Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Defining Acute Mild Head Injury in Adults: A Proposal Based on Prognostic Factors, Diagnosis, and Management
FRANCO SERVADEI,1 GRAHAM TEASDALE,2 and GLEN MERRY,3 on behalf of the NEUROTRAUMATOLOGY COMMITTEE OF THE WORLD FEDERATION OF NEUROSURGICAL SOCIETIES
ABSTRACT The lack of a common, widely acceptable criterion for the definition of trivial, minor, or mild head injury has led to confusion and difficulty in comparing findings in published series. This review proposes that acute head-injured patients previously described as minor, mild, or trivial are defined as mild head injury, and that further groups are recognized and classified as low-risk mild head injury, medium risk mild head injury, or high-risk mild head injury. Low-risk mild injury patients are those with a Glasgow Coma Score (GCS) of 15 and without a history of loss of consciousness, amnesia, vomiting, or diffuse headache. The risk of intracranial hematoma requiring surgical evacuation is definitively less than 0.1:100. These patients can be sent home with written recommendations. Medium risk mild injury patients have a GCS of 15 and one or more of the following symptoms: loss of consciousness, amnesia, vomiting, or diffuse headache. The risk of intracranial hematoma requiring surgical evacuation is in the range of 13:100. Where there is one computed tomography (CT) scanner available in an area for 100,000 people or less, a CT scan should be obtained for such patients. If CT scanning is not so readily available, adults should have a skull x-ray and, if this shows a fracture, should be moved to the high-risk category and undergo CT scanning. High-risk mild head injury patients are those with an admission GCS of 14 or 15, with a skull fracture and/or neurological deficits. The risk of intracranial hematoma requiring surgical evacuation is in the range 610:100. If a CT scan is available for 500,000 people or less, this examination must be obtained. Patients with one of the following risk factorscoagulopathy, drug or alcohol consumption, previous neurosurgical procedures, pretrauma epilepsy, or age over 60 years are included in the high-risk group independent of the clinical presentation. Key words: mild head injury; traumatic hematomas; computed tomography scan; skull x-ray INTRODUCTION
admission, with a loss of consciousness (LOS) of less than 20 min and who were admitted to hospital for less than 48 h. Since then many other methods for defining head injury as minor or mild have been put forward (Table 1). The aim of this paper is to propose criteria for
minor head injury was first used in a paper by Rimel et al. (1981). They applied this term to patients with a Glasgow Coma Score (GCS) of 1315 on
HE TERM
1 WHO 2 University
Neurotrauma Collaborating Center, Ospedale Bufalini, Cesena, Italy. Department of Neurosurgery, Southern General Hospital, Glasgow, United Kingdom. 3 Department of Neurosurgery, University of Queensland, Royal Brisbane Hospital, Queensland, Australia.
657
SERVADEI ET AL.
TABLE 1. VARIOUS DEFINITIONS References Miller et al., 1990; Poon et al., 1992 Miller et al., 1985; Miller, 1986 Feuerman et al., 1988 Jeret et al., 1993; Lee et al., 1995 Borckzuck, 1995; Dacey et al., 1986; Shackford et al., 1992; Stein et al., 1990 Servadei et al., 1993 Mohanty et al., 1991 Gomez et al., 1996; Hsiang et al., 1997 LOC, loss of consciousness. Terminology Minor Minor Minor Mild Minor to mild
OF
MINOR
TO
1315 15 1315
Skull fracture
classifying a patient with a head injury as mild and, within this spectrum, to identify those at higher risk of developing an intracranial complication. The perspective therefore is neurosurgical and concerned with acute management, especially the early detection of intracranial hematomas. Our paper does not concern itself with the controversy about the extent and basis of any neuropsychological sequelae of mild head injury.
