The Practice of Venous Thromboembolism Prophylaxis in the
Major Trauma Patient
Avery B. Nathens, MD, PhD, MPH, Megan K. McMurray, PharmD, Joseph Cuschieri, MD,
Emily A. Durr, PharmD, Ernest E. Moore, MD, Paul E. Bankey, MD, Brad Freeman, MD,
Brian G. Harbrecht, MD, Jeffrey L. Johnson, MD, Joseph P. Minei, MD, Bruce A. McKinley, MD,
Frederick A. Moore, MD, Michael B. Shapiro, MD, Michael A. West, MD, PhD,
Ronald G. Tompkins, MD, ScD, and Ronald V. Maier, MD
Background: The incidence of ve-
nous thromboembolism (VTE) without
prophylaxis is as high as 80% after major
trauma. Initiation of prophylaxis is often
delayed because of concerns of injury-
associated bleeding. As the effect of delays
in the initiation of prophylaxis on VTE
rates is unknown, we set out to evaluate
the relationship between late initiation of
prophylaxis and VTE.
Methods: Data were derived from a
multicenter prospective cohort study eval-
uating clinical outcomes in adults with
hemorrhagic shock after injury. Analyses
were limited to patients with an Intensive
Care Unit length of stay >7 days. The rate
of VTE was estimated as a function of the
time to initiation of pharmacologic pro-
sis and pulmonary embolism might be as high as 40% and
phylaxis. A multivariate stepwise logistic
regression model was used to evaluate fac-
tors associated with late initiation.
Results: There were 315 subjects
who met inclusion criteria; 34 patients
(11%) experienced a VTE within the first
28 days. Prophylaxis was initiated within
48 hours of injury in 25% of patients, and
another one-quarter had no prophylaxis
for at least 7 days after injury. Early pro-
phylaxis was associated with a 5% risk of
VTE, whereas delay beyond 4 days was
associated with three times that risk (risk
ratio, 3.0, 95% CI [1.4–6.5]). Factors as-
sociated with late (>4 days) initiation of
prophylaxis included severe head injury,
absence of comorbidities, and massive
transfusion, whereas the presence of a se-
vere lower extremity fracture was associ-
ated with early prophylaxis.
Conclusions: Clinicians are reticent
to begin timely VTE prophylaxis in criti-
cally injured patients. Patients are with-
out VTE prophylaxis for half of all days
within the first week of admission and this
delay in the initiation of prophylaxis is
associated with a threefold greater risk of
VTE. The relative risks and benefits of
early VTE prophylaxis need to be defined
to better direct practice in this high-risk
Key Words: Venous thromboembo-
lism, Prophylaxis, Deep venous throm-
bosis, Pulmonary embolism, Knowledge
J Trauma. 2007;62:557–563.
enous thromboembolism (VTE) in high-risk trauma pa-
tients is a frequent event. Depending on the cohort and
the diagnostic modality, rates of deep venous thrombo-
20%, respectively.1,2Proximal lower extremity venous
thrombosis has been documented in as many as 10% of
patients, whereas over 1% of patients experience a symptom-
atic pulmonary embolism.1,3Not only is the development of
VTE in the form of a massive pulmonary embolism a fre-
quent contributor to late death in the injured patient, its
development complicates management and leads to late se-
quelae. To reduce the risk of VTE, many practitioners ad-
minister pharmacologic thromboprophylaxis, an approach
supported by several randomized controlled trials and
Pharmacologic thromboprophylaxis is associated with a
low but clearly quantifiable bleeding risk in the range of 1%
to 3%, depending on the agent used.1This risk, however low,
requires the treating physician to balance the risk of bleeding
with the risk of VTE and the perceived benefit of thrombo-
prophylaxis. The ability of the physician to assess the risks
and benefits might be limited, particularly when caring for
patients at high risk for both bleeding and VTE. We postu-
lated that because of the inability to define the risks and
benefits of prophylaxis, many physicians would simply delay
the initiation of pharmacologic VTE prophylaxis until the
bleeding risk had sufficiently abated. To ascertain practice
patterns in relation to this clinical dilemma, we sought to
Submitted for publication October 11, 2006.
Accepted for publication December 14, 2006.
Copyright © 2007 by Lippincott Williams & Wilkins, Inc.
