LSCS in a patient with Eisenmenger's syndrome

performed within 24 h of the onset of the headache. The References

mother is reviewed daily until discharge and is advised, that
if she experiences any further headaches or unexplained
symptoms, to return to the Obstetric Day Assessment Unit
or the Labour Ward where she will be seen by a consultant
anaesthetist. A copy of the patient's discharge summary is
sent to the general practitioner and community midwife,
who visits daily for 10 days. Subsequent to this patient, the
consultant anaesthetist now writes a discharge letter to the
general practitioner providing information about the dural
puncture and its management, and advising that the mother
be referred back to the Obstetric Day Assessment Unit in the
event of further complications. A recent study, which
highlighted the poor understanding of post-dural puncture
headache amongst general practitioners, prompted the
authors' to design pamphlets for general practitioners and
patients with the intention of improving the early recogni-
tion and management of post-dural puncture headache.8
Failure to recognize these rare cases of subdural
haematoma can have permanent and fatal consequences.9 10
Therefore, in the puerperium, it is crucial to investigate
persistent or recurrent headache, particularly those associ-
ated with neurological signs, and a CT or MRI scan should
be performed as appropriate.
Whilst an epidural blood patch usually provides almost
instantaneous relief for a post-dural puncture headache, its
longer-term efcacy is probably only 6070%.11 12 This
case suggests that an epidural blood patch, contrary to
popular belief, may not provide protection against the more
devastating complications of a dural puncture and in
addition highlights the ongoing responsibility anaesthetists
have to mothers who suffer an accidental dural puncture.
1 Vos PE, de Boer WA, Wurzer JA, van Gjin J. Subdural hematoma
after lumbar puncture: two case reports and review of the
literature. Clin Neurol Neurosurg 1991; 93: 12732
2 Reynolds F. Dural puncture and headache: avoid the rst but
treat the second. BMJ 1993; 306: 8746
3 Stocks GM, Wooller DJA, Young JM, Fernando R. Postpartum
headache after epidural blood patch: investigation and diagnosis.
Br J Anaesth 2000; 84: 40710
4 Brownridge P. The management of headache following accidental
dural puncture in obstetric patients. Anaesth Intens Care 1983;
11: 415
5 Kunkle EC, Ray BS, Wolff HG. Experimental studies on
headache: analysis of the headache associated with changes in
intracranial pressure. Arch Neurol 1949; 49: 323
6 Gormley JB. Treatment of post-spinal headache. Anesthesiology
1960; 21: 5656
7 Diemunsch P, Balabaud VP, Petiac C, et al. Hematome sous dural
bilateral apres analgesie peridurale. Can J Anaesth 1998; 45:
32831
8 Schneider GT, Price CM, Thornberry EA. Postdural puncture
headache (PDPH): easily accessible information for the general
practitioner. Int J Obstet Anesth 1997; 6: 208
9 Edelman JD, Wingard DW. Subdural hematomas after lumbar
dural puncture. Anesthesiology 1980; 52: 1667
10 Newrick P, Read D. Subdural haematoma as a complication of
spinal anaesthesia. BMJ 1982; 285: 3412
11 Williams EJ, Beaulieu P, Fawcett WJ, Jenkins JG. Efcacy of
epidural blood patch in the obstetric population. Int J Obstet
Anesth 1999; 8: 1059
12 Taivainen T, Pitkanen M, Tuominen M, Rosenberg PH. Efcacy of
epidural blood patch for postdural puncture headache. Acta
Anaesthesiol Scand 1993; 37: 7025

British Journal of Anaesthesia 86 (5): 7236 (2001)

Incremental spinal anaesthesia for elective Caesarean section in a

patient with Eisenmenger's syndrome
P. J. Cole*, M. H. Cross and M. Dresner

Department of Anaesthesia, Leeds General Inrmary, Great George Street, Leeds LS1 3EX, UK
*Corresponding author

We describe a new approach to anaesthesia for elective Caesarean section in a woman with
Eisenmenger's syndrome. Incremental regional anaesthesia was performed using a microspinal
catheter and haemodynamic monitoring included transthoracic bioimpedance cardiography.
This approach allowed the disadvantages of general anaesthesia and invasive cardiac output
monitoring to be avoided.