PROCESS
A search was conducted of publications listed in Medline from 1978 to 1998 employing the terms minor head injury and mild head injury. A total of 485 papers were identified, of which 42 were considered to be relevant to our purpose. Reports limited to children were excluded, but some papers did not differentiate children from adults and are therefore included. The initial methodology is similar to that used by a group of American neurosurgeons on the guidelines for severe head injury management (Bullock et al., 1996), but we have not classified the papers according to classes of evidence.
the likelihood of a hematoma requiring evacuation has been related to the mechanism of head injury. Such complications are significantly more frequent after an assault, p , 0.01 (Harad and Kerstein, 1992; Jeret et al., 1993) or when pedestrians or cyclists have been struck by motor vehicles, p , 0.0004 (Borkzuck, 1995) and p , 0.001 (Jeret et al., 1993). Harad and Kerstein (1992), in a report on a cohort of mild head-injured patients (GCS 1315), commented that only those who had been assaulted or suffered a gunshot wound required an operation. The differences in causation of injury in different societies must be considered when comparing reports from different areas of the world, in particular in assessing the frequency of occurrence of lesions requiring operation. Age. Increasing age is associated with a higher risk of intracranial lesions on CT scan (p , 0.01, Lee et al., 1995; p , 0.004, Gomez et al., 1996; p , 0.0005, Borzuck, 1995, and Jeret et al., 1993). Arienta et al. (1997) reported that age over 60 was a high-risk factor in patients with minor head injury. Nevertheless, it is unlikely that there is a specific age threshold above which the risk suddenly increases. Indeed, the underlying pattern is more of a gradual progressive increase in intracranial complications, which becomes notable and clinically relevant at around 60 years of age.
Clinical Parameters
Focal neurological deficits. The finding of a focal neurological deficit in an apparently mild head-injured patient is a powerful predictor for the development of an
658
DEFINING ACUTE MILD HEAD INJURY IN ADULTS intracranial hematoma. Miller et al. (1990) reported, in a series of patients whose GCS was 15 when first seen at the hospital but who subsequently developed an intracranial hematoma, that one-third of cases had a focal neurological deficit on arrival at the hospital. Shackford et al. (1992), in a multicenter series of 2,766 patients with a GCS of 1315 with a history of LOC, reported that an abnormal neurological examination at the time of admission increased by a factor of three the risk of a subsequent surgical intervention. The finding of an abnormal neurological examination is more likely with a lower GCS score, particularly with a GCS of 13 (Gomez et al., 1996). However, the circumstances of admission to a busy accident and emergency observation ward do not ensure an accurate or frequent neurological examination (Brown et al., 1994; Ingebritsen et al., 1997). Precise guidelines are therefore required to detect neurological deficits. Headache and vomiting. These symptoms are common in patients with mild head injury. They have been reported as often in cases with or without intracranial lesions (Hsiang et al., 1997; Miller et al., 1990). In contrast, Lee et al. (1995) reported that headache or vomiting was significantly (p , 0.001) related to the occurrence of subsequent clinical deterioration in a group of patients with a GCS of 15 on admission to hospital. Loss of consciousness/amnesia. The occurrence of a transient LOC or a period of amnesia has been taken as a conditio sine qua non to differentiate a brain injury from a scalp bruise or a skull injury. The duration of posttraumatic amnesia (PTA) has been accepted for many years as an index of the brain injury severity (Dickmen and Levin, 1993). In a large study of fully conscious patients, it was estimated