You are on page 1of 6

ORIGINAL ARTICLE

Management of pelvic ring fracture patients with a pelvic blush


on early computed tomography

Diederik O.F. Verbeek, MD, Ijsbrand A.J. Zijlstra, MD, Christaan van der Leij, MD,
Kornelis J. Ponsen, MD, PhD, Otto M. van Delden, MD, PhD,
and J. Carel Goslings, MD, PhD, Amsterdam, the Netherlands

BACKGROUND: The sliding computed tomographic (CT) scanner in our trauma resuscitation room can be used early in the assessment of
pelvic ring fracture patients. We determined the association between the presence of a pelvic blush on CT scan and the need
for pelvic hemorrhage control (PHC). We hypothesized that many pelvic blushes found early in the resuscitation phase can be
safely managed without intervention.
METHODS: Contrast-enhanced CT scans of pelvic ring fracture (pelvic ring disruption) patients admitted from January 1, 2004, to
June 31, 2012, were reviewed for the presence of a pelvic blush. PHC was dened as requiring a surgical or radiologic
intervention for pelvic bleeding. A subanalysis was performed in isolated pelvic fracture/ blush patients (absence of a
major nonpelvic bleeding source).
RESULTS: Overall, 68 (42%) of 162 pelvic ring fracture patients and 53 (40%) of 134 isolated pelvic fracture patients had a pelvic blush.
Of those 32 (47%) and 27 (51%) patients, respectively, required PHC. In the absence of a pelvic blush, 87 (93%) of 94 of all
and 77 (95%) of 81 of isolated pelvic fracture patients did not require PHC. Of all patients with a pelvic blush and of iso-
lated pelvic blush, those with PHC had a higher Injury Severity Score (ISS) (p = 0.01 and p = 0.05), base decit (p = 0.03 and
p = 0.01), as well as 24-hour and any packed red blood cells requirement (p G0.001 and p = 0.05; p G0.001 and p = 0.02).
In isolated pelvic blush patients, there was a trend toward a higher hospital and hemorrhage-related mortality in patients with
PHC (p = 0.06 and p = 0.06).
CONCLUSION: In pelvic ring fracture patients, a pelvic blush on early contrast-enhanced CT is a frequent nding. Many patients with
(particularly isolated) pelvic blushes have stable vital signs and can be managed without surgical or radiologic PHC. The need
for an intervention for a pelvic blush seems to be determined by the presence of clinical signs of ongoing bleeding. (J Trauma
Acute Care Surg. 2014;76: 374Y379. Copyright * 2014 by Lippincott Williams & Wilkins)
LEVEL OF EVIDENCE: Therapeutic study, level IV. Prognostic/epidemiologic study, level III.
KEY WORDS: Pelvic fracture; hemorrhage; contrast extravasation; angiographic embolization; pelvic packing.

I n recent years, the role of the computed tomographic (CT)


scan in the evaluation of trauma patients has changed con-
siderably owing to advances in CT technology and the intro-
use the CT scan as an adjunct to the primary survey without
interrupting ongoing resuscitation.5,6
In patients with a major pelvic fracture, the presence of a
duction of multidetector CT scanners. Contemporary CT pelvic contrast extravasation (or blush) on CT scan is widely
scanners are able to rapidly generate detailed images. This used as an indicator for signicant arterial bleeding that may
feature provides the opportunity to use the CT scanner soon after require (surgical or radiologic) pelvic hemorrhage control.7Y12
patient arrival and may shorten the diagnostic workup and Depending on hospital resources and patient characteristics,
time to denitive treatment.1Y3 Although there is no conclusive either pelvic angiographic embolization or alternatively pelvic
evidence at present, it has been reported that early whole-body packing (or ligation of pelvic arteries) is the preferred method
CT scanning may increase the probability of survival in trauma for immediate hemorrhage control.11,13Y18
patients.4 In most institutions, obtaining a CT scan requires Considering the increasing and earlier use of the CT scan
potentially hazardous patient transport to the radiology depart- and advances in imaging resolution, it may be speculated that
ment. In 2004 a multidetector CT scanner was placed in the at present, more pelvic blushes are detected. Potentially, some
trauma resuscitation room of our institution. It enabled us to of these pelvic blushes do not require hemorrhage control. In
pelvic ring fracture patients both pelvic and nonpelvic (e.g.,
abdominal, thoracic, or extremity) injuries can cause signi-
Submitted: July 14, 2013, Revised: October 18, 2013, Accepted: October 18, 2013. cant bleeding. In determining the clinical outcome specically
From the Trauma Unit, Department of Surgery (D.O.F.V., K.J.P., J.C.G.) and De- related to the presence of a pelvic blush, it is important to
partment of Radiology (I.A.J.Z., C.V.D.L., O.M.V.D.). Academic Medical Center
(AMC), University of Amsterdam (UvA), Amsterdam, the Netherlands.
consider these other sources of hemorrhage as well.
K.J.P. is now with Medisch Centrum Alkmaar (MCA) Alkmaar, the Netherlands. In this study, we examined the management and outcome
Address for reprints: Diederik Verbeek, MD, Trauma Unit, Department of Surgery, of pelvic ring fracture patients who had a contrast-enhanced
Academic Medical Center, Meibergdreef 9, 1100 DD Amsterdam, the Netherlands; CT scan.
email: d.o.verbeek@amc.nl.
We determined the association between the presence of
DOI: 10.1097/TA.0000000000000094 a pelvic blush and the need for pelvic hemorrhage control in all
J Trauma Acute Care Surg
374 Volume 76, Number 2

