ORIGINAL ARTICLE

Retroperitoneal Pelvic Packing in the Management of

Hemodynamically Unstable Pelvic Fractures: A Level I Trauma
Center Experience
Dora K. C. Tai, MRCS, Wing-Hong Li, FHKAM, Kin-Yan Lee, FHKAM, Mina Cheng, MRCS,
Kin-Bong Lee, FHKAM, Lap-Fai Tang, FHKAM, Albert Kwok-Hung Lai, FHKAM, Hiu-Fai Ho, FHKAM,
and Moon-Tong Cheung, FHKAM

Background: Our objective is to evaluate the mortality and outcomes of

hemodynamically unstable patients with pelvic fractures treated with a
protocol that directs the patient to either early pelvic angiography or early
retroperitoneal pelvic packing.
Method: This is a retrospective review of prospectively collected database at
a local trauma center. Hemodynamically unstable pelvic fracture patients
received treatment according to our hospital protocol during two different
time periods. Before June 2008, these patients underwent early angiography
(ANGIO group, n 13), and from June 2008 onward, these patients
underwent early pelvic packing and subsequent angiography if there was
continued hemorrhage from the pelvis (PACKING group, n 11). The
mechanism of injury, physiologic parameters, blood transfusion requirements, time to intervention, trauma scores, and mortality were recorded.
Results: Mean time to intervention in the ANGIO group was longer than that
in the PACKING group, although this was not statistically significant (139.5
minutes vs. 78.8 minutes, respectively, p 0.248). Mortality in the ANGIO
group was higher than that in the PACKING group; however, this was also
not significant (69.2% vs. 36.3%, p 0.107). After univariate analysis,
factors associated with mortality included systolic blood pressure, Glasgow
Coma Score, Injury Severity Score, Revised Trauma Score, Trauma and
Injury Severity Score, pH, and base excess. In the PACKING group, one
patient died of uncontrolled hemorrhage from a liver laceration. In the
ANGIO group, three patients died of uncontrolled hemorrhage from the
pelvic fracture.
Conclusion: Early experience in our institution suggests that early pelvic
packing with subsequent angiography if needed is as good as angiography with
embolization in treating patients with hemodynamically unstable pelvic fractures.
Key Words: Retroperitoneal pelvic packing; Pelvic fracture.
(J Trauma. 2011;71: E79 E86)

he management of hemodynamically unstable pelvic fractures is challenging. Pelvic fractures are often associated
with concomitant hemorrhage from other areas of the body

Submitted for publication August 26, 2010.

Accepted for publication December 22, 2010.
Copyright 2011 by Lippincott Williams & Wilkins
From the Departments of Surgery (D.K.C.T., W.-H.L., K.-Y.L., M.C., L.-F.T.,
M.-T.C.), Orthopaedics and Traumatology (K.-B.L.), Radiology and Imaging
(A.K.-H.L.), and Accident and Emergency (H.-F.H.), Queen Elizabeth Hospital, Hong Kong SAR, China.
Address for reprints: Dr. Dora Kai Chun Tai, Department of Surgery, Queen
Elizabeth Hospital, 30 Gascoigne Road, Kowloon, Hong Kong, China; email:
doratai323@yahoo.com.hk.
DOI: 10.1097/TA.0b013e31820cede0

such as the chest, abdominal organs, and long bones, which

further adds to the difficulty in management.
Mortality of up to 40% has been reported,1 with exsanguinating hemorrhage being the most common cause of death
in the first 24 hours after injury. Persistent hemorrhage and
blood transfusion can lead to the further development of
multiorgan failure and death. Therefore, early identification
and control of pelvic hemorrhage is essential.
Whether hemodynamically unstable patients should undergo angiography or pelvic packing has been a controversial
issue. More than 80% of hemorrhaging from pelvic fractures
originates from low-pressure venous plexi, and arterial bleeding constitutes up to 10% of the hemorrhage. The success rate
of angiography and embolization in stopping arterial bleeding
has been quoted to range from 80% to 100%1,2; however, this
procedure does nothing to address the potentially torrential
venous bleeding associated with unstable pelvic fractures.3
Retroperitoneal pelvic packing is a fast and effective procedure that helps to control venous hemorrhage and can help to
stabilize the patient before angiography.
Our objective is to evaluate the mortality and outcomes
of hemodynamically unstable patients with pelvic fractures
treated with a protocol that directs the patient to either early
pelvic angiography or early retroperitoneal pelvic packing.

