Brain (1984), 107, 15-36

DELAYED DETERIORATION FOLLOWING
MILD HEAD INJURY IN CHILDREN
by J. W. SNOEK1, J. M. MINDERHOUD1 and J. T. WILMINK 2
(From the Departments of Neurology1 and Neuroradiology1, University Hospital,
PO Box 30.001, 9700 RB Groningen, The Netherlands)
SUMMARY

INTRODUCTION

It is a well established fact that children with head injuries who deteriorate following
a lucid interval usually do not have an expanding intracranial haematoma, unlike
adults where secondary deterioration following trauma is a strong indication that
such a haematoma is present (Walton and Brooks, 1897; Pickles, 1949; Lindenberg
et al., 1955; Galbraith and Smith, 1976; James, 1979; Bruce et al., 1979, 1981a).
Secondary deterioration not due to intracranial haematoma has been attributed
to acute cerebral oedema (Pickles, 1949; Biemond, 1970), convulsions (Small and
Woolf, 1957; Livingston and Mahloudji, 1961; Grand, 1974; Reillyefa/., 1975; Rose
et al., 1977), spreading depression of Leao (Oka et al., 1977), a migrainous
mechanism (Haas et al., 1975), a functional disturbance of the rostral brainstem
(Todorow and Feller, 1982), concomitant viral meningoencephalitis (Peters et al.,
1978), complicating meningitis (Rose et al., 1977) or to unknown causes (Plum and
Posner, 1980), whereas recent reports deal with the syndrome of diffuse cerebral
swelling, which is attributed to acute cerebral hyperaemia (Langfitt and Bruce, 1975;
Bruce et al., 1979, 1981a, b). The CT scans of children with this diffuse cerebral
swelling show a diminution or obliteration of the cerebral ventricles and the

Downloaded from by guest on October 22, 2014

A series of 42 children is described who, following a seemingly minor or trivial head injury, developed
neurological signs after a lucid or symptom-free period. This group constitutes 4.34 per cent of 967
consecutive patients aged 2 months to 17 years who were seen by members of the neurological staff
during the years 1978-1981.
Only one patient had an intracranial haematoma. The majority of patients showed a benign
transient syndrome consisting of either convulsive or nonconvulsive signs with a spontaneous and full
recovery. There were, however, 3 deaths in this series, apparently due to severe and uncontrollable
unilateral or diffuse brain swelling, demonstrating the malignant counterpart of this benign syndrome.
The theories seeking to explain these phenomena are reviewed. Special reference is made to the
hypotheses of Bruce and his associates regarding brain swelling as a causative factor. It is considered
that an adequate theory to explain the pathogenesis is still lacking.
It is concluded that the juvenile brain responds to cranial trauma in a manner different from the adult
brain. This implies a different approach in policy to hospital admission.

16

J. W. SNOEK, J. M. M1NDERHOUD AND J. T. WILMINK

perimesencephalic cisterns, whereas the hyperaemia is related to increased CT
attenuation numbers of the deep frontal white matter (Zimmerman et al., 1978).
The present study was undertaken in order to investigate retrospectively the
incidence of lucid or symptom-free periods in a consecutive series of head-injured
children, to describe the different clinical pictures presented by these patients and to
correlate the clinical findings with the results of both EEG and CT studies.
PATIENTS AND METHODS

RESULTS

Fig. 1 shows the characteristics of the group of 967 consecutive juvenile patients.
Of these, 5 were either dead on arrival or died within a few minutes afterwards. Fig. 2
shows the age distribution of the remaining 962 patients. These patients have been
classified according to the Abbreviated Injury Scale (AIS; Ommaya, 1979), which is
summarized in Table 1. Because post-traumatic amnesia is usually difficult to assess
in children (especially in the younger ones), only the estimated period of
unconsciousness has been taken into account.
As fig. 1 shows, 5 patients with linear skull fractures were not admitted, usually
because the head injury had occurred more than 12 h previously. Of all admitted
patients, 22 died (5.7 per cent), death being attributable to primarily extracranial
lesions in 4 cases.
In a total of 40 children a history of a lucid or symptom-free period was found
following a head injury which had caused no immediate loss of consciousness. This
group of patients meets the criteria of a 'trivial injury' according to Jennett (1962,
1975) (no amnesia, haematoma or skull fracture, or linear fracture only). Two cases

Downloaded from by guest on October 22, 2014

The University Hospital of Groningen has approximately 1100 beds and serves a population of
400000 for primary referrals and 2 million for secondary referrals.
In 1972 a Head Injury Research Study was set up in collaboration with neurosurgical centres in
Rotterdam, Glasgow and Los Angeles (Jennett et al., 1977). All injured patients in whom a head injury
is suspected are seen by staff of the Neurology Department at the Accident and Emergency Clinic and
assessed in a uniform way, using the Glasgow Coma Scale (Teasdale and Jennett, 1974). In order to
study the occurrence of lucid or symptom-free periods in children after head injury, the records of all
head-injured children (aged 0 to 17 yrs), seen during the years 1978-1981, were reviewed. During this
period a total of 967 patients in this age group were seen as primary referrals, that is, not having
previously been admitted to another hospital. The neurological assessment usually took place within
1 or 2 h after the accident, in all cases within 24 h.
Patients were included in the present series if, after a head injury not causing unconsciousness or only
a very brief period of unconsciousness (not exceeding 5 min), this symptom-free period was followed by
a subsequent deterioration of the level of consciousness or by the onset of focal neurological signs, or
both. The clinicalfindingsof these patients were compared with the EEGs, skull radiographs and CT
scans if available. Where appropriate, attenuation numbers expressed as EMI units of the deep frontal
white matter were obtained as described by Zimmerman et al. (1978). These attenuation values were
compared with those of an appropriate control group of children. Average CT attenuation numbers of
the deep frontal white matter in a series of control scans, as will be described later in this paper, ranged
from 11.7 EMI units (SD 2.08) to 19.5 EMI units (SD 2.42).

H E A D I N J U R Y IN C H I L D R E N

PATIENTS /
967

17

M 7 rf

<y
»32OS

No skull
fracture 574

No skull fracture 248
/ 1 2 0 Linear
Skull fracture
135
15 Depressed

Skull
fracture 5
AIS 2
183

3 e.p.e.
12 No e.p.e.

AIS 3-5
141

181
2* 20(2)' 121
Survived Died Died Survived
22
FIG. 1. Characteristics of the group of 967 consecutive patients, aged 2 months to 17 years. AIS = Abbreviated
Injury Scale (see text), e.p.e. = early post-traumatic epilepsy. • = death attributable primarily to extracranial
injuries.
Downloaded from by guest on October 22, 2014

No
140120
10080604020-

<1

1

2

3

4

5

6

7
8
9
Age (yrs)

10

11

12

13

14

15

16

FIG. 2. Age distribution of 962 consecutive head-injured children, aged 2 months to 17 years.

17

18

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK
TABLE 1. ABBREVIATED INJURY SCALE

Description of injury

Cerebral injury with headache; dizziness; no loss of
consciousness.
Cerebral injury with or without skull fracture, with
less than 15 min unconsciousness; undisplaced
skull fracture; no post-traumatic amnesia.
Cerebral injury with or without skull fracture, with
unconsciousness more than 15 min, without severe
neurological signs; brief PTA* (less than 3 h).
Cerebral injury with or without skull fracture, with
unconsciousness of more than 15 min with definite
neurological signs; PTA 3-12 h. Compound skull
fracture.
Cerebral injury with or without skull fracture with
unconsciousness of more than 24 h; PTA > 12 h;
intracranial haemorrhage; signs of raised intracranial pressure.

