Page 126 - JSOM Winter 2022

Data Analysis The interventions of protective equipment and handler train-
Descriptive statistics were calculated for demographics, in- ing – the main preventative interventions selected – are pre-
jury event characteristics, and interventions by injury pre- sented in Table 2 by mechanism of injury and anatomical
vention category. Injury event characteristics were compared location for the 100 injuries determined to be either prevent-
between MPCs and conventional MWDs using Fisher’s exact able or potentially preventable. The intervention of change
test. Pairwise comparisons were completed by testing multiple in policy/doctrine was selected for only one injury (data not
sub- tables using the chi-square test or Fisher’s exact test when shown). Handler training was selected as the preventative
more than 20% of the cells of a given table had an expected intervention for 54 (54%) of the injuries. Of these 54, 27
count of less than 5 and adjusting for multiplicity (adjusted (50%) were caused by animal bites/scratches, and the head/
α = α ⁄ √ (# pairwise comparisons)). All analyses were per- neck (59%) were most often impacted. For the remaining 46
formed using Stata version 17.0, and α was set at 0.05. 17 injuries, protective equipment was selected as the preventative
intervention. Thirty-two of these 46 (70%) were caused by a
knife/sharp object, and the body region primarily affected was
Results
the paw (75%).
Descriptive statistics and injury event characteristics by injury
prevention category for the 195 MWD injury events are pre- Comparison of SOF MPCs and Conventional MWDs
sented in Table 1. Most of the injured MWDs were intact males Type of trauma, mechanism of injury, injury prevention cat-
(64%), almost half were Belgian Malinois/Tervurens (46%), egory, and preventive intervention were all statistically as-
and more than one-third were certified in patrol and explo- sociated (p < .001) with dog subpopulation (Table 3). For
sives detection (39%). The majority of injured MWDs were injury prevention, 72% of conventional MWD injuries were
also conventional dogs assigned to a non-SOF unit (67%). preventable (27%)/potentially preventable (45%), whereas
In terms of trauma type, over two-thirds of the injuries were only 10% of SOF MPC injuries were preventable (2%)/po-
penetrating (68%), with knives or sharp objects (24%) being tentially preventable (8%) (p < .001 for both preventable and
the most common mechanism of injury, followed by gunshot potentially preventable injuries). The majority of SOF inju-
wounds (23%) and explosions (22%). Anatomically, many of ries (52%) lacked enough information to determine an injury
the injuries were to the extremities (40%), with 16% of those prevention category compared to just 3% for conventional
occurring to the paw. MWD injuries (p < .001). When considering type of trauma,
conventional MWDs experienced predominately penetrating
Preventable Injury (64%) or blunt (21%) trauma compared to 78% and 2% of
Of the 195 injury events, 101 (52%) were determined to be injuries in SOF MPCs (p = .07 for penetrating trauma and
either preventable (n = 37; 19%) or potentially preventable p < .001 for blunt trauma). Mechanisms of injury also dif-
(n = 64; 33%), whereas 58 (30%) were unpreventable and 36 fered between dog subpopulations. Thirty-four percent of in-
(18%) lacked enough information to determine an injury pre- jury events among conventional MWDs were from a knife/
vention category (Table 1). Regarding trauma type, penetrating sharp object and 23% from animal bites/scratches, while only
injuries were largely considered to be preventable or potentially 3% of SOF MPC injuries were from a knife/sharp object (p <
preventable (81% and 70%, respectively). As for mechanism .001) and no SOF MPC injuries were from traumatic animal
of injury, sixty-eight percent (n=25) of the preventable injuries bites/scratches (p < .001). In contrast, the majority of injuries
were animal bites/scratches, while the potentially preventable sustained by SOF MPCs were from gunshot wounds (67%)
injuries were most frequently caused by a knife/sharp object whereas just 4% of conventional MWD injuries were caused
(59%). Injuries due to explosion were primarily deemed as un- by gunshot wounds (p < .001). Lastly, handler training and
preventable (41%), whereas injuries by gunshot wounds (64%) protective equipment were selected as the prevention interven-
commonly lacked enough information to determine an injury tion for 40% and 31% respectively, of the injuries among con-
prevention category. The most frequent preventable and poten- ventional MWDs compared to 2% and 8%, respectively, of
tially preventable injuries were to the head/neck (46%) and the injuries among SOF MPCs (p < .001 for both handler training
paw (45%), respectively, whereas the most common unprevent- and protective equipment).
able injuries were to the rear limb/tail (28%).
Discussion
Of the 64 MWDs that died from injury, six (9%) of these
deaths were classified as resulting from either a preventable This is the first study to apply the US DoD’s preventable death
or potentially preventable injury. The six deaths were caused review process to MWDs deployed to a combat zone to iden-
by a gunshot wound to the thorax, a knife/sharp object to the tify potentially preventable injury events and identify poten-
hind limb, a fall involving the thorax, motor vehicle accidents tial prevention interventions for this unique population. Over
involving the thorax (two deaths), and smoke inhalation. half of the 195 injury events analyzed were determined to be
either preventable or potentially preventable using this pro-
Preventative Intervention cess. Of these preventable/potentially preventable injuries, six
Consensus of preventative intervention was reached by the were found to have led to the death of the MWD. The most
SME panel on 84% of the 195 injury events reviewed. There common preventative intervention selected was handler train-
was 100% agreement on primary interventions, with two in- ing followed by protective equipment. There were significant
jury events identified as needing both handler training and a differences in injury event characteristics by certification/unit
change in policy/doctrine and one injury event as needing both assignment, with many of the injuries in our sample occur-
handler training and protective equipment. Since the preven- ring in conventional MWDs. More notably, there were more
tative intervention category of change in policy/doctrine was conventional MWDs than MPCs in our sample, and the true
selected as a target intervention for only 1% of the injuries, number of all MWDs deployed during this time is unknown.
these results are not reported. Collectively, these findings indicate that deployed MWDs may

124 | JSOM Volume 22, Edition 4 / Winter 2022