British Journal of Anaesthesia 84 (4): 518–20 (2000)

Cranial subdural haematoma associated with dural puncture in
labour
D. J. A. Vaughan*, C. A. Stirrup and P. N. Robinson
Department of Anaesthesia, Northwick Park and St Mark’s Hospitals, Watford Road, Harrow,
Middlesex HA1 3UJ, UK
*Corresponding author
A 23-yr-old primagravida sustained a dural puncture during epidural catheter insertion and
developed a headache that settled with oral diclofenac and codydramol. On the third day after
delivery, she convulsed twice without warning. As plasma urate was increased, the putative
diagnosis of an eclamptic fit was made, and magnesium therapy was started. A contrast CT
scan revealed that the cause of the patient’s symptoms was a subdural haematoma with
raised intracranial pressure. A coincidental arteriovenous malformation was noted. This case
emphasises the need to consider the differential diagnoses of post-partum headache. The
management of acute intracranial haematoma is described.
Br J Anaesth 2000; 84: 518–20
Keywords: analgesic techniques, epidural; complications, headache; complications, seizure;
complications, subdural haematoma
Accepted for publication: December 6, 1999

Inadvertent dural puncture with subsequent headache is a
recognized complication of epidural analgesia.1 This case
report emphasizes the need for careful and ongoing assessment of these patients in order to ensure that less common
causes of headache are not overlooked.

Case report
A 23-yr-old primiparous Indian woman presented in labour
following spontaneous rupture of membranes at home.
She was distressed and requesting epidural analgesia. The
combination of a high body mass index (height 144 cm,
weight at last antenatal clinic before delivery 82 kg; body
mass index 39.5) and a poorly compliant and distressed
patient made epidural catheter insertion difficult, and in the
process an inadvertent dural puncture occurred. The catheter
was sited at a higher level, and good analgesia was obtained
with a test dose of 0.25% bupivacaine 3 ml, followed by a
further 10 ml with fentanyl 50 µg. Subsequent top-ups by
the anaesthetist with 0.1% bupivacaine 10 ml plus fentanyl
2 µg ml–1 provided good analgesia without side-effects.
The contractions were augmented with an oxytocin
infusion, as meconium was present in the liquor. Four hours
after the epidural, the patient developed low-grade pyrexia
(37.8°C) and was started on amoxycillin and clavulanic
acid (co-amoxiclav 1.2 g) i.v.; a blood culture was taken.
After a further 3 h the patient was fully dilated and started
to push. At this time she began to complain of a continuous,
dull, moderate occipital headache with no radiation. In view

of the pyrexia, meconium and a worsening cardiotocograph
trace, a forceps delivery was performed.
Overnight, the patient complained of severe occipital
headache, although this settled with paracetamol and rest.
This was assumed to be a post-dural puncture headache
(PDPH) because it had a postural component (it was worse
on sitting up or standing). Over the next 2 days the patient
was regularly reviewed and her headache improved rapidly
with oral codydramol and diclofenac. At no time did she
complain of nausea, vomiting, photophobia or neck stiffness.
Her pyrexia settled within a day of delivery. On the third
postnatal day, it was agreed that a blood patch was not
indicated and discharge was planned for the next day.
That evening, the patient had two self-limiting 30 s
tonic–clonic convulsions. After 30 min of post-ictal
drowsiness, the patient was fully alert with a Glasgow coma
score of 15. The only abnormal findings on examination
were symmetrical, moderately brisk reflexes. A putative
diagnosis of eclamptic fits was made, strengthened by a
raised plasma urate level of 438 µmol litre–1 (normal range
100–400 µmol litre–1). Blood pressure was normal, and
there was no significant proteinurea or oedema. A raised
white cell count of 17.5⫻109 ml–1, consisting mainly
of neutrophils, was noted. High dependency care was
initiated and magnesium therapy started as per unit protocol
(magnesium sulphate 4 g i.v. as a slow bolus followed by
an infusion of 1 g hourly).
After a stable night during which the occipital headache
recurred, the obstetric anaesthetic consultant reviewed the

© The Board of Management and Trustees of the British Journal of Anaesthesia 2000

Postpartum presentation of intracranial bleeding

Fig 1 Contrast-enhanced cerebral CT scan showing (A) small subdural
haematoma with an outer, posterior layer of fresh blood and (B) AVM
arising from the basal ganglia.

patient. It was felt that the onset of PDPH during labour,
combined with a tonic–clonic seizure 3 days after delivery,
warranted an urgent computerized tomography (CT) scan
to exclude intracranial pathology. Contrast-enhanced CT
revealed a left-sided subdural haematoma, evidence of
raised intracranial pressure without midline shift, and an
arteriovenous malformation (AVM) in the basal ganglia
(see Fig. 1).
The patient was transferred to a neurosurgical centre
where she returned to a symptom-free state without intervention. A cerebral magnetic resonance imaging (MRI)
scan revealed no further information. It has been decided
to manage her conservatively with regular neurological
follow-up.

