British Journal of Obstetrics and Gynaecology
June 1986, Vol. 93, pp. 554-562

Active management of the unconscious eclamptic
patient
A. M. RICHARDS, J. MOODLEY Consultant Obstetricians and Gynaecologists,
Department of Obstetrics and Gynaecology, University of Natal, Durban, South Africa
D . I. GRAHAM Professor of Neuropathology, Department of Neuropathology, Institute
for Neurological Sciences, Southern General Hospital, Glasgow, UK &
M. R. R. BULLOCK Consultant Neurosurgeon, Department of Neurosurgery, University
of Natal, Durban, South Africa

Summary. Of the many complications which may develop after eclamptic seizures, prolonged unconsciousness is one of the most difficult for
obstetricians to manage as the pathophysiology of this condition
remains largely unknown. Computed axial tomography (CTscan) was
performed on 20 unconscious eclamptic patients, and autopsy was
obtained on an additional two patients. Changes compatable with cerebral oedema were demonstrated in 75% of patients; cerebral haemorrhage occurred in 9%. A programme of intensive neurological
management aimed at optimizing cerebral perfusion and controlling
intracranial pressure is outlined. We have reduced the mortality rate for
unconscious eclamptic patients from 50% to 17% in our institution.
Eclampsia still ranks as one of the most frequent
causes of maternal death throughout the world,
although improvement in antenatal care has
reduced its incidence in Western countries
(Augensen & Bergsjo 1984; Gibbs & Locke
1976).
In South Africa, as in many developing countries, eclampsia remains a common occurrence.
Figures from King Edward VIII Hospital, University of Natal (1983) indicate the development
of eclampsia in approximately 9 out of every
1000 deliveries. Between 5 and 10% of patients
with eclampsia die from the disease process.
Cerebral dysfunction, renal failure and pulmonary oedema are the most common terminal
events (Donnelly et al. 1954).
Cerebral dysfunction, as manifest by a
decreased conscious level, occurs in the majority

Correspondence: Dr A . M. Richards, Department of
Obstetrics and Gynaecology, University of Natal, Faculty of Medicine, PO Box 17039, 4013 Congella, Durban. South Africa.

554

of patients with eclampsia for a variable period
after seizures. Where cerebral dysfunction has
been cited as the cause of death, post-mortem
studies (Melrose 1984; Sheehan & Lynch 1973)
have revealed intracerebral haemorrhage in up
to 60% of patients, but the incidence of cerebral
haemorrhage in non-fatal eclampsia (Beck &
Menezes 1981) and the related pathophysiology
(Souter 1980) remains unknown.
The advent of computed axial tomography
(CT) scan has facilitated the study of living
eclamptic patients with cerebral dysfunction,
and several reports (Beeson & Duda 1981; Benedetti & Quilligan 1980; Drayer & Rosenbaum
1976) have documented CT scan appearances
suggestive of cerebral oedema. Intracerebral
haemorrhage appears to be a less frequent finding (Melrose 1984).
A programme of intensive neurological
management was devised at King Edward VIII
Hospital in an attempt to improve the outcome
for the many eclamptic patients treated
annually. This programme includes CT scanning

Management of the unconscious eclamptic

and intracranial pressure monitoring where
indicated.
The purpose of this paper is to describe the CT
scan changes found in these patients in relation
to intracranial pressure recordings, and to suggest a management protocol aimed at reducing
intracranial pressure and optimizing cerebral
perfusion.
Patients and methods
A total of 192 patients with eclampsia was seen
at King Edward VILI Hospital over the study
period of 14 months (1July 1983 to 30 September 1984). During this time approximately
19000 deliveries were performed at this
hospital.
All eclamptic patients were managed according to a standardized protocol (Moodley et al.
1984). Convulsions were immediately controlled
with magnesium sulphate and blood pressure
was lowered to a diastolic of between 100 and
105 mmHg with intravenous dihydralazine given
in small doses. Dexamethazone was used routinely for the initial 24 h, in the belief that this
may reduce cerebral oedema.
Central venous and radial artery pressures
were monitored in all patients and endotracheal
intubation was undertaken when required to
secure an adequate airway, oxygen therapy and
frequent arterial blood gas determinations were
performed to prevent cerebral hypoxia. Termination of pregnancy was achieved by caesarean
section, if vaginal delivery was not imminent
within 4-6 h of admission. Urgent delivery was a
management priority in comatose patients.
Patients remaining unconscious 6 h postdelivery and those with localizing neurological
signs, underwent CT scanning. Intracranial
pressure monitoring was then performed in
patients with severe brain oedema on CT scan,
where the Glasgow coma scale was 9/15 or less
(i.e those patients who are unable to open eyes,
obey commands or speak) (Teasdale & Jennett
1974). A GE8800 CT scanner was used and
intracranial pressure was monitored through a
ventricular or subdural catheter via a Statham
P50 transducer connected to a continuous bedside recording system.
High intracranial pressure was managed
according to a fixed protocol as follows. All
patients were nursed in a neurosurgical intensive
care unit in the semi-sitting position (30") to
minimize intracranial venous pressure. When