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg
Volume 76, Number 2 Verbeek et al.

pelvic ring fracture patients as well as in a subgroup of iso- bleeding (clinical signs of hypovolemic shock with a continuing
lated pelvic fracture patients (without a major nonpelvic transfusion requirement) were considered for radiologic hem-
bleeding source). We hypothesized that many pelvic blushes orrhage control. Digital subtraction angiography was performed
found early in the resuscitation phase can be safely managed using Philips V3000 (the Netherlands) equipment. Patients with
without intervention. an arterial injury on pelvic angiography received pelvic angio-
graphic embolization. Embolization was performed using a va-
riety of platinum embolization coils, Gelatin sponge particles
PATIENTS AND METHODS (Gelfoam, Upjohn, Kalamazoo, MI) or a combination of both.
All adult patients with a high-energy major pelvic frac- Pelvic blush patients with refractory hemorrhagic shock
ture admitted to the trauma resuscitation room of our aca- (ATLS nonresponders) had immediate surgical hemorrhage
demic Level 1 trauma center between January 1, 2004, and control by pelvic packing or (occasionally) surgical ligation
June 31, 2012, were identied from our prospective trauma of pelvic arteries (followed by pelvic angiography if needed).
registry and the hospitals DRG International Classication of Intraperitoneal pelvic packing was generally performed in pa-
DiseasesV9th Rev. database. Patients who had an intravenous tients who required a laparotomy and preperitoneal packing in
contrast-enhanced CT scan were considered for inclusion. A patients who did not require a laparotomy (rarely at the time of
pelvic ring fracture was dened as a disruption of the pelvic study). Surgical hemorrhage control was also performed in
ring in at least two places. Patients with a single break of the patients with signs of pelvic hemorrhage at immediate lapa-
pelvic ring (i.e., single acetabular, iliac wing, or pubic rami rotomy for abdominal hemorrhage. A pelvic external xator
fractures), transfer patients, and patients declared dead on was placed before laparotomy for unstable fracture patterns. The
arrival were excluded from the analysis. overall patient management was at the trauma surgeons dis-
Medical records were reviewed for age, sex, mechanism cretion in conjunction with the anesthesiologist and interven-
of injury, Injury Severity Score (ISS), systolic blood pres- tional radiologist on call.
sure (SBP) and base decit on arrival, packed red blood cell For the purpose of the study, all CT scans were reviewed
transfusion within 24 hours after arrival, and (hospital and for the presence of a pelvic blush (focal area of high-contrast
hemorrhage-related) mortality. density) by two senior radiology residents and a board-
Pelvic fractures were classied using the Young and certied interventional radiologist in consensus. All pelvic
Burgess classication; major ligamentous disruption (MLD) was angiographies were independently reviewed by a board-certied
dened as anteroposterior Type II and III, lateral compression interventional radiologist for the presence of arterial injury
Type III, vertical shear, and combined mechanism.19Y21 Opera- (contrast extravasation, cutoff, or pseudoaneurysm).
tive and radiology reports were examined for surgical and/or
Denitions
radiologic interventions performed to control pelvic or non-
pelvic bleeding sources. The need for pelvic hemorrhage control was dened as
having received a surgical (pelvic packing or surgical ligation)
Patient Management and Imaging and/or radiologic (angiographic embolization) intervention for
Initial patient assessment followed our institutional pro- pelvic bleeding. A major nonpelvic bleeding source was de-
tocol and advanced trauma life support (ATLS) principles. ned as an abdominal, thoracic, or extremity injury that re-
According to the local imaging protocol, all high-energy trauma quired surgical or radiologic hemorrhage control. A patient
patients receive a chest and pelvic radiography as well as a was considered to have an isolated pelvic fracture or blush in
FAST within 5 minutes of arrival. Pelvic ring fracture patients the absence of a major nonpelvic bleeding source.
also have an abdominopelvic intravenous contrast-enhanced CT
scan. The location of the multislice CT scanner (SOMATOM Statistical Analysis
Sensation 4 and 64 [from 2008], Siemens Medical Systems, Continuous variables are presented as median values
Erlangen, Germany) in the trauma resuscitation room enables us with interquartile ranges (IQRs) and are compared using
to safely perform a rapid CT scan in all patients except for those Mann-Whitney U-test or as mean values with SDs and compared
with refractory hemorrhagic shock despite adequate uid re- with independent t test, depending on data distribution. Cate-
suscitation (ATLS nonresponders). On arrival, patients are gorical values were calculated as percentage of frequency of
placed on the trauma room table, which is radiolucent and also occurrence. Discrete variables were compared using Fishers
acts as the CT table. The CT scanner is able to slide over the exact analyses. Statistical signicance was declared at the
patient without interrupting ongoing resuscitation.5,6 Images are 0.05 level.
viewed in stacked mode using picture archiving and communi- Two-by-two contingency tables were constructed to cal-
cation system. culate the diagnostic indices (sensitivity, specicity, as well as
After the CT scan is obtained, the images are immedi- positive and negative predictive values) for the presence of a
ately reviewed by the attending trauma surgeon and radiologist. pelvic blush on CT scan and the need for pelvic hemorrhage
An interventional radiologist is available on a 24-hour basis. control in all patients as well as in the subgroup of isolated
The decision to proceed for surgical or radiologic hemorrhage pelvic fracture patients. Results are presented with 95% con-
control was typically based on the patients hemodynamic dence intervals.
status and presence of associated injuries. In general, patients All data management and statistical analysis were per-
with a pelvic blush received pelvic angiography on a liberal formed using Statistical Package for the Social Science
basis. Particularly, pelvic blush patients with signs of ongoing (SPSS\, IBM\, Armonk, New York).

