See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318871131

The trauma center is too late: Major limb trauma without a pre-hospital
tourniquet has increased death from hemorrhagic shock

Article in Journal of Trauma and Acute Care Surgery · August 2017

DOI: 10.1097/TA.0000000000001666

CITATIONS READS

109 254

7 authors, including:

Michelle Scerbo Keith S Gates

University of Texas Health Science Center at Houston Tacoma Fire Department
19 PUBLICATIONS 871 CITATIONS 10 PUBLICATIONS 374 CITATIONS

SEE PROFILE SEE PROFILE

Joseph Love Charles Wade

University of Texas Health Science Center at Houston University of Texas Health Science Center at Houston
19 PUBLICATIONS 669 CITATIONS 846 PUBLICATIONS 43,129 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Keith S Gates on 26 January 2023.

The user has requested enhancement of the downloaded file.

 AAST 2016 PLENARY PAPER

The trauma center is too late: Major limb trauma

without a pre-hospital tourniquet has increased death
from hemorrhagic shock
Michelle H. Scerbo, MD, MS, John B. Holcomb, MD, Ethan Taub, DO, Keith Gates, MD, Joseph D. Love, DO,
Charles E. Wade, PhD, and Bryan A. Cotton, MD, MPH, Houston, Texas

BACKGROUND: To date, no civilian studies have demonstrated that pre-hospital (PH) tourniquets improve survival. We hypothesized that late,
trauma center (TC) tourniquet use would increase death from hemorrhagic shock compared to early (PH) placement.
METHODS: All patients arriving to a Level 1, urban TC between October 2008 and January 2016 with a tourniquet placed before (T-PH)
or after arrival to the TC (T-TC) were evaluated. Cases were assigned the following designations: indicated (absolute indi-
cation [vascular injury requiring repair/ligation, operation within 2 hours for extremity injury, or traumatic amputation] or
relative indication [major musculoskeletal/soft tissue injury requiring operation 2–8 hours after arrival, documented large
blood loss]) or non-indicated. Outcomes were death from hemorrhagic shock, physiology upon arrival to the TC, and mas-
sive transfusion requirements. After univariate analysis, logistic regression was carried out to assess independent predictors
of death from hemorrhagic shock.
RESULTS: A total of 306 patients received 326 tourniquets for injuries to 157 upper and 147 lower extremities. Two hundred eighty-one (92%)
had an indication for placement. Seventy percent of patients had a blunt mechanism of injury. T-TC patients arrived with a lower
systolic blood pressure (SBP, 101 [86, 123] vs. 125 [100, 145] mm Hg, p < 0.001), received more transfusions in the first hour of
arrival (55% vs. 34%, p = 0.02), and had a greater mortality from hemorrhagic shock (14% vs. 3.0%, p = 0.01). When controlling
for year of admission, mechanism of injury and shock upon arrival (SBP ≤90 mm Hg or HR ≥120 bpm or base deficit ≤ 4) in-
dicated T-TC had a 4.5-fold increased odds of death compared to T-PH (OR 4.5, 95% CI 1.23–16.4, p = 0.02).
CONCLUSIONS: Waiting until TC arrival to control hemorrhage with a tourniquet was associated with worsened blood pressure and increased trans-
fusion within the first hour of arrival. In routine civilian trauma patients, delaying to T-TC was associated with 4.5-fold increased
odds of mortality from hemorrhagic shock. (J Trauma Acute Care Surg. 2017;83: 1165–1172. Copyright © 2017 Wolters Kluwer
Health, Inc. All rights reserved.)
LEVEL OF EVIDENCE: Level IV.
KEY WORDS: Tourniquet; pre-hospital; major limb trauma; hemorrhagic shock.

I ncreasingly available, accessible, and early tourniquet applica-

tion by the United States Military has delimited isolated limb
exsanguination from the leading cause of preventable wartime
the shortened transport time of mature trauma systems, and the
capability of pre-hospital providers to dedicate efforts to man-
ual compression limit their necessity.7–10 However, the use of
death.1–3 Subsequently, one of the earliest assessments of pene- tourniquets in civilian trauma scenarios has been increasing
trating extremity injuries in civilians postulated that adequate since 2008. The Civilian Tactical Emergency Casualty Care,
hemorrhage control, i.e., with a tourniquet, may have prevented modeled after the United States Military Tactical Combat Ca-
57% of deaths.4 One third of potentially preventable fatal hem- sualty Care, encourages the use of tourniquets by bystanders
orrhages from civilian trauma occur after severe extremity in- and non-medical first responders, as they acknowledge these
jury,5 and direct manual pressure remains the standard of care individuals may be closest to the wounded in hostile environ-
for control of this hemorrhage. Opinions criticizing civilian ments.11,12 The National Association of Emergency Medical
pre-hospital tourniquet use argue the infrequent rate6 and pattern Technicians teaches tourniquet application in courses targeted
of major exsanguinating limb trauma seen in civilian scenarios, at lay persons and police officers.13 The International Liaison
Committee on Resuscitation guidelines now include the use
Submitted: June 8, 2016, Revised: July 13, 2017, Accepted: July 14, 2017, Published of field tourniquets when direct pressure fails to control bleeding
online: August 3, 2017. from a wounded limb.14 Finally, the Hartford Consensus IV, rec-
From the Center for Translational Injury Research (CeTIR) (M.H.S., J.B.H., C.E.W.,
B.A.C.), Department of Surgery (M.H.S., J.B.H., E.T., J.D.L., C.E.W., B.A.C.), ognizing the difficulty of sustained direct manual pressure, now
and Department of Emergency Medicine (K.G.), The University of Texas Health includes the use of tourniquets for hemorrhage control from
Science Center, Houston, Texas; and Life Flight, Memorial Hermann Hospital, any etiology, including motor vehicle collisions. Partnered by
Houston, Texas (J.D.L.).
Address for reprints: Michelle Scerbo, MD, MS, Center for Translational Injury
the Hartford Consensus, the Stop the Bleed campaign especially
Research, 6410 Fannin Street, UTP 1100.26, Houston, TX 77030; email: emphasizes the importance of early tourniquet placement by
[email protected]. non-medical bystanders.15
Meetings presented: 75th Annual Meeting of the American Association for the We began to use pre-hospital tourniquets in 2008 and
Surgery of Trauma, September 14, 2016, in Waikoloa, Hawaii.
have gradually expanded their deployment throughout our
DOI: 10.1097/TA.0000000000001666 trauma system. Similar to observations with military use,16 our
J Trauma Acute Care Surg
Volume 83, Number 6 1165

