Part Two

MEASURING AND MONITORING INJURY

Introduction CT scanning has provided a fast, non-invasive and detailed method of assessing the morphology of brain injury and is now an essential part of any trauma management system. CT scanners are increasingly available at the hospital of first contact and connected to a central trauma center by teleradiology. This has led to an increase in the speed of diagnosis and treatment of surgical hematomas, better planning and retrieval, and more discriminating admission policies for patients with severe head injury or skull fracture. Our limited understanding of the pathophysiological events that occur after brain injury, and the high frequency of secondary delayed neurological deterioration, have stimulated the search for accurate, continuous monitoring techniques, particularly for the first days after injury. Some of these techniques are listed in Table II.1. None of the present techniques are ideal, and academic head injury centers continue to search for better methods. The reasons for monitoring the injured brain are simple:   

to detect harmful pathophysiological events before they cause irreversible damage to the brain; to allow the type of harmful pathophysiological process to be diagnosed and effectively treated; to provide ‘on-line’ feedback to guide therapy directed at these processes.

In addition they should be non-invasive and preferably cheap. In general, monitoring techniques may be divided into two types – those that assess substrate delivery to

the injured brain and those that assess brain function. Brain function is particularly difficult to assess, since most severely head-injured patients are in coma. Monitoring techniques that aim at providing continuous or near continuous information must be differentiated from intermittent measurement techniques (e.g. CT scanning) that provide a static image of the injured brain at intervals. Modern management clearly uses both techniques.

Duration of monitoring The optimal duration of monitoring will vary from patient to patient, depending upon the pathophysiological process in question. In a recent large series from the Medical College of Virginia, the mean duration of ICP monitoring ranged between 5 and 7 days. Often it is not practical to continue to monitor a patient until consciousness returns. On the other hand, it is seldom necessary to continue monitoring once a patient is able to obey commands; the patient can then be followed adequately with the neurological exam. It is desirable to continue monitoring until ICP is beginning to decline and until cerebrovascular autoregulation has been reestablished. Studies have shown that autoregulation is generally recovering by the seventh day after injury and is almost always back to normal for all three physiological stimuli – blood pressure, PaCO2 and PO2 – by 2 weeks after injury. In the following chapters, monitoring of ICP, CBF, transcranial Doppler, brain oxygen uptake and electrophysiological function are highlighted. Many of these techniques are complementary.

144

INTRODUCTION

Table II.1

Advantages and disadvantages of monitoring techniques used for acute brain injury

Method

Advantages

Disadvantages

Approximate cost (US$; 1995)

Invasiveness/ risk

Techniques for assessing substrate delivery ICP Monitoring Ventriculostomy

May drain CSF to lower ICP; may calculate CPP; may use waveform analysis and measure PVI.

Highest infection risk (5–10%). Hemorrhage risk ~ 0.5%. May be difficult to insert.

450

++++

Easy to insert and use. Lower infection/ hemorrhage risk; may use waveform analysis.

Cannot drain CSF. Difficult to recalibrate.

Camino 6500 Each sensor 485

++++

May select the brain region to monitor. May guide hyperventilation therapy.

Measures relative change, not absolute flow. Thermal diffusion requires a craniotomy for optimal insertion.

Instrument 4700–16 500 Sensor 300

+++

May guide hyperventilation and pressor therapy.

Intermittent. Indirect measure of brain O2 extraction and flow. Contaminated by extracranial venous flow.

Catheter ≈ 100 Sample analysis 500/day

+ +

Fiberoptic sensor e.g. Oxymetrix

Continuous.

Inaccurate in up to 40% of readings.

28 000 (instrument) 765 (each sensor)

Transcutaneous near-infrared spectroscopy (NIRS)

Correlates well with brain O2 (Hammamatsu system)

Accuracy/specificity not yet proven in trauma. Not reliable when intracranial bleeding is present.

50 000 (Hammamatsu system)

–

Brain oxygen measurement

Reflects ‘true’ substrate delivery: accurate. (May measure, CO2 , pH and temperature).

Fragile; microregional.

20 000 (system) 300 (each sensor)

+++

Transcranial Doppler

Wave-form analysis may indicate high ICP. Detects vasospasm in ~ 28%.

Qualitative; difficult to fix the head; operatordependent. Significance unclear.

15 000–30 000 (system)

–

Microdialysis

Many analytes available for analysis; highly sensitive.

Not ‘on-line’; requires HPLC for analysis; laborintensive.

20 000 (system) 26 000 (HPLC) 150 (probes)

++

Parenchymal electronic sensors e.g. Camino, Codman CBF flow probes Thermal diffusion or laser Doppler

AVDO2 Jugular catheter

Techniques for assessing brain function Neurological observation (coma scale)

Most sensitive and specific indicator of brain function.

Lost in coma; may be preserved while focal damage is occurring, in ‘silent’ areas?

Nursing time

–

EEG/EEG Spectral array

Sensitive, even in coma; indicates seizures, even when patient is on muscle relaxants. Useful in prognosis.

Difficult to interpret; non-specific; significance uncertain.

system 20 000–80 000

–

Evoked potentials

Loss of evoked potentials correlates well with death/ vegetative outcome. Useful for prognosis. May detect focal deficit or spinal injuries.

Not a useful guide to therapy. Operatordependent.

system 40 000–100 000

–

8

CLINICAL EXAMINATION AND GRADING Donald A. Simpson

8.1

Introduction

8.1.1 ROLES AND LIMITATIONS OF CLINICAL EXAMINATION

Severe closed head injuries are now routinely investigated by early computed tomography (CT), which visualizes most pathological lesions of immediate surgical importance. It is also routine practice to monitor the intracranial pressure and other parameters of cerebral physiology, providing objective data to control the use of artificial ventilation and other forms of conservative therapy. It is therefore legitimate to ask what now are the roles of clinical neurology in the management of head injuries in general, and severe head injuries in particular. The initial clinical evaluation is still crucially important, in triage and as a baseline in assessing progress. The prognosis depends to a large extent on the findings of the initial examination, and the neurological status at a specified time after injury is widely used as a measure of head injury severity. Moreover, valuable as they are, the neuroradiological findings must be interpreted in the light of the clinical findings. Thus clinical examination of the head-injured patient continues to be indispensable. But modern strategies of severe closed head injury management have brought one very important change in the nature of this examination: because it is often wise to perform endotracheal intubation as soon as possible, the first neurological examination is now usually performed at the accident site or in the emergency room, and often by someone with no special training in neurology. This means that the paramedic, the intensivist and the emergency physician must be competent in performing an appropriate neurological examination before intubation and respiratory paralysis are instituted. This does not mean that all the rites of medical neurology should be taught to everyone who may intubate an unconscious patient; it means that such persons must be skilled in making a few basic neurological observations and recording them accu-

rately. These observations are usually done in two phases. In the primary survey, the conscious level and the pupillary reactions are tested (Table 8.1). If resuscitation includes endotracheal intubation, limb movements should be quickly tested before a muscle relaxant is given. In the secondary survey, done after resuscitation, these findings are checked, and in addition the examiner assesses limb motor function and, if possible, vision and cutaneous sensation (Table 8.2). A definitive or tertiary examination by a surgical neurologist retains its value in the late evaluation of a severe head injury. It is often done in collaboration with a neuropsychologist. This definitive examination considers especially the neurological functions that cannot be tested in the unconscious patient, notably speech, mentality, cognitive functions, smell, vision, hearing and sensorimotor function. The definitive assessment also includes a retrospective judgment of the duration of unconsciousness and amnesia (see below). The long-term effects of severe head injury are outside the scope of this book, but the care of a headinjured person should not be compartmentalized: evaluation – like counseling – is a continuous process and the early findings bear important relationships with what is found as the patient emerges from coma. 8.1.2

THE HISTORY

The history retains its importance in the evaluation of severe head injury. It is usually obtained from eyewitnesses of the accident and from the family or friends of the injured person. The site and cause of the impact may give clues to the pathology; blunt weapons, falls and road crashes show more or less characteristic patterns of intracranial damage. The sequence of events after impact may distinguish between primary brain damage and secondary cerebral compression. The health before injury may be relevant. Medicalert bracelets identify serious illness or medication, and are often helpful in unconscious patients. However people

Head Injury. Edited by Peter Reilly and Ross Bullock. Published in 1997 by Chapman & Hall, London. ISBN 0 412 58540 5

146

CLINICAL EXAMINATION AND GRADING

Table 8.1 The primary survey. The A-B-C-D-E summary is a mnemonic for the early evaluation of severe trauma. It gives basic data on the neurological status after head and/or spinal injury (after American College of Surgeons Committee on Trauma, 1993) Survey

Check

Note, record and correct

A Airway

Patent? Noisy?

Obstruction

B Breathing

Effective?

Rate and depth Chest movements Air entry Cyanosis

C Circulation

Adequate?

Pulse rate and volume Skin color Capillary return Hemorrhage Blood pressure

D Disability (= neurological status)

Normal?

Conscious level – AVPU or preferably GCS Pupillary light reactions

E Exposure (= undress)

Other injuries?

Limb movements – on command or on painful stimulus

with serious illnesses may refuse to wear these, especially if the disease is seen as a stigma.

8.2

The initial examination

8.2.1

CONSCIOUS LEVEL

For the clinician, this is the best empirical measure of impaired cerebral function after closed head injury.

Table 8.2 The secondary survey. This is done after resuscitation. It includes a minimal neurological examination, which should be within the competence of any practitioner undertaking the early management of severe trauma. It commonly leads at once to special investigations e.g. X-ray of the cervical spine, CT scan, blood screen for ethanol and other drugs History

From patient and from observers

Reassessment of vital signs

As in primary survey

External signs of injury

Inspection and palpation of scalp, face, eyes and neck

Conscious level

GCS is repeated; this entails evaluation of speech and record of airway

Vision

Pupils: evaluation of size, symmetry and light reaction Acuities and fields: done by covering each eye and checking visual perception of examiner’s face

Limb weakness – lateralized or localized?

Detected from spontaneous movement, movement on commands, movement on finger- and toenail pressure

Impairment of consciousness is stratified in terms of the responses to external stimuli, and serial records of these responses are important in head injury management. The early postinjury conscious level may be preserved prior to secondary deterioration; this means a much better prognosis and also implies that a hematoma may be present. Though the role of nursing records in the early detection of cerebral compression has been diminished by the advent of CT scanning, serial progress records of the conscious level remain standard practice in head-injury observation, especially when the head injury initially appears to be less serious (Figure 8.1). The conscious level is also a valuable index of injury severity. In early evaluation, the depth of impairment of consciousness can be used as a measure of cerebral impairment, provided that the dimension of time after impact is taken into account and provided that confounding causes of impaired consciousness, such as ethanol, drugs or hypoxia, can be excluded. In late evaluation, a retrospective estimate of the duration of loss of consciousness can be used as a definitive measure of cerebral injury, though with certain reservations that are discussed below. For both purposes, it is necessary to define and quantify impairments of consciousness. 8.2.2

COMA SCALES

In 1941, a wartime committee of British clinical neuroscientists recognized the need for a standardized terminology for states of impaired consciousness after head injury, and published a glossary of descriptive terms. This committee included three brilliant neurosurgeons – Hugh Cairns, Geoffrey Jefferson and

THE INITIAL EXAMINATION

147

Figure 8.1 Neurological observation sheet. The record shows rapid deterioration of the consciousness level with appearance of ipsilateral pupillary dilatation and a contralateral hemiparesis. The GCS score falls from 15 to 5. Since the indications for action were evident by 15.30 h at latest, the record would imply unacceptable management! (Hypothetical case recorded on the 15-point version of the Glasgow Coma Scale, by courtesy of Mr M. Fearnside.)