* 2014 Lippincott Williams & Wilkins 375

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg
Verbeek et al. Volume 76, Number 2

RESULTS The sensitivity, specicity, as well as positive and negative


predictive values of a pelvic blush for the need for pelvic hem-
Overall, 172 pelvic ring fracture patients were identied, orrhage control in all patients were 82% (66Y92), 71% (62Y78),
and 166 (97%) had a contrast-enhanced CT scan. Four patients 47% (35Y59), and 93% (85Y97), respectively (Table 3).
were excluded owing to the nondiagnostic quality of the images
or an incomplete CT examination. Pelvic Hemorrhage Control in Isolated Pelvic
In total, 162 pelvic ring fracture patients were included in Fracture Patients
the study. A pelvic blush was present in 68 patients (42%) and In total, 28 patients had a major nonpelvic bleeding
absent in 94 patients (58%). source. Twelve patients required surgical, and 13 required ra-
The characteristics, management, and outcome of all diologic hemorrhage control for abdominal injuries, 2 patients
patients by presence of a pelvic blush on CT scan are presented had a thoracotomy for thoracic injuries, and 1 patient had sur-
in Table 1. Patients with a pelvic blush had a signicantly higher gical hemorrhage control for an extensive extremity injury. In
age and ISS as well as a lower SBP and higher base decit on the remaining 134 isolated pelvic fracture patients, the pelvic
arrival. Furthermore, pelvic blush patients had more complex fracture was the only potential source of major hemorrhage.
pelvic fractures (with MLD) and required more pelvic exter- A pelvic blush was present in 53 patients (40%) (Table 4).
nal xators and pelvic hemorrhage control. The sex, method of Of those, 26 patients (49%) did not have pelvic hemorrhage
injury, and need for nonpelvic hemorrhage control was similar control and 27 patients (51%) had pelvic hemorrhage control.
in both groups. Patients with a pelvic blush also had a signi- Three patients had pelvic packing at laparotomy, and 24 had
cantly higher 24-hour and any packed red blood cells require- pelvic angiographic embolization (4 of whom had both in-
ment and a higher hospital and hemorrhage-related mortality. terventions). Of the 26 patients with a pelvic blush but no pelvic
hemorrhage control, 9 had a negative pelvic angiography for
Pelvic Hemorrhage Control in All Patients arterial injury and 17 had no intervention for hemorrhage control.
Of 68 patients with a pelvic blush, 36 (53%) had no A pelvic blush was absent in 81 patients (60%). Of
pelvic hemorrhage control and 32 patients (47%) had pelvic those, 77 patients (95%) had no pelvic hemorrhage control and
hemorrhage control (Table 2). Five patients had pelvic pack- 4 (5%) did have pelvic hemorrhage control. One patient had
ing at laparotomy, and 27 had pelvic angiographic emboliza- pelvic packing at laparotomy, and three patients had pelvic