AIS code

Degree of injury

1

Minor

2

Moderate

3

Severe (not life threatening)

4

Severe (life threatening; survival
probable)

5

Critical (survival uncertain)

with a history of a brief loss of consciousness (not exceeding 5 min) followed by
a lucid period, were included, bringing the total number to 42 (4.37 per cent of all
patients). Forty of the patients were admitted immediately after the first neurological assessment, 2 had been allowed to go home, one of whom was subsequently
admitted on the same day. In one patient the (short) transient period of
deterioration had occurred several hours previously and at the time of neurological
examination this patient had no symptoms; he therefore was allowed to go home.
Fig. 3 shows the age distribution of the 42 patients. The patients were divided into
two categories (fig. 4), the criterion being the occurrence or absence of convulsions
(Okmetal., 1977).
Group 1. Convulsive Cases
Thirteen children (31 per cent) had early post-traumatic seizures, occurring in the
first week after injury (Jennett, 1962,1969,1975). Table 2 presents the characteristics
of this group. Seven patients developed a focal or generalized status epilepticus,
which necessitated intubation and controlled ventilation in 6. More than half of the
patients developed these seizures within one hour after injury, 12 out of 13 within
two hours, and only one after a longer period (28 h).
Case 1. A 5-year-old-boy fell off his bicycle. He did not lose consciousness and resumed cycling
immediately. One hour after injury he developed focal seizures involving the right hand and arm. On
admission one hour later, the boy was in generalized status epilepticus, which reacted promptly to
intravenous diazepam. A CT scan performed within three hours after injury showed no abnormalities.
The mean CT attenuation numbers of the deep frontal white matter were 15.9 EMI units (SD 2.59).

Downloaded from by guest on October 22, 2014

*PTA = post-traumatic amnesia.

19

HEAD INJURY IN CHILDREN

No
87654-

3
21-

1 2

3

4

5 6

7

8 9 10 11 12 13 14 15 16
Age (yrs)

Downloaded from by guest on October 22, 2014

<1

17

FIG. 3. Age distribution of 42 children with a lucid interval following mild head injury.

.Convulsive easel
n - 13

42 patients

^-Transient syndrome of subacute
onset, 22 patients
Rapid deteriorationn-23

-Fata] case, 1 patient
Slow
deterioration,
required, 1 patient

Nonconvulsive cases
n-29
Slow deterioration
n-6

Slow deterioration, spontaneous
unprovement, 3 patienti
Fatal cases, 2 patients

Fio. 4. Lucid interval scries. Clinical course.

surgery

20

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK
TABLE 2. CHARACTERISTICS OF CONVULSIVE CASES

The ventricles were small, but the basal cisterns appeared normal. An EEG, performed within 24 h,
showed asymmetry with localized slow activity in the left temporal region. There were multifocal
isolated spikes and sharp-slow wave complexes, maximal in both temporal and parietal areas.
Several hours after admission the patient was fully alert, recalling all details of the injury. He was
discharged after two days. Two EEGs, performed one and five months after injury, respectively, were
normal. The past medical history of this patient revealed that he was premature and dysmature at
birth. During the first fourteen days of life cyanosis had been seen on several occasions. Serial EEGs
had not shown any abnormality at that time. He had been treated with anticonvulsants up to the age of
five. Psychomotor development was normal.

The family history for epilepsy was negative in all patients. Five patients,
however, had a past medical history (febrile convulsions, rhesus incompatibility,
mental retardation or prematurity; 1 patient had had an identical episode of
TABLE 3. RELATION OF EEG ABNORMALITIES TO TYPE OF SEIZURE

EEG
Type of seizure

N

P
P+G
G

1

D

F
2

1
1

D+F
1
1
1 1
1

E

F+E

1

Type of seizure. P = partial seizure. PG = partial + generalized seizure. G = generalized seizure.
EEG. N = no abnormalities. D = diffuse abnormalities. F = focal abnormalities. E = epileptic
discharges.

Downloaded from by guest on October 22, 2014

Number: 13
Mean age: 7.3 years (range 1-17)
Cause of injury, fall, 10
traffic accident, 1
struck by object, 2
Skull fracture: 5 (38 per cent)
Onset after injury: within 1 h, 7 (54 percent)
within 2 h, 12 (92 per cent)
Vomiting at onset: 3 (23 per cent)
Types of seizures: partial seizures, 4
partial + generalized seizures, 5
generalized seizures, 4
Status epilepticus: 7 (54 per cent)
Mean age of patients with status: 3.5 years (range 1-8)
Family history of epilepsy: 0
Past medical history: febrile convulsions, 1
rhesus incompatibility, 1
prematurity, 1
mental retardation, 1
early post-traumatic epilepsy, 1
Late epilepsy (follow-up 6 months to 3 years): 1

HEAD INJURY IN CHILDREN

21

convulsions after a previous trivial injury). One patient possibly had a rubella
infection and was readmitted with a second episode of status epilepticus a week later
at which time an increased rubella titre was found. Apart from this patient no
recurrence of seizures was noted in the follow-up period (six months to three years).
One or more EEGs were performed in 11 cases, usually within 24 h after injury
(range 6 h to 9 days). Table 3 shows that there was no relation between the EEG
findings and the type of seizure.
In only 2 patients was CT performed immediately following the convulsions.
Both were normal. The densities measured in the deep frontal white matter were 15.9
EMI units, SD 2.59 (Case 1) and 11.3 EMI units, SD 1.60, respectively.
Group 2. Nonconvulsive Cases
These 29 children did not develop early post-traumatic seizures. In the majority
(23 cases), there was a history of acute or subacute deterioration (i.e. from a level of
no signs, the full clinical picture was reached within 15 min) at varying times after
injury, whereas in the remaining 6 cases there was a slow, gradual deterioration.

TABLE 4. CHARACTERISTICS OF NONCONVULSIVE CASES WITH A TRANSIENT
SYNDROME OF ACUTE OR SUBACUTE ONSET

Number: 22
Mean age: 6 years (range 2-13)
Cause of injury: fall, 15
traffic accident, 5
struck by object, 2
Skull fracture: 5 (23 per cent)
Onset after injury: 5-30 min, 12
30-60 min, 4
1-2 h, 5
36 h, 1
Duration of signs: 15-30 min, 5
30 min-2 h, 2
2-12 h, 15
Vomiting at onset: 21 (95 per cent)
Signs: disturbances of sensory level only, 7
disturbances of sensory level with focal neurological signs, 9
focal neurological signs only, 1
confusion, 5

Downloaded from by guest on October 22, 2014

Cases with rapid deterioration. A. Patients with a transient syndrome of acute or
subacute onset. Patients were included in this group if the following criteria were
met. (1) Acute or subacute onset. (2) Signs ending quickly, sometimes abruptly,
mostly after a short period (not exceeding 12 h). (3) Nature of signs: loss of
consciousness with or without focal neurological signs, focal neurological signs only
or a period of severe restlessness or confusion following a period of normal
behaviour.
This subgroup consists of 22 patients whose characteristics are listed in Table 4.