PATIENTS AND METHODS

This is a retrospective review of patients who were
admitted to our hospital for hemodynamically unstable pelvic
fractures and received treatment according to our hospital
protocol during two time periods. Before June 2008, all
casualties presenting with pelvic ring injuries and hemodynamic instability (defined as persistent systolic blood pressure
90 mm Hg after receiving 2000 mL intravenous crystalloid) underwent management according to our hospital protocol for pelvic fractures (Fig. 1). Sources of hemorrhage
were sought using imaging techniques such as chest roentgenograms, pelvic roentgenograms, and diagnostic peritoneal
lavage or focused abdominal sonography for trauma (FAST).
The pelvis was bound with a pelvic binder or sheet in all
cases with suspected or radiologically confirmed pelvic fracture. Patients with hemodynamic instability and no other obvious bleeding source apart from the pelvis were sent to the
angiography suite for immediate angiography and embolization.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011

E79

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011

Tai et al.

Pelvic Fracture

Hypotension

Resuscitate with 2000ml crystalloid

Wrap Pelvis with Sheet/Binder
Trauma series (CXR, Pelvic X-ray, C-spine X-ray)
Immediate No
ca
on: On-call trauma surgeon, orthopedics
surgeon, interven
onal radiologist

FAST scan/ Diagnos
c Peritoneal Lavage

Grossly Posi
ve

Grossly Nega
ve
Sustained Response to
Ini
al Resuscita
on?

Laparotomy
+ External Fixa
on
Unstable

No

Yes

Stable
Yes
ICU

No
Angiography
+ Emboliza
on

Angiography
+ Emboliza
on

Amenable to
External Fixa
on
No

Yes

Stable
ICU+CT

External
xa
on

CXR- Chest X-ray

FAST scan- Focused Assessment with Sonography for Trauma
ICU- Intensive Care Unit
CT- Computed Tomography

Figure 1. Protocol for the management of hemodynamically unstable casualties with pelvic fractures at Queen Elizabeth Hospital, 2004 2008. ICU, intensive care unit.

All angiography was performed by dedicated interventional

radiologists. Patients with abdominal bleeding confirmed by
FAST or diagnostic peritoneal lavage were first taken to the
operating room to undergo laparotomy. The pelvic binder
was left on the patient throughout the laparotomy. External
fixation was performed after assessment by an orthopedic
surgeon. If patients remained hemodynamically unstable after
laparotomy, pelvic angiography was performed. The cases
treated with angiography before June 2008 were described as
the ANGIO group.
Because uncontrolled bleeding with massive transfusion could lead to complications such as multiorgan failure
and acute respiratory distress syndrome, the option of retroperitoneal packing was discussed in our internal mortality and
morbidity meetings and was supported by evidence from
computed tomography (CT) scans from previous patients.
This information was used as a reference to determine
whether patients would benefit from this technique. From
E80

June 2008, the protocol was altered such that patients with
pelvic fractures who were not responsive to initial resuscitation had a persistent systolic blood pressure 90 mm Hg after
receiving 2,000 mL intravenous crystalloid and who had a
negative result for FAST scan underwent retroperitoneal
pelvic packing, thus comprising the PACKING group
(Fig. 2). These patients were transferred immediately to the
operating room for placement of an external fixator followed
by retroperitoneal pelvic packing. For patients with positive
FAST scans, further surgical procedures such as laparotomy,
which were necessary for each individual patient, were subsequently performed after external fixation of the pelvis by
the on-call orthopedic surgeon and retroperitoneal pelvic
packing. Individuals in the PACKING group who developed
persistent hemorrhage after retroperitoneal packing (as manifested by persistent low blood pressure and persistent need
for blood transfusion) underwent pelvic angiography. In
cases where patients were not stable enough to be transferred
2011 Lippincott Williams & Wilkins

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011 Retroperitoneal Pelvic Packing in Management of Pelvic
Fractures

Pelvic Fracture

Hypotension

Resuscitate with 2000ml crystalloid

Wrap Pelvis with Sheet/Binder
Trauma series (CXR, Pelvic X-ray, C-spine X-ray)
Immediate No
ca
on: On-call trauma surgeon, orthopedics
surgeon, interven
onal radiologist

FAST scan/ Diagnos
c Peritoneal Lavage

Grossly Posi
ve

Grossly Nega
ve
Sustained Response to
Ini
al Resuscita
on?