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

 J Trauma Acute Care Surg
Scerbo et al. Volume 83, Number 6

hypothesis was that late tourniquet application would be associ- the institutional morbidity and mortality process and captured
ated with an increased risk of death from hemorrhage. The pur- by the Trauma Registry. Secondary outcomes included phys-
pose of this study was to demonstrate that pre-hospital (PH) iology on presentation to the TC and massive transfusion of
tourniquet application improves survival compared to presum- blood products. These outcomes were determined by a combi-
ably applying direct pressure. nation of Trauma Registry data and electronic health record
review. Patient demographics, injury mechanism, Injury Sever-
ity Score (ISS), Abbreviated Injury Scale (AIS) scores, length of
METHODS hospital stay (days), length of intensive care unit stay (days),
Study Setting and Selection of Participants PH and TC vital signs (systolic blood pressure, diastolic blood
pressure, heart rate, and Glasgow Coma Score [GCS]), admis-
The University of Texas Health Science Center at Houston
sion laboratories (hemoglobin, hematocrit, base deficit,
and the Memorial Hermann Hospital Institutional Review
platelet count, international normalized ratio, pH, bicarbon-
Boards approved this study. This was a single-center, retrospec-
ate, thromboelastography), total units of blood components trans-
tive cohort study of patients arriving to Memorial Hermann Hos-
fused, and outcome (dead/alive and cause of death) were
pital with a trauma activation, identified using the institution’s
determined from the Trauma Registry. Trauma Registry staff cal-
Trauma Registry of the American College of Surgeons database.
culated ISS and AIS after discharge or death of each patient. Pa-
All trauma patients admitted between October 2008 and January
tients were evaluated for whether or not they required a massive
2016 with a tourniquet listed as a treatment were included in the
transfusion within the first hour of arrival (critical administra-
study. Patients were excluded if no documentation of tourniquet
tion threshold positive CAT+, defined as 3 units of blood per
placement could be corroborated on review of medical records.
hour).19 T-PH or T-TC application and reason were extracted
All patients had an official-issued Combat Application Tour-
from the health record. Type of vascular injury and emergent/
niquet (Composite Resources, Rock Hill, SC), a windlass tour-
urgent operative interventions were confirmed upon review of
niquet recommended by the US Military Tactical Combat
operative reports.
Casualty Care, placed by pre-hospital or in-hospital personnel.
Definitions Statistical Analysis
Indication for Tourniquet Placement Continuous data are presented as medians and 25th,
Tourniquet placement was designated as indicated if 75th percentiles. Comparisons between groups are performed
an a priori absolute or relative indication could be identified using the Wilcoxon rank-sum (Mann–Whitney U test). Categor-
via review of patient hospital record.17 An absolute indication ical data are reported as proportions and, where appropriate,
included a vascular injury requiring repair or ligation, an tested for significance using χ2 or Fisher’s exact tests. After uni-
emergent operation for extremity injury within 2 hours of hos- variate analysis, multivariate logistic regression was carried out
pital arrival, or a traumatic amputation. A relative indication to assess independent predictors of death from hemorrhage.
included a documented significant blood loss at the scene or STATA Statistical software (version 13.1; College Station,
major musculoskeletal/soft tissue injury requiring an urgent TX) was used for the analysis.
(between 2 and 8 hours of hospital arrival) operation. The tour-
niquet was designated as non-indicated if there was no absolute
or relative indication identified. RESULTS

Geographic Location of Tourniquet Placement Three hundred six patients received 326 tourniquets for
All tourniquets placed before arrival to the trauma center injuries to 157 upper and 147 lower extremities. These patients
(TC) were designated as pre-hospital (T-PH). This included tour- were a median of 34 (25, 46) years old with mostly isolated, ma-
niquets placed before arrival (at the scene, in transport, or at an jor limb trauma (AIS extremity 3 [2, 3], ISS 9 [5, 18]) sustained
outside facility), irrespective of whether an additional tourniquet by blunt (70%) mechanisms and transported to our TC by
was required after arrival to the TC. Only de novo tourniquets aeromedical transport (62%). The predominant blunt mecha-
placed after arrival to the TC were considered TC-tourniquet nism of injury was motor vehicle collisions (63%). Table 1
placement (T-TC). displays the demographic and injury data of all patients Over-
all, 21 patients died, 15 from hemorrhagic shock. A total of
281 patients (92%) with a T-PH or T-TC had an indication
Shock for placement (252 T-PH, 29 T-TC). After reaching a steady
Shock was defined upon TC arrival as either a base def- rate, 73 (72, 74) tourniquets on 71 (69, 72) patients per year
icit of ≤4 or stage III or IV hemorrhagic shock, with a heart are used within our trauma system (Fig. 1).
rate of ≥120 beats per minute or systolic blood pressure
(SBP) of ≤90 mm Hg.18 Pre-hospital Tourniquet Placement
Two hundred seventy-seven patients received T-PH for
Outcomes the following injuries: lacerations (41.2%), open fractures
Patients were compared by the geographical location (27.4%), traumatic amputations (18.4%), mangled extremi-
of tourniquet placement—T-PH or T-TC. The primary out- ties (4.0%), large soft tissue defects (3.3%), gunshot wounds
come evaluated was death from hemorrhagic shock. Cause (4.7%), or a deformed extremity and suspected bleeding
of death was determined upon faculty case review through (0.72%). One patient arrived as a trauma activation with a

1166 © 2017 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

J Trauma Acute Care Surg
Volume 83, Number 6 Scerbo et al.

