233
EPIDURAL HAEMATOMA
A RETROSPECTIVESTUDYOF
C. Jonker,*

looPATIENTS

and H. J. 0osterhui.P”

SUMMARY
A retrospective study was made of 100 consecutive patients with an epidural haematoma in
order to establish which clinically demonstrable
factors had influenced the prognosis quoad
vitam et sanationem. Operations were performed on 92 of these patients, 29 (32%~) of whom
died; 8 patients died without operation, and in 2 of these cases the diagnosis was not made
during life.
A lucid interva1 was observed in 57 patients; absence of a lucid interval in combination
with a lowered sensory ievel indicated associated intradural
lesions (cerebral contusion,
acute subdural haematoma),
with consequently
a less good prognosis. The prognosis was
also adversely affected by deeper coma, occurrence of extensor spasms, bilateral stiff pupils,
bilateral pyramidal symptoms and an age over 50. Some 330/o of the patients treated by
operation showed marked symptoms of compression
(extensor spasms and/or bilateral stiff
pupils); although these symptoms are usually described as indicative of a hopeless prognosis,
400/o of the patients in this catagory survived.
Contrary to the data in the literature, the interval between accident and operation within
the first 24 hours did not influence the mortality, which was 500/o. All patients operated on
more than 24 hours after the accident, survived.
A cranial fracture was absent in 13 patients, 11 of whom were under 30; absence of a
cranial fracture was prognostically favourable.
A catamnestic study revealed that 7 of the 58 accessible survivors had residual neurological
dysfunctions; all these patients were up and about. The residual morbidity after early operalion (within 24 hours) was the same as that after later operation.
The data obtained were compared with those on a number of series recently published in the
literature.

INTRODUCTION

Epidurat haematoma
is a vitally dangerous complication
in a small proportion
of
patients with head injuries. The literature reports that it occurs in 1.530/o of all
clinically observed patients with a head injury (MCKISSOCK, TAYLOR, BLOOM, TILL,
1960; JOSEPHSON, 1962; JAMIESON and YELLAND, 1968).
A lowering of the level of consciousness
after a short period of clear consciousness, unilateral
pupillary
dilatation
and pupillary
stiffness, and hemiparesis
conFrom the Dept.

of the Neurology,
Wilhelmina
Gasthuis, Amsterdam,
The Netherlands.
* Present address: Valerius kliniek, Dept. of Psychiatry, Amsterdam, The Netherlands.
** Present address: Dept. of Neurology, University Hospital, Groningen, The Netherlands.
Clin.

Nemo].

Neurosurg.

1975-4

234
tralateral to the pupillary dilatation constitute the classic triad of symptoms which
facilitate diagnosis. Since surgical treatment is not difficult either, the prognosis
of an epidural haematoma diagnosed early ought to be favourable. However, in
recent series of more than 100 patients the mortality varies from 1Y/U to 25@.‘~
(MCKISSOCK
et al, 1960; JAMIESON et al, 1968).
In this study we raise the question as to which clinically demonstrable factors
influenced the prognoses quoad vitam and quoad sanationem in a consecutive
series of 100 patients presented in the neurological and neurosurgical departments
of the Wilhelmina Gasthuis, Amsterdam, between 1961 and 1973.

PATIENTS

AND METHODS

selective

of patients

Working back from 1973, all case histories of patients with an anatomically
verified diagnosis of epidural haematoma were lifted from the files until a total of
100 was attained. We excluded patients with an impression fracture of the skull,
with an abnormal bleeding tendency, and with severe extracranial injuries.
The diagnosis had been made at operation in 92 patients, and at a postmortem
without preceding operation in 8. Patients who showed cerebral contusion in
addition were ltut excluded.
The series comprised 81 males and 19 females, admitted between 1961 and
1973, for the most part directly to the neurosurgical clinic (Head Prof. Dr. W.
Noordenbos). A traffic accident was involved in 63 cases. The group with an
epidurai haematoma amounted to about 3o/0 of the total number of patients
admitted to the neurological and neurosurgical departments of the Wilhelmina
Gasthuis with a head injury during this period.
Owing to the regional function of the neurosurgical clinic, a proportion of
patients came from other hospitals; 68 patients had had their accident in Amsterdam, and the remaining 32 within a radius of 60 km.
Method and procedure