tion (6 of those had both interventions). Of the 36 patients with angiographic embolization for clinical signs of active bleeding.
a pelvic blush but no pelvic hemorrhage control, 10 patients The sensitivity, specicity, as well as positive and neg-
(hemodynamically) stabilized after hemorrhage control for ative predictive values of a pelvic blush for the need for pelvic
major nonpelvic bleeding, 9 had a negative pelvic angiography hemorrhage control in isolated pelvic fracture patients were
for arterial injury, and the remaining 17 had no intervention for 87% (69Y96), 75% (65Y83), 51% (37Y65), and 95% (87Y98),
hemorrhage control. respectively (Table 3).
Of the 94 patients without a pelvic blush, 87 (93%) did
not have pelvic hemorrhage control and 7 (7%) did have pelvic Outcome of Pelvic Blush Patients
hemorrhage control. One patient had pelvic packing at lapa- The characteristics and outcome of all patients with a
rotomy, and six patients had a pelvic angiography (for clinical pelvic blush and patients with an isolated pelvic blush in re-
signs of active bleeding) with subsequent embolization for lation to the need for pelvic hemorrhage control are presented
an arterial injury. in Tables 5 and 6. Both in the overall pelvic blush group as well

TABLE 1. Characteristics, Management, and Outcome of All Patients by Presence of a Pelvic Blush on CT Scan
All Patients Pelvic Blush Pelvic Blush
(n = 162) Present (n = 68) Absent (n = 94) p
Age, mean (SD), y 41 (19) 44 (20) 38 (18) 0.04
Sex, n (%) 124 (77) 54 (79) 70 (75) 0.20
Mechanism of injury, motor vehicle collision, n (%) 35 (22) 12 (18) 23 (25) 0.25
ISS, mean (SD) 27 (14) 31 (16) 24 (12) G0.01
Type fracture, MLD, n (%) 56 (35) 31 (46) 25 (27) 0.01
SBP, mean (SD), mm Hg 119 (33) 113 (40) 124 (25) 0.04
Base decit, median (IQR), mEq/L 4 (5) 4 (6) 3 (4) G0.01
Nonpelvic hemorrhage control 28 (17) 15 (22) 13 (14) 0.12
Pelvic external xator 86 (53) 47 (69) 39 (42) G0.01
Pelvic hemorrhage control 39 (24) 32 (47) 7 (7) G0.01
Packed red blood cells in 24 h, median (IQR), U 2 (10) 8 (19) 2 (4) G0.01
Any packed red blood cell 111 (69) 55 (81) 56 (60) G0.01
Hospital mortality 19 (12) 12 (18) 7 (7) 0.04
Hemorrhage-related mortality 10 (6) 8 (12) 2 (2) 0.01
Italics indicate statistical signicance ( p G 0.05).

376 * 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg
Volume 76, Number 2 Verbeek et al.

TABLE 2. The Association Between the Presence of a Pelvic TABLE 4. The Association Between the Presence of a Pelvic
Blush on CT Scan and the Need for Pelvic Hemorrhage Control Blush on CT Scan and the Need for Pelvic Hemorrhage Control
in all Pelvic ring Fracture Patients (n = 162) in Isolated Pelvic Fracture Patients (With No Other Source of
Major Hemorrhage) (n = 134)
Pelvic Hemorrhage
Control Pelvic Hemorrhage
Yes No Control
Yes No
Blush 32 36 68
Present 7 87 94 Blush 27 26 53
Absent 39 123 162 Present 4 77 81
Absent 31 103 134