Fourteen (5.6%) patients who had an indication for T-PH

TABLE 1. Demographic and Injury Data for all Patients Who
required an additional tourniquet after arrival to the TC. Reasons
Received a Tourniquet
for an additional placement were replacement of an improvised
All Patients tourniquet for a Combat Application Tourniquet (35.7%),
n = 306 documented failure of the T-PH (28.6%), new active bleeding
Age, y 34 (25, 46)
(21.4%), and unknown (14.3%). Sixty-four percent of these
% Male (n) 84 (258)
additional tourniquet placements occurred before 2013.
% Caucasian (n) 46 (141) Trauma Center Tourniquet Placement
BMI, kg/m2 27 (24, 31) Twenty-nine patients had their first tourniquet placed
% Penetrating (n) 30 (93) after arrival to the TC; all of these were indicated. The injuries
% Aeromedical transport (n) 62 (190) associated with T-TC were lacerations (31.0%), open fractures
% Indicated (n) 92 (281) (27.6%), traumatic amputations (20.7%), gunshot wounds
ISS 9 (5, 18) (10.3%), mangled extremities, or a deformed extremity with
AIS—head 0 (0, 2) suspected bleeding (6.9%). Reasons for T-TC included active
AIS—face 0 (0, 0) extremity bleeding with hypotension (56%), active extremity
AIS—chest 0 (0, 3) bleeding without hypotension (26%), ACLS in progress (11%),
AIS—abdomen 0 (0, 2) and unknown (7%). Less than 10 de novo T-TC placements oc-
AIS—extremity 3 (2, 3) curred per year (Fig. 1).
AIS—external 1 (0, 1)
Transport time, min 46 (31, 78) Indicated Patients: Comparison of Pre-hospital
% Tourniquets applied in TC (n) 9 (29) and Trauma Center Tourniquet Placement
% Arterial injury (n) 37 (112) Two hundred eighty-one patients (92%) with a T-PH or
% Venous injury (n) 11 (35) T-TC had an indication for placement (252 T-PH, 29 T-TC).
% Mortality (n) 7 (21) Table 2 displays the demographic and injury data for patients
% Cause of death from hemorrhage (n) 71 (15) with indicated T-PH compared to T-TC. There were no differ-
% in Year (n) ences in age, gender, race, body mass index, mechanism of in-
*2008 1 (4) jury, transport, or duration of transport time in patients who
2009 9 (28) received T-PH or T-TC. Compared to T-PH, patients with T-TC
2010 5 (14) were more severely injured (ISS 9 [5, 17] vs. 20 [9, 27],
2011 4 (11) p < 0.001), with higher median AIS of the head (0 [0, 0] vs. 0
2012 12 (37) [0, 2], p < 0.001), chest (0 [0, 0] vs. 0 [0, 3], p < 0.001), abdomen
2013 23 (71) (0 [0, 2] vs. 0 [0, 2], p = 0.01), and extremity (3 [2, 3] vs. 3 [3, 4],
2014 22 (66) p < 0.001). The rates of vascular injury (42% vs. 41%, p = 0.91),
2015 24 (72) compartment syndrome (2% vs. 0%, p = 0.44), and amputation
*2016 1 (3) (23% vs. 34%, p = 0.17) were similar between T-PH and T-TC.
*Partial years. Values expressed as median (25th, 75th percentile) or n (%).
AIS, Abbreviated Injury Scale; BMI, body mass index; ISS, Injury Severity Score. Location and Management of Injuries
One hundred twelve (40%) patients sustained an arterial
tourniquet in place to control bleeding from a disrupted arte- injury. The majority of arterial injuries occurred in the upper
riovenous fistula.
Two hundred fifty-two (89%) of T-PH tourniquets were
indicated. Each of the 25 patients who did not have an indication
for tourniquet placement had only one tourniquet placed; 10 to
upper extremities and 15 to lower extremities. There were no dif-
ferences in demographics, mechanism of injury, method of
transport, transport time, or scene or arrival physiology and lab-
oratory values for patients with an indicated compared to non-
indicated tourniquet. The patients with indicated tourniquet
placement had a higher ISS (9 [5, 17] vs. 1 [0, 9], p < 0.01)
and AIS of the extremity (3 [2, 3] vs. 0 [0, 2], p < 0.01) com-
pared to those deemed non-indicated. All-cause mortality was
lower in patients with indicated tourniquet placement compared
to those who were non-indicated (5% vs. 20%, p = 0.01); how-
ever, unadjusted mortality from hemorrhagic shock was similar Figure 1. Annual distribution of tourniquet application by
between indicated and non-indicated tourniquets (3% vs. 12%, geographical location. T-TC, de novo tourniquet placed after
p = 0.61). The ISS of these moribund patients with a T-PH arrival to trauma center; T-PH, any pre-hospital tourniquet
was similar between indicated compared to non-indicated (26 application. Data with partial years (2008, 2016) were excluded
[18, 34], p = 0.83). from this figure.