On the basis of a coding scheme we studied a number of factors which, according
to the literature and personal views, could have influenced the course. The data
on the patient’s condition concerned the final examination prior to operation.
Items studied were: age; occurrence of a so-called free interval (defined as
sensory clearing which foIlowed an initial lowering of the level of co~scious~~s
and in turn was foliowed by a lowering of the level of consciousness); variations
in the level of consciousness; the interval between the accident and examination
before operation; the level of consciousness before operation; occurrence of
epileptic convulsions; occurrence of extensor spasms; pulse rate; blood pressure;

235
pupils; ocular fundi; pyramidal symptoms; shift of the midline echo and presence
of a cranial fracture.
Radiographs were not evaluated by the authors; only if the operation report
as well as the radiological report mentioned no fracture was it assumed that no
fracture was present.
All patients who at the last clinical or out-patient follow-up still showed residual
neurological changes were summoned and personally examined (so far as they
were still living in The Netherlands).
The data were transferred to punch cards and analysed by the Mechanical
Administration Unit of the Wilhelmina Gasthuis.
RESULTS

The total mortality was 370/o and the mortality after operation was 32010.
TABLE

I

Age, interval from accident to operation and mortality
Interval from accident to operation
2-7 days
alive

first day
age (Y)

alive

died

o- 9
IO-19
20-29
30-39
40-19
50-

6
7
9
5
7
2

3
4
9
3
7
II

36
--____TABLE

I

1

6
3
I
I
3
_....
_.I---__

37
.“___._____ _ _-_

B 8 days
alive
_ ._. _

I.5

2
2
4
3
_- --..

