The Journal of TRAUMA威 Injury, Infection, and Critical Care
Traumatic Brain Injury in Anticoagulated Patients
David B. Cohen, MD, Charles Rinker, MD, FACS, and Jack E. Wilberger, MD, FACS
Background: Coumadin is widely
used in the elderly population. Despite its
widespread use, little is known about its
effect on the outcome of elderly traumatic
brain-injured patients. This study was undertaken to describe the outcomes of such
a cohort.
Methods: Clinical material was identified from a Level I trauma center prospective head injury database, and a database obtained from the American
College of Surgeons Committee on
Trauma Verification and Review Committee from 1999 to 2002. Both databases
contain many relevant variables, including age, sex, Glasgow Coma Scale (GCS)
score, mechanism of injury, Injury Severity Score, International Normalized Ratio
(INR), computed tomography (CT) findings, operative procedure, time to operating room, complications, length of stay,
and outcome at hospital discharge.
Results: For patients with GCS
scores less than 8, average INR was 6.0,
with almost 50% having an initial value
greater than 5.0. Overall mortality was
91.5%. For the 77 patients with GCS
scores of 13 to 15, average INR was 4.4.
Overall mortality for this group was
80.6%. A subset of patients deteriorated
to a GCS score of less than 10 just hours
after injury, despite most having normal
initial CT scans. Mortality in this group
was 84%.
Conclusions: All patients on warfarin should have an INR performed, and a
CT scan should be done in most anticoagulated patients. All supratherapeutically
anticoagulated patients, as well as any anticoagulated patient with a traumatic CT
abnormality, should be admitted for neurologic observation and consideration
given to short term reversal of anticoagulation. Routine repeat CT scanning at 12
to 18 hours or when even subtle signs of
neurologic worsening occur is a strong
recommendation. A multi-institutional,
prospective trial using these guidelines
would be a first step toward demonstrating improved outcomes in the anticoagulated patient population after head
trauma.
Key Words: Traumatic brain injury,
Anticoagulation, Outcomes.
J Trauma. 2006;60:553–557.
W
arfarin anticoagulation is employed with increasing
frequency for the prevention of thromboembolic
complications of atrial fibrillation, a history of deep
venous thrombosis, extracranial vascular disease, and prosthetic cardiac valves. Because a majority of these medical
problems are diseases of the elderly, as our population ages
the use of anticoagulation is also expected to rise. The risk of
spontaneous hemorrhage in association with anticoagulation
has been well studied, with recent estimates placing the risk
of fatal intracranial hemorrhage at 0.6 to 1.4% per year, major
hemorrhage at 3%, and hemorrhage of any type at 9%. Less
well studied, however, is the role premorbid anticoagulation
may play in the outcome of head-injured patients. This is a
potentially large and growing population, as head injury is the
fifth leading cause of death in the elderly.
It has long been held that mortality and morbidity from
traumatic brain injury (TBI) are significantly increased in
anticoagulated patients. In spite of the impression that this
clinical scenario is increasingly prevalent in the geriatric
population, outcomes data are lacking. Such data has potenSubmitted for publication July 6, 2004.
Accepted for publication June 20, 2005.
Copyright © 2006 by Lippincott Williams & Wilkins, Inc.
From the Department of Neurosurgery, Allegheny General Hospital
(D.B.C., J.E.W), Pittsburgh, Pennsylvania and the Surgical Associates of
Bozeman (C.R.), Bozeman, Montana.
Address for reprints: Jack E. Wilberger, MD, FACS, Department of
Neurosurgery, Allegheny General Hospital, 420 East North Avenue, Suite
302, Pittsburgh, PA 15212; email: jwilberg@wpahs.org.
DOI: 10.1097/01.ta.0000196542.54344.05
Volume 60 • Number 3
tial significance and practical importance in the evaluation
and management of these patients.
The present study was undertaken to determine severity
of head injury in anticoagulated geriatric patients and to
determine the effect of anticoagulation, if any, on outcome.
METHODS
Clinical material was identified from two independent
databases: Level I trauma center prospective TBI database
and a database of prospectively identified variables selected
through chart reviews undertaken by the American College of
Surgeons Committee on Trauma Verification and Review
Committee (VRC) in the course of visits for trauma center
consultations or verifications. Both databases span the years
1999 through 2002.
Variables resident in both databases included age/sex,
Glasgow Coma Scale (GCS) score, mechanism of injury,
International Normalized Ratio (INR), computerized tomographic (CT) scan findings, operative procedure (S), and
outcome at hospital discharge.
Information in the Level I trauma center database also
included Injury Severity Score (ISS), time to operating room
for head injury (HI), complications, and length of stay.