Norman Dott. The committee had great authority and the terminology was widely accepted. Although the original publication (Medical Research Council Brain Injuries Committee, 1941) did not set out a graded hierarchy of levels of impaired consciousness, the

recommended terms were easily used to formulate such a hierarchy, and appeared in graphs of clinical progress in Rowbotham’s (1945) very influential textbook (Table 8.3). Those who used this system soon found that it was often misleading in communication

148

CLINICAL EXAMINATION AND GRADING

Table 8.3 Depth of unconsciousness, simplified from Medical Research Council Brain Injuries Committee, 1941. This terminology for post-traumatic impairments of consciousness is now only of historical interest: it has however influenced later thinking (Starmark, 1988) Coma

Absolute unconsciousness, judged by absence of any psychologically understandable response to external stimuli or inner need Primitive reflexes may or may not be present

Semicoma

Psychologically understandable responses are elicited only by painful or other disagreeable stimuli Primitive reflexes present

Confusion

Impaired capacity to think clearly and rapidly, to perceive and remember current stimuli; also disorientation

Severe

The patient occasionally responds to simple commands, if necessary reinforced by gestures, e.g. ‘put out your tongue’, ‘take my hand’

Moderate

The patient, though out of touch with his surroundings, can be got to give relevant answers to simple questions, e.g. ‘what work do you do?’, ‘where do you live?’

Mild

Confused but capable of coherent conversation and appropriate behaviour

between staff members without neurological training, and many neurosurgeons tried to formulate scales that used explicit grades of response to specified stimuli, given in simple descriptive terms. It was from such endeavors that Teasdale and Jennett (1974) devised the Glasgow Coma Scale (GCS).

This scale has been in use for more than 20 years, and its value is so widely accepted that a description may seem superfluous. However, the scale has its critics, and its application has changed somewhat since it was first reported. The authors described hierarchies in the levels of response for movement of the upper limb, verbal or vocal utterance and eye opening (Table 8.4). In their first paper, they listed only five levels of motor response; they noted that it is possible to distinguish normal and abnormal limb flexion, but concluded that the distinction was not appropriate for general clinical use. As a painful stimulus, nailbed pressure with a pencil was advised. For verbal responses and for eye opening, the scale provided five and four levels respectively. In describing the painful stimulus for eliciting eye opening, the authors specifically warned against using supraorbital pressure or retromandibular pressure, as these stimuli may evoke eye closure. This warning has not always been heeded. Although the value of the scale as a means of communication between different hospitals was emphasized, no attempt was made in this paper to give the findings in an aggregated score. Tests of inter-rater accord were later reported, and showed reasonably close agreement in evaluations by nurses, neurosurgeons and other medical practitioners (Teasdale, Knill-Jones and Jennett, 1974; Teasdale, KnillJones and Van der Sande, 1978). Similar tests have shown good agreement between ratings by emergency physicians and paramedics(Menegazzi et al., 1993). In later papers, the authors formulated six levels of motor responses by including the distinction between abnormal or spastic flexion and flexor withdrawal (Teasdale and Jennett, 1976; Jennett et al., 1977) though

Table 8.4 Variant forms of the Glasgow Coma Scale. In their first publication, Teasdale and Jennett (1974) did not discuss the use of a summated score; in later publications, the 15-point version of the GCS has been used in giving Coma Score 14-point scale (Teasdale and Jennett, 1974)

15-point scale (Teasdale and Jennett, 1976)

Pediatric scale (Simpson and Reilly, 1982)

Eye opening

Spontaneous To sound To pain None

4 3 2 1

The same

The same

Best verbal response

Orientated Confused Inappropriate Incomprehensible None

5 4 3 2 1

The same

Orientated Words Vocal sounds Cries None

5 4 3 2 1

Best motor response

Obeying Localizing Flexing Extending None

5 4 3 2 1

Obeys commands Localizes pain Flexion–withdrawal Flexion–abnormal Extension None

Obeys commands Localizes pain Flexion Extension None

5 4 3 2 1

Maximum sum

14

15

6 5 4 3 2 1

14

THE INITIAL EXAMINATION

they recognized that the original simpler scale might be preferable for clinical purposes (Jennett and Teasdale, 1981). Abnormal flexion was recorded if there were any two of the following: stereotyped flexion posture, extreme wrist flexion, abduction of the upper arm and fisting of the fingers over the thumb. This additional distinction made the GCS somewhat more demanding for the less skilled observer, but increased its analytic power, especially in severe closed head injuries; so much so that Jagger et al. (1984) argued that, as a prognostic guide, the Glasgow motor score alone was more informative, at least in comatose patients. Teasdale and Jennett (1976) also recommended the use of the total score (summed scores for eye opening, verbal and motor responses) for comparison of head injury severity between patients and between series, and as a rough definition of coma. A patient giving no verbal response, not obeying commands and not opening the eyes was judged to be in coma; by this definition, all patients showing a GCS score of 7 or less were comatose, and so were the majority (53%) of those with GCS score 8, when the maximum score was 15.

149

The GCS found many supporters, but also a few critics and skeptics. Some critics were clinicians who already used a measure of consciousness of some kind, and were reluctant to discard it; others devised supposedly more convenient variants of the GCS to meet local needs. Other criticisms related to the admitted limitations of the GCS in cases with periorbital swelling, which may eliminate the eye-opening response, or endotracheal intubation, which eliminates the verbal response. Starmark, Holmgren and Stålhammer (1988) reviewed 96 head injury studies published in the period 1983–1985; in these, GCS data were interpreted or aggregated in many ways, and other methods of grading consciousness were used in 23 papers (24%). Starmark et al. (1988) compared the Swedish form of the GCS (using retromandibular pressure and nailbed pressure as painful stimuli) with their own Reaction Level Scale (RLS 85; Table 8.5) and found better inter-rater agreement with the RLS 85; however, Johnstone et al. (1993) could find no significant differences between these scales in discriminating between grades of head injury severity, though

Table 8.5 The Swedish Reaction Scale (RLS85); in the manual for this scale, the responses are further explained by diagrams (Source: after Starmark et al., 1988, with simplification of the explanatory column. The RLS85 Manual is published by Acta Neurochirurgica (Wien).) Mentally responsive 1. Alert. No delay in response

Alert: not drowsy, orientated (intubated patient: no signs of delay in reaction)

2. Drowsy or confused. Responsive to light stimulation (verbal or touch)

Drowsy: the patients seems drowsy and shows delay in reaction Confused: the patient gives the wrong answer to at least one of three questions: What is your name? Where are you? What is the year and the month?

3. Very drowsy or confused. Responsive to strong stimulation (loud verbal, shaking, pain)

Arousable: performs at least one of the following functions: oral response with words; orientating eye movements; obeying commands; warding off pain

Mentally unresponsive 4. Unconscious. Localizes but does not ward off pain

Unconscious. No mental activity. Cannot perform any of the activities listed above for ‘mentally responsive’ Localizes pain. Examination is done in supine position: retromandibular pressure elicits movement of arm above chin level; nailbed pressure elicits movement of other hand across the midline

5. Unconscious. Withdrawing movements to pain

Withdrawing movements. On retromandibular pressure patient turns face away; on nailbed pressure, patient does not localize the pain but makes clear withdrawing movements

6. Unconscious. Stereotyped flexion movements to pain

Stereotyped flexion movements. On retromandibular pressure or on nailbed pressure, patient makes slow and mechanical flexion movements of elbows and wrists but no localizing or withdrawing movements

7. Unconscious. Stereotyped extension movements to pain

Stereotyped extension movements. On retromandibular pressure or on nailbed pressure, patient makes extension movements, straightening arms or legs. No flexion is seen; if both flexion and extension are seen, the better response (i.e. flexion) is recorded

8. Unconscious. No response to pain

No response to pain. Repeated strong pain from retromandibular or nailbed pressure gives no movement in arms, legs or face

150

CLINICAL EXAMINATION AND GRADING

the RLS 85 was regarded as a simpler procedure. The GCS has been in service for more than 20 years, and it should now be possible to give an appraisal of the present status of the GCS in the three roles for which it has been used: neurological observation, prognosis and severity grading. For the first purpose, it seems that most nurses, ambulance officers and surgeons accept the scale as a means of detecting changes in the conscious level and as a convenient currency in communication. In some centers, other in-house coma scales are used for supposedly greater simplicity, but in general the GCS has been found to be very serviceable. Many standard textbooks recommend the GCS. The American College of Surgeons Committee on Trauma (1993), in its manual Advanced Trauma Life Support Program for Physicians, sanctions the simpler non-quantitative AVPU variant, which has four levels – alert (A), responsive to vocal stimuli (V), responsive to pain (P), and unresponsive (U). The AVPU scale is used in the initial assessment, but the GCS is recommended in the secondary assessment. In modern surgical practice, the chief clinical uses of the GCS are in the early evaluation of the primary effects of a head impact, in routine monitoring of less severe head injuries to detect changes due to complications and in monitoring the progress of recovery. The relative merits of the 14- and 15-point scales have received little discussion in these contexts. In 1990, a questionnaire was sent to senior Australasian neurosurgeons and, of the minority who responded, seven out of 11 preferred the 14-point scale for clinical use. However, the 15-point scale is internationally accepted for research studies and this is a powerful argument for using it routinely. It is less easy to determine the status of the GCS in prognosis. The depth of coma is certainly one factor in clinical decisions on head injury treatment. Jennett (1992) has reviewed the circumstances in which continued treatment of a head-injured person is ‘futile or disproportionately burdensome’ and in this review age and depth of coma stand out as the chief predictors of death. In an earlier study of 1000 severe head injuries, Jennett et al. (1979) had shown the prognostic value of GCS summated scores: death or vegetative survival was the fate of 87% of those whose best score in the first 24 hours was 3/4, whereas these bad outcomes were recorded in only 53% of those with scores of 5/6/7. To exclude the effects of alcohol, hypoxia and other confounding factors, this study included only cases remaining in coma for more than 6 hours. If, as is now often the case, initial resuscitation includes immediate endotracheal intubation and paralysis, the effects of these confounding factors on conscious level cannot always be excluded. The ethanol level can and

should be measured; Jagger et al. (1984) have shown that over a level of 0.20%, ethanol may significantly depress the GCS score, though there is much individual variation in this effect. But even when allowance is made for intoxication, the initial GCS level has to be viewed with reserve as a prognostic factor and, in determining whether to cease supportive treatment, the clinician will usually rely on repeated evaluations of the coma level after temporary cessation of respiratory paralysis and sedation. In establishing the coma scale, and in subjecting it to rigorous statistical analysis, Teasdale, Jennett and colleagues made a very great contribution to clinical neuroscience, and their work has enduring value. But, as these authors presciently forecasted, newer methods of therapy have reduced the availability of GCS scores for prognostic purposes. The value of the GCS as a measure of head injury severity is considered below. 8.2.3

PEDIATRIC COMA SCALES

It is not easy to assess the consciousness level in infants and young children, and mistakes are often made. Sometimes the severity of a head impact is overestimated, but the converse error is much commoner: because an injured infant cries or whimpers, it is thought to be fully ‘conscious’ and serious brain damage may be overlooked. The verbal and motor responses that indicate full consciousness in the GCS are obviously not achievable by preverbal infants. Even after speech is attained, a frightened but fully conscious child may withhold speech or cooperation. There have therefore been many attempts to devise a scale of consciousness appropriate to the first 5 years of life. These have been reviewed by Yager, Johnston and Seshia (1990) and Simpson et al. (1991). Pediatricians and neurological nurses are well aware of the subtlety and scope of preverbal responses, and some of the reported scales try to quantify these. Thus, Seshia, Seshia and Sachdevan (1977) devised an elaborate grading that tested social, adaptive, vocal and motor responses, and also suck/cough responses, both spontaneous and stimulus-evoked; each was given a 0–4 value. Hahn et al. (1988) devised a scale using the Glasgow scale for eye opening and motor responses but with a more complex range of verbal responses, including subscores for smiling, eye orientation, consolability and interaction. Simpson and Reilly (1982) preferred a much simpler system, directly based on the original GCS but with agerelated norms for the verbal and motor responses. This scale expresses the concept that the range of responses in head-injured infants and young children is narrower than is the case over the age of 4 years. Norms

THE INITIAL EXAMINATION

151

Figure 8.2 Norms in the Paediatric Glasgow Coma Scale (PGCS). The expected norms for successive age ranges are set out on a standard ward chart, which shows the modifications of the best verbal responses used in the PGCS. In teaching the use of this scale, it is emphasized that actual performance is often better than the expected norms: many children in the 3–5 year range will demonstrate awareness of place or personal relations.If necessary, a standard adult scale can be used, but it must then be emphasized that adult performance is not to be expected and the record should show what responses are actually elicited.