Pelvic Fixa
on
Pelvic Packing
Laparotomy
Yes

No
Opera
ng Room

Unstable

CT scan

External xa
on
Pelvic packing

Angiography + emboliza
on
Ongoing transfusion
requirements or unstable
ICU
No
ICU

Yes

Angiography +
emboliza
on if
s
ll unstable

External xa
on
Pelvic packing

CXR- Chest X-ray

FAST scan- Focused Assessment with Sonography for Trauma
ICU- Intensive Care Unit
CT- Computed Tomography

Figure 2. Protocol for the management of hemodynamically unstable casualties with pelvic fractures at Queen Elizabeth Hospital, 2008 2009. ICU, intensive care unit.

to the angiography suite, interventional radiologist was called

to perform angiography and embolization in an angiogramcompatible operating room.
Throughout the study period, there was no change in
the diagnostic facilities (e.g., the CT scan machine) or subsequent routine intensive care unit admission/care. The criteria for intervention for hemodynamically unstable pelvic
fractures, such as the definition of shock, also remained
unchanged. No blood transfusion protocol was used throughout
the study period. Recombinant human factor VIIa or any other
fibrinolytics were not administered to these patients.
Medical records were reviewed for patient demographics, mechanism of injury, associated injuries, hemodynamic
2011 Lippincott Williams & Wilkins

status, Injury Severity Score (ISS), Revised Trauma Score

(RTS), Trauma and Injury Severity Score (TRISS), and blood
transfusion requirement. We aimed to compare mortality and
its associated factors during these two different periods. We
defined mortality as death that occurred within the hospital
after admission. Values for continuous variables were presented as the means SD. Differences between the two
groups were compared by the 2 test, Fishers exact test, or
the t test, as appropriate. A p value 0.05 was considered
statistically significant. Statistical analysis was performed using
SPSS software version 16.0 for Windows (SPSS, Inc., Chicago,
IL). The techniques of retroperitoneal pelvic packing adopted
throughout the entire study period are described below.
E81

Tai et al.

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011

Figure 3. An 8-cm incision is made over the lower abdomen.

Techniques of Retroperitoneal Pelvic Packing

The external fixator is placed lower than the iliac crest
to preserve the access for pelvic packing. An 8-cm midline
incision is made at the lower abdomen (Fig. 3).4 The subcutaneous and scarpas fascia as well as the linea alba are
incised.4 Beneath the fascia transversalis is the preperitoneal
space into which the hematoma expands. The peritoneum is
manually freed from the inner aspect of the osseous pubic
symphysis and the pelvic ring. The linea terminalis is followed by palpation to the sacroiliac joint, exposing the inner
aspect of the quadrilateral plate.5 The obturator nerve with
vessels is usually not searched, but if identified, it is pushed
laterally.5 The internal oblique muscle, which drapes the
inner surface of the quadrilateral plate, is left uninterrupted,
and no attempts to free the peritoneum from the sacrum are
needed. Three large radio-opaque long pieces of gauze are
placed on each side in the pelvis in the interspace between the
bony pelvic ring and the peritoneum, starting posteriorly from
the sacroiliac joint and advancing anteriorly to the retropubic
area.4 The gauze pieces are directed toward branches of the
internal iliac artery and the pelvic venous plexus situated in
the retroperitoneal space lateral to the sacrum. After completion of pelvic packing, the linea alba is closed with continuous sutures to achieve an additional tamponade effect
(Fig. 4). No wound drains are necessary.
The rationale is direct access to the bleeding retroperitoneal space. The presacral area and paravesical region are
packed. The key to this maneuver is packing of the true
pelvis, i.e., below the pelvic brim, and not the false pelvis,
i.e., above the pelvic brim.4 Because the major venous bleeding occurs in the plexi of vessels in the true pelvis, packing
above the pelvic brim has minimal tamponade effect.6 The
total time for the packing procedure can be 20 minutes and
operative blood loss is minimal.4,7 The packing is changed or
removed 24 hours to 48 hours after the injury.1,4,7

Angiography Technique
An arterial introducer is positioned via the femoral
artery at the groin. An angiogram is performed by nonselecE82

Figure 4. Postlaparotomy, external fixation, and retroperitoneal pelvic packing.

tive injection of contrast medium just above the aortic bifurcation, followed by selective injection to the branches of the
internal iliac arteries.1,8 Signs of macrovascular lesions such
as false aneurysms, missing arteries, and artery wall irregularities are detected.1,8 Vascular occlusion is performed with
gelfoam, coils, or glues, depending on the size and site of the
lesion. In cases of major hemodynamic instability that persist
despite failure to identify any arterial lesion, nonselective
bilateral embolization of the internal iliac arteries with gelfoam is performed.1,8 At the end of the procedure, angiography can be used to check that the embolized areas are
occluded and to ensure that there are no other sources of
bleeding.