as in the isolated pelvic blush subgroup, patients with pelvic


hemorrhage control had a similar age, sex, mechanism of in- room, we were able to obtain a CT scan in nearly all pelvic
jury, type of fracture, and SBP on arrival compared with pa- ring fracture patients without interrupting ongoing resuscita-
tients who did not have pelvic hemorrhage control. Furthermore, tion. This included hemodynamically compromised patients
in both groups, patients with pelvic hemorrhage control had a who would not be considered for early CT scan in many other
signicantly higher ISS and base decit on arrival. institutions. Second, it may be speculated that the timing of the
In the overall pelvic blush group and in the isolated CT scan (early in the resuscitation phase) contributed to the
blush subgroup, the total packed red blood cells requirement in high rate of pelvic blushes detected. Potentially, some smaller
24 hours after arrival and the need for any packed red blood arterial injuries would not have been detected on a more delayed
cells were signicantly higher in patients with pelvic hemor- CT because these may stop bleeding spontaneously. Lastly, in
rhage control. recent years, advances in CT technology have improved imaging
In the overall pelvic blush group, the hospital and resolution, which may have resulted in the detection of more
hemorrhage-related mortality rates were similar in patients who discrete blushes.
had pelvic hemorrhage control and those who did not. In the Our results suggest that the presence of a pelvic blush
isolated pelvic blush subgroup, there was a trend toward a higher (by itself ) is not reliably associated with the need for pelvic
hospital and hemorrhage-related mortality in patients who had hemorrhage control in all pelvic ring fracture patients as well
pelvic hemorrhage control. as in isolated pelvic fracture patients. In both patient groups,
half of pelvic blushes (53% and 49%, respectively) required
no intervention. This was in part caused by the fact that some
DISCUSSION pelvic blush patients had a negative pelvic angiography for
arterial injury while others showed no clinical signs of active
In this study of pelvic ring fracture patients, a pelvic pelvic bleeding (following nonpelvic hemorrhage control).
blush on early contrast-enhanced CT scan was a frequent The need for pelvic hemorrhage control in pelvic ring
nding and an indicator of severe injury. The percentage of fracture patients was more determined by the presence of
patients with a pelvic blush (42%) found in the present study clinical signs of ongoing bleeding than by the mere presence of a
is signicantly higher than earlier reported. Previous studies
of patients with a variety of (single) pelvic and acetabular
fractures report of an incidence of up to 12%.11,22Y24 Several TABLE 5. Characteristics and Outcome of all Patients With a
factors may have contributed to this discrepancy. First, our Pelvic Blush by the Need for Pelvic Hemorrhage Control (n = 68)
patients had an inherently higher risk for arterial injury con-
Pelvic
sidering that only patients with high-energy pelvic ring frac- Hemorrhage
tures were included while patients with single pelvic or Control
acetabular fractures and those with fractures from low-impact Yes No
falls were excluded from the analysis. Furthermore, because
(n = 32) (n = 36) p
of the location of the CT scanner in the trauma resuscitation
Age, mean (SD), y 47 (20) 44 (19) 0.10
Sex, male n (%) 26 (81) 28 (78) 0.48
TABLE 3. Diagnostic Indices (95% Confidence Intervals) for Mechanism of injury, motor vehicle collision, n (%) 3 (9) 9 (25) 0.10
the Presence of a Pelvic Blush on CT Scan and the Need for Pelvic ISS, mean (SD) 35 (18) 26 (11) 0.01
Hemorrhage Control in All Patients and in Isolated Pelvic
Type fracture, MLD, n (%) 14 (44) 17 (47) 0.48
Fracture Patients
SBP, mean (SD), mm Hg 108 (37) 117 (42) 0.48
All Patients Isolated Pelvic Fracture Base decit, median (IQR), mEq/L) 7 (6) 4 (5) 0.03
(n = 162) (n = 134) Packed red blood cells in 24 h, median (IQR), U 16 (21) 4 (11) G0.01
Any packed red blood cells 29 (91) 26 (72) 0.05
Sensitivity 82% (66Y92) 87% (69Y96)
Hospital mortality 7 (22) 5 (14) 0.30
Specicity 71% (62Y78) 75% (65Y83)
Hemorrhage-related mortality 5 (16) 3 (8) 0.30
Positive predictive value 47% (35Y59) 51% (37Y65)
Negative predictive value 93% (85Y97) 95% (87Y98) Italics indicate statistical signicance ( p G 0.05).