Discussion
This case illustrates the difficulties of diagnosing the cause
of headache and convulsions in the peripartum period. The
differential diagnosis of post-delivery headache in this
case includes PDPH, pre-eclamptic headache, subdural or
subarachnoid haemorrhage, simple tension or migraine
headache, meningitis and cortical vein thrombosis.1 With
the knowledge that dural puncture had occurred earlier in
labour and from the postural component of this headache,
it was assumed that PDPH was the most likely diagnosis.
However, the onset of the headache was early, coinciding
with the second stage of labour and the initiation of pushing,

making the diagnosis of PDPH less likely. The diagnosis
of pre-eclampsia as the causal condition was discounted, as
no other symptoms were present. Subdural or subarachnoid
haemorrhage were not considered at the time of onset of
the headache, as these are extremely rare causes to present
de novo. It was only when the convulsions occurred that
this omission from the differential diagnosis was corrected.
Other than an initial neutrophilia, there were no features to
support a diagnosis of meningitis at this time; the pyrexia
settled and the white count normalized. The history was
not suggestive of cortical vein thrombosis, and this was
excluded later on MRI scanning.
The differential diagnosis of the convulsions in this
case includes eclampsia, intracranial haemorrhage, embolic
episodes and epilepsy. Local anaesthetic toxicity can be
excluded as the convulsion occurred on the third day after
delivery. Subdural haematoma following dural puncture is
a very rare complication of epidural analgesia in labour,
and may be cranial or spinal.2–5 The cause is low cerebrospinal fluid pressure following dural puncture leading to
traction and tearing of thin-walled dural blood vessels.6 In
this case the onset of the headache was very early in relation
to the time of dural puncture. It is unusual for a large CSF
leak to have occurred in this short time. We have assumed
that the tear and subsequent bleeding occurred at or around
the time of onset of headache in labour. However, this may
have been a tension or stress-related headache, and the
pathological events may have occurred later that evening.
It is not possible to pinpoint the time of vessel rupture
definitively.
Cranial subdural haematomas may present acutely,
subacutely or chronically with a variety of complaints
including headache, altered level of consciousness or even
psychiatric symptoms.7 Acute subdural haematomas have
been described as lesions causing signs or symptoms within
7 days of the bleed. Spinal subdural haematomas following
epidural anaesthesia are extremely rare and of complex
presentation.6 8 The presence of a suspected PDPH
complicated by neurological deterioration, as in this case,
should prompt a search for intracranial pathology.
The plain CT scan did not reveal the underlying pathology,
and the subdural haematoma and AVM were only seen
after contrast enhancement. CT may not show subdural
haematomas without contrast, and this has previously been
reported as a cause of delayed diagnosis.9 Haemorrhage
from the AVM at the time of pushing is a possible cause
of the subdural haematoma but, more commonly, ruptured
AVMs cause subarachnoid bleeding.
It was considered by the neurosurgeon and neuroradiologist that the AVM seen on CT was an incidental
finding and that the subdural haematoma was secondary to
the dural puncture. The raised intracranial pressure was a
result of the haematoma and its increasing space-occupying
effect because of oedema. This is in marked contrast to the
low CSF pressure believed to have caused the vessel tear.
Haemorrhage from an AVM is most common in women

519

Vaughan et al.