(Figure 8.2) for best anticipated responses at birth, 6 months, 1 year, 2 years and 3–5 years are derived from accepted developmental milestones (Reilly et al., 1988); actual responses are recorded in ward charts identical with the 14-point GCS except in the verbal scale, in which minor changes have been made to allow a simple grading of preverbal responses (Table 8.4). The Paediatric Glasgow Coma Scale (PGCS), was independently compared by Yager, Johnston and Seshia (1990) with five other systems of quantifying the consciousness level in early life, and found to be one of the two best from the viewpoint of observer disagreement (< 0.10). We believe that the PGCS, in its simplicity and its close resemblance to the GCS, is well adapted to hospital use, though some instruction is needed for nurses used to the adult scale. A video film has been made for this purpose. In Melbourne, G. Klug (personal communication) has successfully developed a 15-point version of the PGCS, in which abnormal limb flexion is recognized as in the adult scale. As a prognostic tool, the PGCS has not been fully tested. In a series of 23 infants and young children with impaired consciousness, bad outcomes were recorded in six of seven cases with summated PGCS scores of 3/4, in three of five in the range 5/6 and in none of 11 cases in the range 7/8 (Simpson et al., 1991). At the lower levels, the PGCS is based on observations identical with those recorded in the adult scale, and it

seems likely that the predictive value is similar. Hofer (1993) compared outcomes with scores derived from the standard 15-point adult scale in a sample of 41 children (age range 2–17 years, mean age 8.8 years) and found that the lowest GCS scores were strong predictors of death. However, this study excluded infants and included only two children less than 4 years old; moreover, it related to observations made 24 hours after operation. There is need for further research on the prognostic significance of post-traumatic coma in infants. 8.2.4

THE PUPILS

Pupillary size, shape and reactions are routinely recorded at the initial examination, and routinely checked at specified intervals thereafter. If the pupillary light reflex is impaired on one side, the consensual light reflex is tested to exclude an optic nerve lesion. Even iconoclastic clinicians approve these timehallowed practices, but one must nevertheless ask what may be learned from them. Pupillary abnormalities may be bilateral or unilateral; they may be present from the time of injury (Figure 8.3(a)), or may appear after an interval of time. If the initial examination shows that both pupils are widely dilated, and if there is no reaction to a strong light – not always available in an emergency room! – then the pathological basis may be an irreparable

152

CLINICAL EXAMINATION AND GRADING

(a)

(b)

(c) Figure 8.3 Oculomotor paralysis. A young girl was admitted in coma after a road accident. (a) The left pupil was dilated (5 mm) and fixed to light. The right pupil was smaller (3 mm) but varied in diameter; initially there was no light reaction, but later this pupil reacted sluggishly to light. (b) 4 weeks later, rotation of the head to the left and right elicited a small change in the deviation of the right eye; the left eye did not move (positive horizontal oculocephalic response in association with left third-nerve paralysis). (c) Flexion and extension of the head elicited no change in the position of the eyes (absent vertical oculocephalic response). Photographs reproduced with permission.

primary midbrain lesion or advanced bilateral transtentorial herniation. There are, however, other causes. The pupils may be fixed and dilated in the aftermath of an epileptic fit, or from inadequate cerebral perfusion (Narayan, 1989), or from local trauma to the iris or its innervation on both sides, or from the use of a mydriatic to view the fundi, a practice to be prohibited in the early period after trauma. (Homatropine was the cause of dilated fixed pupils in an injured motorcyclist admitted in coma after visiting an ophthalmologist, whose use of this mydriatic possibly caused the accident as well as confusing the diagnosis.) In general, the finding of bilateral fixed dilated pupils soon after injury is a very adverse sign, and the appearance of this sign after initial normality often indicates irreversible cerebral compression. However, the finding must be interpreted in its context and with regard to other findings. Bilateral fixed pupils of normal shape and size may indicate a midbrain lesion; bilateral sluggish pupils associated with ptosis and impaired upward gaze are an almost pathognomonic sign of central or posterior transtentorial herniation. Bilateral optic nerve injury may give bilateral fixed or sluggish pupils, sometimes with pupillary escape, and this should be remembered especially in head injury from a frontal impact; in such cases, the pupils typically show spontaneous fluctuation (hippus) in diameter. Bilateral small pupils, often appearing fixed, are a classical sign of a pontine lesion. This is a relatively rare finding in closed head injuries. Large doses of an opiate give similar appearances; in many intensive care units, morphia is infused to control the reflex responses of intubated patients, but in the doses now used the pupillary reactions are usually well preserved, though the diameters may be small. In deep barbiturate coma, the pupils become fixed and nonreactive. Previous neurological disease may be associated with bilateral or unilateral pupillary abnormalities. Neurosyphilis was the confusing cause of small fixed pupils in a workman who fell from a scaffold and sustained compound skull fractures. The sluggish tonic pupils of Holmes–Adie syndrome could be misleading, and it should be kept in mind that tendon areflexia is not an invariable finding in this condition (Bacon and Smith, 1993). Unilateral dilatation and loss of light reflex in one pupil commonly means a third-nerve paralysis, often accompanied by ptosis and a divergent squint (Figure 8.3(a)). This may be a primary effect of the initial head impact (Heinze, 1960), as a traction injury of the nerve or from damage in the skull base or orbit. Delayed onset of a third-nerve paralysis is of course the classical sign of lateral transtentorial herniation. In modern practice this is most often due to an acute

THE INITIAL EXAMINATION

subdural hematoma or massive hemispheric swelling; extradural hematomas are now commonly diagnosed before this dangerous complication has developed. Jones et al. (1993) studied a series of 366 cases of extradural hemorrhage treated by a single neurosurgical service in a 35-year period and found that before the advent of CT scanning, pupillary abnormalities were recorded in 70.4%, but were recorded in only 34.3% of those treated in recent years. A unilateral fixed pupil of normal or fluctuating size may be due to an optic nerve lesion; consensual testing will usually establish this diagnosis. A fixed pupil of large or normal size may result from trauma to the iris and/or ciliary body; other signs of a local impact may or may not be detectable. A more obvious cause of a unilateral fixed pupil is a prosthetic eye. Irregularities of pupil shape are not uncommon in terminal stages of cerebral compression; the mechanism is uncertain. Marshall et al. (1983a) have drawn attention to the finding of an oval pupil, often eccentric, as an early sign of transtentorial herniation (Figure 8.4). An ectopic pupil (corectopia) may also be the result of ocular injury, sometimes long-standing; in a recent case of severe head injury, relatives had known of the abnormal pupil for many years, but had not been questioned about it until its diagnostic significance was discussed. In view of the clinical importance attached to the pupillary light reflex, it might be expected that its prognostic significance would be substantial. This is indeed so in certain categories of head injury, notably extradural hematomas; in the study by Jones et al. (1993) quoted above, death or persistent vegetative states were recorded in 8.3% of cases presenting with normal pupils, but in 26.9% of those with a pupillary abnormality (p = 0.0002). In using the pupils as a

153

general prognostic factor for outcome after severe head impact, difficulties arise from the great diversity of pupillary abnormalities and their variable significance. Nevertheless, Choi et al. (1988) concluded that the pupillary light response was one of the three most accurate predictors of final outcome, the others being age and GCS motor score. In this analysis, these authors recognized three grades of pupillary response on admission: bilaterally absent, unilaterally absent, and normal. Braakman et al. (1980) used a similar grading, but selected the best state of the pupils in the first 24 hours as the prognostic factor: where both reacted, the mortality rate was 29%, rising to 54% when only one reacted and to 90% when both were fixed. Pupillary size has received less attention. In the hope of earlier diagnosis of intracranial mass lesions, Chesnut et al. (1994) considered pupillary inequality, irrespective of reaction to light, in a series of 608 comatose head injuries. They found that inequality of more than 1 mm was present in 35% of these patients; when present, this inequality indicated the presence of an intracranial mass lesion in only 30% – not always ipsilateral and not always an extracerebral clot. Greater asymmetry (more than 3 mm) was more often associated with a mass lesion; nevertheless, in more than half the cases with this degree of asymmetry no such association was found. Commenting on this study, Narayan (1994) drew the conclusion that pupillary inequality is an unreliable sign and no substitute for routine early CT scanning in patients with severe head injury. Nevertheless, pupillary size and reactivity to light are valuable signs if taken in context, both in diagnosis and in prognosis. Pupillary testing is often made impossible by orbital swelling, and this emphasizes the importance

(a) (b) Figure 8.4 Oval pupil. A young man was admitted in coma after a road accident.On admission, the left pupil was dilated (7 mm) and fixed to light. The right pupil was smaller (2–3 mm) and also fixed. (a) 10 weeks after injury, the left pupil was fixed to light as a result of a third-nerve paralysis. The right eye was deviated down and was abducted. (b) The right pupil was oval in shape, and reacted sluggishly to light. Photographs reproduced with permission.

