Page 28 - Journal of Special Operations Medicine - Summer 2016

In this study, we had a data set representing a group of Table 1 Cumulative Sum Results by User Order for Blood Loss
30 persons taught to use junctional tourniquets; each was Difference of
taught one at a time and the results are displayed for each Individual Mean
user in sequence. The mean blood loss (BL) for the entire User Individual Blood Loss and
group was identified as the benchmark value, 200mL. Order, Mean Blood Overall Mean of Cumulative
The mean blood loss for each user over nine junctional No. Loss, mL 200mL, mL Sum, mL
tourniquet applications was compared to the benchmark 1 370 170 170
value (Table 1). Blood loss data came from nine tests per 2 288 88 258
user (three tests for each of three models of junctional 3 304 104 362
tourniquet) so the individual’s mean was their single mea-
sure of performance. The blood loss CUSUM was plotted 4 259 58 421
sequentially for users 1 through 30 (Figure 1). 5 247 47 468
6 215 14 483
Results for mean blood loss volumes for each sequential 7 197 −3 480
user showed that the blood loss trended down as the
instructor gained experience with each newly instructed 8 77 −123 357
user (Figure 1). The user performance continually im- 9 174 −26 331
proved as the instructor gained more experience with 10 158 −42 289
teaching. No plateau effect was observed within the 30 11 198 −2 286
users. The first user had the highest blood loss of the
group, followed a continued decrease in blood loss. Al- 12 205 5 291
though the trend was a gradual decrease, there was vari- 13 162 −38 253
ability without a change in trend detected. 14 266 65 319
15 115 −86 233
The CUSUM plot illustrates a turning point or cusp at
the seventh user. The prior portion of the plot (users 16 138 −63 171
1–7) indicates that the greatest improvement occurred 17 143 −57 114
with the first seven users and then did not improve as 18 168 −33 82
much thereafter. Three of the worst four differences 19 251 50 133
were in the first three users (Figure 1, Table 1). The im-
provement after the seventh user was the only change 20 127 −73 60
detected in the instructor’s trend of performance. 21 162 −38 22
22 99 −101 −78
The main finding of this study is that the instructor’s 23 218 18 −60
teaching experience appeared to directly affect user per-
formance; in a model of junctional hemorrhage, the vol- 24 174 −26 −86
ume of blood loss from the manikin during junctional 25 208 8 −78
tourniquet placement was a useful metric for instructor 26 310 110 32
learning. Two methods of quantifying blood loss had dif- 27 126 −74 −42
ferent usefulness in assessing the learning of the instruc-
tor. The first, a conventional technique, was simple and 28 273 73 32
intuitive: it chronologically showed performance sequen- 29 218 18 49
tially in a way that is familiar to many instructors and 30 158 −42 8
students, but it did not detect a change in trend. The first The overall mean of the data set blood loss was 200mL. Users had an
is a raw technique that was intended to show sequential individual mean blood loss result.
progress and was not intended to show small changes in
trend: it is limited to detecting large changes. However,
the CUSUM technique was less intuitive but was able to may be useful to them and their supervisors. We find
detect a small but meaningful change in trend. CUSUM no method with evidence in the educational literature
detected what the other technique did not. for preparing instructors to be fully effective from the
start. The present study was unique in that it provided
Instructor performance in bleeding control instruction evidence of instructor learning metrics for junctional
may enable organizations like the Army to monitor tourniquet users.
quality of education, such as in its recent introduction
of junctional tourniquet instruction to student medics. The limitation of the present report is its introductory
Allowing instructors to know how effective they are nature. Methods to assess learning curves of instructors

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