._

I2

total

~._

II
I9
23
13
I8
I6
100

2

Free interval and mortality
..~.________

Clear consciousness after initial coma
Free interval
Immediate coma without recovery
~~~
-__
-....~_
Total

._.-.
total

died

13
57
30

0
19
I8

100

37

_.~

_-..___

___

.__~

__

Tables 1 and 2 relate age and free interval to survival. The mortality of patients
whose accident had occurred in Amsterdam or within a radius of 25 km was
about 300/o. But of the 9 patients who had to be transported over a distance
exceeding 25 km, 7 died. Only 10 of the 34 patients who were comatose at the
final examination before operation, survived. The chance of survival was significantly better when the response to pain stimuli was still intact (fig. I).

236
FIG.I
Mortality

and the level of consciousness

before operation

N

q

40-

..‘:.:::g.

:I.::.. .::.

2&g

died

30-

alive

20-

9

10-

I
I

clear

I.
II

III

II responsive, but confused
111 non responsive, restless, reaction to pain
IV no reaction to pain

The pupils were examined in all patients. Of the 6 patients who survived despite
bilateral pupillary dilatation and stiffness, 4 were under 30 (table 3). Two were
operated on immediately after observation of these symptoms. The other two had
no cranial fracture.
TABLE

3

Pupil reactions and mortality

Normal pupils
Anisocoria
Dilated pupils, unresponsive

to light

Total

Extensor

spasms were observed

total

died

30
44
26

3
14
20

100

37

in 28 patients, more than 500/o of whom were
seen within 6 hours after the accident. The mortality in this group was considerable: 18 patients died.
Extensor spasms and/or stiff pupils were found in 38 patients. Since the prognosis was considered hopeless, 8 patients were not operated on; all died. Of the

237
30 patients
better

treated

by operation,

in the 19 patients

under

18 died. The prognosis
30, of whom

was found

9 died, than

not to be much

in the 11 patients

ovet

30, of whom 9 died (FISHER'S test: not significant).
Hemipareses

or other pyramidal symptoms were observed

in 75 patients.

In 6.

the dysfunction
was on the side of the haematoma;
4 of these patients developed
the symptoms of haematoma
within 2 hours of the accident. This contradicts
the
observations
dysfunction

reported
by MCKISSOCK
et al. (1960)
who observed
homolateral
precisely in cases of chronic epidural haematoma
(present longer than

7 days).
Of the patients with bilateral pyramidal
symptoms,
500/o died. Total areflexia
was observed in 6 patients, of whom 5 died.
An epileptic seizure was observed in only 2 of the 100 patients. Both patients
were suffering from cerebral contusion
as well, as established
postmortem
or at
operation.
In only 55 cases the ocular fundi were examined;
choked disc was diagnosed in
24. In 3 patients this was observed within 4 hours of the accident, but the majority
(13 of the 24 cases) were chronic epidural haematomas
which had been present
longer than 24 hours (in 9 of these 13 patients longer than 7 days).
The pulse rate was recorded in 75 patients. A pulse rate < 60/min. was found
in 24 patients (8 of whom died); 40 patients had a pulse rate of 6OlOO/min.
(15 died), and 11 patients had a pulse rate > lOO/min. (9 died).
Blood pressure values were available for only 52 patients. In 5 patients it was
below RR 125/70 (2 died), in 32 it was 150-125 systolic and 90-70 diastolic (14
died); in 15 patients it was over 150/90 (9 died). No relation was established
between mortality
and blood pressure, but a rapid pulse seemed prognostically
unfavourable.
Radiography of the skull was carried out in 94 patients. In 13 patients a fracture
was not found either radiologically
or at operation (table 4). Of these 13 patients,
I I were under 30. Absence of a cranial fracture substantially
enhanced the chance
of survival not only in the total group but also in the group under 30.
Tile midline echo was studied in 70 patients and found to have shifted in 60. Of

TABLE

4

Skull

fracture

and mortality

under age 30

fracture

total

died

total

died

81
13

33
1

38
I1

13

absent
unknown

6

3

4

2

100

37

53

I6

present

I

238
the 10 patients without shift of the midline echo, 2 died. In the 60 patients with a
shift there was no relation between the shift measured and the mortality, which in
all groups was about 30Oi~ In 6 of the 10 patients without shift of the midline
echo the haematoma had been present longer than 7 days.
Within 8 hours of the injury more than 500/o of the patients developed symptoms of epidural haematoma (fig. II). The time factor refers to the duration of the
interval between the injury and the exam~ation which led to operation. It is evident that the prognosis in the first 24 hours after the injury is not dependent on
the interval. All patients in whom a haematoma was not diagnosed until 24 hours
after the injury or later, remained alive.

FIG.

II

Interval from accident to operation and mortality

N
40
died

30
-27

alive
cl

20

fC

o-4

5-8

9-24

1OUfS

In addition to epidural haematoma, operation or postmortem disclosed intradural lesions in 24 patients: subdural haematoma in 12, cerebral contusion with
severe lacerations in 12. There have probably been more patients with lesions of
brain tissue, because cerebral contusion was clinically diagnosed in a total of
38 patients. In a number of these patients there was no indication for opening the