© 2017 Wolters Kluwer Health, Inc. All rights reserved. 1167

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

 J Trauma Acute Care Surg
Scerbo et al. Volume 83, Number 6

extremity, with 46% of these injuries occurring to the radial or

TABLE 2. Demographic and Injury Data and Physiologic for all
ulnar arteries and 19% to the brachial artery. In the lower ex-
Patients Who Received an Indicated Pre-hospital (PH) Compared
to a Trauma Center (TC) Tourniquet tremity, only 16% of arterial injuries were to the femoral ar-
tery; the remaining 19% occurred to the popliteal, tibial, or
T-PH T-TC All peroneal arteries. Thirty-five (12%) patients sustained a ve-
n = 252 n = 29 n = 281 p nous injury. Similarly, venous injuries occurred more com-
Age, y 33 (25, 46) 34 (24, 50) 33 (25, 46) 0.95 monly in the upper extremity (62% vs. 38%). The remaining
% Male 84 93 85 0.14 injuries were to the femoral (23%), saphenous (8%), and popli-
% Caucasian 47 45 47 0.50 teal (7%) veins. The most common procedure for these vascular
BMI, kg/m2 27 (24, 31) 27 (24, 30) 27 (24, 31) 0.86 injuries was arteriorrhaphy/venorrhaphy (45%), followed by li-
% Penetrating 29 38 30 0.30 gation (33%), interposition graft (18%), and shunt (4%).
% Aeromedical transport 64 48 62 0.10
Transport time, min 48 (31, 79) 39 (22, 63) 46 (31, 78) 0.14
Primary Outcome—Death from Hemorrhage
ISS 9 (5, 17) 20 (9, 27) 10 (5, 17) 0.00 Patients with T-PH compared to T-TC had similar rates of
AIS—head 0 (0, 0) 0 (0, 2) 0 (0, 0) 0.00 all-cause mortality (5% vs. 14%, p = 0.07). However, death from
AIS—face 0 (0, 0) 0 (0, 0) 0 (0, 0) 0.46 hemorrhagic shock (3% vs. 14%, p = 0.01) was higher in pa-
AIS—chest 0 (0, 0) 0 (0, 3) 0 (0, 0) 0.00 tients who received T-TC. Time to death from hemorrhagic
AIS—abdomen 0 (0, 0) 0 (0, 2) 0 (0, 0) 0.01 shock was similar between T-PH and T-TC (0.3 [0.1, 7.7] hours,
AIS—extremity 3 (2, 3) 3 (3, 4) 3 (2, 3) 0.00 p = 0.80; Table 3).
AIS—external 1 (0, 1) 1 (0, 1) 1 (0, 1) 0.53 After controlling for year, mechanism of injury, and the
presentation of shock (systolic blood pressure ≤90 mm Hg or
Values expressed as median (25th, 75th percentile) or percent.
AIS, Abbreviated Injury Score; BMI, body mass index; ISS Injury Severity Score.
heart rate ≥120 beats per minute or base deficit ≤4), patients
who had an indication for tourniquet placement had a 4.5-fold

TABLE 3. Demographic and Injury Data and Physiologic for all Patients Who Received an Indicated Pre-hospital (PH) Compared to a
Trauma Center (TC) Tourniquet
T-PH T-TC All
n = 252 n = 29 n = 281 p
Pre-hospital
Systolic blood pressure, mm Hg 119 (92, 139) 100 (83, 113) 114 (91, 138) 0.08
Heart rate, bpm 100 (84, 120) 122 (87, 135) 100 (84, 122) 0.13
GCS 15 (14, 15) 14 (3, 15) 15 (13, 15) 0.01
TC arrival
Systolic blood pressure, mm Hg 125 (100, 145) 101 (86, 123) 122 (98, 144) 0.00
Δ Systolic blood pressure, mm Hg 1 (−12, 28) −10 (−22, 3) 0 (−15, 26) 0.03
Heart rate, bpm 98 (82, 115) 110 (90, 129) 99 (82, 118) 0.10
GCS 15 (14, 15) 15 (3, 15) 15 (13, 15) 0.17
Base excess −4 (−7, −1) −6 (−12, −1) −4 (−7, −1) 0.23
pH 7.31 (7.23, 7.35) 7.23 (7.14, 7.31) 7.3 (7.22, 7.35) 0.06
Hemoglobin, g/dL 13.1 (11.1, 14.3) 12.6 (10.8, 13) 13 (11, 14.3) 0.15
Laboratory values
TEG—activated clotting time 113 (105, 128) 113 (113, 121) 113 (105, 128) 0.93
TEG—angle 73 (68, 76) 74 (68, 76) 73 (68, 76) 0.76
TEG—maximum amplitude 62 (57, 67) 62 (54, 67) 62.1 (57, 67) 0.51
TEG—% lysis 30 min 1 (0, 3) 1.3 (0.4, 2.8) 1.3 (0.2, 3.3) 0.82
Length of stay, d 9 (2, 16) 8 (1, 18) 9 (2, 16) 0.77
ICU length of stay, d 0 (0, 2) 2 (0, 5) 0 (0, 2) <0.01
% Vascular injury 42 41 42 0.91
% Compartment syndrome 2 0 2 0.44
% Amputation 23 34 24 0.17
% Mortality 5 14 6 0.07
Time to death, h 9 (0, 78) 3.6 (0, 7.7) 3.6 (0.1, 24.3) 0.40
% Cause of death—hemorrhage 3 14 4 0.03
Time to death, h 0 (0, 17) 3.6 (0, 7.7) 0.3 (0.1, 7.7) 0.80
Values expressed as median (25th, 75th percentile) or percent. Δ systolic blood pressure calculated as difference between systolic blood pressure on arrival and before arrival.
GCS, Glasgow Coma Scale; TEG, thromboelastography.

1168 © 2017 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

J Trauma Acute Care Surg
Volume 83, Number 6 Scerbo et al.