From 1999 through 2002, the VRC performed 405 adult
trauma center site visits. At each, reviewers varying in number from 2 to 5 routinely reviewed at least 10 charts each with
case summaries appended to the final report. Consequently,
during this time frame, approximately 4,000 charts were
reviewed by team members. Forty-nine patients met the cri553
The Journal of TRAUMA威 Injury, Infection, and Critical Care
Table 1 Demographic Data, Evaluation Results and
Outcomes in Anticoagulated Patients with Severe
Traumatic Brain Injury
Characteristic
Level I Database
n
Age range, years (average)
CT scan in emergency department (abnormal)
Admission
International Normalized Ratio range (average)
Injury Severity Score range (average)
Glasgow Coma Scale score average
Surgery
Average time to operating room (hours)
Outcome
Mortality
Vegetative
Severe disability
49
32–93 (65)
100% (100%)
100%
2.8–14.6 (6.5)
4–46 (22.7)
4.7
55.1%
3
87.8%
10.2%
2%
teria for database inclusion: concurrent warfarin therapy and
documented minor closed head injury (GCS score 13–15).
During the same time frame, the Level I trauma center
database included 110 TBI patients. Of these, 47 presented
with GCS score less than or equal to 8, and 28 patients had
GCS scores of 13 to 15.
RESULTS
For patients with GCS scores less than 8 (N ⫽ 49),
demographic data and outcomes are presented in Table 1. The
average age in this group was 65 years and contusions and
acute subdural hematoma were the most frequent CT abnormalities. Average INR was 6.5 with almost 50% having an
initial value greater than 5.0. Twenty-seven (55.1%) underwent craniotomy, of whom six (12.2%) survived to hospital
discharge—all in vegetative condition or with severe disability. Overall mortality was 87.8%.
For the 77 patients with GCS scores of 13 to 15, demographic data and outcomes are presented in Table 2. The
average age in this group was 68 years and fall was the most
common injury mechanism (67%). An INR was obtained in
57% with an average value of 4.4 and values greater than 3.0
in 47%.
Twenty patients were evaluated and sent home from the
emergency department. Of these, 35% had CT scans, all of
which were normal. Eighteen of these patients returned to the
emergency department and were subsequently diagnosed
with a significant traumatic intracranial abnormality; two
patients died at home, one with autopsy-confirmed acute
subdural hematoma. The overall mortality in these 20 patients
was 88.8%.
Forty-five patients were admitted for observation for the
HI and/or treatment of other injuries. CT scans were obtained
before admission in 70%, with only four showing any traumatic intracranial abnormality—three contusions and one
traumatic subarachnoid hemorrhage. Within 8 to 18 hours of
injury (mean 12 hours), 80% deteriorated to a GCS score of
less than 10 with the following CT abnormalities: acute
subdural hematoma in 31%, contusion in 20%, intracerebral
hemorrhage in 20%, and mixed lesions in 29%. Mortality in
this group was 84%.
Twelve patients presented within hours or days of injury
with neurologic findings of an intracranial mass and CT
evidence of a significant traumatic intracranial abnormality.
All underwent emergent craniotomy with a resultant mortality of 83.3%.
Overall mortality for the entire group of 77 minor head
injury patients was 80.6%.
DISCUSSION
The use of anticoagulation for a multitude of medical
diagnoses appears to be gaining popularity and is especially
prevalent in those over 65 years of age. Given that trauma is
the fifth leading cause of death in this age group, with TBI
from falls an increasingly common occurrence, it is highly
likely that emergency physicians, trauma surgeons, and neurosurgeons will be confronted with the clinical scenario of a
TBI in an anticoagulated patient on a regular basis.
Table 2 Demographic Data, Evaluation Results and Outcomes in Anticoagulated Patients with Minor Traumatic
Brain Injury
554
Characteristic
Level I Database
ASCOT VRC Database
n
Age range, years (average)
Computed tomography scan in emergency
department (abnormal)
Admission
International Normalized Ratio range (average)
Time to deterioration (GCS ⬍10), hours (average)
Surgery
Mortality
Outcome
Mortality
Vegetative
Good recovery
28
40–75 (63)
43% (17%)
49
35–95 (74)
77% (10%)
61%
2.9–9.5 (4.6)
8–18 (12)
71%
84%
57%
1.8–9.0 (4.7)
89%
7%
4%
71.4%
4%
22%
81%
March 2006
Head Injury in Anticoagulated Patients
Various studies have documented that anticoagulation is
associated with a risk of spontaneous intracranial hemorrhage
of up to 1% per year;1 the intensity of anticoagulation is a
highly significant predictor of bleeding risk;2,3 and the response to warfarin is exaggerated with advancing age.4
Several authors have linked increased risk of intracranial
hemorrhage to abnormally prolonged prothrombin times
(PT).5,6 Hylek in 1994 suggested that the rate of intracranial
hematoma in the population is equal to an inherent baseline
risk multiplied by the intensity of anticoagulation, finding a
doubling of risk with each 0.5 increase in PT.7
These clinical observations taken singularly or in combination would appear to support concerns that even a seemingly trivial TBI in an anticoagulated patient might significantly increase the risk of morbidity and mortality. However,
the literature to date is divided. Even though there are a
number of published reports studying this issue, findings as
well as recommendations regarding clinical management
have been widely divergent.