dura during the operation. As was more or less to be expected, not only the total
mortality but also the postoperative mortality increased substantially in the presence of intradural Iesions (table 5). In 8 of the 24 patients with an intradural
lesion, no operation but only a postmortem was performed. Of the remaining 16
who were submitted to operation, 10 died.

239
TABLE

5

Intradural

lesions and mortality

age
---

with intradural
total

died

lesions

10
16
19
12
12
7

2
2
7
2
3
3

1
3
4
1
6
9

i
2
2
1
4
8

76

19

24

18

~~-~

o-9
IO-19
20-29
30-39
40-49
50.

‘CHRONIC’

without intradural lesions
total
died

EPIDURAL

HAEMATOMA

The 27 patients whose operation took place 24 hours after the injury or later,
all survived. Operation records showed that a micro-arterial haemorrhage had
been observed in 6 patients and a tamponaded arterial haemorrhage in 3 patients,
no bleeding artery having been found in 18. The symptoms which had prompted
the diagnosis were generally based on increased intracranial pressure (choked disc,
drowsiness, vomiting) and a shift of the midline echo.
In 20 of the 27 patients neurological examination reveaIed no dysfunction or
only one pathological plantar reflex. In 13 patients no free interval was observed
and the sensory level had not been lowered after the initial loss of consciousness,
The ‘acute’ and the ‘chronic’ cases did not differ in age distribution and
preserice of a fracture. The residual morbidity was also the same for both groups.
The interval between injury and operation can amount to several months, as the
following case history (HM 15152) illustrates,

Case history

In June 1961 a 30-month-old girl feil down about 10 steps of a stairway. After a
free interval of 3 hours she started vomiting and 10% consciousness for severaf
hours. Once the sensory levef was restored to normal she showed left-sided hemiparesis whicn inciuded the facial muscles. The hemiparesis receded in the course
of a week, and after 3 weeks she was free of symptoms. Cranial X-rays showed no
fracture lines. The girl was discharged.
At follow-up 4 months later the EEG showed fairly differentiated features on
the left, but evidence of slight general suppression on the right. In view of this
abnormal EEG she was seen again 6 months later. Because a nodule was palpable
on the head, radiographs of the skull were obtained which revealed a fairly welldefined accumulation of calcium in the right parietal region (cf. photograph).

240

At operation a partly calcified epidural haematoma
after the operation the EEC was symmetrical.

was removed, and 6 monthly

MORBIDITY

morbidity in the 63 surviving patients was studied. Those already released
from follow-up, or those who still complained of mild headaches and dizziness.
were regarded as cured and not summoned to report for a follow-up. Investigation
and case records revealed that only 4 patients were under treatment for epilepsy:
in 3 of these the operation was more than 5 years ago, and in the 4th the followup covered a period of 18 months.
The last out-patient or clinical follow-up had still revealed residual neurological
changes in 17 patients, 4 of whom were foreigners who could not be summoned.
Of the remaining 13, 1 failed to respond to the summons. Of the 12 patients
finally seen, 7 showed neurological changes: spastic hemiparesis in 5, with dysphasia in 3; cerebellar aiaxia and dysarthria in 1; oculomotor paresis and
hypaesthesia of the left side of the body in 1.
We refrained from an evaluation of psychosocial aspects in these cases.
In only 3 patients the follow-up period did exceed 5 years. A striking finding
was that 4 of the 7 patients with residual symptoms had shown extensor spasms
or bilateral pupillary stiffness when examined before operation. At operation
only 1 of these 7 pattents was found to suffer from cerebral contusion.
The

FINDINGS IN EIGHT NON-SURGICAL

PATIENTS

A diagnosis of cerebral contusion was made in 3 patients. The postmortem revealed
marked contusion with a subdural haematoma and a small epidural haernatoma.
In one case epidural haematoma was diagnosed but an operation was reframed

241
from

in view of the clinical

pupils).

The postmortem

only a small epidural

condition

revealed

haematoma.

(bilateral

contusion

dilatation

and

with a large subdural

In 1 patient

the diagnosis

stiffness

of the

haematoma

and

‘epidural

haematoma’

was considered but rejected in view of marked brain stem symptoms (tachycardia,
pyrexia), symmetrical
moderate
dilatation
and stiffness of the pupils, and some
slight shift of the midline echo. The postmortem
disclosed a large epidural haematoma with contrecoup contusion.
In two cases the diagnosis was rejected on the basis of arteriographic
findings
interpreted
as normal and as indicating
cerebral
oedema, respectively.
In one
patient, finally, a burr-hole
was made at the wrong site. Strictly speaking this
patient should not come under the heading non-surgical
patients’.
In these 3
patients the postmortem
revealed an epidural haematoma
without marked evidence
of contusion.

DISCUSSION

The study presented here has the disadvantages
of material analysed in retrospect;
material, moreover, which comprises data recorded by many different physicians.
But in view of the relatively rare occurrence
of epidural haematoma,
in only about
30/o of all head injuries, a prospective