increased risk of death from hemorrhagic shock if tourniquet ap- could have benefited from tourniquets arrived with one in place.
plication was delayed until after arrival to the TC (OR 4.5, 95% In addition, only 36% felt the current civilian research regarding
CI 1.23–16.4, p = 0.02). tourniquet use was adequate. Forty-nine percent of respondents
felt a lack of concomitant civilian research was a barrier to trans-
Secondary Outcomes lation from military to civilian use.27 To date, all civilian studies
Physiology on Presentation to Trauma Center of tourniquet use have been limited by their retrospective nature
Compared to indicated T-PH, T-TC had a lower PH GCS and the lack of a control group.
(15 [14, 15] vs. 14 [3, 15], p = 0.01) and a lower blood pressure In 2005, Dorlac et al.4 reported one of the first investiga-
upon arrival with a larger decrease from PH (TC SBP: 125 [100, tions of isolated civilian penetrating limb injury patterns that
145] vs. 101 [86, 123] mm Hg, p < 0.001; delta SBP 1 [−12, 28] lead to death from hemorrhagic shock. The authors observed a
vs. −10 [−22, 3] mmHg, p = 0.03). This was associated with small incidence of isolated penetrating extremity injuries in
more ICU days (0 [0, 2] vs. 2 [0, 5], p < 0.01) and increased mor- civilians (14/75,000 in 5 years), but a 100% mortality despite
tality from hemorrhagic shock (3% vs. 14%, p = 0.03) (Table 3). attempts at hemorrhage control in 79% of the population stud-
There was no difference in laboratory values in patients who ied. These desperate attempts, which included a combination
had indicated T-PH or T-TC. However, laboratory values were of dressings, direct pressure, military anti-shock trousers, and
missing for up to 35% of these patients. The rate of missing even duct tape, but not tourniquets, failed to rescue all of these
base excess (33% vs. 48%, p = 0.11) and pH (33% vs. 45%, patients. Based on the location of injury and the presence of
p = 0.19) were comparable between patients who had indicated “signs of life” in the field, the authors concluded that early
T-PH compared to T-TC. Compared to T-PH, patients with T-TC and effective hemorrhage control, i.e., with a tourniquet, may
were missing more values for hemoglobin (13% vs. 28%, have prevented at least 57% of these deaths. Passos et al.28 com-
0 = 0.04) and thromboelastography (28% vs. 55%, p = 0.002). pared eight patients who received a pre-hospital tourniquet to six
who did not and died. The authors found a higher rate of mas-
Transfusion of Blood Products sive transfusions in the patients who did not receive a tourniquet
As displayed in Table 4, compared to T-PH, patients with a and died. This was the first civilian publication that suggested
T-TC received more units of plasma (3 [2, 5] vs. 5 [3, 10], the potential ability of tourniquets to prevent but not reverse
p < 0.01), but not red blood cells (3 [1, 6] vs. 4 [2, 9], shock,16,29 and our current study supports their inference.
p = 0.10) or platelets (1 [1, 3] vs. 3 [1, 6], p = 0.11) in the first Because of the low rate of mortality in civilians with major
hour after arrival. However, compared to indicated T-PH, patients limb trauma,30 no civilian study to date has demonstrated an im-
with T-TC had a higher rate of transfusion within the first hour of provement in survival with the use of tourniquets. Similar to
arrival (CAT+ status [34% vs. 55%, p = 0.02]). studies conducted in the US military setting,3,21,23,29 our current
study has demonstrated that waiting until TC arrival to apply a
DISCUSSION tourniquet was associated with lower blood pressure upon ar-
rival, more plasma transfusions, and a higher rate of transfusion
After the demonstration in the United States Military that within the first hour of arrival. In routine civilian trauma pa-
tourniquets are a rapid, safe,20 effective,21,22 and lifesaving23 (es- tients, delaying tourniquet application until arrival at a TC was
pecially when applied before the onset of shock3) method for associated with greater than 4.5-fold increased odds of mortality
hemorrhage control, their use in civilian trauma scenarios has from hemorrhage.
been increasing since 2008. Current recommendations for tour- Combat application tourniquets were supplied to PH pro-
niquet use in civilians range from (1) no use,24 (2) overlooking viders in our trauma system starting in 2008. They were first
pre-hospital use,18,25 (3) use only for penetrating lower extrem- placed in our ED and given to aeromedical transport teams in
ity injuries,25 (4) use only in tactile or hostile scenarios,11–13 or 2008, to 600 ambulances in 2011, and to 5,000 police officers
(5) judicious use by medical and non-medical first responders in 2014. After this implementation, we have observed a steady
for control of hemorrhage from major limb trauma.15,26 The ma- rate of tourniquet use of approximately 70 cases per year.
jority (55%) response of a 2015 survey of United States Trauma The rate of trauma center tourniquet application has also re-
Medical Directors indicated that less than 20% of patients who mained constant at less than 10 per year. Therefore, of over
20,000 trauma admissions over nearly 7 years at our urban
Level I trauma center, less than 0.5% received a tourniquet.
TABLE 4. Resuscitation Differences Between Patients Who had The predominant blunt mechanism was a motorized vehicle
an Indicated Pre-hospital Compared to Trauma Center
Tourniquet Placement
collision whereas the predominant penetrating mechanism
was a laceration with a sharp object; there were no mass casu-
T-PH T-TC All alty or active shooter events in this population. Dissimilar to
Transfusion n = 252 n = 29 n = 281 p military studies,16 we have demonstrated a higher rate of tour-
1h Red blood cells, U 2 (1, 6) 4 (2, 9) 3 (2, 6) 0.10
niquet application for injuries to the upper compared to the
Plasma, U 3 (2, 5) 5 (3, 10) 3 (2, 6) <0.01
lower extremities. A previous study of civilians who sustained
Platelets, U 1 (1, 3) 2 (1, 6) 1 (1, 3) 0.11
penetrating extremity injuries also demonstrated a nearly dou-
% CAT+ 34 55 36 0.02
ble incidence of upper extremity compared to lower extremity
vascular injuries.31 This is likely reflective of the differences
CAT+, critical administration threshold (3 U RBC/1 h). Values are expressed as median in mechanism of injury between wartime and civilian trauma
(25th, 75th percentile).
scenarios. Hence, although civilian tourniquets are applied