22

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

The past medical history was unremarkable in all but 3 cases. One girl (aged 12
years) was known to have behavioural problems and EEG abnormalities (bilateral
synchronous spike and wave complexes) without history of seizures. One boy (aged
5 years) had had febrile convulsions in the past. There was a family history of
epilepsy but no history of seizures in another boy (aged 6 years), whose EEG
subsequently showed spike and wave complexes. One patient was admitted with
fever, the CSF showing no abnormalities.
There were remarkable differences in the severity of the presenting clinical picture
as is illustrated by the following case histories.

A similar short-lasting but alarming picture, consisting of deep coma with
bilaterally unreactive pupils, resembling acute tentorial herniation by a rapidly
expanding mass, was seen in one other patient in this group, a girl aged 12 years. In
this patient these signs occurred 36 h after injury. The CT scan performed at that
time showed small ventricles with obliterated cisterns (fig. 5). The deep frontal white
matter densities were 17.4 EMI units, SD 2.47.
The most frequently observed focal neurological signs were pupillary abnormalities (in 4 out of 10 patients), followed by conjugate eye deviation and hemiparesis.
A combination of focal signs was seen in 2 children. Two patients presented with
transient blindness. In both patients the onset of blindness (some minutes and one
hour after injury, respectively) occurred in combination with sudden vomiting. The
period of blindness lasted less than one hour in both cases and ended abruptly. One
child had struck his forehead, the other had fallen on his occiput.
In this group of patients, 6 EEGs were performed within one to five days after
injury. Three of these showed local slow activity, 1 slight diffuse abnormalities,
whereas bilateral synchronous sharp and slow wave complexes were noted in 2.
A CT scan was performed in only 2 cases {presented above).
B. Fatal case with a rapid deterioration. Case 4. An 8-year-old boy jumped from a slowly
moving cart and fell on his head. He got up immediately, walked towards his father and said 'I feel so
funny in my head'. He then became flaccid. His mother, a nurse, subsequently stated that she had
noticed dilated pupils at the time. On admission one hour later the boy was shocked and respiration
was shallow. He was unresponsive and both pupils were dilated andfixed.No bruises were found on
the head and there was no skull fracture. Echoencephalography showed no midline shift. The boy was

Downloaded from by guest on October 22, 2014

Case 2. A 2-year-old boy fell down some stairs. He cried immediately. One hour later he became
unresponsive within a few minutes. On admission he was obtunded with a left hemiparesis. A CT scan,
performed within three hours of injury, was normal, with frontal white matter attenuation values of
11.7 EMI units, SD 1.57. Within four hours after the injury the hemiparesis disappeared and the level
of consciousness became normal. Two days later the boy was discharged. An EEG, performed on the
same day, showed slow waves over the whole of the right hemisphere.
Case 3. A 5-year-old boy was struck by a car. He was unconscious for less than five minutes. In the
ambulance the right pupil dilated while he was talking. He subsequently became comatose with
bilaterally dilated unreactive pupils. On admission, one hour after the accident, he was awake with
normally reacting pupils. The EEG (day 5) showed only slight diffuse abnormalities. He was
discharged after six days.

HEAD INJURY IN CHILDREN

S

intubated and ventilated. Soon after admission massive pulmonary oedema developed. It was felt that
the situation was hopeless and no attempts were made at further investigation. He died within hours of
admission. Permission for donor nephrectomy was granted by his parents; no attempts were made to
obtain permission for autopsy.
Slow deterioration. A. Slow deterioration, treated surgically. Case 5. A 10-year-old boy
struck his head against a wall while being tossed by his friends. He was briefly unconscious, probably
less than one minute, but recovered quickly. On admission he opened his eyes on request, obeyed
commands and was fully orientated. There was a small occipital fracture. He deteriorated slowly over
the following day. Laboratory screening revealed a clotting disorder, which was found to be a variant
of von Willebrand-Jurgens disease. A CT scan showed a hyperdense area in the posterior fossa, which
on exploration proved to be an extradural haematoma. He made a slow but excellent recovery.
B. Slow deterioration with gradual spontaneous improvement. In 3 patients (aged
10, 12 and 15 years, respectively) the signs developed slowly over a period of time
after a lucid interval and then gradually resolved. In 2 patients these symptoms
consisted of slowly progressive confusion with focal abnormalities in both the EEG
and CT scan, located in the frontal region in one patient and in the left temporal
region in the other. In the third case (Case 6) the deterioration consisted of more
severe confusion, with signs of uncal herniation.
Downloaded from by guest on October 22, 2014

FIG. 5. Example of CT section at the level of the anterior horns and the mesencephalon. Note the narrow aspect of
the ventricular horns which are not significantly displaced from the midline. The CSF cistern behind the
quadrigeminal plate is obliterated.

24

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

Case 6. A boy aged 12 years was hit by a car. He did not lose consciousness. On admission he was
lucid but nauseated. There was no skull fracture. Twelve hours after the accident he started to
deteriorate with confusion and an enlarged pupil on the right. Angiography (because of unavailability
of our CT scanner) showed a shift of the midline vessels with signs of a swollen right temporal lobe. No
avascular mass was discovered. Several hours later his level of consciousness started to improve and the
pupils became normal. The boy remained drowsy for two more days after which he recovered quickly.
An EEG performed four days after injury showed diffuse slowing together with a local area of slow
activity in the right temporal region.
C . Fatal cases with slow deterioration.
Case 7. A 13-year-old girl fell off her bicycle. She did
not lose consciousness at the time. On examination, three hours after the accident, she was lucid but
nauseated. No neurological abnormalities were found. There was no skull fracture. She was sent home.
Several hours later she developed a left hemiparesis and became drowsy. She was admitted to our
hospital. A CT scan showed a marked shift of the compressed ventricles to the left. No abnormalities in
the brain parenchyma were detected. Unfortunately raw CT data are no longer available, so no frontal
white matter attenuation values can be measured in retrospect. An angiogram, performed in order to
exclude an isodense haematoma, showed slowed intracerebral circulation, but no avascular mass. The
girl was intubated and ventilated; high doses of corticosteroids were given, together with mannitol. She
deteriorated over the next few hours and died on the following day. Autopsy was not performed.

DISCUSSION

The largest series of head-injured children, 4465 consecutively admitted cases,
was reported by Hendrick et al. (1964). It was shown that almost 50 per cent of those
who died were not in coma at the time of admission to hospital. Also, a history of a
short or prolonged lucid period was present in 0.8 per cent of the patients, whereas
the level of consciousness deteriorated gradually from the moment of impact in
1.8 per cent. Only the last category was correlated with a high percentage of
intracranial, mostly extradural, haematomas.
In this study we assessed children who, following a minor or trivial injury,
developed neurological signs following a lucid or symptom-free period. As these
included cases with focal neurological signs in conscious children and early posttraumatic convulsions not leading to coma, it is not possible, strictly speaking, to use
the term 'lucid interval', which refers to level of consciousness only, in these cases.

Downloaded from by guest on October 22, 2014

Case 8. A boy aged 9 years fell off his skateboard. He resumed playing immediately, not having lost
consciousness. Later that day he watched television for several hours and then went to bed. The next
morning he complained of a headache and told his mother that he had difficulty in walking. Some time
later he became stuporose and vomited. On admission, 16 hours after the accident, he did not open his
eyes to painful stimuli, but he located these adequately while moaning. A bruise was noted in the right
parietal area. There was no skull fracture. The left pupil was dilated, the right pupil somewhat smaller;
neither reacted to light. While the patient was being examined both pupils started to react to light
spontaneously. As the CT scanner was not available, an angiogram had been scheduled. It was felt,
however, that the spontaneous improvement made an expanding intracranial mass less likely. Without
having further deteriorated in the meantime, the boy suddenly developed fixed dilated pupils and
apnoea. He was immediately taken to the operating theatre where bilateral burr holes revealed
a swollen brain but no haematoma. The boy died 15 hours after admission. Consent for autopsy was
refused.

