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