that a history of altered consciousness increased the risk of traumatic intracranial hematoma by a factor of five, that is, from 1/31,370 to 1/6,663 (Teasdale et al., 1990). However the absolute risk remained quite small. The occurrence and duration of LOC can be difficult to quantify in practice, as a reliable witness is not always available. In one study, the recording of questionable LOC/amnesia was not correlated with a higher incidence of abnormalities on CT scans (Borczuk, 1995). It is recognized that patients without LOC but with a skull fracture may develop an epidural hematoma (Nee et al., 1993). Moreover, in children, an epidural hematoma can develop in the absence of either LOC or skull fracture (Servadei et al., 1993). Glasgow Coma Score. Although Rimel et al. (1981) specified a GCS of 1315, other criteria have been used. Table 2 shows the relationship between the GCS on admission, the finding on CT scans, and the subsequent
TABLE 2. RELATIONSHIP BETWEEN GLASGOW COMA SCORE ON ADMISSION AND PRESENCE OF INTRACRANIAL LESION /NEED FOR SURGERY IN MINOR HEAD INJURED PATIENTS References Cullotta et al., 1996 Study Retrospective No. of cases GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS GCS 13 14 15 13 14 15 13 14 15 13 14 15 13 14 15 13 14 15 13 14 15 176 796 2398 62 142 454 200 512 1454 45 88 2351 40 197 1211 34 103 236 45 138 1177 Percentage of intracranial lesion 28 16 4 40 22 13 33 18 15 27 23 17 27 18 6 38 19 8 58 35 18 Percentage of craniotomy 4.5 1.6 0.4 13 6 3.3 11 3.8 3.2 17 6 0.3 7.5 3.6 0.8 5.8 3.8 0.8 20 5.1 2.2
Retrospective
Prospective
Retrospective
Borczuk, 1995
Retrospective
Retrospective
Prospective
659
SERVADEI ET AL. need for surgical treatment of an intracranial lesion. These two featuresGCS and CTcan each be an index of severity of injury since they are related to outcome (Gomez et al., 1996; Hsinag et al., 1997) as determined by the Glasgow Outcome Scale (Jennet and Bond, 1975). Patients with a GCS of 13 have been shown, in two prospective series (Fearnside and McDougall, 1998; Stein and Ross, 1994), to harbor intracranial lesions at a frequency similar to patients with moderate head injury defined as GCS of 912. It must be noted that the definition of an intracranial lesion has varied. As one of the possible definitions, Shackford et al. (1992) suggested that an abnormal CT scan was one that showed any intracranial abnormality even if it was not related to injury, a positive scan was one that demonstrated lesions of the skull or brain related to injury, and a relevant positive scan was one that showed acute changes of intracranial contents but excluded skull fracture. Obviously, only the presence of a relevant positive CT scan was related to outcome. Master et al. (1987), found no difference in the number of intracranial hematomas detected in patients with or without a skull fracture. It must be pointed out that Masters et al. (1987) included patients of any age (the risk related to the presence of a skull fracture decreases with age, and it is significant only over the age of 10 years; Chan et al., 1990) and with any level of consciousness (including comatose patients where current policy is to perform an urgent CT scan) and where adequate data on follow-up was not provided. Reanalysis of Masters own material in the cases matching our definition of mild head injury showed that the risk of an intracranial hematoma was 6.4% in the patients with a skull fracture but only 0.6% in those without a fracture.
Skull Fracture
Table 3 summarizes the findings reported in several papers. Although the topic is controversial, the prevailing finding in reports made from a neurosurgical perspective is that the presence of a skull fracture in a patient with a mild head injury increases the risk of intracranial lesions by a factor that is both statistically and clinically highly significant. In one study, in patients with a comparable level of consciousness, the finding of a fracture was associated with a 174-fold increased risk (Teasdale et al., 1990). In contrast, radiologists opinions, as summarized by