555

high intracranial pressure persisted a 20% mannitol infusion (0.5 glkg) was employed, followed
by hyperventilation to reduce arterial P C O ~
levels (PcoI 28 k 5 mmHg). Where these
measures failed an infusion of Althesin was commenced at 24 pg/kg/min. This was then titrated
against intracranial pressure in an attempt to
maintain optimal cerebral perfusion pressures.
Intermittent drainage of CSF in 5-ml aliquots
was used in two patients. Althesin (alphaxalonealphadolone), a steroidal short-acting anaesthetic induction agent, has since been withdrawn
from generalized usage, because of hypersensitivity reactions reported in 1 in lo00 patients
(Evans & Keogh 1977). Its value in reducing
high intracranial pressure has been confirmed by
several authors (Stewart et al. 1983; McIlhany et
al. 1983).
The illustrative case histories of two patients
are presented below.
Patient no. 1

Miss N.N., a 20-year-old unbooked primigravid
patient at approximately 26 weeks gestation,
was admitted via a peripheral clinic to a district
Mission Hospital, with a history of generalized
convulsions. A diagnosis of eclampsia was made
on a blood pressure recording of 170/140 mmHg,
mean arterial pressure (MAP) of 150 mmHg and
gross proteinuria. Intravenous diazepam was
administered and transfer arranged.
On arrival at the referral hospital it was estimated that the patient had been fitting intermittently for 18 h. The fetal heart was not detected
and the fetus was felt to be severely growthretarded. There was minimal response to painful
stimuli with flaccidity of all four limbs and no
localizing neurological signs. She had evidence
of early aspiration pneumonia, and gross
generalized oedema.
Control of convulsions with magnesium sulphate and lowering of the blood pressure was
undertaken. Following a hysterotomy, the
patient was electively ventilated for 4 h.
Examination at this stage revealed no change
in her level of consciousness, estimated at
Glasgow coma scale 4/15. Her blood gases were
normal, haemoglobin 9.5 g/l, white cell count
25 000 and crude clotting time 5 min. There was
evidence of renal failure in that her serum
potassium was 6.8 mmol/l and urea 12.6 mmol/l,
with minimal urine output in response to i.v.
frusemide.

556

A . M . Richards et al.

A CT scan was performed and this revealed
diffuse white matter low density consistent with
vasogenic cerebral oedema (Fig. 1). The basal
cisterns were visible at this stage.
An intracranial pressure monitor revealed an

initial pressure of 43 mmHg (normal 10-15
rnmHg). Mannitol infusion (0.5 g/kg) and hyperventilation were commenced.
A clotting profile revealed elevated fibrinogen
degregation products with low haemoglobin and

Fig. 1. (a) Scan series of patient no. 1 (N.N.) performed 8 h after
caesarean section for eclampsia. Widely patent basal cisterns are
noted in cut 2, and moderate diffuse white matter oedema is seen.
(b) CT scan series performed in patient no. 1 (N.N.) 48 h later.
Note complete effacement of basal cisterns in cut 2, and gross
progression of brain oedema.