154

CLINICAL EXAMINATION AND GRADING

of accurate early examination and recording before swelling is established. A slovenly initial examination may miss an optic nerve injury, which may then be detected days later when the swelling subsides. Desmarres’s eyelid retractors sometimes allow exposure of an otherwise inaccessible pupil, but should be used with great caution to avoid corneal abrasion. 8.2.5

EYE MOVEMENTS

In the routine neurological examination of the unconscious patient, spontaneous eye movements should be noted. If there are none, the oculocephalic reflexes are tested by rotating the head fully in the horizontal and vertical planes – the oculocephalic or doll’s eye test (Figure 8.3(b), (c)). This is done only when a cervical spinal injury has been excluded by adequate radiographs demonstrating all vertebrae including C7. The findings relate to the functional integrity of the midbrain, the pons and the third, fourth and sixth cranial nerves. Thus, spontaneous roving eyes with parallel visual axes suggest normal central and peripheral innervation of the extraocular muscles. Lesions of the third and sixth cranial nerves show up as limitation of eye movements effected by the paralyzed muscles. The fourth nerve is untestable in the unconscious patient. Forced downward ocular deviation suggests a midbrain lesion. Absence of upward movement (vertical oculocephalic reflex) has the same significance, but may be hard to elicit in a convincing manner in unconscious patients. Forced lateral gaze suggests an irritative lesion which may be in the brain stem or in the supratentorial brain; absence of lateral gaze may indicate a paralytic lesion in the same sites. Vertical divergence of the visual axes (skew deviation) is usually taken to mean a pontine lesion. In addition to their localizing value, the eye movements have been considered as indices of head-injury severity. Visual fixation and tracking are preserved in relatively mild injuries; the capacity to fix on a target and to follow it is a favorable finding, and is especially useful in examining a preverbal infant or an aphasic at any age. Spontaneous roving eye movements usually indicate a milder impairment of consciousness, in the GCS range 7/8 or better. At the other end of the severity spectrum, absence of eye movements is an ominous finding. Absence of eye movements on irrigating the external auditory canal with up to 100 ml ice-cold water (oculovestibular reflex) is indicative of profound brain-stem failure and is one of the accepted criteria of brain death (Walker, 1985). The test should not be done if there has been anything to suggest a cranioaural fistula, such as cerebrospinal fluid otorrhea, intracranial air or a middle fossa skull-

base fracture. However, between these extremes of severity, there are many ill-defined disturbances of eye movement, and it is questionable whether these have much if any diagnostic or prognostic meaning. Jennett et al. (1977) constructed a composite score for eye movements, but this has not been widely used. From Li`ege, Born et al. (1982) have reported on a composite coma scale combining the 15-point GCS with a five-point reflex scale. In the reflex scale, cephalic reflexes are ordered in a hierarchy corresponding to their supposed clinical significance as a measure of rostrocaudal brain-stem function (Table 8.6). The fronto-orbital blink reflex is elicited by a light tap on the glabella. The pupillary light reflex and the oculocephalic reflexes are elicited as described above. If a cervical spinal injury makes head rotation dangerous, the oculovestibular reflexes are obtained by aural irrigation with iced water – bilaterally for vertical eye movement, unilaterally for horizontal movement. The oculocardiac reflex is obtained by pressure on the globe, failure to demonstrate slowing of heart rate being the lowest level in the 0–5 reflex scale. The sum of the GCS score and the reflex score is termed the Glasgow–Li`ege (GLS) score. Born et al. (1987) have confirmed good inter-rater agreement for the reflex scale. In a comparison of the predictive value of the reflex score with the motor component of the GCS in severe head injuries, Born (1988) found that, in the first 24 hours, the reflex score is superior as a prognostic tool. However, in surviving cases, the reflex score usually returns to normal within 2 weeks, even in patients destined to be severely disabled, whereas the GCS motor score tends to remain low for longer periods, even in cases showing a favorable outcome. This interesting composite scale has not received as much attention as it deserves, perhaps because two of the reflexes – the fronto-orbital and the vertical oculocephalic reflex – are not in general use in most neurosurgical centers. 8.2.6

FUNDI

Fundal abnormalities are not usually of great importance in the early management of severe head injuries, and the examination is often difficult: the pupils may

Table 8.6 Li`ege Reflex Scale. (Source: After Born et al. 1982) Brain-stem reflexes Fronto-orbicular Vertical oculocephalic Pupillary light Horizontal oculocephalic Oculocardiac No response

Score 5 4 3 2 1 0

THE INITIAL EXAMINATION

be small, there may be orbital swelling, and the use of corneal lubricants is often a further impediment. There is however some diagnostic yield from early examination of the fundi. The finding of retinal hemorrhages may indicate a period of sudden increase in intracranial pressure and massive intraocular bleeding may be a threat to vision. In infants, retinal hemorrhages are often due to child abuse, though this is by no means a specific association (Duhaime et al., 1992). The diagnosis of child abuse is of such importance that it may be justifiable to use a mydriatic for fundoscopy in head-injured infants – though only after a CT scan has been done. Early fundoscopy also gives a baseline against which later possibly abnormal findings can be evaluated. Papilledema is not a common finding in severe head injury: Selhorst et al. (1985) found swollen discs in only 15/426 (3.5%) cases. When papilledema does occur, it is usually not gross, and knowledge that the discs were previously normal can be helpful in deciding on the significance of indistinct disc margins. This is true also of the late appearance of optic disc pallor from nerve injury. 8.2.7

LIMB MOVEMENTS AND REFLEXES

Spontaneous and evoked limb movements are studied as part of the GCS examination. This records the best motor response; less good responses are also noteworthy, since asymmetrical or localized impairment of movement may reveal a hemiparesis, monoparesis, paraparesis or limb fracture. Muscle tone is assessed after inspection, by putting the limbs through a full range of passive movement – keeping the possibility of a long bone fracture in mind. The tendon reflexes have very little value in the diagnosis of acute cerebral injuries, but localized absence of tendon jerks may disclose a nerve injury. The plantar reflexes are usually extensor in severe head injuries. With a coexisting acute spinal cord transection, a slow flexor plantar response is often seen, and this can be a useful corroborative finding in an unconscious patient. Absence of sweating may then clinch the diagnosis of cord damage: this is best detected by running the dorsal surface of the examiner’s fingers up the body – the change from dry to moist skin is unmistakable. The motor findings have considerable significance. The prognostic importance of the GCS motor response has been generally confirmed, though in young children the finding of generalized extensor patterns is not so adverse as in adults (Robertson and Pollard, 1955). A lateralized limb weakness may mean a contralateral (rarely ipsilateral) intracranial clot, especially if serial records have shown that movements were previously symmetrical. Other important causes include cerebral infarction from internal carotid occlu-

155

sion. Bilateral flaccid leg weakness very strongly suggests a spinal lesion; bilateral leg weakness is sometimes seen after cranial injury in the vertex region, but is likely to be spastic, and should not show absence of sweating. A flaccid arm weakness means a brachial plexus paralysis until proven otherwise, though in the acute phase of cerebral injury, a hemiplegic arm sometimes shows reduced tone and even depression of tendon reflexes (Russell, 1947). Abnormalities of the motor examination can only be reliably inferred in the absence of neuromuscular paralyzing drugs or sedatives such as barbiturates and benzodiazepines. Use of a nerve stimulator may be helpful, if paralytic drugs have been used. Nurses and residents should avoid repeated painful stimuli for motor-system testing in patients who are pharmacologically paralyzed or heavily sedated for ICP control. 8.2.8

CARDIORESPIRATORY PARAMETERS

Pulse rate, blood pressure and respiratory rate have long been recognized as valuable indicators of raised intracranial pressure (ICP). In 1901, Theodor Kocher of Bern clearly associated higher grades of raised ICP. with slowed pulse and slow, shallow breathing interrupted by deep breaths; thanks to the experimental work of his brilliant young American proteg´e Harvey Cushing he was also aware that when raised ICP compromised the medullary vasomotor centers, the blood pressure would rise (Kocher, 1901). These signs of brain-stem failure now have less diagnostic value in cases of severe head injury, being seen chiefly as late events after failure of treatment. Nevertheless, the pulse, blood pressure and respiration should be monitored routinely and in less severe head injuries may give valuable advance warning of raised ICP. In children especially, a slowing pulse sometimes appears before an obvious fall in the consciousness level. Conversely, tachycardia and a falling blood pressure may be of great importance in detecting an extracranial lesion such as a ruptured spleen or other abdominal or thoracic organ. The prognostic significance of the cardiocirculatory parameters received surprisingly little attention in earlier studies of severe closed head injury (Teasdale and Jennett, 1976; Jennett and Teasdale, 1977). However, with increasing awareness of the critical importance of cerebral perfusion pressure, the significance of a low blood pressure has received more attention. Stening et al. (1986) found records of an arterial blood pressure below 90 mmHg persisting for more than 60 minutes in 90/290 (31.0%) cases of acute subdural hematoma, and analysis showed that this was a strong predictor of bad outcome. These authors saw arterial hypotension as a preventable cause of bad outcome,

156

CLINICAL EXAMINATION AND GRADING

and this is indisputable when hypotension is due to inadequate resuscitation or failure to deal with an extracranial lesion. But arterial hypotension may also be seen as a manifestation of terminal brain-stem failure, and then a bad outcome may be inevitable. Lyle et al. (1986), in a study of severe head injury, found that a systolic arterial blood pressure below 90 mmHg was significantly associated with death on univariant analysis; however, in this study hypotension correlated closely with a low GCS score and on multivariant analysis the only significant variables were the GCS score and the pupillary light reflexes. Choi et al. (1988) also found that blood pressure on admission did not correlate significantly with outcome. The converse finding of hypertension has received less attention, though Robertson et al. (1983) have given prominence to arterial hypertension as an adverse event in severe head injury. Fearnside et al. (1993) considered the prognostic significance of paroxysmal arterial hypertension as a clinical variable in a series of 315 severe head injuries; it was seen in association with profuse sweating and tachycardia and did not adversely affect mortality. Respiratory abnormalities were studied by North and Jennett (1974) in a series of acute neurological disorders. They found that bad outcomes were associated with abnormal breathing patterns, and in particular with tachypnea combined with hyperventilation. With the advent of routine early intubation and control of blood oxygen saturation by pulse oximetry, spontaneous respiratory patterns have lost their diagnostic and prognostic significance; the initial ABC examination always includes an assessment of the adequacy of breathing, but a detailed description of the rhythm does not add much to the overall evaluation of injury severity. 8.2.9

EXTERNAL FINDINGS

In the unconscious acute head injury, concern with the state of neurological function may overshadow the rest of the clinical examination. It is easy to omit an examination for external signs of injury. The scalp and face should be carefully inspected and palpated. Wounds, abrasions and swellings should be recorded on a diagram and also photographed if there are medicolegal implications. If for any reason CT scanning is not possible, the site of the primary impact is an important guide to the siting of an emergency craniotomy or burrhole exploration. The impact site(s) may also be important in future litigation, for example in determining whether a helmet might have given effective protection, or in a victim of child abuse or other form of criminal assault. Orbital swelling has obvious diagnostic importance as a sign of anterior fossa fracture; auscultation may

detect a carotid–cavernous fistula. Bruising behind the ear (Battle’s sign) is a well-known sign of a fractured petrous bone. However, if bruising is seen immediately after injury, it may point to a local impact. If the bruising appears after a time interval, it is likely to be due to a fracture of the skull base. Blood or cerebrospinal fluid (CSF) leakage from the nose or ear should be noted and a sample of fluid (> 0.1 ml) should be sent for immunochemical analysis for 2-transferrin as a marker of the presence of CSF (Ryall, Peacock and Simpson, 1992).