RESULTS
From June 2007 to December 2009, among 632 trauma
patients with an ISS 15 admitted to our hospital, 24 (3.8%)
patients with pelvic fractures with hemodynamic instability
were identified and included in our study. In total, 15 male
patients and 9 female patients were involved. The mean age
was 47.7 years (median, 45 years; range, 18 84 years). All
patients sustained blunt trauma; 13 patients sustained road
traffic accidents and 11 patients sustained falls from substantial heights. The median ISS was 41.2. The overall mortality
rate was 54.2%.
Patient demographics, severity of injury, blood product
transfusion, and mortality were compared in these two periods (i.e., ANGIO group vs. PACKING group); the results are
shown in Table 1. There were 13 patients in the ANGIO
group and 11 patients in the PACKING group. There were no
significant differences noted between groups in terms of age,
blood transfusion at the accident and emergency department
(AED), time spent at the AED, systolic blood pressure on
arrival, Glasgow Coma Score (GCS), ISS, RTS, TRISS, initial
blood results (including hemoglobin level, platelet count, creatinine, international normalized ratio, pH, and base excess), or
transfusion of blood products within the first 24 hours after
angiography or packing (p 0.05). In the PACKING group,
2011 Lippincott Williams & Wilkins

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011 Retroperitoneal Pelvic Packing in Management of Pelvic
Fractures

TABLE 1. Overview of Age, Injury Severity, Physiological

Parameters, and Mortality in the Two Study Groups
(Mean SD) (N 24)
ANGIO
(N 13)

PACKING
(N 11)

Age (yr)
44.8 24.7.7
51.2 19.0
Blood transfusion in AED (units)
3.2 2.3
2.0 1.2
Combination of treatment
0/13
5/11
Time spent in AED (min)
61.9 33.5
69.7 15.6
SBP on arrival (mm Hg)
61.2 52.3
99.0 40.9
GCS
11.2 4.5
13.2 3.5
ISS
42.3 18.8
40.0 12.5
RTS
5.5 2.5
6.9 2.0
TRISS
0.5 0.4
0.7 0.3
Hemoglobin (g/dL)
11.6 1.6
11.8 2.3
209.4 48.3
188.5 76.4
Platelet (109/L)
INR
1.3 0.2
1.2 0.1
Blood creatinine (mol/L)
93.5 29.4
84.4 27.0
Blood pH
7.1 0.2
7.2 0.2
Blood base excess (mmol/L)
14.8 9.0
10.8 6.2
Transfusion of packed cells within
5.0 4.4
9.0 8.0
24 h after the procedure (units)
Transfusion of platelet concentrate
5.3 3.5
8.4 7.8
within 24 h after the procedure
(units)
Transfusion of FFP within 24 h
6.7 5.5
8.8 8.8
after the procedure (units)
Time to intervention (min)
139.5 95.0
78.8 23.5
Mortality
9/13 (69.2%)
4/11 (36.3%)

TABLE 2.

Fracture Type

p
0.490
0.486
0.002
0.711
0.065
0.239
0.732
0.147
0.056
0.780
0.424
0.243
0.442
0.366
0.231
0.243
0.425

0.582
0.248
0.107

SBP, systolic blood pressure; INR, international normalized ratio; FFP, fresh frozen
plasma.

5 of the 11 patients underwent angiography after pelvic

packing; however, in the ANGIO group, none of the patients
underwent retroperitoneal packing afterward (p 0.002).
The mean time to intervention for the ANGIO group was
longer than that observed for the PACKING group, although
this was not statistically significant (139.5 minutes 95.0
minutes vs. 78.8 minutes 23.5 minutes, respectively, p
0.248). Mortality in the ANGIO group was higher than that in
the PACKING group; however, this was also not significant
(69.2% vs. 36.3%, p 0.107). According to Giannoudis and
Pape,7 there were several fracture types (as defined using the
Young and Burgess classification) that were associated with
major ligament disruption and bleeding. These fracture types
included lateral compression type III; anteroposterior compression types II and III; vertical shear fracture types I, II, and
III; and combined mechanism fractures. We defined major
fractures as those that included these fracture types. Minor
fractures included lateral compression types I, IIa, and IIb and
anteroposterior compression fracture type I. Our results
showed that there was no difference in terms of severity of
fracture according to the Young and Burgess classification
(p 0.423) (Table 2).
Univariate analysis was performed on patient parameters to compare survivors and nonsurvivors; the results are
shown in Table 3. The mean systolic blood pressure on
arrival among the nonsurvivors (55.5 mm Hg 51.7 mm Hg)
2011 Lippincott Williams & Wilkins

Pelvic Fracture Type in the 2 Study Groups

ANGIO

PACKING

Total

8
3
11

7
6
13

15
9
24

Major
Minor
Total

p 0.423.
Major fractures include lateral compression type III; anteroposterior compression
types II and III; vertical shear fracture types I, II, and III; and combined mechanism
fractures. Minor fractures include lateral compression types I, IIa, and IIb and anteroposterior compression fractures type I.