Management of the unconscious eclamptic

557

Fig. 2. CT scan series of patient no. 2. (K.H.) demonstrating
‘watershed’ pattern brain oedema. Basal cisterns are uneffaced.
Low density areas involve the white matter of the temperoparietal-parieto-occipital and high parietal regions.

platelet levels, suggestive of a consumptive
coagulopathy.
Over the next 12 h intracranial pressure rose
to reach levels around 60 mmHg in spite of large
doses of diuretics, steroids, mannitol and subsequently an Althesin infusion. A repeat CT
scan revealed massive progression of cerebral
oedema with obliteration of basal cisterns (Fig.
1).
Shortly afterwards the patient developed fixed
dilated pupils and her blood pressure dropped to
100160 mmHg (MAP 73). A simultaneous fall in
intracranial pressure was noted. Twelve hours
later formal testing for brain function revealed
the patient to be brain dead. Ventilation was
therefore discontinued.
Post-mortem revealed haemorrhages in the
liver and adrenal medulla, pulmonary oedema
and a soft swollen brain with signs of uncal transtentorial herniation (coning). Neurohistology
revealed fibrinoid arterial wail necrosis, perivascular microinfarcts and brain oedema. Widespread hypoxidischaemic neuronal damage was
found.

Patient no. 2.
Mrs K.H., a 37-year-old para 7, gravidity 8, was
delivered vaginally at home of a normal healthy

male infant weighing 3-1 kg. Shortly afterwards
she became confused and developed seizures.
The patient was admitted to hospital within
2 h of delivery. On admission her blood pressure
was 160/110 mmHg (MAP 126) and moderate
proteinuria was demonstrated. She did not obey
commands, and her Glasgow coma scale was
8/15.
Routine anticonvulsant and hypotensive
therapy was commenced and the patient was
referred for CT scanning which revealed patchy
areas of low density indicative of white matter
oedema (Fig. 2). Following insertion of an intraventricular pressure monitor the pressure was
found to be 25 mmHg. This rapidly rose to 50
mmHg. A right hemiparesis was present.
Steroids, mannitol, Althesin infusion and intermittent cerebrospinal fluid aspirations were
carried out to control intracranial pressure.
On the third day after delivery, the intracranial pressure began to settle to normal levels
and spontaneous eye opening occurred, but
during the course of the afternoon the intracranial pressure suddenly rose again. A repeat
CT scan showed progression of the cerebral
oedema.
Over the following 24 h the patient’s level of
consciousness improved with spontaneous eye
opening and movement. The intracranial pressure fell rapidly to near normal levels, and there

558

A . M . Richards et al.

Fig. 3. Occipital oedema.

Clinical variable

Mean

(Range)

for CT scan. The clinical features of these 22
patients are outlined in Table 1. Most were
young primigravida around 36 weeks gestation;
the majority were unbooked or had received
minimal antenatal care before their first presentation with seizures. Average blood pressure for
the group measured at the time of first hospital
admission was 1661110 mmHg (MAP 129). The
interval between the first and last seizure ranged
from 25 min to 20 h and reflects the long and
often slow referral chain which operates in certain rural areas in Natal, which may be up to
300 km from King Edward VIII Hospital. This
interval was significantly correlated with the
severity of changes on CT scan (Table 2). The
most marked CT changes were seen in those
patients with the longest duration of seizures.
Five patients died, two of whom were found to
have an intracerebral haemorrhage (Table 3 ) .

Age (years)
Gestation at onset of
seizures (weeks)
Parity
Glasgow coma scale
Time taken for seizure
control (h)
Blood pressure on first
hospital admission
(mmHg)
Mean arterial pressure

21.3

(15-37)

CT scan features

36

(30-40)
W7)*

6

(3-9)

5.6

(0.3-20)

Twenty patients underwent CT scanning (Table
2). In five patients CT scan appearances were
regarded as normal. A brain stem haemorrhage
was diagnosed in one patient. The remaining
patients all demonstrated CT scan low density
areas interpreted as cerebral oedema of varying
degrees.

was no evidence of the previously recorded right
sided weakness. The patient was extubated and
discharged from the intensive care unit.
A 6-week follow-up visit revealed normal
blood pressure, renal function and neurological
status. A tuba1 ligation was advised.

Results

Clinical features
Twenty-two eclamptic patients remained in
coma 6 h or more following delivery, two of
these patients died before inter-hospital transfer
Table 1. Clinical features in 22 patients in coma after
eclampsia

(mmHd

166/110(130/90-210/128)
128

(103-141)

*Eighteen patients were in their first pregnancy.