Br J Anaesth 2001; 86: 7236
Keywords: complications, Eisenmenger's syndrome; anaesthesia, obstetric; anaesthetic
techniques, subarachnoid
Accepted for publication: January 4, 2001

The Board of Management and Trustees of the British Journal of Anaesthesia 2001
 Cole et al.
Eisenmenger's syndrome includes any condition in which a
communication between the systemic and pulmonary
circulations gives rise to pulmonary vascular disease,
which in turn causes a right to left shunt. It is a rare
condition that poses a signicant risk of maternal death,
with the mortality of between 30 and 70%, having changed
little over the last 50 yr.14 Eighty per cent of deaths occur
between the 2nd and 30th post-natal day,1 but it remains
unclear whether the choice of anaesthetic technique inu-
ences the outcome. The primary anaesthetic goal is to avoid
any haemodynamic change that might increase the right to
left shunt and thereby increase hypoxaemia. With this in
mind, many practitioners have avoided regional anaesthetic
techniques in favour of general anaesthesia. We describe a
regional technique that, to our knowledge, has not been
previously reported in this context.
Case report
A 28-yr-old woman with complex cyanotic heart disease
was referred at 27 weeks gestation to our Antenatal
Anaesthetic Clinic. Her cardiac pathology had been classi-
ed as a form of Eisenmenger's syndrome and consisted of
mitral atresia, a single ventricle, atrial septal defect,
transposition of great arteries with pulmonary hypertension,
and subsequent pulmonary vascular obstructive disease. Her
condition had not been considered amenable to surgery,
other than by heart/lung transplant. This was contra-
indicated by her marked thoracic scoliolis, which had
resulted in signicant restrictive respiratory disease. Her
medical management consisted of nifedipine and frusemide,
and she underwent regular venesection for secondary
polycythaemia to maintain haemoglobin at approximately
15 g dl1. Despite counselling and advice that pregnancy
carried a signicant risk of death, she was determined to
have children. After four miscarriages, her fth pregnancy
had progressed without signicant problems apart from
some fetal growth retardation.
Her condition at 27 weeks gestation had deteriorated from
her non-pregnant state, with central cyanosis and dyspnoea,
worsening with speech. At that time, the obstetricians had
made a provisional plan to perform an elective Caesarean
section (LSCS) at 35 weeks, so the anaesthetic choices and
risks were explained to the patient and her family at an
Anaesthetic Antenatal Clinic consultation. A formal man-
agement plan was reserved until nearer the delivery date in
order to take changes in her condition into consideration.
She was subsequently admitted at 29 weeks because of
increasing dyspnoea, and her diuretic and thromboprophy-
lactic treatments were optimized. Because of continuing
dyspnoea and poor fetal growth, it was decided to perform
the LSCS at 32 weeks.
In the anaesthetic room, the patient was dyspnoeic with
an oxygen saturation of 77% in air and an arterial pressure
of 120/60. Peripheral venous and radial arterial lines were
sited followed by a right internal jugular triple lumen
724
catheter that revealed a central venous pressure (CVP) of 13
mm Hg. Meticulous attention was paid to the avoidance of
bubbles in lines and syringes because of the risk of
paradoxical embolus.
With the patient in the sitting position, a 32-gauge
Ruschke catheter was inserted through a 24-gauge Sprotte
spinal needle at the third to fourth lumbar spinal interspace.
In order to avoid maldistribution of the injected solution, the
catheter was pushed only 23 cm past the needle tip. Before
giving any drugs through the catheter, the patient was placed
in the supine position with left lateral tilt, and oxygen
therapy was instituted via a Hudson mask. This increased
the oxygen saturation from 77 to 80%. Non-invasive cardiac
output monitoring was then established using the transthor-
acic bioimpedence method (NCCOM3, BoMED
Manufacturing, Irvine, CA, USA), indicating a cardiac
output of 3.0 litre min1. Antibiotic prophylaxis was
administered. No formal uid preload was performed, but
a slow infusion of 0.9% saline was commenced.
After rst giving diamorphine 300 mg through the spinal
catheter, incremental doses of 0.25 ml 0.5% hyperbaric
bupivacaine were titrated against anaesthetic and haemo-
dynamic effects. Over a period of 30 min, a total of 2.25 ml
produced a block height to the eighth thoracic dermatome
tested to touch with blunt forceps. During this time
oxygenation remained unchanged, arterial pressure fell by
10 mm Hg, cardiac output rose to 3.7 litre min1, and CVP
gradually fell to 6 mm Hg. This latter change was corrected
with a 300 ml bolus of saline. No maternal symptoms or
fetal distress occurred.