* 2014 Lippincott Williams & Wilkins 377

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg
Verbeek et al. Volume 76, Number 2

vital signs and a minimal transfusion requirement.7,8 We found


TABLE 6. Characteristics and Outcome of Patients With an
Isolated Pelvic Blush (With No Other Source of Major
that many pelvic blush patients have no signs of ongoing
Hemorrhage) by the Need for Pelvic Hemorrhage Control bleeding and have a favorable outcome without the need for
(n = 53) pelvic hemorrhage control. This nding was particularly ap-
parent in the absence of major nonpelvic bleeding sources
Pelvic
Hemorrhage
that can have a signicant impact on clinical outcome.
Control While the CT scan can be a valuable tool for the over-
Yes No
all assessment of pelvic fracture patients, the benets of this
modality should be carefully considered in light of the potential
(n = 27) (n = 26) p
risk of transporting the patient to the CT scanner. It is evident
Age, mean (SD), y 46 (20) 44 (21) 0.63 that a CT scan is only indicated if it can be safely obtained in
Sex, male, n (%) 21 (78) 19 (73) 0.47 an environment with close patient monitoring and resuscita-
Mechanism of injury, motor vehicle collision, n (%) 3 (11) 3 (12) 0.65 tion capabilities. A specic note of caution is for pelvic fracture
ISS, mean (SD) 33 (18) 25 (11) 0.05 patients who remain unresponsive to adequate uid resusci-
Type fracture, MLD, n (%) 12 (44) 13 (50) 0.45 tation. In these unstable patients, an immediate intervention
SBP, mean (SD), mm Hg 109 (39) 127 (36) 0.85 is required and should not be delayed by performing a CT
Base decit, median (IQR), mEq/L 5 (6) 3 (3) 0.01 scan.27Y29
Packed red blood cells in 24 h, median (IQR), U 12 (24) 1 (6) G001
Any packed red blood cells 24 (89) 16 (62) 0.02 Limitations
Hospital mortality 6 (22) 1 (4) 0.06 This study is retrospective in design with the inherent
Hemorrhage-related mortality 4 (15) 0 (0) 0.06 limitations regarding data collection and selection bias. Although
there was general agreement on the management strategy, no
pelvic blush on CT scan. It seems that further treatment should set treatment algorithm for pelvic ring fracture patients was in
therefore be guided primarily by hemodynamic parameters place to standardize decision making. The need for an inter-
and the patients response to resuscitation and not by radiologic vention was determined early in the resuscitation phase based
parameters only. primarily on clinical signs of active bleeding and the presence
In the current literature, there is no agreement on whether of associated injuries; no patients had a delayed intervention for
a pelvic blush predicts the need for pelvic hemorrhage control hemorrhage control. We were unable to examine the vital signs or
and more specically for pelvic arterial embolization. The transfusion requirements in detail to determine the effect certain
limited predictive value found in our study is in concurrence interventions had on the patients hemodynamic status. Lastly,
with several earlier reports on pelvic fracture patients. These the design and power of the study did not permit us to formulate
studies found that angiographic embolization was required in reliable predictors for the need for pelvic hemorrhage control.
only 29% to 41% of pelvic fracture patients with a pelvic To dene more accurately what early parameters are
blush.22,25,26 However, other studies came to different con- reliable predictors for the need for pelvic hemorrhage control,
clusions and suggested that a pelvic blush on CT scan is a a further prospective study would be needed.
sign of arterial injury that requires angiographic embolization
(in 69% and 80% of the patients).23,24 In a study that described CONCLUSION
patients that either required angiographic embolization or
In pelvic ring fracture patients, a pelvic blush on early
emergency surgery for pelvic hemorrhage control, it was found
contrast-enhanced CT scan is a frequent nding. Many patients
that of 15 pelvic fracture patients with a pelvic blush, 9 (60%)
with (particularly isolated) pelvic blushes have stable vital
required an intervention.11 The contradictory ndings in current
signs and can be managed without surgical or radiologic pelvic
literature are potentially a result of small sample sizes, differ-
hemorrhage control. The need for an intervention for a pelvic
ences in inclusion criteria, and variation in treatment strategies.
blush seems to be determined by the presence of clinical signs
Our data also show that the absence of a pelvic blush on
of ongoing bleeding.
CT scan dismisses the need for pelvic hemorrhage control in
the great majority of all major pelvic fracture patients as well as AUTHORSHIP
isolated pelvic fracture patients (93% and 95%). Earlier studies D.O.F.V. contributed to the literature search, study design, data col-
concur with this nding and report of a negative predictive lection, data analysis, data interpretation, and writing. I.A.J.Z. and
value of 98% to 100%.11,22Y24 It is however important to note C.V.D.L. performed the data collection. K.J.P., O.M.V.D., and J.C.G.
the small number of patients in our study that did require contributed to the study design, data analysis, data interpretation,
(mainly radiologic) pelvic hemorrhage control in the absence and writing.
of a pelvic blush. The false-negative results in these patients
were potentially caused by intermittent bleeding and were DISCLOSURE
detected because of an ongoing transfusion requirement
The authors declare no conflicts of interest.
In pelvic ring fracture patients with a pelvic blush on CT
scan, the need for pelvic hemorrhage control is evident in
patients with signs of ongoing pelvic bleeding (in the absence REFERENCES
of major nonpelvic bleeding). However, there is no consensus 1. Chan O. Primary computed tomography survey for major trauma. Br J Surg.
regarding the management of pelvic blush patients with stable 2009;96(12):1377Y1378.