under 25 yr and is most likely during the second trimester
and in the immediate post-partum period.10 The clinical
presentation is sudden onset of headache, often accompanied
by nausea and vomiting, photophobia, seizure activity, and
focal motor and sensory disturbances.11 In this case, the
onset of headache coincided with initiation of pushing,
arousing suspicion—in retrospect—of an intracranial event.
Subarachnoid haemorrhage only occurs in one in 10 000
pregnancies, but is associated with a high maternal
mortality.12
The management of subdural haematoma is either
conservative (clinical observation and possible intracranial
pressure monitoring) or surgical evacuation. For this reason
the patient was transferred to a neurosurgical unit. In view
of the small size of the haematoma coupled with a rapid
return to normal neurological status, conservative management was deemed appropriate. Haematomas under 5 mm
often spontaneously resolve. In cases requiring surgery,
preoperative neurological status affects postoperative
survival. Early haematoma evacuation and young age are
favourable prognostic factors. This patient had no midline
shift on CT scan although there was evidence of raised
intracranial pressure. Parenchymal injury causes midline
shift and is associated with decreased survival.13
This case also raises the issue of when a blood patch for
PDPH should be performed. Reynolds recommends that,
‘Headache after dural puncture with a large needle should
be treated promptly with an epidural blood patch’.14 Our
unit policy is to assess all known dural punctures for blood
patching after 24 h bed rest, hydration and analgesia. In
this case, the patient improved rapidly with conservative
measures and it was felt that a patch was not necessary. It
could be argued that early blood patching may decrease the
risk of subdural bleeding by preventing a fall in CSF
pressure. In this patient such action probably would not
have helped, as we believe the bleeding occurred during
labour. It is our practice to discharge patients with PDPH
24 h after a successful blood patch, or when they are mobile
in conservatively managed cases. All patients are reviewed
in the obstetric anaesthetic clinic 1–2 weeks later to ensure
that symptoms have not recurred or worsened, and are
advised to contact the obstetric anaesthetist if any symptoms
occur or recur. We feel that this follow-up minimizes the
risk of late complications being missed.

The management of PDPH is often complex and is an
important role for the obstetric anaesthetist. While we
emphasize the rarity of subdural haematoma as a complication of dural puncture, this case highlights the importance
of including it in the differential diagnosis. We suggest the
urgent use of contrast-enhanced CT in all cases of seizure
after dural puncture to exclude intracranial bleeding.

References
1 Berger CW, Crosby ET, Grodecki W. North American survey of
the management of dural puncture occurring during labour
epidural analgesia. Can J Anaesth 1998; 45: 110–4
2 Diemunsch P, Balabaud VP, Petiau C, Marescaux C, Muller A.
Bilateral subdural haematoma following epidural anaesthesia. Can
J Anaesth 1998; 45: 328–31
3 Skoldefors EK, Olofsson CI. Intracranial subdural haematoma
complicates accidental dural tap during labour. Eur J Obstet Gynecol
Reprod Biol 1998; 81: 119–21
4 Pavlin DJ, McDonald JS, Child B, Rusch B. Acute subdural
haematoma—an unusual sequel to lumbar puncture. Anesthesiology
1979; 51: 338–40
5 Jack TM. Post partum intracranial haematoma. A possible
complication of epidural analgesia. Anaesthesia 1979; 34: 176–80
6 Russell IF. Post partum neurological problems. In: Russell IF, Lyons
G, eds. Clinical Problems in Obstetric Anaesthesia. London: Chapman
and Hall Medical, 1997; 149–60
7 Campbell DA, Varma TRK. Chronic subdural haematoma after
lumbar subarachnoid analgesia for caesarean section, presenting
as puerperal psychosis. Br J Obstet Gynaecol 1993; 100: 782–4
8 Roscoe MWA, Barrington TW. Acute spinal subdural haematoma:
a case report and review of the literature. Spine 1984; 9: 672–5
9 Wyble SW, Bayhi D, Webre D, Viswanathan S. Bilateral subdural
haematomas after dural puncture: delayed diagnosis after false
negative CT scan without contrast. Reg Anesth 1992; 17: 52–3
10 Bromage PR. Neurologic complications of regional anesthesia in
obstetrics. In: van Zundert A, Ostheimer GW, eds. Pain Relief
and Anesthesia in Obstetrics. New York: Churchill Livingstone 1996;
993–1004
11 Robinson JL, Hall CS. Arteriovenous malformations and
aneurysms. J Neurosurg 1974; 41: 63–70
12 Yih PS, Cheong KF. Anaesthesia for caesarean section in a patient
with an intracranial arteriovenous malformation. Anaesth Intensive
Care 1999; 27: 66–8
13 Ogilvy CS, Barker FG. Head injury. In: Bogousslavsky J, Fisher M,
eds. Textbook of Neurology. Boston: Butterworth–Heinemann 1998;
711–25
14 Reynolds F. Dural puncture and headache. Br Med J 1993; 306:
874–6

520