‘Diffuse white matter oedema’ (Fig. l a , b). Five
patients demonstrated a low density pattern
diffusely distributed throughout the white

Management of the unconscious eclamptic

559

Table 2. CT scan data in 20 patients in coma after eclampsia
~~~~~~

~

CT scan appearance
‘Diffuse white
‘Watershed ‘Occipital white
matter oedema’ pattern oedema’ matter oedema’
Number of patients
Basal cistern effacement
Intracranial pressure
High
Low
(No. monitored)
Mean period before
control of seizures (h)

Normal

Brainstem
haemorrhage

5
1

5
0

1

1
(1)

-

-

-

6.3

3.0

0.8

5

5

4

4

0

3
1
(4)

9.4*

1

High intracranial pressure, greater than 20 mmHg for >2 h.
*Significantly prolonged when compared to normal CT patients. P<0.05. *Kruskal-Wallis test and Dunn’s
multiple comparison procedure for paired groups.
matter. This was regarded as the most severe
form of oedema in that the majority of these
patients showed associated basal cistern effacement which may indicate raised intracranial
pressure (Teasdale et al. 1984). In addition,
three of the four patients in this group, who
underwent intracranial pressure monitoring,
were found to have significantly raised
pressure.

‘Watershed oedema’ (Fig. 2). A further four
patients demonstrated the so-called watershed
pattern oedema, maximal between middle cerebral and posterior cerebral artery distribution
territories. The one patient in this group who
underwent intracranial pressure monitoring was
found to have high pressure.
Table 3. Details of the five patients who died

Patient

Cause of death

N.N.

Gross cerebral oedema, high ICP and
renal failure
Intra-pontine bleed (demonstrated on CT
scan)
Died before (3T scan, autopsy revealed
oedematous brain with severe ischaemicl
hypoxic neuronal damage
Occipital oedema, recovered
consciousness, but died from adult
respiratory distress syndrome and renal
failure
Died before CT scan, autopsy revealed a
large frontal intracerebral haematoma

D.Z.
T.N.
P.B.

H.S.

‘Occipital oedema’ (Fig.3). In the remaining five
patients small patchy areas of low density were
noted in the white matter of the occipital region.
Intracranial pressure monitoring.
Of the five patients who underwent intracranial
pressure monitoring, four exhibited pressures in
excess of 20 mmHg for >2 h duration. In three of
these patients intracranial pressures rose to
levels at which cerebral perfusion was jeopardized and at this stage Althesin infusion was
employed (Fig. 4).

Cerebral haemorrhage
Two patients were found to have cerebral haemorrhages and both these patients subsequently
died. The first patient, a 35-year-old para 4, fitted for 5 h before caesarean section, after which
her level of consciousness remained at Glasgow
coma scale 4/15. CT scan revealed an intrapontine bleed. The second patient, a 15-year-old
primigravida, was admitted moribund having
fitted for many hours during inter-hospital transfer. Autopsy revealed a large frontal intracerebral haematoma.

Discussion
Eclampsia is still a relatively common occurrence in developing countries, and its
pathophysiology remains largely unknown.
CT scans revealed low density areas suggestive of cerebral oedema in 70% (14 out of 20) of

560

A. M.Richards et al.

i'i
Moved for CT scan

,./

- 60-

.

-

.

-'

.'

N.N. (Died)

0

I

K.H.
I

I

12

24

36

48 V.K. 60

7
72

Time ( h )

Fig. 4. Intracranial pressure changes in five patients in coma after
severe eclampsia. Arrows indicate Althesin mannitol ventilation.

our patients remaining unconscious 3 6 h after
delivery. Three patterns of cerebral oedema
were demonstrated which, when correlated with
basal cistern effacement on CT scan (Teasdale et
al. 1984), measurement of intracranial pressure
and duration of seizures, represent a spectrum of
severity (Table 2). The changes being most
marked in patients with the longest duration of
seizures.
The pattern of oedema demonstrated on CT
scan involved predominantly cerebral white
matter (Fig. 1). Vasogenic brain oedema is
known to produce this pattern of white matter
low density (Miller 1979). The reasons for a
vasogenic pattern of oedema are unclear but
may be related to the generalized vasculopathy found in eclamptic patients. Vessel wall
damage may allow transudation of plasma, this
would be exacerbated by hypertension and
seizures.