H E A D I N J U R Y IN C H I L D R E N

25

Therefore the criterion 'symptom-free' has been added. It does not seem illogical to
use these criteria because in this way it is possible to focus attention on all those
patients who demonstrated disturbed cerebral function after an injury which at first
had appeared to be minor or even trivial.
In 4.37 per cent of 962 patients such a symptom-free period was encountered. In
only one of these patients was an intracranial haematoma subsequently found. The
remaining 41 patients showed a syndrome consisting of either convulsive or
nonconvulsive signs.

Downloaded from by guest on October 22, 2014

Post-traumatic Epilepsy
Early post-traumatic epilepsy was noted in 16 head-injured children (1.6 per cent
of our total series of 962). The 3 patients not having had a lucid interval all had
a depressed fracture (fig. 1). According to Jennett (1962, 1973, 1975), early posttraumatic epilepsy is seen in about 5 per cent of all head-injured patients admitted to
hospital. In children under 5 years of age, however, the incidence is twice as high.
Hendrick and Harris (1968) also give an incidence of about 10 per cent in this age
group. In patients under 5 years of age we found a percentage of 3.2 per cent. The
difference between the incidence found by Jennett and by Hendrick and Harris, and
by us, probably reflects the distinction between patient populations in neurosurgical
and neurological departments.
According to Jennett, epilepsy rarely follows a trivial injury except in children
under 5 years. The only adults encountered with epilepsy after trivial injury are
those with 'immediate' epilepsy, when a seizure occurs at the moment the head
sustains an impact. In the present series, the occurrence of early post-traumatic
epilepsy was not limited to the under 5 age group, as 6 of the 13 patients were
between 6 and 17 years of age. A similar age distribution has been described by
Grand (1974).
Two explanations are possible for the relatively high frequency of early posttraumatic epilepsy following trauma in children as compared to adults (Jennett,
1962, 1975). The first is that children are predisposed to epilepsy, related to
underlying but hitherto undetected brain damage. The second, favoured by Jennett
and by Oka et al. (1977), is that children are somehow more liable to react to stress of
different kinds by a seizure. Our finding that in almost half of the convulsive cases
there was, albeit ill defined, a past medical history {see Table 2), which may have
resulted in a liability to develop convulsions, gives some support to the first
possibility expressed by Jennett. According to this author, the major significance of
early post-traumatic epilepsy is the risk of late epilepsy, which he found to be 25 per
cent. Focal early epilepsy in children was the only kind of early epilepsy in his series
which did not significantly increase the risk of late epilepsy. Except for the child that
was readmitted in a second episode of status epilepticus one week after the accident,
no other cases of late epilepsy were encountered during the follow-up (three months
to three years) of our patients with early epilepsy following trivial injury (Table 2).
Oka et al. (1977) did not find a single case of late epilepsy in their series.

26

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

In the literature several explanations have been offered for the delayed
deterioration in children following minor injuries. Several authors considered both
convulsive and nonconvulsive signs as part of one syndrome and have attempted to
give an explanation for its occurrence (Biemond, 1970; Haas et al., 1975; Oka et al.,
1977); others have confined themselves to giving an explanation for the occurrence
of convulsions only (Livingston and Mahloudji, 1961) or of nonconvulsive signs
only (Walton and Brooks, 1897; Pickles, 1949; Gjerris and Mellemgaard, 1969;
Bruce et al, 1979, 1981a, b; Todorow and Feller, 1982).
Spreading Depression of Leao

Downloaded from by guest on October 22, 2014

By dividing our patients into two groups, convulsive and nonconvulsive cases, we
follow Oka et al. (1977). They described a series of 37 children in whom transient
neurological disorders occurred in the acute stage of trivial head injury and for
whom no surgical treatment was subsequently required. The ages of their patients
ranged from 10 months to 21 years, but the majority were under 14 years of age. Of
these 37 children, 28 developed convulsive attacks and 9 children demonstrated
nonconvulsive signs following a head injury without initial loss of consciousness. As
in our series, most of the convulsive patients were under 8 years of age. Our group of
patients with a transient syndrome of subacute or acute onset (Table 4) shows the
greatest resemblance to their group of nonconvulsive cases, although our patients
tended to be younger. They considered the nonconvulsive signs to be the primary
and basic disturbance and they regarded the convulsive attacks as secondary
phenomena. According to these authors, both phenomena can be explained as
manifestations of the experimental phenomenon designated as spreading depression
of Leao. In rabbits, Leao (1944) found that weak faradic or mechanical stimulation
of the exposed cerebral cortex elicits a characteristic response, consisting of
a marked enduring reduction of the spontaneous electrical activity of the cortex.
Typical discharges of experimental epilepsy were noted in cortical regions when
these were reached by a spreading wave of depression. Oka et al. suggested that the
fact that the nonconvulsive disturbance was followed by convulsions in the younger
children could be ascribed to a maturational factor, but that the primary process,
the spreading depression, was identical in both groups. Thus, according to these
authors, only the age factor determines whether the syndrome of transient
neurological disorders, which include headache, nausea and vomiting, pallor,
somnolence, irritability and restlessness, stupor, hemiparesis and aphasia, is
followed by convulsive attacks. We noted some differences between our convulsive
and nonconvulsive patients, however, which may indicate that the two phenomena
are not as closely linked as these authors suggest. First there was, as stated earlier,
a past medical history in almost half of the convulsive patients, which may have
resulted in a liability to develop convulsions. Secondly, although a nonspecific sign
such as vomiting was noted at the onset of the deterioration of almost all patients in
the nonconvulsive group (Table 4), this sign was only rarely seen in patients with
early post-traumatic convulsions (Table 2). We were unable to confirm the finding

HEAD INJURY IN CHILDREN

27

of Oka et al. (1977) that the convulsive attacks did not develop directly from a lucid
interval, but were nearly always preceded by nonconvulsive signs to a greater or
lesser degree.