© 2017 Wolters Kluwer Health, Inc. All rights reserved. 1169

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

 J Trauma Acute Care Surg
Scerbo et al. Volume 83, Number 6

rarely and for very different mechanisms of injury than in mil- tourniquet placement. Although the AIS of the other anatomic
itary scenarios, the efficacy over manual direct pressure is still regions were higher, the 75th percentile for head and abdominal
implied. The infrequent rate of use should not overshadow the injuries was only 2, which is a moderate injury (mortality of
impact of lives saved. Further, this infrequent use of tourni- head AIS ≤2 is <1%37). The 75th percentile for chest injuries
quets (similar to that of other major Level I US trauma cen- was 3, which is a serious injury. However, the more injured
ters)27,32,33 makes it unlikely that a randomized, controlled status of these patients who did not receive a tourniquet until
trial evaluating PH tourniquet versus manual compression in after arrival to the TC depicts a scenario where a patient has
civilians with major limb trauma will ever be conducted.34 multiple possible sources of hemorrhagic shock and, if any-
The purpose of this study was to evaluate the effect of thing, would benefit from adequate control of a bleeding limb
early compared to late tourniquet application on the risk of death and full team dedication to all pertinent issues in transport
from hemorrhagic shock. Because of deficits in the collection (i.e., not dedicating a team member to direct pressure).
and recording of data, direct pressure can only be implied as The safety of these tourniquets has been previously pub-
the method of hemorrhage control until a tourniquet was applied lished in a subset (n = 105) of the patients included in this
as it is a standard of care within our system. For direct pressure to study.17 One hundred five of the current 306 patients were eval-
be effective, both hands should be available to apply direct and uated for potential complications associated with tourniquet use,
sustained force on a firm surface that can provide counterpres- including acute renal failure, amputation, compartment syn-
sure.13 This means that there must be dedicated personnel to ap- drome, nerve palsy, and venous thromboembolism. After ad-
ply pressure and the patient should not be moved during this judication by surgical staff, none of the complications were
application, which is not realistically feasible while also extricat- attributed to the use of a tourniquet, even in patients without
ing, transporting, and caring for the rest of the patient with a one- an indication for tourniquet placement. These results are consis-
to two-person medic team. As more patients died from hemor- tent with other military and civilian reports of few tourniquet-
rhage that did not have a tourniquet placed until arrival to the associated complications,22,31,38 if any at all.17,39–41 This is the
TC, the results from this study suggest that direct pressure may largest single-center study to date that reports the use of tour-
not be the most effective method of controlling bleeding limbs. niquets for civilians with severe extremity injuries. Similar to
This could be either because of difficult execution as a result military studies,3,16 we have demonstrated a benefit of early
of the limited team composition, environment where the injury tourniquet application and the devastating consequences of
occurred, transport, or the wound itself, as open fractures, trau- waiting until TC arrival. Nonetheless, pre-hospital providers
matic amputations, and mangled extremities can bleed substan- were incorrect in 17.6% of patients, either by placing a tourni-
tially from both injured vessels and the soft tissue. quet that was later deemed non-indicated or by failing to place
Both the American College of Surgeons Committee on a tourniquet when the patient ultimately needed one. This study
Trauma (ACS-COT) and the American College of Emergency should serve as additional encouragement for civilian trauma
Physicians (ACEP) now recommend that tourniquets be used systems to employ proper education for tourniquet use to all
when extremity hemorrhage presents a threat to life,35,36 empha- members of the trauma system.
sizing that tourniquet use should be avoided if the bleeding from This study is limited because of its retrospective and
an extremity injury is minor. single-center nature. As a result, tourniquet times could not be
We defined designated tourniquet placement a priori to abstracted from the patient record, so it is difficult to conclude
this study17 as either an absolute indication, relative indication, that the limited observation of complications would be observed
or non-indicated. However, it is reasonable to assume that all of regardless of tourniquet time or in a setting with a trauma system
the tourniquets were applied with the intent to control significant with prolonged transport times. Assuming that tourniquet and
hemorrhage, so deeming the 25 non-indicated tourniquets as transport times are nearly equivalent, our recorded transport
“inappropriate” is difficult to judge in this retrospective study times took 46 (31, 78) minutes, which is considerably shorter
and discounts the impression of the care provider. The predom- than the 2 hours used for tourniquets in the operating room42,43
inant injuries associated with tourniquet application in patients and was not associated with any potential complications in an
who had a non-indicated placement were lacerations, open frac- evaluation of a subset of these patients.17 In addition, Kragh
tures, and gunshot wounds, which can bleed considerably from et al.38 reported no loss of limbs resulting from tourniquet is-
the soft tissue, even in the absence of a vascular injury. In ad- chemia in studies of tourniquet use in combat trauma despite
dition, the unadjusted mortality was higher in patients with evacuation times that were often much longer than those seen
non-indicated compared to indicated PH tourniquet place- in urban trauma systems. However, the translation of our find-
ment. These moribund patients with non-indicated PH tourni- ings into trauma systems with prolonged transport times de-
quet placement had a median ISS of 26.20,34 Three arrived serves further study.
with cardiopulmonary resuscitation in progress. Therefore, use Missing data made some values difficult to evaluate. Up
in these patients is consistent with teaching from the ACS- to 35% of laboratory values were missing, and most often these
COT and ACEP, despite designation of “non-indicated” by our were in the most seriously injured patients who succumbed to
definitions defined a priori. Finally, use in this setting is congru- their injury. For example, despite the higher rate of massive
ent with studies that have demonstrated the influence of the ap- transfusions and hypotension in patients who received a tourni-
pearance of the wound for tourniquet use20 and the implication quet in the TC, laboratory values did not indicate that these pa-
that it serves as a “threat-to-life” indicator.29 tients were more coagulopathic or acidotic (Table 3). When
The patients who had a tourniquet applied after arrival identifying the limitations of this study, several areas for
to the TC were more severely injured than those with PH improvement regarding documentation of tourniquet use

1170 © 2017 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

J Trauma Acute Care Surg
Volume 83, Number 6 Scerbo et al.