Rapid control of seizures is necessary to prevent widespread hypoxic cerebral damage. Maximal brain metabolism during seizure activity
with consequent depletion of glucose and oxygen reserves, in addition to the absence of
normal respiration, results in hypoxic damage.
Recurring seizures constitute a particularly
dangerous situation, for worsening cerebral
hypoxic damage.
Maintenance of an adequate airway and respiration during seizures should be an immediate
priority in eclamptic patients. Frequent arterial
blood gas measurements are essential to quantify oxygen therapy. An arterial Po2 level of at
least 100 mmHg is recommended for all unconscious patients.
Optimal control of blood pressure in eclamp-

tic patients requires considerable judgment and
caution. The need to reduce blood pressure to
prevent intracerebral haemorrhage is well
known. However, our post-mortem data have
shown that a major cause of brain damage in
eclamptics is cerebral ischaemidhypoxia, and
this may be greatly worsened by lowering the
blood pressure below the level necessary to
maintain cerebral perfusion in the face of raised
intracranial pressure (Graham 1975; Ledingham
& Rajagopalan 1979). Assuming a cerebral perfusion pressure around 40-60 mmHg is required
to prevent brain ischaemia, and that intracranial
pressure may be around 20-30 mmHg in severe
eclamptics (Fig. 4), then a minimum mean
arterial pressure around 80 mmHg would be
required (Fig. 5).
On the other hand, when the mean arterial
pressure exceeds 150 mmHg, the upper limit of
cerebral autoregulation is breeched, and widespread cerebral vasodilatation occurs, leading to
engorgement, probable further oedema, and
further elevation of the intracranial pressure
(Fig. 5).
Injudicious rapid reduction of blood pressure
maybe as dangerous as uncontrolled hypertension, therefore the optimal mean arterial
pressure in eclamptic patients with suspected
cerebral oedema would be between 80 and 100
mmHg.
Delivery of the fetus is desirable to terminate
the eclamptic vasculopathic process. Caesarean
section is usually the most rapid route available.
Scrupulous anaesthetic technique aimed at
avoiding blood pressure fluctuations is essential
to optimize cerebral perfusion. Where general
anaesthesia is employed, extubation should be

\

Management of the unconscious eclamptic
561
HIGH BLOOD PRESSURE
)regulation
Loss of autoregulation
>1?0 mmHg)
?+ VASCULOPATHY
* (MAP >150
b
Cerebral engorgement
. of
lowering
Iatrogenic
Epileptogenic/
focus

Seizures

<\
Cerebral
haemorrhage

''

blood pressure

( M A P < 80 mmHg)

nypuma
Hypoxia

\

ISCHAEMIC

c

//

Cerebral
ereDral oedema
+ raised ICP

Inadequate cerebral
perfusion

/

BRAIN DAMAGE

Cerebral perfusion pressure = mean arterial pressure (MAP) - intracranial pressure (ICP)
Fig. 5. Mechanisms for the development of brain damage in eclampsia.
delayed until the patient is fully awake and able
to maintain an adequate airway.
Fluid replacement, predominantly with colloids, should be undertaken against central
venous pressure measurement. Large volumes
of crystalloids are held by certain authors to
worsen brain oedema (Miller 1979).
We have empirically administered steroids to
our patients for CT scan in the belief that the
development of vasogenic oedema maybe
retarded, although no concrete evidence exists
that steroids are of benefit in eclampsia.
We have found that the decision to refer
eclamptic patients should be delayed for several
hours after delivery, except where localizing
neurological signs are present, or where deterioration in consciousness occurs. Many apparently
comatose eclamptic patients recover rapidly and
uneventfully after delivery. However, we
believe CT scanning is desirable for all those
who are persistently unconscious 4-6 h after
delivery. Our data suggests that the majority of
these women will have cerebral oedema.
Intracranial pressure monitoring should only
be considered in persistently comatose patients
with marked CT scan abnormalities. We have
used basal cisternal effacement on CT scan as a
criterion of raised intracranial pressure
(Teasdale et al. 1984), and this was an accurate
predictor of high pressure in 60% of our cases in
whom intracranial pressure was monitored.
Intracranial pressure monitoring should be
carried out in an intensive care unit, with the
intention of controlling high pressure and