TABLE 3. Overview of Age, Injury Severity, Physiological

Parameters Between Survivors and Nonsurvivors (Mean SD)
Parameters
Age (yr)
SBP (mm Hg)
Lowest SBP (mm Hg)
Time spent in AED (min)
GCS
ISS
RTS
TRISS
Hemoglobin (g/dL)
Platelet (109/L)
INR
Blood creatinine (mol/L)
Blood pH
Blood base excess (mmol/L)
Transfusion of packed cells
within 24 h after the
procedure (units)
Transfusion of platelet
concentrate within 24 h
after the procedure (units)
Transfusion of FFP within
24 h after the procedure
(units)
Time to intervention (min)

Survivors
(Mean SD)

Nonsurvivors
(Mean SD)

51.9 18.7
110.8 24.9
87.2 14.5
83.8 82.9
14.5 1.0
29.2 12.2
7.5 0.4
0.9 0.1
11.8 2.0
203.6 84.4
1.1 0.2
76.4 24.7
7.3 0.9
7.9 4.0
4.2 3.2

44.7 24.4
55.5 51.7
55.7 42.2
54.1 15.7
10.4 4.7
49.9 12.5
5.2 2.7
0.4 0.3
11.7 1.9
197.1 43.5
1.3 0.2
98.5 27.5
7.1 0.2
16.6 8.2
7.7 5.5

0.443
0.005
0.034
0.202
0.012
0.001
0.017
0.000
0.922
0.806
0.108
0.055
0.019
0.006
0.316

4.8 2.5

7.3 5.5

0.512

4.4 3.2

9.1 4.8

0.212

105.6 75.5

75.1 52.7

0.256

SBP, systolic blood pressure; INR, international normalized ratio; FFP, fresh frozen
plasma.

was significantly lower than that for the survivors (110.8 mm

Hg 24.9 mm Hg, p 0.005). The lowest systolic blood
pressure recorded was also significantly lower in the nonsurvivors (55.7 mm Hg 42.2 mm Hg) than that in the
survivors (87.2 mm Hg 14.5 mm Hg, p 0.034). The GCS
(nonsurvivors 10.4 4.7 vs. survivors 14.5 1.0, p
0.012), ISS (nonsurvivors 49.9 12.5 vs. survivors 29.2
12.2, p 0.001), RTS (nonsurvivors 5.2 2.7 vs. survivors
7.5 0.4, p 0.017), and TRISS (nonsurvivors 0.4 0.3 vs.
survivors 0.9 0.1, p 0.000) were all significantly poorer
for the patients who died than for the patients who survived.
The nonsurvivors were found to have significantly lower
blood pH levels (7.1 0.2 vs. 7.3 0.9, p 0.019) and base
excess (16.6 8.2 vs. 7.9 4.0, p 0.006) than the
survivors. However, after further multivariate analysis, none
E83

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011

Tai et al.

TABLE 4.

Characteristics of Non-survivors

Age

ISS

Associated Injuries

Other Operations Performed

M/25

57

Multiple rib fracture

M/36

66

M/35

43

M/47

50

M/55

38

F/84
M/22

33
50

F/68
M/20

66
66

M/46

50

M/82
M/26
M/72

45
50
59

Liver laceration, fracture right

humerus
Fracture sternum; fracture T8, L1
Exploratory laparotomy
vertebral body; fracture right
transverse process of L1, L5;
fracture left acetabulum
Capsular splenic tear; liver lacerations; Exploratory laparotomy; packing of
complete rupture of membranous
liver and spleen
urethra; fracture bilateral tibia and
fibula
Fracture right femur, tibia; fracture left Exploratory laparotomy
tibia; fracture left fourth rib;
fracture T7 vertebra; left kidney
laceration
Exploratory laparotomy
Liver laceration; fracture, T12 to L2
Exploratory laparotomy with liver
vertebral bodies; fracture, right
packing
maxilla; fracture, anterior arch of
C1
SAH
Right subdural hematoma, SAH;
External ventricular drainage
fracture, right occipital skull;
fracture, L1 and L5 vertebrae
SAH fracture, transverse process of
External ventricular drainage
left L1-5 and right L2-3
ICH, SAH spleen laceration
Splenectomy
SAH fracture, T12, L1, L2 vertebrae
SAH, ICH fracture; right occipital
External ventricular drainage
bone fracture; C5, 6, 7 spinous
processes