Although it has been intuitively held that severe TBI
(GCS score ⬍8) in the setting of premorbid anticoagulation is
a fatal combination, no prior published studies could be found
specifically addressing this issue. The findings of the present
study would appear to support such clinical intuition— overall mortality greater than 90% and mortality greater than
80%, even with aggressive neurosurgical treatment. However, the significance of the seemingly excessive mortality
must be questioned, considering the average age of this study
group (79 years).
Age has consistently proven to be one of the most sensitive and reliable predictors of outcome from TBI. Patients
over 65 years of age will have a mortality rate twice that of
patients under 65 years, even when matched for GCS score
and intracranial pathology.8,9 This is the likely reason that
more attention has been focused on minor TBI (GCS score
13–15) anticoagulated patients in attempts to define the scope
of the problem and establish guidelines for acute management
issues such as CT scanning, hospital admission, and reversal
of anticoagulation.
Published reports began appearing in the mid-1990s in
an attempt to address this issue. In 1995, Saab10 raised significant concern reporting on two patients, both over 65 years
old with fall as a mechanism of TBI. One patient deteriorated
from GCS score of 15 to 5 from an intracerebral hematoma
(ICH) and died. Based on this limited data, recommendations
were made to admit all anticoagulated minor TBI patients,
check, “correct” overly prolonged INRs, and to maintain a
low threshold for CT scanning. The following year, Volans,11
using three personal cases and reviewing eight cases collected
from literature reports, developed a theoretical risk index for
intracranial hemorrhage in anticoagulated patients with minor
TBI. Utilizing the population base prevalent to anticoagulation for atrial fibrillation and minor TBI in the over 65 age
group, and assuming a risk of approximately 10% of a spontaneous intracranial hematoma in an anticoagulated patient,
Volume 60 • Number 3
the risk of a posttraumatic intracranial abnormality in an
anticoagulated patient was calculated to be increased tenfold,
given the same clinical presentation, over a nonanticoagulated patient. Recommendations were forthcoming to check
an INR in this setting and to have a low threshold for CT
scanning. If any traumatic CT abnormalities were identified,
reversal of the anticoagulation was suggested.
Garra et al.12 in 1998 studied 65 anticoagulated patients
with minor TBI and no loss of consciousness. CT was undertaken in 60% without any abnormalities identified. Patients who did not have a CT were followed up by phone and
none reported any subsequent problem. It was concluded that
this population was not at increased risk and CT is rarely
indicated for routine evaluations. However, in the 38 patients
in whom PT was checked, none was above 30 seconds and
almost one-third were less than 14 seconds, indicating that
even though these patients were on warfarin, few were anticoagulated.
CT was recommended in every anticoagulated patient
after TBI by Li13 based on a retrospective review of 144
elderly patients. Utilizing information from large cohort studies, a 4% incidence of intracranial injury was anticipated
while a 7% incidence was found. Although GCS scores were
not reported, patients presented with dizziness, headache, or
no symptoms. Mean INR was 2.1 both in patients with and
without an abnormal CT. No outcome information was provided.
In a large retrospective study of 2,142 patients taken
from the Pennsylvania Trauma System Registry, Wojcik et
al.14 identified 1,986 patients with preinjury anticoagulation.
No statistically significant difference in outcome was found
between the Warfarin and no anticoagulation patients in both
the TBI and non-TBI groups. Likewise, when stratified by
GCS score (3– 8, 9 –12, 13–15), there was no difference in
mortality between the anticoagulated and nonanticoagulated
groups. Mean GCS score was 14.13, indicating predominantly minor TBI was studied. INR data were not available
from the registry, thus making it impossible to determine the
degree of anticoagulation in these patients, if at all.