8.3

The definitive examination

8.3.1

TIMING

This depends on the speed and degree of recovery. Ideally, a full neurological examination is done when the patient is conscious, cooperative and fully oriented. Since recovery from a severe head injury is usually slow and often incomplete, more or less selective serial neurological examinations are usually done at intervals before full cooperation is complete, and when the patient is still in the phase of posttraumatic amnesia (see below). These progress examinations can be very informative. It is very important in planning and assessing rehabilitation, and also for medicolegal purposes, to ensure that all residual neurological disabilities are documented before the patient goes home or is transferred to another hospital. In particular, residual amnesia should be assessed and the senses of sight, smell and hearing should be tested. Therefore, a definitive neurological examination is mandatory before discharge or transfer. 8.3.2

ORGANIZATION

Who does the definitive examination? In many centers, the neurosurgeon is a surgical neurologist and examines the patient in person or delegates to a properly trained proxy. In some centers, a medical neurologist may be consulted. Increasingly and beneficially, parts of the examination are now subcontracted to a neuropsychologist, neuro-ophthalmologist, neurootologist or specialist in neurorehabilitation; pediatric neurosurgeons may consult a specialist in developmental neurology. Whatever the division of labor, there needs to be a final common synthesis and evaluation, preferably made by a single person with final clinical responsibility. 8.3.3

ORIENTATION AND AMNESIA

Orientation is always confirmed as part of the definitive examination. As a minimum, the patient should be asked to name the day of the week, date and

THE DEFINITIVE EXAMINATION

place. The patient should also be asked to describe the last event recalled before the injury and the first event recalled after the injury. These questions should define and quantify the periods of retrograde amnesia (RA) and post-traumatic amnesia (PTA). In patients emerging from a prolonged amnesic state, one of the amnesia questionnaires described below can be used. 8.3.4

SPEECH, MENTAL STATE AND COGNITION

Speech is assessed in conversation, both for impaired articulation and for fluency and thought content. Minor degrees of dysarthria can be brought out by tongue-twisting words, such as ‘hopping hippopotamus’. A tape recording of speech may be made for future comparison. The preferred hand is recorded, though it must be remembered that many left-handed persons have full or partial left-hemisphere dominance. If there is reason to suspect injury to the dominant hemisphere, the patient is asked to name a series of test objects of increasing complexity: for adults, the final challenge can be the parts of a watch and for children a toy. This test for nominal dysphasia can also be used to test recent visual memory, by asking the patient to recall as many as possible of the objects immediately after the naming test; 8/10 is a good score. Other standard tests of memory include recall of a name, an address and a flower after 5 minutes, digit retention forwards and backwards, and timed serial subtraction of 7 from 100. Tests of cognition are best done by a neuropsychologist; Walsh (1985) and Wood and Woodroffe (1995) have reported on tests found useful in head injury practice. If the services of a neuropsychologist are not available, there are many simple tests that require little special expertise; literacy can be checked with an agegraded word list and non-verbal intelligence can be assessed by the Raven colored matrix test (Raven, 1986), which also probes function in the non-dominant parietal lobe. The emotional state and the degree of insight are noted in conversation and in discussions of future plans for rehabilitation. For head-injured children, cognitive capacities and mental attitudes can be assessed in play and games appropriate to the age; a child’s attempts to draw a man are very informative, and so are games based on family relationships. At some stage, a formal developmental assessment (Griffiths, 1970) should be done; this requires considerable pediatric experience. 8.3.5

VISION

This is always assessed in severe head injuries, though the depth and scope of the assessment vary with the nature of the injury. Covering one eye may bring out

157

subjective visual loss or blurring. The Snellen and/or reading test types should always be used in the definitive examination of a severe head injury. For illiterates, the E tests can be used; alternatively, the STYCAR toys provide a simple and quick way of estimating visual acuity (Sheridan, 1976). Peripheral visual fields are tested by confrontation with the examiner’s fingers as stimulus. Central fields can be tested with a small white or red object, such as the head of a mapping pin, or a bead on a black stick. It is also possible to assess the central fields very effectively by asking the patient to fix on the examiner’s nose and to say if any feature is missing or blurred. If there is a field defect, formal perimetry is done – usually by a neuro-ophthalmologist. The optic fundi are always examined; as noted above, a mydriatic should not be used in the early period after a head injury, but may be used later when definitive fundoscopy is done. 8.3.6

SMELL AND TASTE

Olfaction must always be tested. Tar or phenol is a good strong test odor, but should be complemented by a milder odor such as banana or raspberry, or cloves. Each nostril is tested separately and, to exclude guessing, the patient is warned that the test bottle may be empty. A more objective system of smell testing is provided by the University of Pennsylvania Identification Test (Doty et al., 1984), a quantitative smell test that permits the examiner to determine whether there is normal olfaction, microsmia, anosmia or malingering. Taste is rarely of importance, but may be tested when there is a facial paralysis, by dropping strong syrup or salt solution on each side of the tongue; electrical tests of taste are not always reliable, as the patient may report tingling as a taste. 8.3.7

HEARING

This is tested with particular care when there is evidence of a skull-base fracture. A simple check is done by whispering words or numbers into each ear, hearing by the opposite ear being masked by gentle circular rubbing with a finger tip pressed on the opposite tragus to occlude the external auditory meatus. If deafness is found, a 1024 or 512 Hz tuning fork is used to distinguish inner- and middle-ear deafness by the Weber and Rinne tests. The external auditory canals are examined with an otoscope and the color of the ear drum is noted. An otologist should be consulted if hearing is impaired, or if there is a hemotympanum. An audiogram should also be performed, both to aid prognosis for recovery and for medicolegal reasons.

158

CLINICAL EXAMINATION AND GRADING

8.3.8

CRANIAL NERVES

The eye movements are tested in lateral, vertical and oblique planes, and note is made of ptosis, diplopia, squint or nystagmus. The trigeminal nerve is tested in head injuries associated with facial or skull-base fractures. The corneal reflex is tested with a wisp of cotton wool. Mild trigeminal hypesthesia may be brought out by stroking parts of the face and asking the patient if there is any qualitative difference. Differences in pinprick perception may be checked, especially if there has been a facial injury; a blunted disposable needle is used. Two-point testing on the lips is sometimes useful. The facial nerve is tested by asking the patient to screw up the eyes, whistle or smile; an emotional facial weakness of upper motor neuron type may be brought out by watching the spontaneous smile. If there is any facial weakness, it is most important to know whether it is of late onset. In unconscious patients, unless in deep coma, facial movements can be elicited by strong pressure on the supraorbital nerve or mandibular ramus. Schirmer’s test of lacrimation is very useful in cases of peripheral facial paralysis: thin strips of filter paper are placed in the conjunctival sac for 30 seconds or more, and the extent of saturation is measured (Trott and Cooter, 1995). The lower cranial nerves (nerves IX–XII) are tested by examining the movements of the palate, pharynx, tongue, trapezius and sternomastoid muscles. Peripheral paralyses of these nerves are occasionally seen after closed head injuries. Much commoner are dysphagias and dysarthrias from brain-stem damage. For these an assessment by a speech pathologist is necessary and a radiological swallow study is advisable. 8.3.9

SENSORIMOTOR LIMB FUNCTIONS

Limb function is tested with respect to muscle tone, power against graded resistance, and coordination; ataxia is a very common sequel in severe closed head injuries, presumably from injuries of the superior cerebellar peduncle. Quantitative measures of limb function are desirable if cooperation is good. In most units, these are done by physiotherapists and/or occupational therapists, but their value and limitations should be understood by all clinicians concerned in head injury evaluation. Dynamometry should be used to give an objective measure of hand grip (Mathiowetz, 1990); hand-held dynamometers can be used to test other muscle groups. Manual facility and coordination can be quantified by finger tapping or by pegboard tests, and for these consultation with an occupational therapist is advisable. Wood and Hammerton (1995) have reported on the Purdue pegboard

test in evaluating head injury, both in adults and in children over the age of 7 years. This test was devised for selecting industrial workers (Tiffin, 1968); Gardner and Broman (1982) found it to be an excellent test of minimal brain damage. In the simplest form of the test, the subject inserts metal pegs in a row of holes in a standard board as rapidly as possible. A score is obtained for each hand over a period of 30 seconds. The reproducibility of pegboard testing is impressive, and the test is of value as a guide in determining when a stable level of recovery has been attained. Other quantitative tests of hand–arm function are described by Gloss and Wardle (1982). If the patient can walk, the gait is described, and the ability to hop on either leg is noted; this is a good quick test of lower-limb motor competence, though limb or spinal injury may falsify the interpretation. In patients unable to walk, the degree of mobility in bed or in a wheelchair is recorded. The progress of motor recovery can be documented by video. The tendon and plantar reflexes are again tested, and with more attention: persisting reflex abnormalities have much more significance than the evanescent reflex changes seen in the acute phase. Sensation, except in the trigeminal area, is rarely affected in closed head injuries, but occasionally one sees what appears to be a spinothalamic sensory loss in cases of brain-stem damage. A full sensory examination is needed if there is an associated spinal injury or suspicion of a lesion in the parietal lobe or basal ganglia.

8.4

Evaluation of injury severity

8.4.1

PROSPECTIVE GRADING

Estimates of head-injury severity may be made prospectively, as aids in triage, prognosis and family counseling. For these purposes, the estimate can take into account many factors and nuances; most clinicians will admit that intuition enters assessments done for prognosis. But when the estimate is done for statistical purposes, as in therapeutic trials, the criteria should be as few as possible, and they should be based on observations that have good inter-rater reliability. The consciousness level, assessed at a specified period after injury, has been widely used in definitions of a severe head injury, both for prognosis and for research purposes. In most reports, the chief criterion of severity is a GCS score of 8 or less. For the US National Coma Data Bank the definition for inclusion as a severe head injury is:  

GCS score of 8 or less following resuscitation, which may include endotracheal intubation; or GCS score deteriorating to 8 or less within 48 hours of injury (Marshall et al., 1983b).

EVALUATION OF INJURY SEVERITY

In this definition the use of endotracheal intubation could reduce the best verbal score to 1. In theory this might result in the inclusion of less deeply unconscious patients, and it is reasonable to use the best motor score to control the reliability of the GCS summated score. If this is done, the six-level Glasgow motor scale should be used. Data from the six-level motor scale should be routinely recorded for audit and research studies. This definition of head-injury severity is widely accepted, and is being used as a basis for inclusion in therapeutic trials. Thus, in a well designed phase 2 trial of an oxygen radical scavenger, Muizelaar et al. (1993) accepted for entry into the trial cases with a GCS score of 8 or less who were unable to follow commands after resuscitation; the time at which the GCS score was estimated was not specified. The criteria for exclusion included ‘the likelihood of brain death after resuscitation’; presumably this exclusion would remove cases with GCS score of 3 and other adverse signs. The GCS has also been used to stratify cases within a more broadly inclusive trial. In a randomized trial of nimodipine therapy for head injury, Bailey et al. (1991) accepted all patients who were unable to obey commands, thus including GCS scores as high as 13. For statistical analysis, the best motor response was used, the scale being collapsed into three classes – nil or extending; flexing; localizing. (These classes were further subdivided by the presence or absence of an intracranial lesion requiring operation.) This trial protocol shows the flexibility of the GCS in stratifying injury severity and it seems likely that, in further trials of neuroprotective agents, components of the GCS will be used in various ways, depending on the level of severity at which the agent is expected to be beneficial and the size of the series to be analyzed. 8.4.2

RETROSPECTIVE GRADING: COMA DURATION

Head injury severity may also be assessed retrospectively, for epidemiological and other research studies, especially in correlation with measures of outcome (see below). The duration of impaired consciousness has been much used as a retrospective measure of injury severity. In contemporary neurosurgical practice, this is commonly done in two different ways. The duration of coma can be measured on the basis of serial clinical observations of responsiveness. Thus, Bricolo, Turazzi and Feriotti (1980), in a very thoughtful study, reported on 135 cases who were in coma 14 days after head injury, coma being defined as ‘unresponsive . . . or incapable of obeying simple commands or showing any rapport with their environment’. Outcomes were assessed at 1, 3, 6 and 12