From the Department of Surgery, University of Toronto, Toronto,
Canada (A.B.N.); the Department of Surgery, University of Washington,
Seattle, Washington (M.K.M., J.C., R.V.M., E.A.D.); the Department of
Surgery, University of Colorado, Denver, Colorado (J.L.J., E.E.M.); the
Department of Surgery, University of Texas at Southwestern, Dallas, Texas
(J.P.M.); the Department of Surgery, University of Rochester, Rochester,
New York (P.E.B.); the Department of Surgery, Washington University, St.
Louis, Missouri (B.F.); the Department of Surgery, University of Pittsburgh,
Pittsburgh, Pennsylvania (B.G.H.); the Department of Surgery, Methodist
Hospital, Houston, Texas (B.A.M., F.A.M.); Division of Burn Surgery,
Massachusetts General Hospital, Boston, Massachusetts (R.G.T.).
Supported by funding from the National Institutes of Health (NIH
NIGMS U54 GM062119-1).
Presented at the 65th Annual Meeting of the American Association for
the Surgery of Trauma, September 28–30, 2006, New Orleans, Louisiana.
Address for reprints: Avery B. Nathens, MD, St. Michael’s Hospital,
3-073 Queen Wing, 30 Bond Street, Toronto, Canada, M5B 1W8; email:
The Journal of TRAUMA?Injury, Infection, and Critical Care
Volume 62 • Number 3
evaluate (1) the timing of pharmacologic prophylaxis; and (2)
the effects of delayed prophylaxis in a severely injured cohort
of patients admitted to the intensive care unit.
This is a secondary analysis of data collected in the
context of an ongoing large multicenter prospective cohort
study whose primary objective is to evaluate the relationship
between the inflammatory response to injury and posttrau-
matic multiple organ failure.6In this analysis, the exposure
represented the time to pharmacologic VTE prophylaxis and
the primary outcome was the development of VTE within 28
days of injury.
The Inflammation and the Host Response to Injury is a
collaborative program supported by the National Institute of
General Medical Sciences (NIGMS) to better define the pro-
teomic and genomic response in injured patients at high risk
for posttraumatic multiple organ failure. This report focuses
on patients admitted to any one of seven participating insti-
tutions during the interval April 2004 through February 2006.
Criteria for enrollment were age ?16 years, blunt trauma,
arrival to hospital within 6 hours of injury, either hypotension
(?90) or an elevated base deficit (?6), blood transfusion
within 12 hours of injury, any body region exclusive of the
brain with an Abbreviated Injury Scale score (AIS) ?2, and
an intact cervical spinal cord to exclude those with isolated
severe head injuries or spinal cord injuries, respectively. As
the primary objective of this analysis was to evaluate the
practice of pharmacologic VTE prophylaxis in the Intensive
Care Unit (ICU) and subsequent VTE, we further limited
inclusion to subjects surviving at least 48 hours and who were
in the ICU for at least 7 days. Standard operating procedures
(SOP) for clinical care were developed to minimize variation
across centers.6During the time interval of this analysis, the
SOP for VTE prophylaxis had not yet been finalized. The
institutional review board of each participating center ap-
proved the cohort study.
Clinical data were obtained by trained nurse abstractors
and entered into TrialDb, a Web-based data collection
platform.7Injuries were coded by body region using the AIS
with AIS ?3 considered a severe injury.8The Injury Severity
Score (ISS) was used as a global measure of injury severity.9
Data integrity was evaluated centrally through an assessment
of missing values, range checks, evaluation for implausible
values and internal consistency. Additionally, data were val-
idated through an external review of a random sample of
charts by a physician and independent chart abstractor.
Timing of VTE Prophylaxis
We evaluated the timing of initiation of pharmacologic
VTE prophylaxis during the first 7 days of ICU admission.
Agents were considered prophylactic if administered as one
of the following regimens: heparin 5000 IU bid or tid; enox-
aparin 30 mg bid; or dalteparin 5000 IU, each administered
by subcutaneous injection. Additionally, when applicable, the
date of insertion of an inferior vena caval filter was recorded.
VTEs were diagnosed by pulmonary angiogram, com-
puted tomographic pulmonary angiography, or duplex ultra-
sonography (in the case of deep venous thrombosis). All
subjects were followed for at least 28 days or until hospital
discharge, whichever occurred first.