The triad of anticoagulation, age over 65, and TBI was
considered lethal by Karni et al.15 on retrospective review of
278 patients with TBI and CT documented intracranial hemorrhage. Sixteen of these patients were on warfarin with an
average INR of 3.0. Thirty-day mortality was 50% in this
group compared with 20% in a matched nonanticoagulated
cohort. In those with INR greater than 3.5, mortality was
75%. However, average GCS score in this study was 11, with
a median age of 78. Thus, a confounding effect of age on
outcome after more severe HI cannot be excluded.
Mina et al.16 found a four- to fivefold increased risk of
mortality from TBI in anticoagulated patients. Ten percent of
380 anticoagulated patients admitted to a Level I trauma
center with intracranial injuries were identified. Of these, 12
were on warfarin, the rest on antiplatelet medications. Mean
GCS score was 11.8 and average INR was 2.37. Compared
555
The Journal of TRAUMA威 Injury, Infection, and Critical Care
with a matched control group of anticoagulated patients without TBI, mortality was 38% versus 8% ( p ⫽ 0.006). Mortality was 33% in those on warfarin and 50% in patients on
aspirin.
Although all the previously cited studies primarily focus
on “minor” TBI, only two provide a specific breakdown by
GCS score.10,14 Thus, the present study provides the most
comprehensive information to date on patients with GCS
scores of 13 to 15; the results appear to clarify and amplify
concerns expressed in prior studies. Interpretation of this
study is, however, hampered by lack of a matched control.
Nevertheless, two concerning findings dominate. The majority of patients were supratherapeutically anticoagulated and,
of those undergoing CT on initial presentation, only slightly
more than 30% had any evidence of traumatic intracranial
abnormality.
It has been noted that advancing age appears to exaggerate the anticoagulant effect of warfarin. However, many other
confounding factors may be involved. The most well known
are concomitant medications that may potentiate the anticoagulant effects of Warfarin. Liver disease, hypoproteinemia,
and a diet deficient in vitamin K may all result in increased
anticoagulant response.17 Patient compliance with warfarin
dosing regimens would obviously effect the degree of anticoagulation. How or whether these factors may have been
operative in the supratherapeutically anticoagulated patients
in this study cannot be determined. However, the established
relationship between anticoagulation intensity and risk of
intracranial hemorrhage cannot be overlooked given the findings in this study.
The “delayed” appearance of traumatic intracranial abnormalities documented in this study is of concern. Delayed
appearance of intracranial hematomas is well recognized.
Delayed traumatic intracerebral hematomas were originally
described by Bollinger et al. in 1891.18 Mortality associated
with this entity ranges from 50 to 75% with a high number of
surviving patients having a poor outcome.19,20 Delayed acute
subdural hematoma makes up approximately 0.5% of operatively treated acute subdurals and are typically associated
with other parenchymal lesions. Although delayed appearance of intracranial hematomas most frequently occurs in the
setting of severe TBI, they have been associated with
coagulopathy.21
Conversely, because the majority of patients in this study
did not have a CT until clinical deterioration, they may have
harbored an intracranial hemorrhage that “enlarged” secondary to underlying anticoagulation. It is well known that minor
TBI patients have a clear risk of significant pathologic CT
scan findings— up to 10% with GCS scores of 15; 20% with
GCS scores of 14; and 35% with GCS scores of 13.22–24
Although a number of guidelines have been developed
for the clinical approach to the evaluation and treatment of
minor TBI, our findings would indicate a need to significantly revise these in a setting of anticoagulation. As part of
the initial evaluation, all patients on warfarin should have
556
their INR checked. A CT scan should be obtained in all
anticoagulated patients presenting with GCS scores of 13 or
14, given the relatively high risk of potential traumatic intracranial abnormalities with these GCS scores, as well as in all
patients with GCS scores of 15 who are supratherapeutically
anticoagulated because of the known increased bleeding risk
with increasing levels of anticoagulation. All supratherapeutically anticoagulated patients, as well as any anticoagulated
patient with a traumatic CT abnormality, should be admitted
for neurologic observation and consideration given to short
term reversal of the warfarin. Supratherapeutic anticoagulation should at least be reversed to therapeutic levels.
Strong consideration must be given to routine repeat CT
scanning at 12 to 18 hours or when even subtle signs of
neurologic worsening occur. The data in this study indicates
that when neurologic decline occurs, it is precipitous and
devastating. A multi-institutional, prospective trial using
these guidelines would be a first step toward demonstrating
improved outcomes in the anticoagulated patient population
after TBI.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Hart RG, Boop BS, Anderson DC. Oral anticoagulants and
intracranial hemorrhage. Facts and hypotheses. Stroke. 1995;
26:1471–1477.