159

months by the Glasgow Outcome Scale (Jennett and Bond, 1975): by the categorization given in that scale, 13.3% made good recoveries, 17.7% were left with moderate disabilities, 31.1% were left with severe disability, 8.1% remained in persistent vegetative states (Jennett and Plum, 1972) and 29.6% were dead. This study illustrates both the usefulness and the complexity of the duration of coma as a measure of unconsciousness. Recovery from coma was identified in terms of the three components of the GCS. Spontaneous or evoked eye opening appeared after 1 month in 76.3%. Response to commands came later, and was achieved after 3 months in only 52%. Speech restoration was achieved by 3 months in only a third of cases, rising to 51% by the end of the study. There was a correlation between duration of coma and quality of final outcome. In this study, each measure of responsiveness was separately correlated with quality of recovery. In theory, it should always be possible to do this if accurate GCS records are kept, but many reports on the prognosis of traumatic coma have made the simpler distinction between comatose and not comatose on the basis of the summated GCS score. Lyle et al. (1986), who did this, noted that the GCS score is a relatively insensitive measure of recovery, since the early return of spontaneous eye opening has little prognostic value. The GCS descriptors can be combined with terms such as akinetic mutism, apallic state, and persistent vegetative state, but the significance of these terms is not always clear and they can be dangerous labels if given a prognostic importance during the first few months after injury. 8.4.3

RETROSPECTIVE GRADING: AMNESIA

The other widely used measure of duration of impaired consciousness is the period of PTA. Ritchie Russell, in his pioneering study of the neurology of head injury (Russell, 1932), argued that the duration of unconsciousness is best estimated as the period before return of memory. He found that the return of memory could be timed by the patient’s recollection of when he woke up. Russell believed that this wake-up time could be estimated with fair accuracy long after the accident. On this basis, he graded surviving cases into three groups – those unconscious for less than 1 hour, those unconscious for 1–24 hours and those unconscious for longer periods. With later experience, Russell became aware that the first clear recollection could be followed by a further period of amnesia, and the PTA was therefore measured by the return of continuous memory (Russell and Nathan, 1946). The PTA was correlated with return to full wartime duties after head injury and appeared to be a robust prognosticator. Russell and his colleagues also studied the period of amnesia before the injury, retrograde

160

CLINICAL EXAMINATION AND GRADING

amnesia (RA); this was found to have less significance as a measure of injury severity. The RA is still generally recorded, as it has some diagnostic value; RA is often important in medicolegal issues, both as confirmation of a cerebral insult and because it obviously affects the victim’s capacity as an accident witness. But the duration of RA often shrinks with the passage of time; Richardson (1990), reviewing the abundant literature, concluded that the RA has no practical value as an indicator of injury severity or in prognosis. There has been general agreement that the PTA is a very valuable measure, especially for less severe injuries, but there has been doubt as to the reliability of amnesia endpoints ascertained by simple retrospective questioning. In the first place, the period of PTA may be interrupted by islands of recollection; Gronwall and Wrightson (1980) found such islands in 26 (39%) of 67 minor head injuries. Awareness of this phenomenon led to the definition of the PTA as the period of continuous memory loss after injury (see above), but this also has proved to be hard to define by retrospective interrogation. Much of Russell’s very productive work was done on British soldiers transferred to Oxford for assessment and rehabilitation. In their passages from accident site or battlefield to evaluation, they had usually experienced a series of well-defined and well-documented events that made retrospective estimation of return of continuous memory easier than in cases where all treatment has been undertaken in a single institution. Furthermore, in Russell’s earlier reports the proportion of cases with very prolonged periods of amnesia was not high (Russell and Nathan, 1946; Russell, 1954). In later studies on a larger database, Russell accepted that factors other than the severity of injury influenced the PTA, notably the presence of focal brain lesions, severe associated extracranial injuries, and the age of the injured person (Russell and Smith, 1961; Richardson, 1990). Present-day practice is especially concerned with cases slowly emerging from prolonged coma or confusional states, and retrospective interrogation is often done after the patient has been (quite rightly) briefed by family members on the course of events: the patient may then confuse what is remembered with what has been told, giving a falsely short PTA. This is especially likely to happen in children, and retrospective PTA measurements in children under the age of 8–10 years are very unreliable. On the other hand, a retrospective PTA evaluation done weeks or months later may give a falsely long measure, since the patient may have forgotten some landmark event or may have become confused between true recollections and second-hand information. For these and other reasons, efforts have been made to determine the PTA by more reliable means.

Richardson (1990) and Forrester and Geffen (1996) have reviewed the development of prospective measurements of the PTA – prospective in the sense that the aim is to detect by ongoing assessments the time at which the return of continuous memory can be demonstrated objectively. These tests embody standardized questionnaires, which are presented to the patient at regular (usually daily) intervals until the answers are considered to indicate that the patient has emerged from PTA. The best known is the Galveston Orientation and Amnesia Test (GOAT) designed by Levin, O’Donnell and Grossman (1979). This tests orientation in considerable detail (Table 8.7), allotting error points for disorientation. The test also gives error points for PTA and RA; for each, five error points are deducted for inability to recall a verifiable, or at least plausible event before or after injury, and an additional five error points when the patient cannot give details of this event. A final score is made by subtracting the sum of the error points from 100; a score of 75 or more is said to be within normal limits. The GOAT is open to the obvious objection that this normal score is in theory obtainable when a patient is

Table 8.7 Galveston Orientation and Amnesia test (Source: after Levin, O’Donnell and Grossman, 1982) Questions What is your name? Where do you live? Where were you born? Where are you now? City Hospital (need not be correctly named) On what date were you admitted to this hospital? How did you get here? What is the first event you remember after the injury? Describe event in detail, e.g. date, time, companions What is the last event you remember before the injury? Describe event in detail, e.g. date, time, companions What time is it now? (1 error point per half hour removed from correct time) What day of week is it now? (1 error point per day removed from correct day) What day of month is it now? (1 error point per day removed from correct day to maximum of 5) What is the month? (5 error points per month removed from correct month to maximum of 15) What is the year? (10 error points per year removed from correct year to maximum of 30)

Maximum no. error points 2 4 4 5 5 5 5 5 5 5 5 5 3 5 15 30

161

EVALUATION OF INJURY SEVERITY

fully oriented but still in an amnesic state; Gronwall and Wrightson (1980) found that a head-injured person may be oriented but amnesic, or vice versa. Nevertheless, Levin, O’Donnell and Grossman (1979) found that GOAT scores correlated well with the duration of GCS impairment and with the final outcome. Ewing-Cobbs et al. (1990) have devised a pediatric version of the GOAT, the Children’s Orientation and Amnesia Test (COAT). This does not attempt to test the PTA in any way, but does include a quantitative evaluation of temporal orientation, using the classic test of forward digit retention and the more modish capacity to recall television programs. The COAT is applicable in full to children in the age range 8–15, and without the memory tests in children as young as 3 years. Baryza and Haley (1994) have used the COAT as a screening test for otherwise undetected impairments in memory and orientation in children who appear to have recovered from head injury, and this may be its most valuable application. Two leading Australian neurorehabilitation units have designed simple questionnaires for evaluation of the PTA. From Sydney, Shores et al. (1986) have reported on the Westmead PTA scale. This embodies two biographical questions, five questions related to time date and place and three questions assessing recollection of pictures of objects (Table 8.8). The ability to remember the examiner’s face and name are also tested. The patient is deemed to be out of the PTA when able to give correct answers in all components of the scale on three consecutive days. Shores (1989) compared this scale with the duration of coma measured with the GCS and concluded that the Westmead scale was a better predictor of outcome. Haslam et al. (1994) have further explored the relation between long-term cognitive impairments and PTA established with the Westmead scale. These authors have reported on a new variable, the post-coma disturbance (PCD). This is the period of confusion after emergence from coma and is derived by subtracting the duration of coma from the duration of PTA. In this study, it appeared that the PCD was a significant predictor of impairment in recent memory 12 months after injury, whereas the PTA better predicted poor performance in information processing. With both the PCD and the PTA, the relations between duration and cognitive impairment were non-linear. Forrester and Geffen (1996) have criticized the Westmead scale on practical grounds, and advocate the scale used in the Julia Farr Centre, Adelaide. In this scale (Table 8.9), the questionnaire has six orientation items and five memory items. In the memory items, the patient is asked to memorize the name attached to a photograph, a gesture, and the names of three objects shown in photographs. Memory is not tested

Table 8.8 Westmead PTA Scale questionnaire – this is presented daily until the patient achieves a perfect score of 12 on three successive days; the PTA is deemed to have ended on the first of the three days (after Shores et al., 1986) Questions

Maximum No. points

1. How old are you?

1

2. What is your date of birth?

1

3. What month are we in?

1

4. What time of day is it?

1

5. What day of the week is it?

1

6. What year are we in?

1

7. What is the name of this place? If the patient does not know, a multiple choice is given – home, name of hospital, name of another hospital

1

8. The patient is asked to remember the examiner’s face. On the following day, he/she is shown three photographs, one of the examiner, and asked to identify the examiner

1

9. The patient is asked to remember the examiner’s first name. On the following day, he/she is asked to recall this name; if unable to do so, he/she is asked to select the name from a series including this name and two phonologically similar names or names with an equal number of syllables

1

10. Pictures I, II and III: the patient is shown three colored pictures of common objects and asked to name them. On the following day, the patient is asked to name the pictures. If unable to do so, he/she is asked to identify these pictures in a series of 12 pictures, being a random assortment of the three original pictures and nine distractor pictures

3

Total

12

until full orientation is confirmed. In this test also, PTA is deemed to be ended when the patient scores correctly in all orientation and memory tests on three consecutive days. G. Geffen (personal communication) finds that the GOAT and Julia Farr Centre PTA scale correlate closely, and the choice of scale may be a matter of unit preference rather than theoretical advantage. 8.4.4

APPLICATIONS OF CLINICAL EVALUATION

It seems that for most clinical trials, severe head injuries are best defined prospectively by the GCS score, either summated or as the best motor score,

162

CLINICAL EXAMINATION AND GRADING

Table 8.9 Julia Farr Centre Post-traumatic Amnesia Scale®. Orientation tests: These include four autobiographical questions, one question on time of day (cf. Question 4 in Westmead Scale) and one on place (cf. Question 7 in Westmead Scale). Orientation is tested daily until a score of 6 or more is achieved on three successive days. Memory tests: When oriented in person, time and place, the patients is taught to memorize a gesture, the name of a person seen in a black-and-white photograph and three objects seen in black-and-white photographs – a cup, a comb and an umbrella. On the following and successive days, the patient is asked to recall these, first freely and if unable to do so, after cued prompting by showing the test gesture or photograph together with a distractor. When the patient is oriented and achieves a minimum memory score of 5 (gesture ≥ 1, name ≥ 1, picture ≥ 1) on three successive days, the PTA endpoint is recorded for the first of the three days. (Source: from Forrester and Geffen, 1996) Orientation tests Questions

Answers

Score

a. Personal orientation: 1. What is your name?

