 0.3475). By the Dunnett Glove liners and leather gloves: 4.90 .3400
glove liners
method, three glove-group means (cold gloves layered under Leather gloves: flight gloves 4.80 .3499
mittens, mittens, and cold gloves) were significant (p < .0103,
all three; Table 2). Times were slowed by wearing gloves as Mittens: leather gloves 4.20 .4133
compared with bare hands. Glove effects on bleeding control Cold gloves: glove liners and 4.20 .4133
leather gloves
for these three groups were longer than for bare hands by 26
seconds, 18 seconds, and 17 seconds, respectively. Glove liners and leather gloves: 3.45 .5014
flight gloves
Cold gloves: leather gloves 2.85 .5786
For devices, results came in three levels. Devices U4-1 and
U4-2 constituted the slow level, whereas U1-1 was alone in Flight gloves: examination gloves 2.85 .5786
the fast level. Devices U2-1, U2-2, and U3-1 were in the mid- Flight gloves: glove liners 1.45 .7774
level. Device effects when parsed for intrauser results, there- Glove liners: examination gloves 1.40 .7849
fore, were consistently in the same level. Among 15 pairwise Cold gloves: mittens 1.35 .7924
comparisons of difference between device means, 11 were sig- Glove liners and leather gloves: 1.35 .7924
nificant (p < .0306, all 11 pairs). leather gloves
a Mean of one glove group compared with a mean of another glove
Trial Time group (one minus another).
The standard error of the mixed-model analysis of variance was 5.12
b
The mean trial time was 50 ± 19.0 seconds (median, 45.5 sec- for the mean difference.
onds; minimum, 21 seconds; maximum, 123 seconds; range,
102 seconds). Thus, there was a sixfold range in trial times. and flight gloves, whereas cold gloves layered under mittens
were alone in the slowest level, repeating the pattern seen in
For users, results had the same pattern as time to determina- longer time to hemorrhage control with thicker gloves. By the
tion of bleeding control. The analysis showed that 55% of the Dunnett method, five glove-group means (cold gloves layered
variance of results in trial time could be attributed to the users. under mittens; mittens; cold gloves; leather gloves; and leather
gloves and glove liners) were significantly slower than the
For gloves, results came in five levels. Bare hands and exami- mean for bare hands (p < .0295, all five). The gloves in these
nation gloves were in the fastest level along with glove liners five glove groups were the thickest.

User, Glove, and Device Effects on Tourniquet Use | 31