A Functional Disturbance of the Rostral Brainstem
Todorow and Feller (1982) have recently described 49 children with mild posttraumatic stupor. This 'sleepy state', which lasted for 3 to 5 h, occurred in 9 per cent
of all children admitted after a minor head injury. The lucid interval ranged from 15
min to 3 h. In almost one-third of the patients minor neurological signs were noted
(pupillary abnormalities, pyramidal tract signs or ataxia), while EEG abnormalities
were found in more than half. These EEG abnormalities mostly consisted of slight
to moderate slowing of background frequency and, in many cases, persisted for
a considerable time (from 5 days to 6 weeks) after the clinical signs had subsided.
The authors postulate that both the secondary disturbance of consciousness and the
EEG abnormalities can be explained by a functional disturbance of the rostral brain
stem, probably related to the lability of arousal mechanisms in children.
In our series we have not regarded sleepiness in itself as a sign of secondary
deterioration. Many children will fall asleep some time after the shock of an injury
and the subsequent examination and treatment in an unfamiliar hospital setting. We
cannot exclude the possibility, however, that the children in the Todorow and Feller
series who became somnolent without developing focal neurological signs may
represent the mildest form of the 'delayed deterioration syndrome'.
Direct Contusion of the Cerebral Cortex
Livingston and Mahloudji (1961) reported 4 patients between 2 and 4 years of age
who developed convulsive seizures after a latent interval following mild head injury.
They postulated that, since this picture has only been seen in young children at an

Downloaded from by guest on October 22, 2014

A Migrainous Mechanism
Haas et al. (1975), in a study of 25 patients, described 50 attacks consisting of
transient neurological signs following mild head trauma; 40 of the attacks occurred
in children under 14 years of age. All attacks developed after a latent interval,
generally of one to ten minutes. Attacks were grouped into 4 clinical types: (1)
hemiparesis; (2) somnolence, irritability and vomiting; (3) blindness and (4)
brainstem signs. Two patients had an epileptic seizure. Five of the children later had
spontaneous attacks which resembled their triggered attacks closely enough to be
considered the same experience. The authors concluded that the temporary posttraumatic syndromes mentioned by them can be seen as diverse manifestations of
a common underlying process which involves the cerebral vasculature, and that they
resemble spontaneous classical migraine attacks in juveniles closely enough to
postulate an underlying mechanism similar to migraine. Several authors {see Oka et
al., 1977; Shinohara et al., 1979) regard this mechanism as analogous or identical
with the mechanism underlying migraine.

28

J. W. SNOEK, J. M. M I N D E R H O U D AND J. T. W I L M I N K

age when the cranium is rather malleable, deformation of the cranium at the
moment of impact produces direct contusion of the cortex.
Focal or Generalized Brain Swelling, Due either to Rapid Oedema Formation or to
Cerebral Hyperaemia

Downloaded from by guest on October 22, 2014

For almost a century there have been reports of clinical findings in head-injured
children and adolescents strongly suggesting the presence of a rapidly expanding
intracranial haematoma, but with thesefindingsapparently being due to some other
condition. The first case reports were published by Walton (1898) and Walton and
Brooks (1897).
Pickles (1949) concluded that transient acute cortical oedema, with associated
capillary anaemia, best explained the rapid and complete recovery without
operation. Biemond (1970) also considered rapid oedema formation to be the cause
of the transient cortical dysfunction accompanied by loss of consciousness. He
differentiated between the more common benign transient syndrome and the rare
fatal cases, which he saw only three times during his long career. These three
children died after an exploratory craniotomy yielding negative results. Postmortem examination showed diffuse cerebral swelling with microscopicfindingsof
hyperaemia in the smaller arteries and precapillaries. He concluded that there was
evidence both of brain swelling and brain oedema and he considered the brain
swelling to be the primary event, probably attributable to diencephalic lesions.
Lindenberg et al. (1955) described the post-mortem findings of children dying
from blunt head injuries. Diffuse cerebral swelling was the most consistent finding.
They concluded that this post-traumatic brain swelling is markedly more common
in children than in adults and that it may develop after a seemingly minor head
trauma without subsequent loss of consciousness. Similar findings have been
recorded by other pathologists (Adams and Graham, 1972; Adams, 1975), who also
reported that they could notfinddistinct abnormalities on histological examination.
Until recently, post-traumatic brain swelling has always been attributed to brain
oedema, but the crucial evidence for this cause is lacking. According to Miller and
Corales (1981), direct measurements of water content of the brain have not been
reported in patients with head injury. The concept of brain oedema causing diffuse
cerebral swelling has now been replaced by the alternative explanation of increased
cerebral blood volume. Bruce et al. (1979, 1981a, b) and Zimmerman et al. (1978)
described the CT appearance of diffuse cerebral swelling, which they considered to
be the commonest CT finding in head-injured children. This CT picture consists of
obliteration or narrowing of the lateral and third ventricles and perimesencephalic
cisterns, while cerebral attenuation values are higher than on follow-up scans.
According to these authors, these attenuation numbers (measured in the deep
frontal white matter) are also higher than the normal range defined from values
measured in normal paediatric CT scans and certainly higher than would be
expected if brain oedema were the cause of this cerebral swelling, as an increase in
water content of the brain is linearly related to a decrease in attenuation numbers.

HEAD INJURY IN CHILDREN

29

Downloaded from by guest on October 22, 2014

The finding of increased cerebral blood flow (CBF) in the patients who showed
this CT pattern of diffuse cerebral swelling is taken to suggest a relationship between
the increased attenuation at CT, and this increased CBF and the cerebral swelling
is therefore ascribed to cerebral hyperaemia. The association of diffuse cerebral
swelling and hyperaemia has also been reported by Obrist et al. (1979) from the same
institution. Cerebral metabolism studies in their hyperaemic patients yielded a very
low oxygen uptake and arteriovenous oxygen difference, indicating that the high
blood flow is a true 'luxury perfusion'. The cause of the cerebral vasodilatation,
which allegedly leads to cerebral swelling, is as yet unknown. Meyer et al. (1971)
found in their animal model that the areas which caused increases in CBF when
stimulated were located in the pontine and midbrain reticular formation, the
thalamus and the hypothalamus. Raichle et al. (1978) assumed that locus coeruleus
stimulation may change both CBF and cerebral capillary permeability, probably
through central vascular aminergic pathways. Recently, Mies et al. (1981), using
autoradiographic techniques, observed an increase in cerebral blood flow during
spreading depression in animals. This finding thus appears to link the theories of
Bruce et al. (1979, 1981a, b) with those of Oka et al. (1977). The increase in CBF
found by Mies et al. (1981), however, was linked with increased metabolism, which
means that apparently the increase in CBF during spreading depression cannot be
interpreted as 'luxury perfusion'.
There are some unresolved problems relating both to the CT appearance of
diffuse cerebral swelling and to the theory ascribing this swelling to vascular
engorgement. Until clearly defined criteria for minimal ventricular size in normal
children are set, caution is needed in interpreting the sign of'small ventricles' (Snoek
et al., 1979). The perimesencephalic cisterns are normally quite evident in children,
but the same caution is needed in attaching any value to an alleged 'compression or
narrowing'; in our opinion only clear absence of these cisterns on the CT scan
warrants a conclusion of cerebral swelling.
The unresolved problems in the theory ascribing the CT picture of diffuse cerebral
swelling to vascular engorgement are threefold. The first is the reliability of cerebral
attentuation coefficients determined by CT. We feel that relatively small changes in
CT attenuation numbers, as described by Zimmerman et al. (1978), should be
interpreted with circumspection. We have attempted to reproduce the normal range
of values established by these authors (14.6 to 16.6 EMI units in the deep frontal
white matter of 18 normal hemispheres). We determined average attenuation
numbers in the same fashion in circular fields of 150 to 350 pixels each in the deep
frontal white matter of 6 hemispheres of children aged 1 to 12 years, in whom no
signs of cerebral swelling could be found at CT. All the examinations had been
performed on the same scanner (EMI 5005) using a 160 x 160 matrix, 10 or 13 mm
slice thickness. In some subjects a scan time of 70 s per slice was selected, in others
20 s. The same variable scanning parameters were used in the other subjects whose
CT characteristics are mentioned elsewhere in this study. In our 6 subjects white
matter CT attenuation numbers ranged from a maximum average of 19.5 EMI units