No answer or wrong answer Correct answer

0 1

2. Are you married/Do you live with a partner?

No answer or wrong answer Correct answer

0 1

3. Do you have any children?

No answer or wrong answer Correct answer

0 1

4. What is your job?

No answer or wrong answer Correct answer

0 1

No answer or wrong answer Prompted correct answer Unprompted correct answer

0 1 2

No answer or wrong answer Prompted correct answer Unprompted correct answer

0 1 2

Maximum score

8

b. Orientation in time: 5. What time of day is it? (Is it morning, afternoon or night?)

c. Orientation in place: 6. Where are we now? (Are we at home, in a hospital or a hotel?)

Orientation daily total Memory tests

Maximum score

Test

Response

1. Gesture

Free recall Cued recall Recognition

3 2 1

2. Name of person in photograph

Free recall Cued recall Recognition

3 2 1

3. Pictures – (i)

Free recall Recall after cue Recall when shown and reject the distractor

3 2 1

– (ii)

Free recall Recall after cue Recall when shown and reject the distractor

3 2 1

– (iii)

Free recall Recall after cue Recall when shown and reject the distractor

3 2 1

Picture total Memory total

(i) + (ii) + (iii)

9 15

EVALUATION OF INJURY SEVERITY

determined after resuscitation. For prognostic purposes, the GCS score can be strengthened by taking into account other predictors, notably age, arterial blood pressure and the absence of pupillary light reflexes on one or both sides. Choi et al. (1988) have tried to refine the prognosis after severe head injury by preparing three graphs relating outcome to motor score, pupils, and age. The prognostic importance of increasing age is now well documented. Jennett and Teasdale (1981) studied outcomes in severely head-injured persons who remained in coma for at least six hours. There was a linear relationship between age and bad outcome (death or vegetative survival) and over 70 years there were no good recoveries. Luerssen, Klauber and Marshall (1988) also found an increase in mortality rates with advancing age; for comatose (GCS < 8) patients, the death rate rose steeply in the 45–49-year age group, and remained between 60% and 80% thereafter. Conversely, comatose children over the age of 5 years tend to have better outcomes. Infants and young children do not show this favorable tendency, but the difficulties in grading coma below the age of 5 years make it necessary to be cautious in using the consciousness level for prognoses in this age group (Simpson et al., 1991), and especially in extrapolating adult experience to the prognosis in injured infants. For retrospective classification of head-injury severity, it is still debatable whether duration of coma or length of PTA is the better yardstick. Much effort has gone into this debate, and some of this represents quests for precision in a field where in reality precision is impossible. For patients who emerge from coma to become responsive and cooperative, the time of recovery from confusion and amnesia has pragmatic importance in rehabilitation and prognosis and a simple standardized endpoint test is desirable; to establish this, a questionnaire is certainly useful. Wilson et al. (1993) have emphasized that the PTA is of value as a measure of injury severity, even when ascertained by the traditional method of retrospective questioning; when compared with coma duration, the PTA correlated better with lesion severity measured in MRI scans. Nevertheless, duration of coma is also important. There appears to be agreement that coma persisting after 14 days is a very adverse finding and usually predicts a severe disability. Clinical records should be maintained to ensure that this period is well documented for future reference. It is also desirable to record separately the return of eye opening, responsiveness to commands and capacity to communicate. This book is concerned only with severe head injuries, but it should be noted that Rimel et al. (1981, 1982) have used the GCS within 1 hour of admission

163

to identify minor (GCS score 13–15) and moderate (GCS score 9–12) head injuries. This has an attractive simplicity, but Johnstone et al. (1993) found poor interrater consistency in defining a moderate head injury on the basis of a coma scale, especially when used by a relatively inexperienced observer. For less severe head injuries, PTA estimated at discharge seems to be a preferable criterion. When an exact definition of a relatively short period of PTA is needed for research purposes, it may be advisable to administer a simple orientation questionnaire at short intervals, taking accurate answers to a standard set of questions as the endpoint (Gronwall and Wrightson, 1980). 8.4.5

CLINICAL EXAMINATION AND OUTCOME

Outcome grading is discussed in Chapter 22. The assessment of outcome requires a synthesis of medical and social data, and the whole spectrum of posttraumatic neurological disability, unique in each individual, has to be taken into account. An appropriate clinical assessment is therefore an essential part of outcome evaluation. The purpose of the assessment determines its depth and scope. Epidemiological outcome studies require broad categorizations based on social functional evaluations; these categorizations are discussed in Chapter 22. Rehabilitation requires an atomistic analysis of function, repeated over time. A comprehensive neurological examination is an essential prelude to any rehabilitation program. A preliminary and selective neuropsychological evaluation is also desirable, and when a planned program has been concluded, a full reassessment is essential. It is often desirable to repeat such assessments at scheduled times. Victims of head injury often express resentment that rehabilitation has been discontinued too soon, and this may in part stem from failure to promise a future review of progress and perhaps a further cycle of rehabilitation. In the USA, Australia and many other countries, the legal system requires definitive or interim medical assessments of disability. These have to be both atomistic and holistic. In assessing the outcome of a severe head injury for medicolegal purposes, the impact of the cerebral injury must be evaluated in terms of physical, cognitive and behavioral effects, but also in its effects on the victim’s social status and quality of life. Table 8.10 sets out the chief complaints voiced after head injuries; each requires detailed analysis and objective verification where possible. It should be noted that the causes of some of these complaints include non-neurological injuries of the facial skeleton, the facial viscera or even extracranial structures. Efforts are being made to formulate the effects of injury in percentages of total impairment (American

164

CLINICAL EXAMINATION AND GRADING

Table 8.10 Medicolegal checklist – the first column lists the chief complaints made after severe head injury (the list is not exhaustive), the second sets out the chief impairments or other conditions usually associated with each complaint Complaint

Common causes

Personality change

Cerebral damage, especially frontal lobe damage Loss in lifestyle, depression

Poor memory, poor concentration, reduced intelligence

Cerebral damage Depression

Speech defect

Cerebral damage Cerebellar damage Cranial nerve injury

Blackouts, giddiness and ‘funny turns’

Epilepsy Vasovagal attacks Postural vertigo

Headache, facial pain, scalp pain

Tension states Nerve injury Migraine (rare)

Loss of smell/taste

Olfactory nerve injury

Deafness, noise in ears

Injury to ear and/or auditory nerve Cranial bruit

Visual loss

Injury to eye and/or visual pathways

Double vision

Injury to cranial nerves III, IV, VI Eye injury (rare)

Impaired swallowing and/or chewing

Injury of lower cranial nerves Maxillofacial injury Injury to dentition

Limb weakness, tremor, unsteadiness, gait change

Cerebral injury affecting motor or sensorimotor pathways Cerebellar injury Spinal or limb injury

Incontinence and/or impotence

Spinal injury; cerebral injury – especially frontal lobe injury

Disfigurement

Scar, loss of hair, cranial or facial deformity, eye injury

Medical Association, 1993). These formulations are difficult to apply in a realistic way when the cognitive and behavioral outcomes of head injury are under consideration, but can be meaningful if given as broad assessments of social incapacity. The purpose of outcome evaluation also determines the timing of the final examination. For research purposes, a relatively early evaluation may be acceptable. Choi et al. (1994) have argued for an assessment of outcome at 6 months after injury. For medicolegal evaluations, a later period is necessary, both because much functional improvement may become evident after 6 and even 12 months, and also because all

parties should be satisfied that a plateau in recovery has been reached. As a rule, 2 years or more should elapse. In children, the final examination is usually deferred until adolescence, so that the impact of the head injury on the child’s educational experience can be assessed and quantified by neuropsychological tests. Such a prolonged deferral may have adverse financial consequences for the child’s parents, and there should be legal provision for interim evaluation at an earlier date, if the child’s upbringing is in any way dependent on a monetary settlement.

8.5

References

American College of Surgeons Committee on Trauma (1993) Advanced Trauma Life Support® Program for Physicians. Instructor manual, 5th edn, American College of Surgeons, Chicago, IL. American Medical Association (1993) Guides to the Evaluation of Permanent Impairment, 4th edn, American Medical Association, Chicago, IL. Bacon, P. J. and Smith, S. E. (1993) Cardiovascular and sweating dysfunction in patients with Holmes–Adie syndrome. Journal of Neurology, Neurosurgery and Psychiatry, 56, 1096–1102. Bailey, I., Bell, A., Gray, J. et al. (1991) A trial of the effect of nimodipine on outcome after head injury. Acta Neurochirurgica (Vienna), 110, 97–105. Baryza, M. J. and Haley, S. M. (1994) Use of the Children’s Orientation and Amnesia Test at hospital discharge for children with neurological and nonneurological traumatic injuries. Brain Injury, 8, 167–173. Born, J. D. (1988) The Glasgow–Li`ege scale. Prognostic value and evolution of motor response and brain stem reflexes after severe head injury. Acta Neurochirurgica (Vienna), 91, 1–11. Born, J. D., Hans, P., Dexters, G. et al. (1982) Evaluation pratique du dysfonctionnement enc´ephalique chez le traumatis´e cranien. Neurochirurgie, 28, 1–7. Born, J. D., Hans, P., Albert, A. and Bonnal, J. (1987) Inter-observer agreement in assessment of motor response and brainstem reflexes. Neurosurgery, 20, 513–517. Braakman, R., Gelpke, G. J., Habbema, J. D. F. et al. (1980) Systematic selection of prognostic features in patients with severe head injury. Neurosurgery, 6, 362–370. Bricolo, A., Turazzi, S. and Feriotti, G. (1980) Prolonged posttraumatic unconsciousness. Therapeutic assets and liabilities. Journal of Neurosurgery, 52, 625–634. Chesnut, R. M., Gautille, T., Blunt, B. A. et al. (1994) The localizing value of asymmetry in pupillary size in severe head injury: relation to lesion type and location. Neurosurgery, 34, 840–845. Choi, S. C., Narayan, R. K., Anderson, R. L. and Ward, J. D. (1988) Enhanced specificity of prognosis in severe head injury. Journal of Neurosurgery, 69, 381–385. Choi, S. C., Barnes, T. Y., Bullock, R. et al. (1994) Temporal profile of outcomes in severe head injury. Journal of Neurosurgery, 81, 169–173. Doty, R. L., Shaman, P., Kimmelman, C. P. and Dann, M. S. (1984) University of Pennsylvania Smell Identification Test: a rapid quantitative olfactory function test for the clinic. Laryngoscope, 94, 176–178. Duhaime, A.-C., Alario, A. J., Lewander, W. J. et al. (1992) Head injury in very young children: mechanisms, injury types, and ophthalmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics, 90, 179–185. Ewing-Cobbs, L., Levin, H. S., Fletcher, J. M. et al. (1990) The Children’s Orientation and Amnesia Test: relationship to severity of acute head injury and to recovery of memory. Neurosurgery, 27, 683–691. Fearnside, M. R., Cook, R. J., McDougall, P. and McNeil, R. J. (1993) The Westmead Head Injury Project outcome in severe head injury. A comparative analysis of pre-hospital, clinical and CT variables. British Journal of Neurosurgery, 7, 267–279. Forrester, G. and Geffen, G. (1996). The Julia Farr Services Post Traumatic Amnesia Scales (PTAS), in ACER Psychological Tests and Materials Catalogue 1996–7, Australian Council for Educational Research, Sydney, NSW. Gardner, R. A. and Broman, M. (1982) The Purdue Pegboard: normative data on 1334 school children. Journal of Clinical Child Psychology, 8, 156–162. Gennarelli, T. A. (1984) Emergency department management of head injuries. Emergency Medicine Clinics of North America, 2, 749–760. Gloss, D. S. and Wardle, M. G. (1982) Reliability and validity of American Medical Association’s guide to ratings of permanent impairment. Journal of the American Medical Association, 248, 2292–2296. Griffiths, R. (1970) The Abilities of Young Children: A Comprehensive System of Mental Measurement for the First Eight Years of Life, Young, Chard.