Chest drain insertion, exploratory

laparotomy
Exploratory laparotomy

Time from
Intervention
to Death (d)

Intervention

Cause of Death

Angiography Uncontrolled hemorrhage

Packing

Uncontrolled hemorrhage from

liver laceration
Angiography Uncontrolled hemorrhage

Packing

Angiography DIC, ARF, ARDS

23
1

Packing
Packing

DIC, MOF

ACS, sepsis
DIC, ARF, ARDS

4
1

Angiography MOF
Angiography MOF, DIC

Angiography DIC, MOF

116
1
1

Angiography Sepsis, ARF

Angiography Uncontrolled hemorrhage
Angiography DIC, MOF

SAH, subarachnoid hemorrhage; ICH, intracerebral hemorrhage; DIC, disseminated intravascular coagulopathy; MOF, multiorgan failure; ARF, acute renal failure; ARDS, acute
respiratory distress syndrome; ACS, acute coronary syndrome.

of the parameters showed any statistically significant difference (p 0.05).

Table 4 shows the characteristics of the nonsurvivors.
Most of the patients died due to disseminated intravascular
coagulopathy and multiorgan failure. Of the four patients who
underwent pelvic packing, one died of uncontrolled hemorrhage
from a liver laceration within 24 hours of the intervention. Of the
nine patients who underwent angiography, three patients died of
uncontrolled hemorrhage from the pelvic fracture. All the three
patients died within 24 hours of intervention.

DISCUSSION
Pelvic fractures continue to pose a challenge in terms of
mortality and morbidity and are indicators of severe trauma.
Hemodynamically unstable pelvic fractures have a mortality
rate of up to 40% to 60%,6 with exsanguinating hemorrhage
being the major cause of death in the first 24 hours after
injury and multiorgan failure being the most common cause
of death thereafter.1 To add to the difficulty of management,
pelvic fractures are often associated with severe injuries to
other organs. This is because high-energy impact is needed to
E84

disrupt the pelvic ring, and these high-energy transfers are

absorbed by the rest of the body.6 Up to 90% of patients with
unstable pelvic fractures have associated injuries, and 50% of
patients have sources of major hemorrhage other than pelvic
fractures.1 In this study, the trauma scores including ISS,
RTS, and TRISS for both the ANGIO and PACKING groups
indicate that our patients sustained severe injuries.
Two different modalities are currently suggested to stop
hemorrhage in unstable pelvic fractures. Angiography and
embolization are used to control arterial bleeding, and pelvic
packing is used to control venous bleeding. There is still
controversy about the use of these modalities.
Angiography and embolization of internal iliac arteries
for pelvic fractures were first reported in 1972.1 The main
indications for angiography and embolization are arterial
extravasation of contrast as seen on CT and persistent hemodynamic instability after fluid or blood product resuscitation.
Early angiography and embolization have been recommended
by many authors to improve patient outcomes.1 Successful
embolization rates have been reported to range from 80% to
100%.1,2 However, even with successful embolization, pa 2011 Lippincott Williams & Wilkins

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011 Retroperitoneal Pelvic Packing in Management of Pelvic
Fractures