30

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

Downloaded from by guest on October 22, 2014

per measuring field (SD 2.42) to a minimum of 11.7 EMI units (SD 2.08). This is
a much larger spread of normal values than is presented in the material of
Zimmerman et al. (1978).
A number of factors may influence the assessment of tissue attenuation
coefficients as expressed in CT numbers.
(1) Scanner characteristics. These vary from one manufacturer to another with
regard to acquisition of data (scanner generation, presence or absence of a water
bag), as well as the method of reconstruction. Extreme caution must be exercised in
any comparison of different types of scanners often functioning under dissimilar
operating conditions (Speller et al., 1981). With regard to a single CT scanner,
consistency of performance may be influenced by the state of maintenance
(McCullough, 1977). The EMI 5005 whole body scanner mentioned in our study is
serviced by contract with the manufacturer. It is also interesting to note that, as the
linear attenuation coefficient of water decreases with increasing temperature, there
is a difference in CT attenuation numbers of 2.5 to 3.0 EMI units between water at
room temperature and water at body temperature. This should be taken into
account in calibration of the scanner (Bydder and Kreel, 1979). Factors such as
voxel size (determined by slice thickness and area represented by a pixel) may be
expected to influence the statistical accuracy of the attenuation measurement. It
is also our experience that in the EMI 5005 scanner CT attenuation numbers
determined in a 20 s rapid scan are usually significantly higher than in a 70 s slow
scan performed in the same region. We have no explanation for this discrepancy.
Finally, changes in position of the head within the scanner ring can cause variations
in CT numbers measured in the same cerebral region in serial examinations, as can
changes in kilovoltage between examinations (Levi et al., 1982).
(2) Factors related to the patient. The most important of these appears to be skull
thickness. As the x-ray beam traverses tissue, its composition is altered. The lowerenergy photons are filtered out, and while the x-ray beam as a whole is attenuated,
the mean kinetic energy of the remaining photons is higher. This beam-hardening
effect is especially marked in substances with a high linear attenuation coefficient
such as bone, and when a large thickness of skull has to be traversed, this will
influence attenuation measurement of skull contents (Di Chiro et al., 1978). Shifts
of some 7.5 EMI units have been noted between maximum and minimum skull
thickness (Payne and Latchaw, 1978).
(3) Partial volume effect. Cerebral grey matter has a mean attenuation number
some 3 EMI units higher than white matter (Weinstein et al., 1977; Phelps et al.,
1975). If the field of measurement located in the frontal white matter should
inadvertently contain grey matter, this will induce an upward bias in the mean CT
attenuation value through a partial volume effect. This could happen if thefieldof
measurement were located too near to the cerebral surface in the CT section, but
also if a CT section were selected for measurement which contains in its basal
portion some of the frontobasal grey matter.
The minor increase in CT attenuation number of 1.6 EMI units described by

HEAD INJURY IN CHILDREN

31

Downloaded from by guest on October 22, 2014

Zimmerman et al. (1978) in their group of patients with post-traumatic brain
swelling achieves statistical significance when serial measurements are performed in
the same patient using the same CT scanner under identical operating conditions.
We would like to warn, however, against attributing diagnostic significance to
marginal variations in mean CT attenuation numbers measured in the first CT
examination in an individual patient.
The second unresolved problem in relating the CT picture of diffuse cerebral
swelling to hyperaemia is as follows. Although the increase of 1.6 EMI units in ET
attenuation numbers mentioned above is relatively small in the context of other
variable factors influencing the measurements, it represents a large shift if it is to be
explained solely on the basis of increase in blood content. The blood concentration
of cerebral white matter is reported to be 2.2 per cent and of grey matter 5.6 per cent
(Ladurner, 1978). The CT number for blood is given as 27 EMI units (New and
Scott, 1975), and for subfrontal white matter as 16.5 EMI units (Arimitsu et al.,
1977). In order to cause an increase of the value for white matter in the order of 1.6
EMI units, its blood content would have to increase very substantially. By the same
token doubling the cerebral blood volume would result in an increase in total
cerebral volume of only 3.5 per cent, and an increase of this amount of blood volume
is hardly ever seen (Raichle, 1979).
The third unresolved problem is that the crucial measurement is cerebral blood
volume rather than cerebral blood flow, as measured by Bruce et al. (Miller and
Corales, 1981). Determination of cerebral blood volume (CBV) by emission
tomography in 4 head-injured children belonging to the series of Bruce et al. (1981 a)
failed to demonstrate an increase in CBV in 3 of them (Kuhl et al., 1980). These
authors state that the CT pattern characteristic of diffuse brain swelling may occur
in head-injured children without discernible increases in CBV and that this picture
therefore cannot be explained by increases in CBV alone.
The clinical significance of the finding of a CT appearance suggesting diffuse
cerebral swelling caused by cerebral hyperaemia is as yet unresolved. The crucial
question is whether such a swelling (leading to severely increased intracranial
pressure only in exceptional cases) is the cause of the deterioration in head-injured
children or whether in most patients it is an epiphenomenon caused by the same
processes which lead to the clinical deterioration. Cold and Jensen (1980) found that
hyperaemia following head injury is a common phenomenon in children, that this
hyperaemic phase lasts for some days to several weeks and that it is of no prognostic
importance. The CT pattern of diffuse cerebral swelling is seen in 29 per cent of
head-injured children and adolescents. There is a higher incidence in patients with a
low score on the Glasgow Coma Score (41 per cent with a score of 8 or less) than in
patients with a score greater than 8 (15 per cent) (Bruce et al., 1981a). These authors
do not specifically mention the rate of occurrence of this CT pattern in children with
a lucid interval. Conversely, in the clinical description of their group of 63 juvenile
patients with the CT pattern of diffuse cerebral swelling, 37 per cent had experienced
some form of lucid interval. Within the latter group two subgroups are described. In

32

J. W. SNOEK, J. M. M I N D E R H O U D A N D J. T. W I L M I N K

Downloaded from by guest on October 22, 2014

a typical case of thefirstsubgroup (8 cases) there was a lucid interval with a period of
talking and complete consciousness from minutes to hours after injury, followed by
the onset of vomiting, headache and frequently pallor and sweating, associated with
decreased alertness. These patients exhibited evidence of decreased spontaneous
motor activity, and loss of spontaneous speech and eye opening. No mention is
made of focal neurological signs. None of these patients progressed to coma
according to the criteria of Teasdale and Jennett (1974) and the intracranial pressure
when measured in these patients remained normal. All these patients made a good
recovery. The other subgroup consisted of 15 patients who were unconscious
following trauma and then had an interval associated with recovery of eye opening,
occasional words and more spontaneous motor function. This was followed by
rapid deterioration leading to coma (Glasgow Coma Score of less than 8). One of
these patients died as a result of delayed brain swelling. The others all made a good
recovery. The patients of our series are by definition only comparable with their first
group of patients. As most of our patients showed a benign, often shortlasting
syndrome, few CT scans were made. In only one out of 4 scans, performed in
children with a rapid deterioration consisting of either convulsive or nonconvulsive
signs, was the typical CT appearance consistent with diffuse cerebral swelling seen
(fig. 5), which leads us to the conclusion that in the majority of cases deterioration
apparently is not associated with such cerebral swelling and that, therefore, there
must be other explanations for this rapid deterioration.
The CT scan of Case 7, a girl with a slow deterioration, was consistent with
swelling of one hemisphere, a finding which has been described in CT scans
performed in the period immediately after injury (Kobrine et at., 1977; Waga et al.,
1979). In both cases this swelling was, on the basis of attenuation measurements,
also attributed to vasodilatation and not to oedema. The results of the angiogram in
our Case 6, also a child with a slow deterioration, support the interpretation that
local damage can be followed by local swelling, associated with clinical deterioration. The clinical course, radiological and (in one case) surgicalfindingsin our fatal
cases, although not confirmed by autopsy, strongly suggest that 'malignant brain
swelling' following trivial head injury in children does exist. Excluding the rare
occurrence of intracranial haematoma, we feel, however, that the syndrome of
delayed deterioration following mild head injury in children in fact covers a variety
of pathological states and that neither the clinical nor the CTfindingssupport the
view that brain swelling caused by cerebral hyperaemia is the only or even the usual
underlying cause for such deterioration.
We conclude that trivial injuries in children are not infrequently followed by
a deterioration which is preceded by a lucid or symptom-free period and which is
usually transient. It is very likely that only in a minority of affected children signs of
such severity occur that these children will come under medical care. On clinical
grounds it is possible to differentiate between several distinct pictures. At this
moment it can only be speculated whether or not the different clinical pictures
presented by these patients reflect different pathophysiological entities. With the

