International Journal of Obstetric Anesthesia (2004) 13, 110–113

2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.ijoa.2003.10.005

CASE REPORT

Anaesthetic dilemma: spinal anaesthesia in

an eclamptic patient with mild thrombocytopenia
and an “impossible” airway
O. O. Nafiu, R. A. Salam, E. O. Elegbe
Department of Anaesthesia, Korle Bu Teaching Hospital, Accra, Ghana

SUMMARY. We present our anaesthetic management of a 27-year-old woman with antepartum eclampsia, mild
thrombocytopenia, difficult airway and clinical evidence of impending upper airway obstruction. She required urgent
delivery by caesarean section, which was conducted uneventfully under spinal anaesthesia. We discuss the man-
agement conundrums presented by this case and why we chose spinal anaesthesia over other anaesthetic options.

2004 Elsevier Ltd. All rights reserved.

INTRODUCTION section in an eclamptic, thrombocytopenic parturient at

the Korle Bu Teaching Hospital Accra, Ghana.
The incidence of airway-related problems with the at-
tendant risk of morbidity and mortality is much higher in
obstetrics than in the general surgical patient.1 Patients CASE REPORT
with eclampsia have a high incidence of instrumental
deliveries and the anaesthetist is often called to provide A 27-year-old obese, unbooked, primigravida was ad-
analgesia/anaesthesia for these procedures.2 The choice mitted at 36 weeks’ gestation with a history of recurrent,
of anaesthesia for caesarean section in patients with pre- generalised tonic-clonic seizures. Pregnancy was pre-
eclampsia/eclampsia remains controversial.3 The disad- sumably uneventful until the day before presentation
vantages of general anaesthesia in these patients have when she complained of headaches and upper abdominal
been reviewed extensively.4;5 Some anaesthetists also pain. These were not relieved by local herbal ointments
consider regional anaesthesia unsafe for a variety of rea- and other oral concoctions. She was given some para-
sons.6 Choosing between spinal or epidural blockade, cetamol and antimalarials with minimal relief. She also
however, is more controversial. Epidural blockade is of- vomited twice. A few hours before presentation, she
ten preferred by anaesthetists because it provides greater complained of “seeing things” and shortly after had the
maternal haemodynamic stability.7 Another important first episode of generalised tonic-clonic seizures. There
consideration is the risk of spinal haematoma.8 While in was no past history of epilepsy and no symptoms to
most cases the risks of regional can be weighed against suggest acute intracranial infection. She subsequently
those of general anaesthesia, occasionally the anaesthetist had four more episodes of seizures each lasting ap-
is presented with a patient in whom the available anaes- proximately 20 min. She was observed to be bleeding
thetic options are severely limited. We present our di- from the mouth having bitten her tongue during the
lemma in providing anaesthesia for emergency caesarean second bout of fits. Delay in presentation to the hospital
was attributed to unavailability of transportation.
Clinical examination revealed an acutely ill, drowsy
patient with generalised oedema, dyspnoea and inspi-
ratory stridor. She had a grossly swollen, woody hard,
––––––––––––––––––––––––––––––––––––––––––––––––––––––––– erythematous tongue that filled the entire oral cavity as
Accepted October 2003 well as protruding out of the mouth, preventing mouth
closure (Figs. 1 & 2). There was associated bleeding
Correspondence to: O.O. Nafiu, MD FRCA, Department of from the mouth and nostrils. Mallampati assessment was
Anesthesiology, University of Michigan, Room UH 1H247,
1500 East Medical Drive, Ann Arbor, MI 48109- 0048, USA. impossible but her neck was noted to be thick and obese.
Fax: +1-734-9369091; E-mail: [email protected] She had full-volume peripheral pulses and a blood
110
 Thrombocytopenia and an impossible airway 111