tient mortality remains high. Many authors have reported a

mortality of about 50% despite successful embolization to
control arterial bleeding.1,9 Results from anatomic and clinical studies have shown that the cause of death in the majority
of pelvic fracture patients is due to hemorrhage without major
arterial injury.1,10 12 Therefore, this implies that angiography
and embolization cannot always stop the hemorrhage arising
from pelvic fractures. In addition, angiography and embolization can be time consuming, and simultaneous treatment of
other associated injuries during the procedure is delayed. The
overall time for performance of embolization has been reported to range from 50 minutes to 5.5 hours.1,5,13,14 In this
study, the mortality in the ANGIO group was 69.2% compared with 36.3% in the PACKING group (p 0.107), with
a trend toward significance.
Pelvic packing was first used as a damage-control
surgery for pelvic fractures in the 1960s. In the early stages,
pelvic packing was performed transabdominally after exploratory laparotomy. However, incision of the intact peritoneum
and disruption of the pelvic hematoma can disrupt the tamponade effect of the retroperitoneal space,1 leading to aggravation of
the pelvic hemorrhage. In addition, there was an increased risk
of infection in the pelvic hematoma after transabdominal pelvic
packing.1 Finally, packing was often performed as a last resort
late in the course of resuscitation, and patients were often
already coagulopathic. Therefore, the results were poor and
transabdominal packing was largely abandoned.
The technique for pelvic packing was later refined so
that packing was performed in the retroperitoneal space and
the peritoneum was left intact. This procedure was quick and
easy to perform, operative blood loss was minimal, and this
method helped to reduce unnecessary angiography.1 Cothren
et al. compared two different management protocols, one
advocating early pelvic packing and the other advocating
early angiography. Matched hemodynamically unstable cases
of pelvic fractures were compared, and there were no deaths
due to acute blood loss in the group of patients treated with
pelvic packing. In addition, the time to intervention was
significantly shorter in the packing group. Only 4 of 24
patients (16.7%), the nonresponders, required subsequent
embolization.15 Totterman et al.5 found that patients demonstrated a significant increase in systolic blood pressure immediately after pelvic packing.
During initial resuscitation, it is difficult to accurately
identify the major source of pelvic hemorrhage. Arterial
bleeding accounts for 10% to 15% of the hemorrhage,
whereas the primary source of bleeding is posterior pelvic
venous plexus or fractured bone surfaces.1,7,8 Angiography
and embolization are ineffective for the control of bleeding
from veins or fractured cancellous bone surfaces. Therefore,
if angiography fails to reveal a significant source of bleeding,
time has been wasted, i.e., time during which other treatment
interventions could have been performed. However, venous
bleeding and bleeding from fractured bone always accompany
pelvic fractures and can cause exsanguination. Therefore, it is
reasonable that pelvic packing should be the first-line treatment
for hemodynamically unstable patients with pelvic fractures.1
2011 Lippincott Williams & Wilkins

Thus, since June 2008, we have revised our hospital protocol so

that early pelvic packing is advocated.
In cases where the patient remains hemodynamically
unstable despite pelvic packing, arterial bleeding should be
suspected and angiography may be required. Pelvic packing
should be considered a quick, effective tool to reduce venous
bleeding while waiting for the angiographic team to set up. It can
help to rapidly control bleeding and reduce the need for blood
transfusion, which can ultimately cause coagulopathy. Angiography can then be supplemented with no significant time
delay between the two procedures, as angiography requires
an average of 2 hours to set up. In our institution, the
equipment for angiography can be set up within the operating
room, which helps to further reduce the risk associated with
transferring the unstable patient from the operating room to
the angiography suite.
Our findings indicate that the ANGIO group and the
PACKING group are quite comparable, as there is no significant
difference between the two groups regarding age, blood transfusion, time spent in AED, systolic blood pressure on arrival,
trauma scores, or transfusion of blood products within 24 hours
after the procedure. The time to arrange for pelvic packing was
shorter than the time taken to arrange for angiography
(PACKING group 78.8 minutes 23.5 minutes vs. ANGIO
group 139.5 minutes 95.0 minutes, p 0.248). However,
this difference was not statistically significant (Table 1).
Five of 11 patients in the PACKING group underwent
angiography and embolization after retroperitoneal packing,
but none of the patients in the ANGIO group underwent
retroperitoneal packing. The mortality rate for the PACKING
group was 36.3% compared with 69.2% in the ANGIO group.
Although there is no statistically significant difference in
mortality between the two groups, there was a trend toward
better results in the PACKING group. Therefore, retroperitoneal packing combined with angiography and embolization
may yield better results (Table 1).
It is now advocated that fracture patterns are inconsistent in predicting arterial bleeding and that exsanguinating
hemorrhage can occur in all fracture patterns.1,4,5,16 In our
study, a total of eight patients in the ANGIO group and seven
patients in the PACKING group sustained major pelvic fractures, and a total of three patients in the ANGIO group and six
patients in the PACKING group sustained minor pelvic
fractures. This supports the notion that even minor pelvic
fractures can cause massive hemorrhage, and we should not
underestimate the need for aggressive management for such
patients (Table 2).
When looking at the results in Table 3, it is clear that
the trauma scores (including GCS, ISS, RTS, and TRISS) for
the nonsurvivors were significantly worse than that for the
survivors. This shows that the nonsurvivors sustained more
severe injuries than the survivors and probably accounts for
the significantly lower systolic blood pressure on arrival, pH,
and base excess levels. As these patients generally presented
with a more critical condition, it is understandable that these
are the patients who failed to survive. This is further supported by data in Table 4, which show that all the nonsurvivors sustained multiple injuries to organs located outside the
E85

The Journal of TRAUMA Injury, Infection, and Critical Care Volume 71, Number 4, October 2011

Tai et al.