The relative risk of VTE as a function of the timing of
initiation of pharmacologic prophylaxis was evaluated using
a Poisson model. The results obtained identified an inflection
point allowing the classification of timing of initiation into
early or late, where the latter was associated with an in-
creased risk of VTE. We then determined the relationship
between late initiation of VTE prophylaxis and several pa-
tient and/or injury characteristics using the ?2test for cate-
gorical variables and the t test for continuous variables. All
variables with ? ? 0.1 on univariate analyses were entered
into a forward stepwise regression model to identify risk
factors for late initiation of prophylaxis. Model parameters
for entry and exit into the model were ? ? 0.1 and ? ? 0.05,
respectively. All analyses were performes using Stata Ver-
sion 8 (College Station, Tex).
There were 731 patients enrolled during the interval of
study. Of this cohort, only 315 spent the first 7 days of their
admission in the ICU. Of these 315 patients, 34 (11%) had a
diagnosis of VTE: 16 DVT, 14 PE, and 4 with both. One-
third of events occurred in the first week after injury (Fig. 1).
Fig. 1. The day of diagnosis of any VTE (deep venous thrombosis or
pulmonary embolism) within the first 28 days after injury.
The Journal of TRAUMA?Injury, Infection, and Critical Care
Four patients had evidence of DVT within the first 48 hours
after admission and were excluded from further analysis
given the lack of available opportunity for prophylaxis, leav-
ing a final cohort of 311 patients.
Pharmacologic prophylaxis was initiated within 48 hours
of injury in one-quarter of all patients, and another one
quarter went without prophylaxis for at least 7 days after
injury (Fig. 2). The delay in prophylaxis was not because of
mitigation of the risk of pulmonary embolism with the place-
ment of IVC filters, as only 37 (12%) patients had filters
inserted during the course of their admission. Three-quarters
(29) of filters were placed before VTE prophylaxis was
The proportion of patients developing VTEs increased
significantly with delays in initiation of pharmacologic pro-
phylaxis (Fig. 3). Early prophylaxis was associated with a 5%
risk of VTE, whereas a delay beyond 4 days was associated
with three times that risk (risk ratio: 3.0, 95% confidence
interval [1.4–6.5]). There was no further increase in risk with
delays beyond this time. Specifically, beginning prophylaxis
on day 7 was no worse than on day 5. All the increased risk
was assumed through delays in excess of 4 days.
To better understand the reasons for delays in the initi-
ation of pharmacologic prophylaxis, we evaluated patient and
injury characteristics among patients receiving early (?4
days) prophylaxis compared with the patients receiving late
prophylaxis (?4 days). Almost half of all patients (n ? 137,
44%) fell into the latter category. Although there were no
differences in age, gender, race, or mechanism of injury
across the two groups, patients without significant comor-
bidity (exclusive of liver disease) appeared to be over-
represented in the late prophylaxis group (Table 1).
There were several injury-specific factors that were re-
lated to the timing of prophylaxis (Table 2). Although severe
lower extremity injuries were significantly associated with
early prophylaxis, severe head injuries and massive early
transfusion (?6 units/12 hours) were associated with delays
in prophylaxis. Coagulopathy as evidenced by either an ele-
vated INR or the need for fresh frozen plasma was also more
frequent in the late prophylaxis group, although this did not
reach statistical significance.
To evaluate the independent factors associated with the
timing of prophylaxis, we performed a forward stepwise
logistic regression model assessing the contributions of co-
Fig. 2. The day of initiation of pharmacologic thromboprophylaxis
(black) and the day of initiation of prophylaxis operating room
placement of an inferior vena caval filter (gray). The latter is
demonstrated to show that in this particular cohort, the delays in
initiation of prophylaxis are not because of a significant use of vena
Fig. 3. The proportion of patients with VTE in relation to the day of
initiation of pharmacologic prophylaxis. Gray shading represents
the 95% confidence intervals.
Table 1 Baseline Patient Characteristics Associated
With Late Pharmacologic Thromboprophylaxis
N ? 174N ? 137
Mean age (SD)
Male gender (%)
Chronic renal failure
Prior history or VTE
Mechanism of injury
Motor vehicle crash
COPD, Chronic obstructive pulmonary disease.
VTE Prophylaxis in the Trauma ICU
Volume 62 • Number 3
morbidity status, highest Glasgow Coma Scale motor score,
severe head or lower extremity injury, increasing blood trans-
fusion requirements, and elevated INR. Each of severe head
injury, absence of comorbidities, and massive transfusion
were associated with a twofold increased odds of delayed
prophylaxis (Table 3). Only the presence of a severe lower
extremity fracture was associated with early prophylaxis.