REFERENCES Gronwall, D. and Wrightson, P. (1980) Duration of post-traumatic amnesia after mild head injury. Journal of Clinical Neuropsychology, 2, 51–60. Hahn, Y. S., Chyung, C., Barthel, M. J. et al. (1988) Head injuries in children under 36 months of age: demography and outcome. Child’s Nervous System, 4, 34–40. Haslam, C., Batchelor, J., Fearnside, M. R. et al. (1994) Post-coma disturbance and post-traumatic amnesia as nonlinear predictors of cognitive outcome following severe closed head injury: findings from the Westmead Head Injury Project. Brain Injury, 8, 519–528. Heinze, J. (1969) Cranial nerve avulsion and other neural injuries in road accidents. Medical Journal of Australia, 2, 1246–1249. Hofer, T. (1993) Glasgow Scale relationships in pediatric and adult patients. Journal of Neuroscience Nursing, 25, 218–227. Jagger, J., Fife, D., Vernberg, K. and Jane, J. A. (1984) Effect of alcohol intoxication on the diagnosis and apparent severity of brain injury. Neurosurgery, 15, 303–306. Jennett, B. (1992) Severe head injuries: ethical aspects of management. British Journal of Hospital Medicine, 47, 354–357. Jennett, B. and Bond, M. (1975) Assessment of outcome after severe brain damage. A practical scale. Lancet, i, 480–484. Jennett, B. and Plum, F. (1972) Persistent vegetative state after brain damage. A syndrome in search of a name. Lancet, i, 734–737. Jennett, B. and Teasdale, G. (1977) Aspects of coma after severe head injury. Lancet, i, 878–881. Jennett, B. and Teasdale, G. (1981) Management of head injuries, F. A. Davis, Philadelphia, PA. Jennett, B., Teasdale, G., Galbraith, J. et al. (1977) Severe head injuries in three countries. Journal of Neurology, Neurosurgery and Psychiatry, 40, 291–298. Jennett, B., Teasdale, G., Braakman, R. et al. (1979) Prognosis of patients with severe head injury. Neurosurgery, 4, 283–289. Johnstone, A. J., Lohlun, J. C., Miller, J. D. et al. (1993) A comparison of the Glasgow Coma Scale and the Swedish Reaction Level Scale. Brain Injury, 7, 501–506. Jones, N. R., Molloy, C. J., Kloeden, C. N. et al. (1993) Extradural haematoma: trends in outcome over 35 years. British Journal of Neurosurgery, 7, 465–471. Kocher, T. (1901) Hirnerschutterung, ¨ Hirndruck und chirurgische Eingriffe bei Krankheiten, Holder, ¨ Vienna. Levin, H. S., O’Donnell, V. M. and Grossman, R. G. (1979) The Galveston Orientation and Amnesia Test: a practical scale to assess cognition after head injury. Journal of Nervous and Mental Disease, 167, 675–684. Luersson, T. G., Klauber, M. R. and Marshall, L. F. (1988) Outcome from head injury related to patient’s age. A longitudinal prospective study of adult and paediatric head injury. Journal of Neurosurgery, 68, 409–416. Lyle, D. M., Pierce, J. P., Freeman, E. A. et al. (1986) Clinical course and outcome of severe head injury in Australia. Journal of Neurosurgery, 65, 15–18. Marshall, L. F., Barba, D., Toole, B. M. and Bowers, S. A. (1983a) The oval pupil: clinical significance and relationship to intracranial hypertension. Journal of Neurosurgery, 58, 566–568. Marshall, L. F., Becker, D. P., Bowers, S. A et al. (1983b) The National Traumatic Coma Data Bank. Part 1: design, purpose, goals, and results. Journal of Neurosurgery, 59, 276–284. Mathiowetz, V. (1990) Grip and pinch strength measurements, in Muscle Strength Testing, (ed. L. F. Amundsen), Churchill Livingstone, New York, ch. 7. Medical Research Council Brain Injuries Committee (1941) A glossary of psychological terms commonly used in cases of head injury. MRC War Memorandum No. 4. HMSO, London. Menegazzi, J. J., Davis, E. A., Sucov, A. N. and Paris, P. M. (1993) Reliability of the Glasgow Coma Scale when used by emergency physicians and paramedics. Journal of Trauma, 34, 46–48. Muizelaar, J. P., Marmarou, A., Young, H. F. et al. (1993) Improving the outcome of severe head injury with the oxygen radical scavenger polyethylene glycol-conjugated superoxide dismutase: a Phase II trial. Journal of Neurosurgery, 78, 375–382. Narayan, R. K. (1989). Emergency room management of the head -injured patient, in Textbook of Head Injury, (eds D. P. Becker and S. K. Gudeman), W. B. Saunders, Philadelphia, PA, ch. 2. Narayan, R. K. (1994) Comments on Chesnut et al., 1994. Neurosurgery, 34, 845–846. North, J. B. and Jennett, S. (1974) Abnormal breathing patterns associated with acute brain damage. Archives of Neurology, 31, 338–344. Raven, J. C. (1986) Coloured Progressive Matrices, H. K. Lewis, London. Reilly, P. L., Simpson, D. A., Sprod, R. and Thomas L. (1988) Assessing the conscious level in infants and young children: a paediatric version of the Glasgow Coma Scale. Child’s Nervous System, 4, 30–33. Richardson, J. T. E. (1990) Clinical and Neuropsychological Aspects of Closed Head Injury, Taylor & Francis, London. Rimel, R. W., Giordani, B., Barth, J. T. et al. (1981) Disability caused by minor head injury. Neurosurgery, 9, 221–228.

165

Rimel, R. W., Giordani, B., Barth, J. T. and Jane, J. A. (1982) Moderate head injury. Completing the clinical spectrum of brain trauma. Neurosurgery, 11, 344–351. Robertson, R. C. L. and Pollard, C. (1955) Decerebrate state in children and adolescents. Journal of Neurosurgery, 12, 13–17. Robertson, C. S., Clifton, G. L., Taylor, A. A. and Grossman, R. G. (1983) Treatment of hypertension associated with head injury. Journal of Neurosurgery, 59, 455–460. Rowbotham, G. F. (1945) Acute Injuries of the Head, 2nd edn, E. & S. Livingstone, Edinburgh, pp. 104, 144. Russell, W. R. (1932) Cerebral involvement in head injury. A study based on the examination of two hundred cases. Brain, 55, 549–603. Russell, W. R. (1947) The neurology of brain wounds. British Journal of Surgery War Supplement, 1, 250–252. Russell, W. R. (1954) Studies on head injury. British Medical Bulletin, 10, 65–68. Russell, W. R. and Nathan, P. W. (1946) Traumatic amnesia. Brain, 69, 280–300. Russell, W. R. and Smith, A. (1961) Post-traumatic amnesia in closed head injury. Archives of Neurology, 5, 4–17. Ryall, R. G., Peacock, M. K. and Simpson, D. A. (1992) Usefulness of 2-transferrin assay in the detection of cerebrospinal leaks following head injury. Journal of Neurosurgery, 77, 737–739. Selhorst, J. B., Gudeman, S. K., Butterworth, J. F. et al. (1985) Papilledema after acute head injury. Neurosurgery, 16, 357–363. Seshia, S. S., Seshia, M. M. K. and Sachdeva, R. K. (1977) Coma in childhood. Developmental Medicine and Child Neurology, 19, 614–628. Sheridan, M. (1976) STYCAR Vision Tests, NFER, Windsor. Shores, E. A. (1989) Comparison of the Westmead PTA Scale and the Glasgow Coma Scale as predictors of neuropsychological outcome following extremely severe blunt head injury (letter). Journal of Neurology, Neurosurgery and Psychiatry, 52, 126–127. Shores, E. A., Marosszeky, J. E., Sandanam, J. and Batchelor, J. (1986) Preliminary validation of a clinical scale for measuring the duration of posttraumatic amnesia. Medical Journal of Australia, 144, 569–572. Simpson, D. A. and Reilly, P. L. (1982) Paediatric coma scale (letter). Lancet, ii, 450. Simpson, D. A., Cockington, R. A., Hanieh, A. et al. (1991) Head injuries in young children: the value of the Paediatric Coma Scale. Child’s Nervous System, 7, 183–190. Starmark, J.-E. (1988) Analysing ‘Coma Scales’. The Introduction of a New Scale – The Reaction Level Scale, University of Goteborg,. ¨ Goteborg. ¨ Starmark, J.-E., Holmgren, E. and Stålhammer, D. (1988) Current reporting of responsiveness in acute cerebral disorders. A review of the neurosurgical literature. Journal of Neurosurgery, 69, 692–698. Starmark, J.-E., Stålhammer, D., Holmgren, E. and Rosander, B. (1988) A comparison of the Glasgow Coma Scale and the Reaction Level Scale (RLS85). Journal of Neurosurgery, 69, 699–706. Stening, W. A., Berry, G., Dan, N. G. et al. (1986) Experience with acute subdural haematomas in New South Wales. Australian and New Zealand Journal of Surgery, 56, 549–556. Teasdale, G. and Jennett B. (1974) Assessment of coma and impaired consciousness. A practical scale. Lancet, ii, 81–84. Teasdale, G. and Jennett, B. (1976) Assessment and prognosis of coma after head injury. Acta Neurochirurgica (Vienna), 34, 45–55. Teasdale, G. M., Knill-Jones, R. and Jennett, W. B. (1974) Assessing and recording conscious level. Journal of Neurology, Neurosurgery and Psychiatry, 37, 1286. Teasdale, G., Knill-Jones, R. and Van der Sande, J. (1978) Observer variability in assessing impaired consciousness and coma. Journal of Neurology, Neurosurgery and Psychiatry, 41, 603–610. Tiffin, J. (1968) Purdue Pegboard Examiner Manual, Science Research Associates, Chicago, IL. Trott, J. and Cooter, R. (1995) Systematic clinical assessment, in Craniomaxillofacial Trauma, (eds D. J. David and D. A. Simpson), Churchill Livingstone, Edinburgh, ch. 6. Walker, A. E. (1985) Cerebral Death, 3rd edn, Urban & Schwarzenberg, Baltimore, MD, ch. 3. Walsh, K. W. (1985) Understanding Brain Damage. A Primer of Neuropsychological Evaluation, Churchill Livingstone, Edinburgh. Wilson, J. T. L., Teasdale, G. M., Hadley, D. M. et al. (1993) Post-traumatic amnesia: still a valuable yardstick. Journal of Neurology, Neurosurgery and Psychiatry, 56, 198–201. Wood, M. and Hammerton, M. (1995) Impairments and disabilities, in Craniomaxillofacial Trauma, (eds D. J. David and D. A. Simpson), Churchill Livingstone, Edinburgh, ch. 22. Wood, M. and Woodroffe, P. (1995) Neuropsychological testing, in Craniomaxillofacial Trauma, (eds D. J. David and D. A. Simpson), Churchill Livingstone, Edinburgh, appendix III. Yager, J. V., Johnston, B. and Seshia, S. S. (1990) Coma scales in pediatric practice. American Journal of Diseases in Children, 144, 1088–1091.