Page 44 - 2020 JSOM Winter

bleed-to-death time of 300 seconds. The rate fell into three 387 ± 66.6mmHg and 995 ± 479.6mL, respectively. The mean
levels: uncontrolled, partly controlled during mechanical ap- field tourniquet trial time was 2.8 ± 0.9 minute (Table 2). Ease
plication of a tourniquet, and controlled when an application of use in field tourniquet trials was either 4 or 5, with a me-
succeeded. The manikin was laid supine on a bed sheet at gur- dian of 4.
ney height atop two footlockers stacked atop a flatbed cart.
TABLE 2 Mean Results of Trials
The manikin autonomously measured times (time to stop Time to Stop
bleeding and trial time). It determined patient status (bleeding Trial Time, Bleeding, Pressure, Blood Loss,
or stable), trial status (“go” [satisfactory] or not), tourniquet Trial Type min* min* mmHg* mL*
placement (good or incorrect), pressure (in mmHg and cate- Field 2.8 ± 0.9 1.9 ± 0.8 387 ± 66.6 995 ± 479.6
tourniquet
gory [good or loose]), and blood loss. The caregiver collected
ease-of-use data (1, very difficult; 2, difficult; 3, neutral; 4, Conversion 3.1 ± 0.9 1.7 ± 1.1 247 ± 117.6 467 ± 264.8
(pneumatic
easy; or 5, very easy) and test times (start, end, time of day). tourniquet)
Donning time was not measured directly but occurred in the *Mean ± standard deviation.
tourniquet trial during the time to stop bleeding. Donning
delay times were separately calculated as differences in mean In conversion trials, status results showed that all 30 pneu-
times to stop bleeding compared with the control group. Each matic tourniquet uses were effective to stabilize the patient.
test had a moment in both trials when bleeding was controlled, Fifteen pneumatic tourniquets were placed at the correct site;
and the time periods thereafter within a trial were called tour- 15 were placed incorrectly. Pressures applied by pneumatic
niquet trial posttime or conversion trial posttime. The care- tourniquets were sufficient in 21 uses and insufficient in 9
giver assessed bleeding control and secured the tourniquet uses. Overall, 15 trials were satisfactory and 15 were unsat-
during these periods. Other calculations included posttime isfactory (incorrect placement in 15, of which 9 were loose).
(trial time minus time to stop bleeding) and test blood loss The mean pressure and blood loss were 247 ± 117.6mmHg
(sum of tourniquet trial and its conversion trial). Test num- and 467 ± 264.8mL, respectively. The mean conversion trial
ber, a surrogate of accrued healthcare experience, assessed test time was 3.1 ± 0.9 minute. Ease of conversion varied between
time. The data collection time (test time minus the sum of its 3 and 5, with a median of 4. Conversion trials began with
trial times) was a surrogate of healthcare recordkeeping. the field tourniquet in place at a pressure that had controlled
hemorrhage, yet the transilluminated manikin wound served
Descriptive statistics were used to portray results (Excel 2003; as feedback that signified uncontrolled hemorrhage.
www.microsoft.com). Analyses were performed using SAS, v.
9.4 (SAS Institute; www.sas.com). Continuous variables were Control Group:
tested for normality. Normally distributed data were presented No Personal Protective Equipment (PPE0)
as mean ± standard deviation (SD), and analysis of variance Field tourniquet trial status results showed that all 10 tour-
was used with the F test to assess mean differences among niquet uses without PPE stabilized the patient. There were
groups, followed by pairwise comparisons with post hoc Bon- nine satisfactory tourniquet results with correct placement
ferroni correction. Non-normally distributed data were pre- and sufficient pressure. There was one unsatisfactory result,
sented as median and interquartile ranges. Nonparametric with incorrect placement as well as insufficient pressure. The
methods for testing whether samples originated from the same median field tourniquet pressure was 372mmHg. The median
distribution were used. The Kruskal-Wallis test by ranks was blood loss was 403mL (Figure 1). Ease of use in field tourni-
used for comparing the groups for time data, followed by a quet trials in the PPE0 group was either 4 or 5, with a median
Mann-Whitney test for post hoc comparisons. A coefficient of of 5.
determination, denoted as r , was used as the portion of the
2
variance in the dependent variable that was predictable from Without PPE, all 10 conversions to pneumatic tourniquet sta-
the independent variable. Significance was determined at the bilized the patient. However, only 5 of these 10 conversions
two-sided P < .05 level. were satisfactory and placed at the correct site. Six of the 10
conversions had sufficient pressure, and 4 of the 10 conver-
Results sions were applied too loosely. Median conversion pressure
and blood loss were 266mmHg and 365mL, respectively. Ease
Test Results of conversion was 4 in each conversion trial in PPE0.
The first test time lasted 6.0 minutes. On average, test times
did not change significantly [in a power law of practice: test PPE1: Personal Protective Equipment Group 1
time = 6.0863 × (test number) −0.026 ; r = 0.0135]. Test blood The field tourniquet trial status results for PPE1 showed that
2
loss averaged 1462 ± 554.5mL and ranged from a minimum all 10 tourniquet uses stabilized the patient and had sufficient
of 624mL to a maximum of 2391mL. pressures. Nine tourniquets were placed at the correct site,
with satisfactory results. One tourniquet was placed incor-
Trials of Tourniquet Application and Conversion rectly with unsatisfactory results. In PPE1, median tourniquet
In field tourniquet trials, results showed that all 30 uses were pressure and blood loss were 386mmHg and 1234mL, respec-
effective in bleeding control to stabilize the patient. Field tively. Ease of field tourniquet use in PPE1 was either 4 or 5,
tourniquets were placed correctly in 28 of 30 uses and incor- with a median of 4.
rectly twice. Pressures applied by tourniquets were sufficient
in 29 uses and loose in 1. Overall, there were satisfactory In conversion to pneumatic tourniquet use, patient status
results in 28 uses, with 2 unsatisfactory results involving in- results showed that all PPE1 conversions stabilized the pa-
correct placement both times, compounded by insufficient tient. Four conversions were satisfactory. There was incor-
pressure in 1 case. The mean pressure and blood loss were rect placement in six conversions, three of which were also

42 | JSOM Volume 20, Edition 4 / Winter 2020