were identified. Specifically, documentation of direct pressure, 13. National Association of Emergency Medical Technicians (U.S.) P-HTLSC,
actual tourniquet time, the reason for and level of tourniquet American College of Surgeons, Committe on Trauma. PHTLS: Prehospital
Trauma Life Support. St. Louis, MO: Jones & Bartlett Publishers; 2013.
placement, the adequacy of tourniquet placement, and a 14. American Heart Association and American Red Cross Guidelines for
formal Mangled Extremity Severity Score assessed in real time First Aid. Web-based Integrated Guidelines for First Aid. Available at:
by the clinicians evaluating the injury would have improved firstaidguidelines.heart.org. Accessed May 27, 2016.
the quality of this study. 15. Jacobs LM. The Hartford Consensus IV: a call for increased national resil-
ience. Bull Am Coll Surg. 2016;101:17–24.
16. Kragh JF Jr, Littrel ML, Jones JA, Walters TJ, Baer DG, Wade CE, Holcomb
JB. Battle casualty survival with emergency tourniquet use to stop limb
CONCLUSION bleeding. J Emerg Med. 2011;41:590–597.
17. Scerbo MH, Mumm JP, Gates K, Love JD, Wade CE, Holcomb JB, Cotton
Our study demonstrates that in patients with an a priori BA. Safety and appropriateness of tourniquets in 105 civilians. Prehosp
designated indication for tourniquet placement, waiting until Emerg Care. 2016;20(6):712–722.
arrival to the TC to control hemorrhage with a tourniquet 18. ATLS Subcommittee; American College of Surgeons’ Committee on
was associated with worsened blood pressure upon arrival to Trauma; International ATLS working group. Advanced Trauma Life
Support (ATLS®): the ninth edition. J Trauma Acute Care Surg. 2013;
the TC, increased plasma transfusion, increased transfusion
74:1363–1366.
within the first hour of arrival, and an increased risk of death 19. Savage SA, Sumislawski JJ, Zarzaur BL, Dutton WP, Croce MA, Fabian TC.
from hemorrhagic shock. The new metric to define large-volume hemorrhage: results of a prospective
study of the critical administration threshold. J Trauma Acute Care Surg.
AUTHORSHIP 2015;78(2):224–229; discussion 9–30.
The author contribution to this work is as follows: data collection—K.D.G. 20. Beekley AC, Sebesta JA, Blackbourne LH, Herbert GS, Kauvar DS, Baer
and M.H.S.; literature search—M.H.S. and B.A.C.; data analysis and DG, Walters TJ, Mullenix PS, Holcomb JB. Prehospital tourniquet use
interpretation—M.H.S. and B.A.C.; drafting of article—M.H.S.; critical in Operation Iraqi Freedom: effect on hemorrhage control and outcomes.
review—E.A.T., J.D.L., C.E.W., J.B.H., and B.A.C. J Trauma. 2008;64(2 Suppl):S28–S37; discussion S37.
21. Lakstein D, Blumenfeld A, Sokolov T, Lin G, Bssorai R, Lynn M, Ben-Abra-
DISCLOSURE ham R. Tourniquets for hemorrhage control on the battlefield: a 4-year accu-
The authors report no conflicts of interest. The authors alone are responsi- mulated experience. J Trauma. 2003;54(5 Suppl):S221–S225.
ble for the content and writing of the article. 22. Kragh JF Jr, Walters TJ, Baer DG, Fox CJ, Wade CE, Salinas J, Holcomb
Funding for M.H.S. was provided through NIH grant 5T32GM008792. JB. Practical use of emergency tourniquets to stop bleeding in major limb
trauma. J Trauma. 2008;64(2 Suppl):S38–S49; discussion S49–50.
23. Eastridge BJ, Mabry RL, Seguin P, Cantrell J, Tops T, Uribe P, Mallett O,
Zubko T, Oetjen-Gerdes L, Rasmussen TE, et al. Death on the battlefield
REFERENCES (2001–2011): implications for the future of combat casualty care. J Trauma
1. Holcomb JB, McMullin NR, Pearse L, Caruso J, Wade CE, Oetjen-Gerdes L, Acute Care Surg. 2012;73(6 Suppl 5):S431–S437.
Champion HR, Lawnick M, Farr W, Rodriguez S, et al. Causes of death 24. Sasser SM, Hunt RC, Faul M, Sugerman D, Pearson WS, Dulski T, Wald
in U.S. Special Operations Forces in the global war on terrorism: MM, Jurkovich GJ, Newgard CD, Lerner EB, et al. Guidelines for field triage
2001–2004. Ann Surg. 2007;245:986–991. of injured patients: recommendations of the National Expert Panel on Field
2. Kelly JF, Ritenour AE, McLaughlin DF, Bagg KA, Apodaca AN, Mallak CT, Triage, 2011. MMWR Recomm Rep. 2012;61(RR-1):1–20.
Pearse L, Lawnick MM, Champion HR, Wade CE, et al. Injury severity 25. Fox N, Rajani RR, Bokhari F, Chiu WC, Kerwin A, Seamon MJ, Skarupa
and causes of death from Operation Iraqi Freedom and Operation Endur- D, Frykberg E. Evaluation and management of penetrating lower extrem-
ing Freedom: 2003–2004 versus 2006. J Trauma. 2008;64:S21–S26; ity arterial trauma: an Eastern Association for the Surgery of Trauma prac-
discussion S6–7. tice management guideline. J Trauma Acute Care Surg. 2012;73(5 Suppl 4):
3. Kragh JF Jr, Walters TJ, Baer DG, Fox CJ, Wade CE, Salinas J, Holcomb JB. S315–S320.
Survival with emergency tourniquet use to stop bleeding in major limb 26. Pons PT, Jerome J, McMullen J, Manson J, Robinson J, Chapleau W. The
trauma. Ann Surg. 2009;249:1–7. Hartford Consensus on Active Shooters: implementing the continuum of
4. Dorlac WC, DeBakey ME, Holcomb JB, Fagan SP, Kwong KL, Dorlac GR, prehospital trauma response. J Emerg Med. 2015;49(6):878–885.
Schreiber MA, Persse DE, Moore FA, Mattox KL. Mortality from isolated 27. Haider AH, Piper LC, Zogg CK, Schneider EB, Orman JA, Butler FK,
civilian penetrating extremity injury. J Trauma. 2005;59:217–222. Gerhardt RT, Haut ER, Mather JP, MacKenzie EJ, et al. Military-to-civilian
5. Davis JS, Satahoo SS, Butler FK, Dermer H, Naranjo D, Julien K, Van Haren translation of battlefield innovations in operative trauma care. Surgery.
RM, Namias N, Blackbourne LH, Schulman CI. An analysis of prehospital 2015;158(6):1686–1695.
deaths: who can we save? J Trauma Acute Care Surg. 2014;77:213–218. 28. Passos E, Dingley B, Smith A, Engels PT, Ball CG, Faidi S, Nathens A, Tien
6. Smith ER, Shapiro G, Sarani B. The profile of wounding in civilian public H. Canadian Trauma Trials Collaborative. Tourniquet use for peripheral vas-
mass shooting fatalities. J Trauma Acute Care Surg. 2016;81:86–92. cular injuries in the civilian setting. Injury. 2014;45(3):573–577.
7. Husum H, Gilbert M, Wisborg T, Pillgram-Larsen J. Prehospital tourniquets: 29. Kragh JF Jr, Nam JJ, Berry KA, Mase VJ Jr, Aden JK 3rd, Walters TJ,
there should be no controversy. J Trauma. 2004;56:214–215. Dubick MA, Baer DG, Wade CE, Blackbourne LH. Transfusion for shock
8. Welling DR, Burris DG, Hutton JE, Minken SL, Rich NM. A balanced in US military war casualties with and without tourniquet use. Ann Emerg
approach to tourniquet use: lessons learned and relearned. J Am Coll Med. 2015;65(3):290–296.
Surg. 2006;203:106–115. 30. Schroll R, Smith A, McSwain NE Jr, Myers J, Rocchi K, Inaba K, Siboni S,
9. Lee C, Porter KM, Hodgetts TJ. Tourniquet use in the civilian prehospital Vercruysse GA, Ibrahim-Zada I, Sperry JL, et al. A multi-institutional anal-
setting. Emerg Med J. 2007;24:584–587. ysis of prehospital tourniquet use. J Trauma Acute Care Surg. 2015;79(1):
10. Doyle GS, Taillac PP. Tourniquets: a review of current use with proposals for 10–14; discussion 4.
expanded prehospital use. Prehosp Emerg Care. 2008;12:241–256. 31. Inaba K, Siboni S, Resnick S, Zhu J, Wong MD, Haltmeier T, Benjamin E,
11. Callaway DW, Smith ER, Cain J, Shapiro G, Burnett WT, McKay SD, Mabry Demetriades D. Tourniquet use for civilian extremity trauma. J Trauma Acute
R. Tactical emergency casualty care (TECC): guidelines for the provision of Care Surg. 2015;79(2):232–237; quiz 332–3.
prehospital trauma care in high threat environments. J Spec Oper Med. 2011; 32. Fisher AD, Callaway DW, Robertson JN, Hardwick SA, Bobko JP, Kotwal
11:104–122. RS. The Ranger First Responder program and Tactical Emergency Casualty
12. Smith R, Callaway DW. Tactical emergency casualty care. The need for & Care implementation: a whole-community approach to reducing mortality
evolution of civilian high threat medical guidelines. Jems. 2014:10–15. from active violent incidents. J Spec Oper Med. 2015;15(3):46–53.