Petitti DB, Strom BL, Melmon KL. Prothrombin time ratio and other
factors associated with bleeding in patients treated with warfarin.
J Clin Epidemiol. 1989;42:759 –764.
Landefeld CS, Rosenblatt MW, Goldman L. Bleeding in outpatients
treated with warfarin, relation to the prothrombin time and important
remediable lesions. Am J Med. 1989;87:153–159.
Gurwitz JH, Avorn J, Ross-Degnan D, et al. Aging and the
anticoagulant response to warfarin therapy. Ann Int Med. 1992;
116:901–904.
Kase CS, Robinson RK, Stein RW, et al. Anticoagulant related
intracerebral hemorrhage. Neurology. 1985;35:943–948.
Wintzen AR, deJonge H, Loeliger EA, et al. The Risk of
Intracerebral Hemorrhage During Oral Anticoagulant Treatment: A
Population Study. Ann Neurol. 1984;16:553–558.
Hylek EM, Singer DE. Risk Factors for Intracranial Hemorrhage in
Outpatients Taking Warfarin. Ann Int Med. 1994;120:897–902.
Braakman R, Gelpke GJ, Habbema JDF, et al. Systematic Selection
of Prognostic Features in Patients with Severe Head Injury.
Neurosurgery. 1980;6:362–370.
Teasdale G, Skeene A, Spiegelhalter D, et al. Age, Severity and
Outcome of Head Injury. Grossman RG, Gildenberg PL, eds. Head
Injury: Basic and Clinical Aspects. New York: Rabin Press, 1982:
213–220.
Saab M, Gray A, Hodgkinson D, et al. Warfarin and the Apparent
Minor Head Injury. J Accid Emerg Med. 1996;13:208 –209.
Volans AP. The Risks of Minor Head Injury in the Warfarinized
Patient. J Accid Emerg Med. 1998;15:159 –161.
Garra G, Nashed AH, Capobianco L. Minor Head Trauma in
Anticoagulated Patients. Acad Emerg Med. 1999;6:121–124.
Li J, Brown J, Levine M. Mild Head Injury, Anticoagulants and
Risks of Intracranial Injury. Lancet. 2001;357:771–772.
Wojcik R, Cipolle MD, Seislove E, et al. Preinjury Warfarin Does
Not Impact Outcome in Trauma Patients. J Trauma. 2001;51:1147–
1152.
Karni A, Holtzman R, Bass T, et al. Traumatic Head Injury in the
Anticoagulated Elderly Patient: A Lethal Combination. Am Surg.
March 2006
Head Injury in Anticoagulated Patients
2001;67:1098 –1100.
Mina AA, Knipfer JF, Park DY, et al. Intracranial Complications of
Pre-Injury Anticoagulation and Trauma Patients With Head Injury.
J Trauma. 2002;53:668 – 672.
17. Gilman AG, Rall TW, Nies AS, Taylor P. Goodman and Gilman’s
The Pharmacologic Basis of Therapeutics. New York: Pergamon
Press, 1998.
18. Bollinger O, Levenjhah RES, Berlan A: Ueber Traumatische
Spatapoplexia. Ein Beitrag zor Lehre von der Hirnerschutterung.
Berlin: Hirschwald. 1891;70:457– 470.
19. Diaz FG, Yock DH Jr., Larson D, et al. Early Diagnosis of Delayed
Posttraumatic Intracerebral Hematomas. J Neurosurg. 1979;50:217–
223.
20. Rockswold GL, Leonard PR, Nagib MG. Analysis and Management
of Thirty-Three Closed Head Injury Patients Who “Talked and
16.
Volume 60 • Number 3
Deteriorated.” Neurosurgery. 1987;21:51–55.
Reynolds FD, Dietz PA, Higgins D, et al. Time to Deterioration of
the Elderly, Anticoagulated, Minor Head Injury Patient Who
Presents without Evidence of Neurologic Abnormality. J Trauma.
2003;54:492– 496.
22. Culotta VP, Sementilli ME, Gerold K, et al. Clinicopathological
Heterogeneity in the Classification of Mild Head Injury.
Neurosurgery. 1996;38:245–250.
23. Jeret JS, Mandell M, Anziska B, et al. Clinical Predictors of
Abnormality Disclosed by Computed Tomography after Mild Head
Trauma. Neurosurgery. 1993;32:9 –16.
24. Shackford SR, Wald SL, Ross SE, et al. The Clinical Utility of
Computed Tomographic Scanning and Neurologic Examination in
the Management of Patients with Minor Head Injuries. J Trauma.
1992;33:385–394.
21.
557