The LSCS proceeded uneventfully without pain or
discomfort and no further drugs were administered through
the spinal catheter. The usual bolus of oxytocin was omitted
in favour of uterine massage and a slow oxytocin infusion.
This produced adequate uterine contraction and blood loss
was minimal. During uterine repair, an episode of dizziness
coincided with a decrease in arterial pressure to 93/50.
A bolus of ephedrine 3 mg restored the arterial pressure,
increased the cardiac output to its highest level of 4.9 litre
min1, and resolved the dizziness. This was the largest
haemodynamic change during the whole procedure. A total
of 1500 saline was given. Haemodynamics and oxygenation
in air were almost identical to pre-operative values at the
completion of surgery.
Post-operative analgesia consisted of rectal diclofenac
100 mg at the end of surgery, followed by regular 8 hourly
doses of 50 mg supplemented by a codeine/paracetamol
combination on request. The patient was transferred to the
Obstetric High Dependency Unit from where, after 36 h, she
was transferred to a routine post-natal ward. At 48-h follow-
up, she gave maximal scores for peri-operative comfort and
overall satisfaction, experiencing no nausea, vomiting, or
spinal headache. After 10 days she was discharged from
hospital, and 1 yr later both the patient and child are doing
well. Ignoring medical advice, she is now planning another
pregnancy.
 LSCS in a patient with Eisenmenger's syndrome
Discussion
The exact mortality for Eisenmenger's syndrome in preg-
nancy is unknown, as cases that are unsuccessful are often
not reported in the literature. Two recent reviews found the
mortality was found to be 36 and 40%1 3 whilst another
quoted a gure of 70%.4 Pregnancy prevention or early
termination of pregnancy is the preferred measure for
improving long term survival in women of childbearing age.
Despite this, 80% of women with Eisenmenger's syndrome,
who are pregnant, have been given the diagnosis before the
pregnancy.1
Whatever anaesthetic technique is chosen the principle
remains the same. The cardiac output must be maintained
and the systemic vascular resistance (SVR) must not be
allowed to fall. This should ensure that there is minimal
change in the amount of right to left shunt.
Several factors affected our anaesthetic plan. First, the
patient wanted to be awake at the time of delivery with her
partner present. Second, whilst traditional teaching has been
that general anaesthesia is to be preferred to a regional
technique we felt that general anaesthesia posed clear risks
and disadvantages. These included the potential for increas-
ing pulmonary vascular resistance in response to catecho-
lamine release after laryngoscopy, during anaesthesia and in
recovery given the relatively poor pain control achievable
with systemic opioids. Intermittent positive pressure venti-
lation increases intrathoracic pressure, reduces venous
return and increases pulmonary arterial pressure. This
would in turn increase the right to left shunt in this patient.
Also, general anaesthesia has the potential to exacerbate her
respiratory disease, thereby increasing post-operative hy-
poxia. These hazards are avoided by regional anaesthesia,
although the level of block required using a regional
technique might produce excessive sympathetic block and
an uncontrolled decrease in the SVR.2 Epidural anaesthesia
has been used successfully in this condition57 and, because
of its slow onset, this technique reduces the chances of
precipitous haemodynamic changes. However, epidural
anaesthesia can produce patchy or incomplete sensory
block, which may result in undesirable sympathetic stimu-
lation, or the need to convert to general anaesthesia. In
addition, the large amount of local anaesthetic that is
required when an epidural is used may result in blood
concentrations high enough to cause myocardial depression
in such a vulnerable patient. Spinal anaesthesia is, we
believe, more reliable, but a single shot approach is too
haemodynamically unstable.
Our experience with incremental spinal anaesthesia using
spinal catheters8 gave us condence that haemodynamic
stability could be maintained, particularly given that
aortocaval compression would be less of a problem at 32
weeks with a growth retarded fetus.