HEAD INJURY IN CHILDREN

33

REFERENCES
ADAMS J H (1975) Neuropathology of head injuries. In: Handbook of Clinical Neurology. Edited by
P. J. Vinken and G. W. Bruyn. Amsterdam: North Holland, Volume 23, pp. 35-65.
ADAMS J H, GRAHAM D I (1972) The pathology of blunt head injuries. In: Scientific Foundations of
Neurology. Edited by M. Critchley, J. L. O'Leary and B. Jennett. London: Heinemann,
pp. 478^91.
ARIMITSU T, D I CHIRO G, BROOKS R A, SMITH P B (1977) White-gray matter differentiation in

computed tomography. Journal of Computer Assisted Tomography, 1, 437-442.
BIEMOND A (1970) Brain Diseases. Amsterdam: Elsevier, pp. 771-785.
BRUCE D A, RAPHAELY R C, GOLDBERG A I, ZIMMERMAN R A, BILANIUK L T, SCHUT L, KUHL D E

(1979) Pathophysiology, treatment and outcome following severe head injury in children. Child's
Brain, 5, 174-191.
BRUCE D A, ALAVI A, BILANIUK L T, DOLINSKAS C, OBRIST W, UZZELL B (1981a) Diffuse cerebral

swelling following head injuries in children: the syndrome of'malignant brain edema'. Journal of
Neurosurgery, 54, 170-178.
BRUCE D A, SUTTON L N, SCHUT L (19816) Acute brain swelling and cerebral edema in children. In:
Brain Edema. Edited by M. de Vlieger, S. A. de Lange and J. W. F. Beks. New York: John Wiley,
pp. 125-145.
BYDDER G M, KREEL L (1979) The temperature dependence of computed tomography attenuation
values. Journal of Computer Assisted Tomography, 3, 506-510.

Downloaded from by guest on October 22, 2014

exclusion of the relatively rare extracerebral haematoma, these clinical pictures
point either to focal or generalized cerebral dysfunction, both with and without
seizures. Although the recent theories developed by Bruce et al. relate the occurrence
of cerebral dysfunction with diffuse cerebral swelling to hyperaemia, insufficient
data are available so far to discriminate between this swelling being a causative
factor for the deterioration or just an epiphenomenon in the majority of cases. We
feel that caution is needed in interpreting the findings of marginally increased
attenuation values in CT scans with small ventricles and compressed perimesencephalic cisterns as a state of diffuse cerebral hyperaemia and also in basing
a therapeutic regimen solely on this interpretation. The rarity of this CT pattern in
our patients does not give support to the assumption that in the majority of children
diffuse cerebral swelling was the cause of the transient deterioration. There is now
some evidence, however, that for unknown reasons some patients develop severe
and often uncontrollable unilateral or diffuse brain swelling, as indicated by the
radiological and surgical evidence in 2 out of our 3 fatal cases.
The fact that not infrequently deterioration following a trivial head injury may
occur in children makes it mandatory in our opinion to accept a rather strict
admission policy for head-injured children, especially if there is an indication in the
past medical history that cerebral damage has been suffered in earlier life. In
a deteriorating head-injured child that was symptomless before, the possibility of an
expanding intracranial mass will always have to be excluded, preferably by CT
scanning, although few haematomas will be found. The remaining patients usually
show a benign and shortlasting syndrome followed by a spontaneous and full
recovery in most, but not in all cases.

34

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

JENNETT W B, TEASDALE G, GALBRAITH S, PICKARD J, GRANT H, BRAAKMAN R, AVEZAAT C, MAAS A,

MINDERHOUD J, VECHT C J, HEIDEN J, SMALL R, CATON W, KURZE T (1977) Severe head injuries

in three countries. Journal of Neurology, Neurosurgery and Psychiatry, 40, 291-298.
KOBRINE A I, TIMMINS E, RAJJOUB R K, RIZZOLI H V, DAVIS D O (1977) Demonstration of massive

traumatic brain swelling within 20 minutes after injury. Case report. Journal of Neurosurgery, 46,
256-258.
KUHL D E, LAvi A, HOFFMAN E J, PHELPS M E, ZIMMERMAN R A, OBRIST W D, BRUCE D A,

GREENBERG J H, UZZELL B (1980) Local cerebral blood volume in head-injured patients.
Determination by emission computed tomography of 99m Tc-labeled red cells. Journal of
Neurosurgery, 52, 309-320.
LADURNER G (1978) Die Bestimmung des zerebralen Blutvolumens mit der Computer-tomographie in
grauer und weisser Substanz. Fortschritte der Neurologic-Psychiatric und ihrer Grenzgebiete, 46,
369-381.
LANGFTTT T W, BRUCE D A (1975) Microcirculation and brain edema in head injury. In: Handbook of
Clinical Neurology. Edited by P. J. Vinken and G. W. Bruyn. Amsterdam: North Holland,
Volume 23, pp. 133-161.
LEAO A A P (1944) Spreading depression of activity in cerebral cortex. Journal of Neurophysiology, 7,
359-390.
LEVI C, GRAY J E, MCCULLOUGH E C, HATTERY R R (1982) The unreliability of CT numbers as

absolute values. American Journal of Roentgenology, 139,443-447.
LINDENBERG R, FISHER R S, DURLACHER S H, LovnT W V JR, FREYTAG E (1955) The pathology of the

brain in blunt head injuries of infants and children. In: Proceedings of the Second International
Congress of Newopathology. Part II. Amsterdam: Excerpta Medica, pp. 477-479.
LIVINGSTON K. E, MAHLOUDJI M (1961) Delayed focal convulsive seizures after head injury in infants
and children. A syndrome that may mimic extradural hematoma. Neurology, Minneapolis, 11,
1017-1020.

