Page 102 - JSOM Summer 2018

heart and thoracic vessels; and no vascular injury was identi- TABLE 1 Demographics of Patients Who Met Selection Criteria
fied that would have prevented safe acquisition of common ResQFoam REBOA AAJT Total
femoral artery access. AAJT use was classified as potentially (n = 351) (n = 384) (n = 35) (N = 402)
beneficial if injuries were located below the aortic bifurcation Age, y
and no vascular injuries were noted above the aortic bifurca- Mean 38 38 39 38
tion. ResQFoam potential benefit criteria focused on the in- SD 16.67 16.60 14.25 16.48
tegrity of the abdominal cavity. The anatomic injury needed Sex
be located inside the abdominal cavity or retroperitoneum, the
abdominal cavity needed to remain intact without any defect Male, % 75 74 83 75
in the abdominal wall or the diaphragm, greater than 1cm in Female, % 25 26 17 25
length, and without any previous abdominal surgery and the Mechanism
resulting potential for intra-abdominal adhesions. 20 Blunt, % 60 59 74 58
Penetrating, % 40 41 26 42
To delineate the patients with the most substantial physiologic Abdominal AIS
derangement, we identified those who required transfusion of Mean 3 3 3 3
3 or more units of RBCs and therefore had either had substan- SD 1.05 1.06 0.97 1.06
tial blood loss or were continuing to bleed. 21,22 Patients requir- ISS
ing this level of transfusion have a higher mortality rate and
represent a subset of patients who would likely benefit most Mean 24 24 23 24
significantly from early hemorrhage control. SD 14.34 14.43 12.34 14.51
GCS (mean)

We reviewed resuscitation bay records and identified the sur- PH GCS 12 12 13 12
face location of injuries noted on secondary survey. For pene- ED GCS 12 12 13 12
trating trauma, the location of skin break was considered to be FAST exam
the external site of injury, while for blunt trauma, the location Positive, % 55 54 50 53
of ecchymosis was used. The injury locations were described as Time to laparotomy
positive or negative for each of the four abdominal quadrants: ≤30 min (deaths) 54 (11) 60 (13) 4 (1) 64 (14)
right upper quadrant (RUQ), right lower quadrant (RLQ), left ≤30 min + ≥3
upper quadrant (LUQ), and left lower quadrant (LLQ). We RBC units (deaths) 41 (11) 44 (12) 4 (1) 47 (13)
compared the surface quadrant of injury to the quadrant of ≤60 min (deaths) 154 (28) 173 (32) 12 (1) 185 (35)
the internal injuries identified at laparotomy.
≤60 min + ≥3 86 (27) 96 (30) 8 (1) 105 (33)
RBC units (deaths)
The cause of death of patients who met selection criteria and >60 min (deaths) 194 (24) 210 (24) 23 (1) 216 (27)
died in hospital was abstracted from both Trauma Registry
data and records of the center’s weekly morbidity and mortal- >60 min + ≥3 80 (18) 86 (18) 8 (1) 89 (21)
RBC units (deaths)
ity conferences. Base deficit
Average –4.68 –4.60 –6.29 –4.66
Because all three interventions were individually evaluated for
potential benefit on each patient, we considered the outcomes SD 5.82 5.83 6.70 5.91
to be matched by patient. We performed all possible pairwise REBOA, resuscitative endovascular balloon occlusion of the aorta;
comparisons (REBOA vs ResQFoam, AAJT vs ResQFoam, AAJT, Abdominal Aortic and Junctional Tourniquet; AIS, Abbre-
viated Injury Scale (0–6); ISS, Injury Severity Score (0–75); GCS,
and REBOA vs AAJT) among the three methods by using the Glasgow Coma Scale (3–15); PH, prehospital; ED, emergency depart-
McNemar test. We used the Bonferroni correction to adjust ment; FAST, focused assessment with sonography in trauma; time to
for multiple comparisons yielding a corrected α = .017. laparotomy measured from ED admittance to OR; RBC, red blood
cell; SD, standard deviation.

Results
reason for REBOA not being potentially helpful was due to
Between 1 January 2013 and 1 January 2015, there were 9,608 additional injuries sustained above the diaphragm.
trauma admissions with 402 patients requiring a trauma lap-
arotomy. The age, sex, mechanism of injury, and other demo- AAJT could have been potentially helpful in 26 (11%) cases
graphic data for each method are given in Table 1. Patients of blunt trauma and 9 (5%) cases of penetrating trauma, for a
were predominantly male (75%) and had a mean Injury Se- total of 35 (9%) cases. The most common reason for AAJT to
verity Score of 24. not be potentially helpful was due to the injury being located
above the aortic bifurcation.
Blunt trauma was more common than penetrating trauma,
accounting for 232 (58%) cases. The most common organ ResQFoam could have been potentially helpful in 351 (87%)
injuries identified included liver (34%), colon and mesoco- patients. It could have potentially been helpful in 209 (90%)
lon (33%), small intestine and mesentery (31%), and spleen cases of blunt injury and 142 (86%) cases of penetrating in-
(28%) (Table 2). jury. The most common reasons for foam to not be potentially
helpful were prior abdominal surgery and defects in the ab-
REBOA could have been potentially helpful in 384 (96%) dominal wall or diaphragm greater than 1cm in length.
of the 402 cases that met the selection criteria. It could have
potentially been helpful in 228 (98%) cases of blunt injury Diaphragm injuries occurred in 58 (14%) patients with foam
and 156 (92%) cases of penetrating injury. The most common being potentially helpful in 36 (62%) of them. In those 36

100 | JSOM Volume 18, Edition 2/Summer 2018