Given the injury characteristics associated with early and
late initiation of prophylaxis, severe lower extremity and
severe head injury, respectively, we postulated that their
co-existence might alter practice by compelling the physician
to initiate prophylaxis earlier. To evaluate whether the pres-
ence of a lower extremity fracture affected decisions regard-
ing the timing of prophylaxis in patients at high risk for
intracranial bleeding, we added an interaction term to the
regression model. This term effectively asks the question
“Does the presence of a severe lower extremity fracture affect
the timing of prophylaxis in patients with a head injury?”.
This interaction term was not significant (p ? 0.53), indicat-
ing that the lower extremity fracture does not modify the
approach in head injured patients.
The risk of VTE and their clinical significance in the
severely injured patient is well established. Our ability as
clinicians to balance this risk with the potential for hemor-
rhagic complications is limited. In this report, we demon-
strated that less than half of all patients admitted to the
intensive care unit receive pharmacologic thromboprophy-
laxis within 4 days of injury and another 25% receive no
prophylaxis during the first 7 days. It is evident that this delay
is associated with a higher likelihood of VTE. Patients in
whom prophylactic therapy is initiated beyond day 4 have a
threefold greater risk of VTE. It is important to note that these
data do not address whether earlier thromboprophylaxis
might have reduced this risk, as prophylaxis is typically
delayed in higher risk patients. These data simply demon-
strate the practice of pharmacologic prophylaxis and how this
practice might be associated with a higher risk of VTE; it
cannot infer causality.
Patients with severe head injury, those having received
greater than six units of packed red blood cells within 12
hours of injury, and those without significant comorbidity are
twice as likely to have pharmacologic prophylaxis delayed
beyond day 4. By contrast, the presence of a severe lower
extremity fracture was associated with early pharmacologic
prophylaxis. Importantly, the presence of lower extremity
fractures, a strong risk factor for VTE, did not modify the
timing of prophylaxis among patients with severe head inju-
ries, suggesting that the concern over bleeding took prece-
dence. The VTE risk was not obviated by the placement of an
IVC filter among patients with this injury combination, as
only 8 of 29 filters (28%) placed before the initiation of
prophylaxis involved patients with both severe head and
lower extremity fractures.
The timing of initiation of VTE prophylaxis after major
trauma has been poorly addressed in consensus guidelines
and suggested protocols, with most advising the use of me-
chanical prophylaxis (e.g. intermittent pneumatic compres-
sion or venous foot pumps) until the risk of bleeding has
abated.4,5,10However, the assessment of bleeding risk in an
individual patient is difficult and in many cases entirely
subjective, leaving it up to physician judgment as when it
might be safe to initiate pharmacologic thromboprophylaxis.
Unfortunately, the evidence supporting the efficacy of me-
Table 2 Injury Characteristics Associated With Late
N ? 174N ? 137
Mean Injury Severity Score
Maximum AIS score
Shock (SBP ?90)*
Highest GCS motor score
in ED (SD)
Apache II score (SD)
Severe (AIS ?3) injury by
Blood transfusion ?6
Blood transfusion in first
12 h (mL, SD)
?10 l/12 h
Base deficit ?8 mEq/l in
first 12 h
Highest INR in
Fresh frozen plasma in
first 12 h (mL, SD)
34 (12)34 (13) 0.86
30 (6.2)30 (6.4)0.46
2,623 (2550)3,594 (3467) 0.005
109 (62)80 (57)0.36
129 (74)112 (80)0.19
1.4 (0.5)1.6 (1.5)0.10
1,109 (1,305)1,351 (1,559)0.14
* Refers to hypotension in the field or within 1 hour of hospital
Table 3 Independent Predictors of Delays (>4 d) in the
Initiation of Pharmacologic VTE Prophylaxis
Odds Ratio for Delayed
Prophylaxis* (95% CI)
Severe head injury
Blood transfusion ?6 u/12 h
Severe lower extremity injury
* Odds of delayed prophylaxis compared to patients without that
The Journal of TRAUMA?Injury, Infection, and Critical Care
chanical prophylaxis in this patient population is weak and
compliance associated with these devices is relatively
low.11–13This challenge is further magnified when one con-
siders that many VTE events might occur within the first few
days of injury, suggesting that if there is to be any benefit,
prophylaxis should be initiated as early as possible.14,15
We demonstrate that at least a half of severely injured
patients receive no pharmacologic thromboprophylaxis
within the first 4 days after injury and 25% receive no such
prophylaxis within the first week. This later time point is
clearly beyond the time at which significant bleeding
might occur, even in patients with intracranial bleeding.