378 * 2014 Lippincott Williams & Wilkins

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg
Volume 76, Number 2 Verbeek et al.

2. Wurmb TE, Fruhwald P, Hopfner W, Keil T, Kredel M, Brederlau J, Roewer N, 15. Bif WL, Smith WR, Moore EE, Gonzalez RJ, Morgan SJ, Hennessey T,
Kuhnigk H. Whole-body multislice computed tomography as the rst Offner PJ, Ray CE, Jr., Franciose RJ, Burch JM. Evolution of a
line diagnostic tool in patients with multiple injuries: the focus on time. multidisciplinary clinical pathway for the management of unstable patients
J Trauma. 2009;66(3):658Y665. with pelvic fractures. Ann Surg. 2001;233(6):843Y850.
3. Sierink JC, Saltzherr TP, Reitsma JB, Van Delden OM, Luitse JS, Goslings JC. 16. Eastridge BJ, Starr A, Minei JP, OKeefe GE, Scalea TM. The importance
Systematic review and meta-analysis of immediate total-body computed of fracture pattern in guiding therapeutic decision-making in patients with
tomography compared with selective radiological imaging of injured hemorrhagic shock and pelvic ring disruptions. J Trauma. 2002;53(3):
patients. Br J Surg. 2012;99(Suppl 1):52Y58. 446Y450; discussion 450Y451.
4. Huber-Wagner S, Lefering R, Qvick LM, Korner M, Kay MV, Pfeifer KJ, 17. Velmahos GC, Chahwan S, Falabella A, Hanks SE, Demetriades D. An-
Reiser M, Mutschler W, Kanz KG. Working Group on Polytrauma of giographic embolization for intraperitoneal and retroperitoneal injuries.
the German Trauma Society. Effect of whole-body CT during trauma re- World J Surg. 2000;24(5):539Y545.
suscitation on survival: a retrospective, multicentre study. Lancet. 2009; 18. Gansslen A, Giannoudis P, Pape HC. Hemorrhage in pelvic fracture: who
373(9673):1455Y1461. needs angiography? Curr Opin Crit Care. 2003;9(6):515Y523.
5. Fung Kon Jin PH, Goslings JC, Ponsen KJ, van Kuijk C, Hoogerwerf N, 19. Burgess AR, Eastridge BJ, Young JW, Ellison TS, Ellison PS Jr, Poka A,
Luitse JS. Assessment of a new trauma workow concept implementing Bathon GH, Brumback RJ. Pelvic ring disruptions: effective classication
a sliding CT scanner in the trauma room: the effect on workup times. system and treatment protocols. J Trauma. 1990;30(7):848Y856.
J Trauma. 2008;64(5):1320Y1326.
20. Sarin EL, Moore JB, Moore EE, Shannon MR, Ray CE, Morgan SJ, Smith
6. Saltzherr TP, Bakker FC, Beenen LF, Dijkgraaf MG, Reitsma JB, Goslings WR. Pelvic fracture pattern does not always predict the need for urgent
JC. REACT Study Group. Randomized clinical trial comparing the effect embolization. J Trauma. 2005;58(5):973Y977.
of computed tomography in the trauma room versus the radiology de-
21. Hamill J, Holden A, Paice R, Civil I. Pelvic fracture pattern predicts pelvic
partment on injury outcomes. Br J Surg. 2012;99(Suppl 1):105Y113.
arterial haemorrhage. Aust N Z J Surg. 2000;70(5):338Y343.
7. Davis JW, Moore FA, McIntyre RC Jr, Cocanour CS, Moore EE, West MA.
22. Brasel KJ, Pham K, Yang H, Christensen R, Weigelt JA. Signicance
Western trauma association critical decisions in trauma: management
of pelvic fracture with hemodynamic instability. J Trauma. 2008;65(5): of contrast extravasation in patients with pelvic fracture. J Trauma.