study would have to cover a long period
(about 10 years). In the literature such a study is to be found only with regard to
cerebral injuries in general (OVERGAARD,
HVID-HANSEN,
LAND, PEDERSEN,
CHRTS'TANSEN,HAASE,HEIN
and TWEED, 1973; JENNETT,
1972).
Our study has shown that each of the following factors relates to an unfavourable prognosis:
depth of coma, presence of extensor spasms, bilateral
pupillary
stiffness; bilateral pyramidal symptoms, areflexia, a pulse rate exceeding 1OO/min..
absence of a free interval and a more advanced age (over 50).
The absence of a free interval is generally
indicative
of an intradural
lesion
(cerebral contusion,
with or without acute subdural haematoma).
The free interval
was absent in 16 of the 24 patients with an intradural
lesion. The very high
mortality
intradural

over age 50 is based
lesions (table 6).

on, among

other

things,

the higher

frequency

of

The total mortality in this series was 370/o and the mortality after operation
was
320/o. In patients who underwent
an operation within 24 hours of the accident the
mortality was as high as 500/o (fig. II). The total mortality in our series exceeds
that in any other series published
in the literature
since 1960 (table 6). An explanation
does not seem to lie in the higher frequency of intradural
lesions, which
is generally reported as 15-250/o. In any case the exactness of these data seems but
relative,
because
anatomical
verification
cannot
always have taken place. 111
addition, the mortality depends on the age distribution
and on the interval between
accident and operation (table 1).
JOSEPHSON
(1962), MCLAURIN
and FORD (1964) reported a mortality
of 60’)/,,
within the first 6-8 hours, and of 250/o after the first 8 hours.

242
TABLE

6

Mortality of the epidural hematoma.

Data from literature

period

no. of
patients

died after
operation

intradural
lesion at
operation

1949-1959

125

23%

I 1 Y,;,

1916-1961

71

23%

25:/;,

1957-1962

115

19%

199;,

1956-1967

167

15.6%

47.3 %

Bock et al. (1973) Vienna

19641971

65

20%

?

Jonker and Oosterhuis (1976)
Dept. Neurosurg. Neurol.
Wilhelmina Gasthuis, Amsterdam

1961-1973

100

32%

16%

McKissock et al. (1960)
Neurosurg. Dept. St. Georg’s
Hosp., London
Heyser and Weher (1961)
Neurochirurg. Klinik Ziirich
Philips and Azariah (1965)
South Western Neurosurg.
Unit., Bristol
Jamieson and Yelland (1965)
Dept. Neur. Neurosurg.
Royal Brisbane Hosp.,
Queensland (Austr.)

B&K, BRENNER and W~~BER (1973) reported a total mortality of 200/o,
Z/l8
patients dying after an operation within 12 hours of the accident and 6/19 after
operations performed 12-24 hours after the accident. It is not immediately clear
why the time interval within the first 24 hours did not influence the chance of
survival in our series {fig. II).
It may be that the interval between accident and operation is not relevant, but
rather the interval between the not readily traceable first symptoms of cerebral
compression and the operation. When ‘the symptoms of cerebral compression
are rapidly progressive, the chance of survival is of course smaller when the
patient is not in a (neuro)surgical centre.
In this context it may be mentioned that 330/o of the patients showed symptoms
of compression (extensor spasms and/or stiff pupils) before the operation; in the
series of PHILIPS and AZARIAH(1965) and JAMIESONet al. (1968) these symptoms
were observed in 210/o and 14.50/o of patients, respectively. The JAMIESONseries
moreover, comprised only surgical patients. There may have been more patients
with severe symptoms of compression in whose cases operation was refrained from.
Although the published series are not readily comparable, the relatively high
mortality in our series would seem to be determined in part by the large percentage of patients with compression symptoms and the large percentage of
patients (710/o) who had to be operated on within 24 hours of a rapidly developing
haematoma.

243
Of the patients transported over a distance exceeding 25 km, 7 died although
the prognosis was considered to be fair as transport started.
Since the mortality in our patients operated on after more than 24 hours was
zero, comparison with other series can really only be made if the material is
divided on the basis of this criterion. According to HOOPER (1959) a mortality of
over 250,$ in patients with an epidural haematoma indicates faulty management
and treatment; he maintains that the unavoidable mortality is only lO”/o.
Apparently the situation in Amsterdam and the surrounding region of the
Province of North-Holland has been less than optimal. In any case, only one of
the patients transported to Amsterdam had been given a decompressing burr-hole
by a general surgeon at ‘the site of first aid a procedure recommended by such
authors as PHILIPS et al.(1965).
Epidural haematoma is probably less quickly diagnosed when the free interval
described as characteristic, is absent. This was the case in 47 patients of our series
(table 2). They included on the one hand 13 patients without a lowering of the
sensory level after initial loss of consciousness, all whom were operated on more
than 24 hours later, and on the other hand the group with other intradural lesions.
MCKISSOCK et al. (1960) observed a free interval in only 50 of his 125 patients.
For the sake of completeness it should be pointed out that a free interval occurs