For example, in trials of pharmacologic thromboprophy-
laxis in patients undergoing elective neurosurgery, prophy-
laxis initiated within the first 24 hours after injury is not
associated with an increased risk of bleeding.16Although
the risk of bleeding might be slightly higher in those with
traumatic intracranial hemorrhage, it is very unlikely that
this risk remains significantly elevated for as long as 7
days after injury.
A retrospective analysis of patients with splenic injuries
treated nonoperatively demonstrated no increased risk of
bleeding among those receiving early (?48 hours) versus late
pharmacologic prophylaxis with low molecular weight hep-
arin. Although this study design has significant limitations, it
does suggest that earlier prophylaxis might be safe even in
those with solid organ injuries.17
It is plausible that the failure to initiate early thrombo-
prophylaxis shown in this analysis might reflect an overesti-
mate of the efficacy of mechanical compression devices.
Alternatively, it might be failure on the part of clinicians to
recognize the risk of VTE. If so, this is not exclusive to
trauma. For example, in a multicenter study of critically ill
medical patients, 90% met prespecified criteria based on
international consensus guidelines for thromboprophylaxis.18
However, some form of prophylaxis was administered to
only 23% of patients, and in only 16% was this deemed to
be appropriate for the level of risk. In contrast to the
findings noted in the current report, the presence of risk
factors for bleeding did not influence the use or choice of
Our analysis has several potential limitations. First, chart
abstraction did not routinely capture the use of mechanical
prophylaxis, thus we cannot estimate the extent to which
physicians relied on this modality as the sole means of throm-
boprophylaxis. This might not have any effect, given the
questionable efficacy of sequential compression devices. Sec-
ond, this analysis focused primarily on the day on which
pharmacologic prophylaxis was initiated. Using this ap-
proach, we estimate that approximately half of all patient-
days within the first week are unprotected. We did not probe
further to estimate the frequency at which doses were held
after prophylaxis was initiated, suggesting that our data are
likely an overestimate of the exposure of patients to throm-
boprophylaxis. Lastly, it would have been desirable to eval-
uate the precise rationale and reasons among the physicians
electing to withhold prophylaxis, as this would provide in-
formation on how to better educate physicians to translate
evidence into practice.
Taken together, our data suggest a need to better estimate
the risks and benefits of early pharmacologic thrombopro-
phylaxis in major trauma patients. These data can then be
incorporated into the development of future consensus guide-
lines to better direct the physician at the bedside.
Additional participating investigators in the Large Scale Collaborative
Research Agreement entitled Inflammation and the Host Response to Injury
include: Henry V. Baker, Ph.D., Ulysses Balis, M.D., Timothy R. Billiar,
M.D., Bernard H. Brownstein, Ph.D., Steven E. Calvano, Ph.D., David G.
Camp II, Ph.D., George Casella, Ph.D., Irshad H. Chaudry, Ph.D., J. Perren
Cobb, M.D., Ronald W. Davis, Ph.D., Asit K. De, Ph.D., Constance Elson,
Ph.D., Celeste C. Finnerty, Ph.D., Richard L. Gamelli, M.D., Nicole S.
Gibran, M.D., Douglas L. Hayden, M.A., Laura Hennessy, R.N,. David N.
Herndon, M.D., Jureta W. Horton, Ph.D., Marc G. Jeschke, M.D., Ph.D.,
Matthew B. Klein, M.D., James A. Lederer, Ph.D., Stephen F. Lowry, M.D.,
John A. Mannick, M.D., Philip H. Mason, Ph.D., Grace P. McDonald-Smith,
M.Ed., Carol L. Miller-Graziano, Ph.D., Michael N. Mindrinos, Ph.D., Lyle
L. Moldawer, Ph.D., Grant E. O’Keefe, M.D., M.P.H., Laurence G. Rahme,
Ph.D., Daniel G. Remick, Jr. M.D., David A. Schoenfeld, Ph.D., Geoffrey M.