pelvis. An important point to note is that of the nonsurvivors

who underwent pelvic packing, only one died of uncontrolled
hemorrhage from a liver laceration. In contrast, among the
nonsurvivors who underwent angiography, three died of uncontrolled hemorrhage from the pelvic fracture. Overall,
these results imply that pelvic packing with subsequent angiography if needed may be just as effective if not better than
angiography for controlling bleeding.
The major limitation of our study is that the sample size
may be to too small to reveal a genuine difference, should
there be one. We are considering further modifications to our
existing protocol to try to improve our results. For patients
who are too hemodynamically unstable to undergo CT, we
are considering the implementation of retroperitoneal packing
and subsequent angiography regardless of whether the patient
is stabilized after packing; this would allow us to confirm the
absence of arterial bleeding before leaving the operating
room. However, we will have to consider the increased
resources necessary for such modifications.
In conclusion, the presence of exsanguinating hemorrhage in patients with pelvic ring disruption represents a
complex and difficult challenge to the trauma team. It demands aggressive yet balanced surgical, orthopedic, and angiographic management as soon as the patient is admitted to
the emergency department. Here, we describe our initial
experience with the use of pelvic packing in patients with
hemodynamically unstable pelvic fractures. Our initial results
seem promising, and future prospective studies are underway
to confirm that by integrating this rapid and easily learned
technique, we can further reduce the mortality of patients
with pelvic fracture in the future.
ACKNOWLEDGMENTS
We thank Ms. Annice Chang, Trauma Nurse, Department of Accident and Emergency, Queen Elizabeth Hospital,
for her support.
REFERENCES
1. White CE, Hsu JR, Holcomb JB. Hemodynamically unstable pelvic
fractures. Injury. 2009;40:10231030.

E86

2. Papakostidis C, Giannoudis PV. Pelvic ring injuries with hemodynamic

instability: efficacy of pelvic packing, a systematic review. Injury.
2009;40(Suppl 4):S53S61.
3. Osborn PM, Smith WR, Moore EE, et al. Direct retroperitoneal pelvic
packing versus pelvic angiography: a comparison of two management
protocols for hemodynamically unstable pelvic fractures. Injury. 2009;
40:54 60.
4. Smith WR, Moore EE, Osborn P, et al. Retroperitoneal packing as a
resuscitation technique for hemodynamically unstable patients with
pelvic fractures: report of two representative cases and a description of
technique. J Trauma. 2005;59:1510 1514.
5. Totterman A, Madsen JE, Skaga NO, Rise O. Extraperitoneal pelvic
packing: a salvage procedure to control massive traumatic pelvic hemorrhage. J Trauma. 2007;62:843 852.
6. Suzuki T, Smith WR, Moore EE. Pelvic packing or angiography:
competitive or complementary? Injury. 2009;40:343353.
7. Giannoudis PV, Pape HC. Damage control orthopedics in unstable
pelvic ring injuries. Injury. 2004;35:671 677.
8. Geeraerts T, Chhor V, Cheisson G, et al. Clinical review: initial management of blunt pelvic trauma patients with hemodynamic instability.
Crit Care. 2007;11:204 212.
9. Heetveld MJ, Harris I, Schlaphoff G, Sugrue M. Guidelines for the
management of hemodynamically unstable pelvic fracture patients. ANZ
J Surg. 2004;74:520 529.
10. Huittinen VM, Slatis P. Postmortem angiography and dissection
of the hypogastric artery in pelvic fractures. Surgery. 1973;73:
454 462.
11. Kataoka Y, Maekawa K, Nishimaki H, Yamamoto S, Soma K. Iliac vein
injuries in hemodynamically unstable patients with pelvic fracture
caused by blunt trauma. J Trauma. 2005;58:704 710.
12. Papadopoulos IN, Knakaris N, Bonovas S, et al. Auditing 655 fatalities
with pelvic fractures by autopsy as a basis to evaluate trauma care. J Am
Coll Surg. 2006;203:30 43.
13. Agolini SF, Sah K, Jaffe J, Newcomb J, Rhodes M, Reed JF III. Arterial
embolization is a rapid and effective technique for controlling pelvic
fracture hemorrhage. J Trauma. 1997;43:395399.
14. Cook RE, Keating JF, Gillespie I. The role of angiography in the
management of hemorrhage from major fractures of the pelvis. J Bone
Joint Surg Br. 2002;84:178 182.
15. Cothren CC, Osborn PM, Moore EE, Morgan SJ, Johnson JL, Smith
WR. Preperitoneal pelvic packing for hemodynamically unstable pelvic
fractures: a paradigm shift. J Trauma. 2007;62:834 842.
16. Sarin EL, Moore JB, Moore EE, et al. Pelvic fracture pattern does not
always predict the need for urgent embolization. J Trauma. 2005;58:
973977.

2011 Lippincott Williams & Wilkins