1012Y1015. 2007;62(5):1149Y1152.
8. Cullinane DC, Schiller HJ, Zielinski MD, Bilaniuk JW, Collier BR, Como 23. Stephen DJ, Kreder HJ, Day AC, McKee MD, Schemitsch EH,
J, Holevar M, Sabater EA, Sems SA, Vassy WM, et al. Eastern Associa- ElMaraghy A, Hamilton P, McLellan B. Early detection of arterial bleed-
tion for the Surgery of Trauma practice management guidelines for ing in acute pelvic trauma. J Trauma. 1999;47(4):638Y642.
hemorrhage in pelvic fractureVupdate and systematic review. J Trauma. 24. Pereira SJ, OBrien DP, Luchette FA, Choe KA, Lim E, Davis K Jr, Hurst
2011;71(6):1850Y1868. JM, Johannigman JA, Frame SB. Dynamic helical computed tomogra-
9. White CE, Hsu JR, Holcomb JB. Haemodynamically unstable pelvic phy scan accurately detects hemorrhage in patients with pelvic fracture.
fractures. Injury. 2009;40(10):1023Y1030. Surgery. 2000;128(4):678Y685.
10. van Vugt AB, van Kampen A. An unstable pelvic ring. The killing fracture. 25. Diamond IR, Hamilton PA, Garber AB, Tien HC, Chughtai T, Rizoli SB,
J Bone Joint Surg Br. 2006;88(4):427Y433. Tremblay LN, Brenneman FD. Extravasation of intravenous computed
11. Mohseni S, Talving P, Kobayashi L, Lam L, Inaba K, Branco BC, Oliver tomography scan contrast in blunt abdominal and pelvic trauma. J Trauma.
M, Demetriades D. The diagnostic accuracy of 64-slice computed to- 2009;66(4):1102Y1107.
mography in detecting clinically signicant arterial bleeding after pelvic 26. Ryan MF, Hamilton PA, Chu P, Hanaghan J. Active extravasation of ar-
fractures. Am Surg. 2011;77(9):1176Y1182. terial contrast agent on post-traumatic abdominal computed tomography.
12. Verbeek D, Sugrue M, Balogh Z, Cass D, Civil I, Harris I, Kossmann T, Can Assoc Radiol J. 2004;55(3):160Y169.
Leibman S, Malka V, Pohl A, et al. Acute management of hemodynami- 27. Howell GM, Peitzman AB, Nirula R, Rosengart MR, Alarcon LH, Billiar
cally unstable pelvic trauma patients: time for a change? Multicenter re- TR, Sperry JL. Delay to therapeutic interventional radiology postinjury:
view of recent practice. World J Surg. 2008;32(8):1874Y1882. time is of the essence. J Trauma. 2010;68(6):1296Y1300.
13. Cothren CC, Osborn PM, Moore EE, Morgan SJ, Johnson JL, Smith WR. 28. Neal MD, Peitzman AB, Forsythe RM, Marshall GT, Rosengart MR,
Preperitonal pelvic packing for hemodynamically unstable pelvic frac- Alarcon LH, Billiar TR, Sperry JL. Over reliance on computed tomog-
tures: a paradigm shift. J Trauma. 2007;62(4):834Y839; discussion raphy imaging in patients with severe abdominal injury: is the delay worth
839Y842. the risk? J Trauma. 2011;70(2):278Y284.
14. Heetveld MJ, Harris I, Schlaphoff G, Balogh Z, DAmours SK, Sugrue M. 29. Clarke JR, Trooskin SZ, Doshi PJ, Greenwald L, Mode CJ. Time to lap-
Hemodynamically unstable pelvic fractures: recent care and new guide- arotomy for intra-abdominal bleeding from trauma does affect survival
lines. World J Surg. 2004;28(9):904Y909. for delays up to 90 minutes. J Trauma. 2002;52(3):420Y425.

* 2014 Lippincott Williams & Wilkins 379

Copyright 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.