Silver, M.D., Richard D. Smith, Ph.D., John Storey, Ph.D., Robert
Tibshirani, Ph.D., Mehmet Toner, Ph.D., H. Shaw Warren, M.D., Wenzhong
1. Geerts WH, Jay RM, Code KI, et al. A comparison of low-dose
heparin with low-molecular-weight heparin as prophylaxis against
venous thromboembolism after major trauma. N Engl J Med. 1996;
2. Schultz DJ, Brasel KJ, Washington L, et al. Incidence of
asymptomatic pulmonary embolism in moderately to severely injured
trauma patients. J Trauma. 2004;56:727–731.
3.Schuerer DJ, Whinney RR, Freeman BD, et al. Evaluation of the
applicability, efficacy, and safety of a thromboembolic event
prophylaxis guideline designed for quality improvement of the
traumatically injured patient. J Trauma 2005;58:731–739.
4.Rogers FB, Cipolle MD, Velmahos G, Rozycki G, Luchette FA.
Practice management guidelines for the prevention of venous
thromboembolism in trauma patients: the EAST practice
management guidelines work group. J Trauma. 2002;53:142–164.
5.Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous
thromboembolism: the Seventh ACCP Conference on Antithrombotic
and Thrombolytic Therapy. Chest. 2004;126:338S–400S.
6.Maier RV, Bankey P, McKinley B, et al. Inflammation and the host
response to injury, a large-scale collaborative project: patient-
oriented research core-standard operating procedures for clinical
care: foreword. J Trauma. 2005;59:762–763.
7. Brandt CA, Deshpande AM, Lu C, et al. TrialDB: a web-based clinical
study data management system. AMIA Annu Symp Proc. 2003;794.
8. Association for the Advancement of Automotive Medicine. The
Abbreviated Injury Scale, 1990 revision. Des Plaines, IL; 1990.
9. Baker SP, O’Neill B, Haddon W Jr, Long WB. The injury severity
score: a method for describing patients with multiple injuries and
evaluating emergency care. J Trauma. 1974;14:187–196.
10.Knudson MM, Ikossi DG, Khaw L, Morabito D, Speetzen LS.
Thromboembolism after trauma: an analysis of 1602 episodes from
the American College of Surgeons National Trauma Data Bank. Ann
VTE Prophylaxis in the Trauma ICU
Volume 62 • Number 3
11.Velmahos GC, Kern J, Chan LS, et al. Prevention of venous
thromboembolism after injury: an evidence-based report–part I:
analysis of risk factors and evaluation of the role of vena caval
filters. J Trauma. 2000;49:132–138.
Huk M, Lynsky D, O’Callaghan T. Compliance of sequential
compression device for deep vein thrombosis prophylaxis in the
adult trauma patient: surgical intensive care unit vs. intermediate
care. Crit Care Med. 1998;26(suppl):A47.
Geerts WH, Jay R, Code K, et al. Venous foot pump as
thromboprophylaxis in major trauma. Thromb Haemost. 1999;
Owings JT, Kraut E, Battistella F, Cornelius JT, O’Malley R.
Timing of the occurrence of pulmonary embolism in trauma patients.
Arch Surg. 1997;132:862–866.
Sing RF, Camp SM, Heniford BT, et al. Timing of pulmonary
emboli after trauma: implications for retrievable vena cava filters.
J Trauma. 2006;60:732–734.
Agnelli G, Piovella F, Buoncristiani P, et al. Enoxaparin plus
compression stockings compared with compression stockings alone
in the prevention of venous thromboembolism after elective
neurosurgery. N Engl J Med. 1998;339:80–85.
Alejandro KV, Acosta JA, Rodriguez PA. Bleeding manifestations
after early use of low-molecular-weight heparins in blunt splenic
injuries. Am Surg. 2003;69:1006–1009.
Kahn SR, Panju A, Geerts W, et al. Multicenter evaluation
of the use of venous thromboembolism prophylaxis in acutely
ill medical patients in Canada. Thromb Res. 2007;110:
Dr. Selwyn O. Rogers (Boston, Massachusetts): This is
a descriptive study reviewing the current practice of venous
thrombo-embolism prophylaxis for critically ill trauma pa-
tients. As clearly stated in the conclusions, there is a nested
cohort retrospective study with an inflammation and host
response to injury study within it.
The authors found a significant delay in initiation of
venous thrombo-embolism and prophylaxis and correlated
this delay with a more than threefold higher venous thrombo-
embolism risk. Interestingly, almost a quarter of the patients
had no prophylaxis for greater than seven days. By definition,
this population is a high-risk patient population with immo-
bility, prolonged ICU stay, and abnormal coagulation.