© 2017 Wolters Kluwer Health, Inc. All rights reserved. 1171

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

 J Trauma Acute Care Surg
Scerbo et al. Volume 83, Number 6

33. Bobko J, Lai TT, Smith ER, Shapiro GL, Baldridge RT, Callaway DW. 38. Kragh JF Jr, O'Neill ML, Walters TJ, Jones JA, Baer DG, Gershman LK,
Tactical emergency casualty care pediatric appendix: novel guidelines for Wade CE, Holcomb JB. Minor morbidity with emergency tourniquet use
the care of the pediatric casualty in the high-threat, prehospital environment. to stop bleeding in severe limb trauma: research, history, and reconciling ad-
J Spec Oper Med. 2013;13(4):94–107. vocates and abolitionists. Mil Med. 2011;176(7):817–823.
34. Kragh JF Jr, Aden JK 3rd, Walters TJ, Baer DG, Nam JJ, Berry KA, Mase VJ 39. Ode G, Studnek J, Seymour R, Bosse MJ, Hsu JR. Emergency tourni-
Jr, Dubick MA, Wade CE, Blackbourne LH. In reply. Ann Emerg Med. 2015; quets for civilians: can military lessons in extremity hemorrhage be
66(3):340–341. translated? J Trauma Acute Care Surg. 2015;79(4):586–591.
35. Bulger EM, Snyder D, Schoelles K, Gotschall C, Dawson D, Lang E, Sanddal 40. Callaway DW, Robertson J, Sztajnkrycer MD. Law enforcement-applied
ND, Butler FK, Fallat M, Taillac P, et al. An evidence-based prehospital guide- tourniquets: a case series of life-saving interventions. Prehosp Emerg Care.
line for external hemorrhage control: American College of Surgeons Commit- 2015;19(2):320–327.
tee on Trauma. Prehosp Emerg Care. 2014;18(2):163–173. 41. Zietlow JM, Zietlow SP, Morris DS, Berns KS, Jenkins DH. Prehospital
36. American College of Emergency Physicians. Out-of-hospital severe hemor- use of hemostatic bandages and tourniquets: translation from military expe-
rhage control. Available at: https://www.acep.org/Physician-Resources/ rience to implementation in civilian trauma care. J Spec Oper Med. 2015;15
Policies/Policy-statements/EMS/Out-of-Hospital-Severe-Hemorrhage- (2):48–53.
Control/. Accessed May 26, 2016. 42. Fitzgibbons PG, Digiovanni C, Hares S, Akelman E. Safe tourniquet use:
37. Demetriades D, Kuncir E, Murray J, Velmahos GC, Rhee P, Chan L. Mortal- a review of the evidence. J Am Acad Orthop Surg. 2012;20(5):310–319.
ity prediction of head Abbreviated Injury Score and Glasgow Coma Scale: 43. Wilgis EF. Observations on the effects of tourniquet ischemia. J Bone
analysis of 7,764 head injuries. J Am Coll Surg. 2004;199(2):216–222. Joint Surg Am. 1971;53(7):1343–1346.

1172 © 2017 Wolters Kluwer Health, Inc. All rights reserved.

View publication stats Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.