TABLE 3. SOME OF THE PUBLISHED STUDIES ABOUT THE V ALUE OF THE SKULL FRACTURE AS A RISK FACTOR OF INTRACRANIAL H EMATOMAS IN MINOR HEAD -INJURED PATIENTS References Dacey et al., 1986 Servadei et al., 1988 Fuermann et al., 1988 Teasdale et al., 1990 Location Virginia Italy California U.K. Departments involved Neurosurgery Neurosurgery, Casualty Neurosurgery Accident and Emergency Neurosurgery Neurosurgery, Orthopedics, Pediatrics Neurosurgery Trauma Centers Emergency Department Neurosurgery Study Prospective Prospective Retrospective Retrospective/ prospective Prospective Statistical analysis x2 x2 No statistics Log regression Risk factor p , 0.001 p , 0.01 No risk factors p , 0.0000
Denmark
x2
No risk factors
Stein and Ross, 1992 Shackford et al., 1992 Gomez et al., 1996 Hsiang et al., 1997
660
DEFINING ACUTE MILD HEAD INJURY IN ADULTS 13; recommended for GCS 14 and 15 (Sahckford et al., 1992) 6. Any one or more of the following: (a) focal neurological deficits, (b) basilar skull fracture, (c) scalp injury, and (d) age over 60 years (Borczuk, 1995) These varying recommendations lead to controversy and clinical confusion. The establishment of a consistent policy is the main reason for wishing to develop a coherent risk-based approach to classification. Furthermore, the application of the categories proposed for investigation need to be related to the availability of CT scanning. Patients with a GCS of 14 and 15 seen within 12 h of injury are defined as having an acute mild head injury. With this population, three levels of risk can be identified (Table 4). Patients with a GCS of 15 on admission but without a history of LOC, amnesia, vomiting, or diffuse headache are at low risk (Hsiang et al., 1997; The Study Group, 1996). The risk of intracranial hematoma requiring surgical evacuation is definitively less than 0.1:100. Patients with a GCS of 15 but with one or more of the foregoing symptoms are at medium risk (Borczuk, 1995; Stein and Ross, 1990), that is, one or more of LOC, amnesia, vomiting, or diffuse headache. This includes patients who have recovered from an altered state of consciousness (so-called concussion). The risk of intracranial hematoma requiring surgical evacuation is in the range of 13:100. Patients with an admission GCS of 14, or a GCS of 15 with a skull fracture and/or neurological deficits are at high risk (Harad and Kerstein, 1992; Servadei et al., 1988; Shackford et al., 1992). The risk of intracranial hematoma requiring surgical evacuation is in the range 610:100. Patients with an admission GCS of 15 with or without clinical feedings, absent neurological deficits, and absent skull fracture, but with one of the following risk factorscoagulopathy, drug or alcohol consumption, previous neurosurgical procedures, pretrauma epilepsy, or age over 60 yearsare included in the high-risk group independent of clinical presentation (Arienta et al., 1997; Borczuk, 1995; Miller et al., 1986; The Study Group, 1996).
Proposed Definitions
A mild head injury is characterized as having a GCS of 14 or 15. Patients who, on arrival, have a GCS of 13 are excluded from the group of mild head-injured patients because their risk of intracranial lesions is similar to that in the moderately (GCS 912) head-injured patients (Fearnside and McDougall, 1998; Stein and Ross, 1994).
TABLE 4. LEVEL Definition Low risk Medium risk High risk High risk High risk High risk
aOne b Coagulopathy, drug or
OF
RISK
FOR
IN
Neurological deficits
Absent Absent Present Absent May or not be associated with other clinical or Present/absent Present Present/absent Absent Present/absent Absent
Absent Absent Absent Absent radiological findings. Absent Absent Present Absent Absent Present
or more of loss of consciousness, amnesia, vomiting, and diffuse headache. alcohol consumption, previous neurosurgical procedures, pretrauma epilepsy, and age over 60 years.