This study clearly describes our current practice varia-
tion but does not answer the crucial question of why. I have
several questions that the authors may be able to answer,
given the data available.
First, were all the CVTs symptomatic? That is, were any
of the trauma centers routinely performing surveillance
Doppler ultrasounds in this high-risk patient population?
Secondly, data on mechanical prophylaxis were not able
to be collected. Do the authors believe that clinicians consid-
ered non-pharmacological VTE prophylaxis adequate for this
Third, given the modern era of retrievable IVC filters,
why was this intervention not more widely used for this very
high-risk patient population?
Finally, it is difficult to understand why VTE prophy-
laxis was withheld for such a long period of time in these
critically ill patients, because by the time a week has elapsed,
you would imagine that the bleeding risk should be minimal.
Can you speculate on why VTE prophylaxis was withheld for
such a prolonged period of time?
Dr. Avery B. Nathens (Toronto, Ontario, Canada): The
first question pertained to screening for DVTs, and there is
probably some variability across the seven participating in-
stitutions as to whether they screen asymptomatic patients. I
know that Harborview Medical Center screens selected pa-
tients who are considered high risk and can’t be anticoagu-
lated and my sense is that other centers might as well but this
is clearly not protocolized.
Your second question is on the use of mechanical pro-
phylaxis. I, personally, don’t think mechanical prophylaxis
have a benefit in this high-risk population. The evidence for
their efficacy is questionable and we know that when they are
ordered, they are used appropriately in only 20% of patients.
As a result, we did not evaluate mechanical prophylaxis,
because compliance is poor and efficacy is unproven.
Interestingly, retrievable filters should be changing our
practice, but unfortunately, there are recent data suggesting
that when these filters are placed, the likelihood of removal is
not high. Perhaps when be become more aware of who can
and cannot benefit from a retrievable filter, our practice might
So why was prophylaxis withheld in patients with mas-
sive transfusion for as long as four days? That is a very good
question. I agree that this practice is probably irrational and I
think if we examine our practices closely, we would come to
realize that many of us don’t practice as we should. I know
most patients are not bleeding on day three or day four. They
probably can be started earlier but we tend to find reasons to
hold off. For example, we might be holding off because
they’re going for an orthopedic procedure, or perhaps there
was a drop in the hematocrit.
Dr. Truswing (Richmond, Virginia): Two questions.
One is, did you look at the role of the need for operative
intervention as a cause for delay of initiating therapy,
particularly orthopedic procedures, since our orthopedic
colleagues sometimes don’t want us to start therapy prior
to their intervention?
No. 2 is, we have an algorithm that stratifies patients into
moderate, mild, and severe risk for VTE. Do you have that?
If you do, do you think we all should have one that would
prompt us to start therapy sooner?
Dr. Avery B. Nathens: We actually have the data to look
at the timing of procedures. This is a worthwhile question to
pursue; we just have not addressed it at this point. Our data
day two or day three. My assumption is that they are continued
on for the next seven or more days. However, it is likely that the
actual exposure to daily prophylaxis is probably much worse
than we think it is, because were often holding it because of
major orthopedic procedures.
The Journal of TRAUMA?Injury, Infection, and Critical Care
While I don’t think it’s necessary to hold off because of
those procedures, the orthopedic surgeons are insistent upon
it. This consortium is developing standard operating proce-
dures for DVT prophylaxis in this patient population. There
are guidelines available, and I think the guidelines are helpful
in directing our practice. But as you know, it is difficult to
convert guidelines into clinical practice, and that’s where the
Dr. Charles Yowler (Cleveland, Ohio): That was the
basis of my question. Were any of these participating insti-
tutions using guidelines during this study? In other words,
was this sort of a uniform distribution seen at all centers, or
did certain centers have more delay than other centers? Be-
cause that could skew, obviously, the data.
Dr. Avery B. Nathens: We did not look at variation
across centers. This is a critical question, as the objective
within the consortium is to create uniformity in care across
centers. Having said that, we haven’t gone to the standard
operating procedure for DVT prophylaxis yet, so I imagine
there is quite a bit of variability across centers, and that
variability might impact the data presented. So, yes, there is
variability but, no, we haven’t assessed its impact.
VTE Prophylaxis in the Trauma ICU
Volume 62 • Number 3