We proceeded with the LSCS with a block height to the
8th thoracic dermatome (T8) tested to touch with blunt
forceps. A study investigating the block height required for
725
Caesarean section, concluded that absent touch sensation to
the level of T4 was necessary when using a solution of plain
local anaesthetic.9 This study needs to be repeated with
spinals using lipid soluble opioids as our audit results have
shown T6 to be adequate in the presence of diamorphine 300
mg. We would not normally recommend proceeding with a
block as low as T8, but in this case we were anxious to
minimize the sympathetic block and were condent the
level could be rapidly increased if pain ensued.
Our choice of monitoring requires some comment. Pulse
oximetry was the most practical way of continuously
assessing the degree of right to left shunt during the peri-
operative period. Invasive arterial pressure monitoring was
considered mandatory simply because of the immediacy of
the information provided. The assessment of cardiac lling
and output was not so straightforward given the unusual
anatomy of the heart. CVP may not have correlated well
with left ventricular end diastolic pressure, but central
access was certainly required in case of the need for
resuscitation. In fact, initial readings and waveforms were
entirely appropriate for the patient's clinical condition, and
subsequent changes in CVP followed the clinical picture.
Cardiac output monitoring was not considered of critical
importance, given that we expected the heart to be of
relatively xed, low output. However, the BoMED, with
which we have extensive experience, is a simple non-
invasive method for cardiac output monitoring that is
quicker, less technically demanding,10 and more comfort-
able than Doppler echocardiography. Once again, this
monitor gave information and trends that tted the clinical
picture. The anatomical anomolies in this patient (mitral
atresia and transposition of the great arteries) would have
rendered the pressures measured with a pulmonary artery
otation catheterization (PAFC) of doubtful or potentially
misleading values. Thus, the decision to avoid the potential
hazards of a PAFC was justied.
We chose not to give oxytocin as a bolus because it
causes direct vasodilatation and reduces SVR with a
compensatory increase in heart rate and cardiac output in
rst trimester pregnant women.11 12 We have conrmed
these ndings in healthy term pregnant women under spinal
anaesthesia in our own as yet unpublished work. Signicant
hypotension after 10 units oxytocin was prevented by a
mean increase in cardiac output of 80% measured with the
BoMED. It was our assumption that our patient would be
unable to mount such a response. Indeed, in the survey by
Weiss and colleagues1 two of the patients who had a
complicated course after systemic hypotension did so after
oxytocin was administered. The policy of using uterine
massage followed by a slow oxytocin infusion proved safe
and effective in this case.
Whether regional or general anaesthesia is used the
importance of maintaining SVR has already been empha-
sized. The prophylactic and therapeutic use of vasocon-
strictor drugs, therefore, seems logical and attractive.
Ephedrine is the most commonly used vasoconstrictor in
 Cole et al.
obstetric practice, but the tachycardia it can produce would
be undesirable in a patient with Eisenmenger's syndrome.
Whilst a norepinephrine infusion has been documented to
maintain SVR throughout the peri-operative period,13 an
excessive dose of metaraminol has caused near catas-
trophe,14 presumably by increasing pulmonary vascular
resistance and increasing the degree of right to left shunt.
Directly acting vasoconstrictors also have the potential to
compromise placental perfusion. These agents, therefore,
need to be used with caution. It was decided to proceed
initially with our familiar agent, ephedrine, and change to a
direct acting vasoconstrictor if necessary. In the end, by
using an incremental spinal approach to anaesthesia, only 3
mg of ephedrine was required during the procedure and
post-operative period.
Although there is general agreement that patients with
signicant cardiac disease should have their obstetric care at
a centre where expert cardiological and cardiac surgical
help is available, debate exists about whether elective LSCS
should occur on the labour ward or in the cardiac operating
theatre. Successful conversion of a planned cardiac oper-
ation during the third trimester to a combined emergency
LSCS and cardiac operation is well documented. With the
advent of improved neonatal care, physicians managing
these patients are tending to move to an elective combined
approach if patients decompensate during the third trimester
with a surgically correctable lesion.
However, it is particularly uncommon for patients with
signicant cardiac disease, undergoing an elective LSCS
(without concomitant cardiac surgery), to require emer-
gency cardiac surgery. For this reason it is probably safe for
the operation to be conducted on the labour ward.15 The
theatre staff on the labour ward may be unfamiliar with
some of the monitoring required and with some of the drugs
used and for this reason we may opt to carry out the LSCS in
a cardiac operating theatre. This is despite the knowledge
that cardiac surgery will not be an option and that the
obstetric and neonatal staff are working in a strange
environment. The risks and benets of each operating
theatre are always considered before making a nal decision
about the place of delivery.