also in cerebral fat embolism (usually 12-24 hours) and in so-called infantile contusion in which cerebral oedema develops a few hours after a sometimes slight
head injury. Differential diagnosis from epidural haematoma can be difficult in
such cases.
Absence of a cranial fracture has been found to improve the prognosis substantially (table 5). Thi,s also applies to patients under 30. In this respect our
findings corroborate GALBRAITH'S (1973) conclusion from a study of the literature
that 910/o of patients with an epidural haematoma without cranial fracture were
under age 30. It is conceivable that under that age the efasticity of the skull gives
less marked local deformation, as a result of which on the one hand less force is
exerted on the cerebrum, while on the other hand the lesion of the middle
meningeal artery is less severe {FORD and MCLAURIN, 1963; P A, 1964).
Absence of a crania1 fracture was not more frequent in patients not operated on
within 24 hours. These data indicate that a diagnosis of epidural haematoma in
patients under 30 cannot be excluded on the basis of absence of cranial fracture.
A 24-hour observation period therefore seems advisable in all patients under 30
admitted with a head injury.
It seems likely, although not statistically demonstrable in this material, that
immediate surgical intervention is sensible even when serious symptoms of compression of the brain stem (stiff pupils, extensor spasms) are already present. This
applies in particular to patients under 30. Even the time required for arteriography
can be fatal to the patient in this situation. JAMIESON et al. (1968) in fact entirely
omitted radiological examination of the skull in a large proportion of his patients.
BoCK, BRENNER and WGBER (1972), OR the other hand, maintained that burr-

244
holes

without preceding carotid angiography may be made only in extreme cmc‘rbecause there is a risk that the haematoma is missed.
In our series ‘chronic’ epidural haematomas proved to carry a good prognosis,
although diagnosis and operative treatment might take place after days to weeks.
The residual morbidity after the later operations did not exceed that after acute
operations (24 hours after the accident). Again it is difficult to decide whethet
the data in the literature are comparable.
MCKISSOCK et al. (1960) reported 130/u mortality after 24 hours, ~AM~ESON el al.
(1968) reported 80/o; of his patients, 45oi0 were operated on after 24 hours. B&x
et al. (1973) operated on 27164 patients after 24 hours: 4 died.
The borderline between acute and non-acute therefore seems to be an arbitrary
one; in any case the literature shows no agreement on this point. SPARACIO.
KWATIB, CHXJ and COOK (1972) counted the first 48 hours as acute phase; HEYSEK
and WEBER (1964) counted the first 7 days as such. MCKISSOCK
et al. (1960),
PHILIPS et al. (1965) and JAMIESON et al. 1968) made no distinction between acute
and chronic, and reported a gradually decreasing mortality with an increasing
interval between accident and operation. The very favourable course in our 27
patients who were operated on after 24 hours is therefore somewhat exceptional.
gencies,

The morbidity
7 patients
assisted.

in the 58 survivors

showed
B&K

moderately

residual

et al. (1973)

with a follow-up

symptoms
also

severe residual symptoms

was relatively

of any significance,

reported

that

200/o

low. Only

but could

of their

walk

patients

un-

showed

but were able to walk, and that none showed

akinetic mutism.

ACKNOWLEDGEMENTS:

We thank Prof. Dr. W. Noordenbos,

Dr. L. v. Trotsenburg

and Prof.

Dr. W. A.

den Hartog Jager for their critical comments.

REFERENCES
BtiK, F., BRBNNERH. and WijBER,0. (1972) Welche Maasnabmen kiinnen die Ergebnisse beim
epiduralen Wimatom verbessern? Mschr. Unfallheilk, 75,408.
B&K, F., BRENNER, H. and WaBER, G. (1973) Zur Symptomatologie
und Prognose der traumatischen epiduralen Blutungen. Mschr. Unfailheilk. 76, 300.
FORD, L. E. and MCLAURIN, R. L. (1963) mechanisms
of extrad~al
hematomas. J. neurosurgery
20, 760.
GALBRAITH, s. L. (1973) Age-distribution
of extradural hemorrhage
without skull fracture.
Lancet I, 1217.
HEYSER, J. and WEBER,G. (1964) Die epiduralen Hlmatome. Schweiz. med. Wschr. I, 2.
HOOPER,
R., (1959) Observations on extradural hemorrhage. Brit. J. Surg. 47, 71.
JAMIESON,K. G. and YELLAND,J. D. N. (1968) Extradural haematoma. Report of 167 cases.
J. Neurosurgery 29, 12.
JENNETT,B. (1972) Prognosis of severe head injury. Clin. Neurosurg. 19,200.
JOSEPHSON, s. (1962) Epidural haematoma.
A lo-year series. Acta chir. stand. 12#,26,
MCKISSOCK,
w., TAYLOR,J. c., BLOOM, W. H. and TILL, K. (1960) Extradurat haematoma.
Observations on 125 cases. Iancet II, 167.

245
MCLAURIN,

R. L., FORD, L. E. (1964) Extradural

OVERGAARD,

J., HVID-HANSEN,

0.

LAND,

haematoma.

A., PEDERSEN,

J. Neurosurg.

21, 364.

K. K., CHRISTANSEN,

S., HAASE,

J.,

and TWEED, w. J. (1973) Prognosis after head injury based on early clinical
examination. Lancet II, 631.
PHILIPS, D. G. and AZARIAH, R. c. s. (1965) Acute intracranial
haematoma from head injury: A
study in prognosis. Brit. J. Surg. 52, 218.
PIA, H. w. (1964) Die traumatische
Hirnblutingen des Kindesalters. Acta Neurochir. II, 582.
SPARACIO, R. R., KHATIB, R., CHILI, J. and COOK, A. w. (1972) Chronic epidural haematima.
J.
Trauma 12, 435.
HEIN, o.