661
SERVADEI ET AL. hematoma is not zero, but the costbenefit ratio is against any radiological examination. Medium risk. Medium risk patients have a higher risk of intracranial hematoma. Where there is one CT scanner available in an area of 100,000 people or less, a CT scan should be obtained for such patients. Published data show an incidence of 200350 cases of hospital admissions for mild head injury per 105 inhabitants per year (WHO Collaborating Centers, 1995), resulting in an average of one to two examinations per day for patients in this group. The CT examination must include bone windows; if these show a skull fracture, the patient should be admitted for 24 h observation since cases of delayed epidural hematomas have been reported (Riesgo et al., 1997). If CT scanning is not so readily available, adults should have a skull x-ray and, if this shows a fracture, should be moved to the high-risk category and undergo CT scanning. High risk. In the high-risk group, the risk of an intracranial lesion is 1520% and the risk of a surgically significant lesion 26% (Table 2). If a CT scan is available for 500,000 people or less, this examination must be obtained. If CT scanning is not available, admission and close clinical observation for at least 48 h is essential in such patients. Guidelines for early detection of an intracranial hematoma, in addition to identifying at-risk patients, must also indicate when to perform a CT examination. Clinical deterioration related to an epidural hematoma occurs mainly in the first 6 h after injury (Knuckey et al., 1989; Servadei et al., 1995). In a series of delayed epidural hematomas, only two out of 31 (6%) appeared in mild head-injured patients (Poon et al., 1992). The risk of missing a delayed hematoma is overshadowed by the benefits of early hematoma detection. A CT scan should be obtained as soon as possible, at most within 4 h of the injury. Most cases of delayed epidural hematomas are identified on admission as small collections that subsequently enlarge (Smith and Miller, 1991). The detection and interpretation of the importance of small collections and other findings (such as lucent areas within the hematoma and the presence of air) require expertise, and we advise that the scan of these patients should be reviewed by a neurosurgeon/neuroradiologist (also with a teleradiology system) because transfer may be needed, even if there is not an indication for immediate operation. In areas where neurosurgery is centralized, early transfer can be life saving because any subsequent clinical deterioration and/or hematoma enlargement will occur where the patient can be treated surgically most expeditiously. In conclusion, Teasdales (1990) expectation that CT scanning to all head injured patients in coma or alert with a skull fracture should achieve early detection of 95% of intracranial hematoma has been shown to be close to reality. Other expectations that mortality for pure epidural hematoma could be as low as zero (Bricolo and Pasut, 1984) have also been proved to be feasible (Servadei et al., 1995). Further substantial reductions in mortality for head injury, along with spectacular economic savings, are less likely to be achieved by increasingly elaborate medical management in sophisticated institutions than by the prevention of deterioration in patients who initially appear to be at low risk (Klauber et al., 1989) and who are often located outside neurosurgical units.
ACKNOWLEDGMENTS
We wish to thank Prof. Lawrence F. Marshall, Department of Neurosurgery, UCSD, San Diego, for his valuable advice in preparing the manuscript.
REFERENCES
ARIENTA, C., CAROLI, M., and BALBI, S. (1997). Management of head-injured patients in the emergency department: a practical protocol. Surg. Neurol. 48, 213219. BORCZUK, P. (1995). Predictors of intracranial injury in patients with mild head trauma. Ann. Emerg. Med. 25, 731736. BRICOLO, A.P., and PASUT, L.M. (1984). Extradural hematoma: toward zero mortality. A prospective study. Neurosurgery 14, 812. BROWN, S.R., RAINE, C., ROBERTSON, C.E., et al. (1994). Management of minor head injuries in the accident and emergency department: the effect of an observation ward. J. Acc. Emerg. Med. 11, 144148. BULLOCK, R., CHESNUT, R.M., CLIFTON, G., et al. (1996). Guidelines for the management of severe head injury. J. Neurotrauma 13, 643645. CHAN, K.H., MANN, K.S., YUE, C.P., et al. (1990). The significance of skull fracture in acute traumatic intracranial hematomas in adolescents: a prospective study. J. Neurosurg. 72, 189194. CULOTTA, V.P., SEMENTILLI, M.E., GEROLD, K., et al. (1996). Clinicopathological heterogeneity in the classification of mild head injury. Neurosurgery 38, 245250. DACEY, R.G., ALVES, W., RIMEL, R., et al. (1986). Neurosurgical complications after apparently minor head injury.
662
663
SERVADEI ET AL.
THE STUDY GROUP ON HEAD INJURY OF THE ITALIAN SOCIETY FOR NEUROSURGERY. (1996). Guidelines for minor head injured patients management in adult age. J. Neurosurg. Sci. 40, 1115. WHO COLLABORATING CENTERS FOR NEUROTRAUMA. (1995). Prevention, Critical Care and Rehabilitation of Neurotrauma. Perspectives and Future Strategies. WHO: Geneva.
Address reprint requests to: Franco Servadei, M.D. WHO Neurotrauma Collaborating Center Division of Neurosurgery Ospedale M. Bufalini 47023 Cesena, Italy E-mail: fservade@ausl-cesena.emr.it
664
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.