Downloaded from by guest on October 22, 2014

COLD G E, JENSEN F T (1980) Cerebral blood flow in the acute phase after head injury. I. Correlation to
age of the patients, clinical outcome and localisation of the injured region. Ada Anaesthetica
Scandinavica, 24, 245-251.
Di CHIRO G, BROOKS R A, DUBAL L, CHEW E (1978) The apical artifact: elevated attenuation values
toward the apex of the skull. Journal of Computer Assisted Tomography, 2, 65-70.
GALBRAJTH S, SMITH J (1976) Acute traumatic intracranial haematoma without skull fracture. Lancet,
1,501-502.
GJERRIS F, MELLEMGAARD L (1969) Transitory cortical blindness in head injury. Ada Neurologica
Scandinavica, 45, 623-631.
GRAND W (1974) The significance of post-traumatic status epilepticus in childhood. Journal of
Neurology, Neurosurgery and Psychiatry, 37, 178-180.
HAAS D C, PINEDA G S, LOURIE H (1975) Juvenile head trauma syndromes and their relationship to
migraine. Archives of Neurology, Chicago, 32, 727-730.
HENDRICK E B, HARRIS L (1968) Post-traumatic epilepsy in children. Journal of Trauma, 8, 547-556.
HENDRICK E B, HARWOOD-HASH D C F , HUDSON A R (1964) Head injuries in children: a survey of 4465
consecutive cases at the Hospital for Sick Children, Toronto, Canada. In: Clinical Neurosurgery.
Proceedings of the Congress of Neurological Surgeons. Baltimore: Williams and Wilkins,
pp. 46-65.
JAMES H E (1979) Analysis of therapeutic modalities for head-injured children. Child's Brain, 5,
263-271.
JENNETTW B(1962,1975) Epilepsy after Non-missile Head Injuries. London: Heinemann. First edition
1962; second edition 1975.
JENNETT W B (1969) Early traumatic epilepsy. Definition and identity. Lancet, i, 1023-1025.
JENNETT W B (1973) Trauma as a cause of epilepsy in childhood. Developmental Medicine and Child
Neurology, 15, 56-62.

HEAD INJURY IN CHILDREN

35

MCCULLOUGH E C (1977) Factors affecting the use of quantitative information from a CT scanner.
Radiology, 124,99-107.
MEYER J S, TERAURA T, SAKOMOTO K, KONDO A (1971) Central neurogenic control of cerebral blood
flow. Neurology, Minneapolis, 21, 247-262.
MIES G, NIEBUHR I, HOSSMANN K-A (1981) Simultaneous measurement of blood flow and glucose
metabolism by autoradiographic techniques. Stroke, 12, 581-588.
MILLER J D, CORALES R L (1981) Brain edema as a result of head injury: fact or fallacy? In: Brain
Edema. Edited by M. de Vlieger, S. A. de Lange and J. W. F. Beks. New York: John Wiley,
pp. 99-115.
NEW P F J, SCOTT W R (1975) Computed Tomography of the Brain and Orbit (EMI Scanning).
Baltimore: Williams and Wilkins, pp. 263-267.
OBRIST W D, GENNARELLI T A, SEGAWA H, DOLINSKAS C A, LANGFITT T W (1979) Relation of cerebral

PETERS A C B, VERSTEEG J, LINDEMAN J, BOTS G T H A M (1978) Viral meningoencephalitis and head

injury. Acta Neurologica Scandinavica, 57, 77-87.
PHELPS M E, HOFFMAN E J, TER-POGOSSIAN M M (1975) Attenuation coefficients at various body
tissues, fluids, and lesions at photon energies of 18 to 136 keV. Radiology, 117, 573-583.
PICKLES W (1949) Acute focal edema of the brain in children with head injuries. New England Journal of
Medicine, 240, 92-95.
PLUM F, POSNER J B (1980) Supratentorial lesions causing coma. In: The Diagnosis of Stupor and Coma.
Philadelphia: F. A. Davis. Third edition, pp. 87-151.
RAICHLE M E (1979) Discussion. In: Neural Trauma. Edited by A. J. Popp, R. S. Bourke, L. R.
Nelson and H. It. Kimelberg. New York: Raven Press, p. 49.
RAICHLE M E, GRUBB R L JR, PHELPS M E, GADO M H, CARONNA J J (1978) Cerebral hemodynamics

and metabolism in pseudotumor cerebri. Annals of Neurology, 4, 104-111.
REILLY P L, GRAHAM D I, ADAMS J H, JENNETT B (1975) Patients with head injury who talk and die.
Lancet, ii, 375-377.
ROSE J, VALTONEN S, JENNETT B (1977) Avoidable factors contributing to death after head injury.
British Medical Journal, 2, 615-618.
SHINOHARA M, DOLLINGER B, BROWN G, RAPOPORT S, SOKOLOFF L (1979) Cerebral glucose utilization:

local changes during and after recovery from spreading cortical depression. Science, 203,
188-190.
SMALL J M, WOOLF A L (1957) Fatal damage to the brain by epileptic convulsions after a trivial injury
to the head. Journal of Neurology, Neurosurgery and Psychiatry, 20, 293-301.
SNOEK J, JENNETT B, ADAMS J H, GRAHAM D I, DOYLE D (1979) Computerised tomography after recent

severe head injury in patients without acute intracranial haematoma. Journal of Neurology,
Neurosurgery and Psychiatry, 42, 215-225.
SPELLER R D, WHITE D R, SHOWALTER C K, ROTHENBERG L N, PENTLOW K S, MORGAN T J, SHOPE T B

(1981) An evaluation of CT systems from ten manufacturers. British Journal of Radiology, 54,
1053-1061.
TEASDALE G, JENNETT B (1974) Assessment of coma and impaired consciousness. A practical scale.
Lancet, ii, 81-84.

Downloaded from by guest on October 22, 2014

blood flow to neurological status and outcome in head-injured patients. Journal of Neurosurgery,
51, 292-300.
OKA H, KAKO M, MATSUSHIMA M, ANDO K (1977) Traumatic spreading depression syndrome. Review
of a particular type of head injury in 37 patients. Brain, 100, 287-298.
OMMAYA K (1979) Indices of neural trauma: an overview of the present status. In: Neural Trauma.
Edited by A. J. Popp, R. S. Bourke, L. R. Nelson, and H. K. Kimelberg. New York: Raven Press,
pp. 205-216.
PAYNE J T, LATCHAW R (1978) Variation and nonuniform CT number response for intracranial
contents as a function of skull thickness and head size. Journal of Computer Assisted Tomography,
2,509.

36

J. W. SNOEK, J. M. MINDERHOUD AND J. T. WILMINK

TODOROW S, FELLER A-M (1982) Ueber das Vorkommen benigner sekundarer Bewusstseinstriibung—
das Einschlafsyndrom—nach Schadel-Hirn-Trauma bei KJndern. Zeitschrift fur Kinderchirurgie,
36, 83-87.
WAGA S, TOCHJO H, SAKAKURA M (1979) Traumatic cerebral swelling developing within 30 minutes
after injury. Surgical Neurology, 11, 191-193.
WALTON G L (1898) Subarachnoid serous exudation productive of pressure symptoms after head
injuries. American Journal of the Medical Sciences, 116, 267-275.
WALTON G L, BROOKS W A JR (1897) Observations on brain surgery suggested by a case of multiple
cerebral hemorrhage. Boston Medical and Surgical Journal, 136, 301-305.
WEINSTEIN M A, DUCHESNEAU P M, MACINTYRE W J (1977) White and gray matter of the brain
differentiated by computed tomography. Radiology, 122, 699-702.
ZIMMERMAN R A, BILANIUK L T, BRUCE D, DOLINSKAS C, OBRIST W, KUHL D (1978) Computed

tomography of pediatric head trauma: acute general cerebral swelling. Radiology, 126, 403-408.
(Received February 8, 1983. Revised April 26,1983)

Downloaded from by guest on October 22, 2014