improving cerebral perfusion. Measures we
have found useful include. posture (30" of head
up tilt to improve venous drainage), 20% man-

nitol(O.5 g/kg) given as an i.v. bolus, and hyperventilation to reduce arterial Pco, to 28 ?
4 mmHg. Where these measures failed, we
employed Althesin, a steroidal anaesthetic
induction agent which rapidly reduces brain
activity and metabolism. We believe Althesin to
be a safer agent than barbiturates, although
barbiturate coma is employed in many neurosurgical centres to reduce high intracranial
pressure.
In all of our surviving patients with high intracranial pressure, a dramatic fall occurred around
24-36 h after delivery (Fig. 4). We believe this
corresponds to resolution of the vasculopathy
rather than the brain oedema itself. We have
observed persistent low-density on CT scan up to
7 days after delivery.
Five out of 192 eclamptic patients in our series
died, a marked improvement in our previously
reported maternal mortality. Early CT scanning
allows differentiation between cerebral oedema
and cerebral haemorrhage. Aggressive neurological management directed towards lowering
intracranial pressure and preserving cerebral
perfusion appears to greatly improve the prognosis for severe eclamptic patients.

Acknowledgments
We would like to thank the Natal Provincial
Ambulance Service for the efficient and expert
transportation of the patients in this study. We
are also very grateful to Professor Van Dellen
and the staff in both the CT scanning and Neurosurgical I.C.U. departments at Wentworth
Hospital.

$62

A . M.Richardsetal.

References
Augensen, K. & Bergsjo, P. (1984) Maternalmortality
in the Nordic Countries. Acta Obstet Gynecol
Scand 63, 115-121.
Beck, D. W. & Menezes, A. H. (1981) Intracerebral
hemorrhage in a patient with eclampsia. J A m Med
Assoc 246, 1442-1443.
Beeson, J. H. & Duda, E. E. (1981) Computed axial
tomography scan demonstration of cerebral
oedema in eclampsia preceded by blindness.
Obstet Gynecol60, 529-532.
Benedetti, M. D. & Quilligan, E. J. (1980) Cerebral
oedema in severe pregnancy-induced
hypertension. A m J Obstet Gynecoll37, 860-862.
Donnelly, J. F., Lock, F. R. & Winston-Salem, N. C.
(1954) Causes of death in five hundred and thirty
three fatal cases of toxemia of pregnancy. A m J
Obstet Gynecol68, 184-189.
Drayer, B. P. & Rosenbaum, A. E. (1976) Brain
oedema defined by cranial computed tomography.
J Computer Assisted Tomography 3, 317-323.
Evans, J . M. & Keogh, J. A . M. (1977) Adverse
reactions to intravenous anaesthetic induction
agents. Br Med J ii, 735-736.
Gibbs, C. E. & Locke, W. E. (1976) Maternal deaths
in Texas, 1969 to 1973. A m J Obstet Gynecol 125,
687-691.
Graham, D. I. (1975) Ischaemic brain damage of cerebral perfusion failure type after treatment of severe
hypertension. Br Med J iv, 739.
Ledingham, J. G. G. & Rajagopalan, B. (1979) Cerebral complications in the treatment of accelerated
hypertension. Quart J Med 48,25-41.
McIlhany, M., Rapport, R.,Dunn, R., Namazie, M.

& Delilican, A. (1983) Althesin in the management of head injuries. A preliminary report. Neurosurgery 12, 71-19.
Melrose, E. B. (1984) Maternal deaths at King
Edward VIII Hospital, Durban. S Afr Med J 65,
161-165.
Miller, J . D. (1979) Management of cerebral oedema.
Br J Hosp Med 21, 152-157.
Moodley, J., Norman, R. J. & Reddi, K. (1984)
Central venous concentrations of immunoreactive
prostaglandins E, F, and 6-keto-prostaglandin F, in
eclampsia. Br Med J 288, 1487-1489.
Sheehan, H. L. & Lynch, J. B. (1973) Pathology of
toxaemia of pregnancy. Williams and Williams Co,
Baltimore.
Souter, W. P. (1980) The haemodynamic pathophysiology of pre-eclampsia. S Afr Med J 48, 351-355.
King Edward VIII Hospital, University of Natal
(1983) Statistics from the Department of Obstetrics
and Gynaecology.
Stewart, G. O., Dobb, G. J . & Graig, I. A. (1983)
Clinical trial of continuous infusion of Alphaxalone-Alphadolone in intensive care patients.
Anaesth Intens Care 1 1 , 107-112.
Teasdale, G . & Jennett, B. (1974) Assessment of coma
and impaired consciousness: A practical scale.
Lancet ii, 81-83.
Teasdale, E., Cardose, E., Galbraith, S . & Teasdale,
G. (1984) CT scan in severe diffuse head injury:
physiological and clinical correlations. J Neurol
Neurosurg Psychiatry 41, 600-603.

Received 25 June 1985
Accepted 7 October 1985