The FDA withdrew approval to use small-bore catheters
for continuous spinal anaesthesia in April of 1992 after
several case reports of cauda equina syndrome.16 The view
in the UK was that this problem was because of the pooling
of excessive doses of hyperbaric 5% lidocaine rather than
the catheters per se. Although microspinal catheters were
never withdrawn in the UK, the dwindling world market led
to manufacturers abandoning their production. The Ruschke
kit used for this case report is, therefore, no longer available.
Larger catheters may introduce an unacceptable incidence
of post-dural puncture headache, as they have to be passed
726
through an even larger spinal needle. An alternative is the
Braun `catheter-over-needle' system17 where a 22-gauge
spinal catheter covers a 27-gauge Quinke spinal needle with
the sharp point outside the catheter.
References
1 Weiss BM, Zemp L, Seifert B, Hess OM. Outcome of pulmonary
vascular disease in pregnancy: a systematic overview from 1978
through 1996. J Am Coll Cardiol 1998; 31: 16507
2 Foster JM, Jones RM. The anaesthetic management of the
Eisenmenger syndrome. Ann R Coll Surg Engl 1984; 66: 3535
3 Yentis SM, Steer PJ, Plaat F. Eisenmenger's syndrome in
pregnancy: maternal and fetal mortality in the 1990s. Br J
Obstet Gynaecol 1998; 105: 9212
4 Ramin SM, Maberry MC, Gilstrap LC, 3rd. Congenital heart
disease. Clin Obstet Gynecol 1989; 32: 417
5 Spinnato JA, Kraynack BJ, Cooper MW. Eisenmenger's syndrome
in pregnancy: epidural anaesthesia for elective Cesarean. N Engl J
Med 1981; 304: 12157
6 Rosenberg B, Simon K, Peretz BA, Roguin N, Birkhahn HJ.
Eisenmenger's syndrome in pregnancy: controlled segmental
epidural block for Cesarean section. Reg Anaesth 1984; 7: 1313
7 Tibaldi G, Marchi L, Huscher M, Forlini G. Anaesthesia for
Cesarean section in a pregnant woman with Eisenmenger's
syndrome: description of a clinical case. Minerva Ginecol 1988; 40:
1456
8 Dresner M, Maclean A. Anaesthesia for Caesarean section in a
patient with Klippel-Feil syndrome. The use of a microspinal
catheter. Anaesthesia 1995; 50: 8079
9 Russell IF. Levels of anaesthesia and intraoperative pain at
Caesarean section under regional block. Int J Obstet Anesth 1996;
5: 2369
10 Northridge DB, Findlay IN, Wilson J, Henderson E, Dargie HJ.
Non-invasive determination of cardiac output by Doppler
echocardiography and electrical bioimpedence. Br Heart J 1990;
63: 937
11 Weis FR, jnr, Markello R, Mo B, Bochiechio P. Cardiovascular
effects of oxytocin. Obstet Gynecol 1975; 46: 2114
12 Secher N, Arnsbo P, Wallin L. Haemodynamic effects of oxytocin
and methyl ergometrine on the systemic and pulmonary
circulations of pregnant anaesthetized women. Acta Obstet
Gynecol Scand 1978; 57: 97103
13 Sammut MS, Paes ML. Anaesthesia for laparoscopic
cholecystectomy in a patient with Eisenmenger's syndrome. Br
J Anaesth 1997; 79: 8102
14 Bird TM, Strunin L. Anaesthesia for a patient with Down's
syndrome and Eisenmenger's complex. Anaesthesia 1984; 39:
4850
15 Parry AJ, Westerby S. Cardiopulmonary bypass during
pregnancy. Ann Thor Surg 1996; 61: 18659
16 Rigler M, Drasner K, Krejcie TC, et al. Cauda equina syndrome
after continuous spinal anaesthesia. Anesth Analg 1991; 92:
27581
17 Mollmann M, Van Steenberge A, Sell A, et al. Spinocath, a new
approach to continous spinal anaesthesiapreliminary results of
a multicenter trial. Int Monitor 1996; 8: 74