emergency caesarean section was planned and anaes-

thetic assessment was requested. An urgent full blood
count, coagulation profile and serum biochemistry were
requested. All were within normal limits apart from a
haematocrit of 27% and a platelet count of 85  109 /L.
Following review by the anaesthetic consultant, a
decision was made to proceed with spinal anaesthesia.
The possible risk of epidural/spinal haematoma was
discussed with the husband and other relatives as the
patient was too drowsy to give informed consent. We
obtained consent from the husband to photograph the
patient for educational purposes.
The patient was transferred to the operating theatre
and the circulation preloaded with Ringer’s lactate so-
lution 500 mL. Subarachnoid block was established with
the patient supported in the sitting position. A 26-gauge
Quincke needle with the bevel turned laterally was used
and once free-flow of cerebrospinal was obtained, 2.5
mL of 0.5% hyperbaric bupivacaine was injected. The
patient was turned into the left lateral, wedged, semi-
recumbent position. The height of block to pinprick was
T6.
Fig. 1 Showing the swollen tongue filling the entire oral cavity,
making Mallampati assessment impossible.
The blood pressure was monitored non-invasively
every 2 min for the first 10 min and thereafter every 5
min until the end of surgery. The pre-induction blood
pressure was 170/100 mmHg. Following induction of
subarachnoid block, the systolic blood pressure dropped
to 100 mmHg for which ephedrine 3 mg i.v. and in-
creased i.v. fluids were given. Thereafter, the blood
pressure stabilised between 150/95 and 130/80 mmHg
for the rest of the surgery.
A live female baby was delivered by caesarean sec-
tion with Apgar scores of 5 at 1 min and 9 at 5 min. The
surgery lasted 45 min and the estimated blood loss
was 500 mL. The total intra-operative fluid volume was
1.5 L and the urine output 250 mL. The abdominal
wound was infiltrated with 0.5% bupivacaine and the
mother was given rectal paracetamol 1 g at the com-
Fig. 2 Side view of same patient showing tongue and circum-oral pletion of surgery and six-hourly thereafter. We did not
swelling and her short neck.
administer any postoperative opioids as we were still
concerned about sedatives and her compromised airway.
pressure of 190/120 mmHg. She was not in heart failure The patient made a steady postoperative recovery
and her lung fields were clear. She had global hypertonia with no further seizures and the swollen tongue com-
and hyperreflexia with sustained ankle clonus. Visual pletely subsiding by the third day. She suffered no
field assessment was impossible as the patient was too morbidity from the spinal anaesthetic and was dis-
drowsy to co-operate, although she had normal direct charged home on the seventh postoperative day with
and consensual pupillary light response. She also had 3+ normal blood pressure and a healthy baby.
proteinuria. She was not in labour. The fetal heart rate
was 150 beats/min.
While the patient was being admitted into the labour DISCUSSION
ward, she had another episode of convulsions which
responded to intravenous diazepam and magnesium Eclampsia remains an important cause of maternal and
sulphate. Intravenous hydralazine 5 mg was given to be perinatal morbidity and mortality in developing coun-
followed by 5–10 mg p.r.n. for blood pressure control. tries.9 As in the present case, poor antenatal supervision
The patient was given oxygen via nasal cannula, an may delay referral and multiple seizures before hospital
112 International Journal of Obstetric Anesthesia

admission are common presenting features in eclamptic not think we would be able to achieve adequate topical
patients in developing countries. The parturient with anaesthesia of her airway. Inadequate airway anaesthe-
severe preeclampsia/eclampsia presents a significant sia in a drowsy patient could lead to a severe hyper-
challenge to the anaesthetist. The presence of labile tensive response, possible laryngeal spasm and total loss
hypertension, depleted intravascular volume and multi- of airway control. We discussed the possibility of awake
ple organ dysfunction requires a technique that favours tracheostomy (option 5) with the ear, nose and throat
haemodynamic stability.4 Also, seizures (isolated or surgeons but they were rightly worried about potential
recurrent) are a constant threat to the airway. for airway haemorrhage and the possibility of severe
This patient’s mode of presentation (unbooked pri- hypertensive response to airway manipulation. They did
migravida, recurrent antepartum fits, and futile attempts however agree to stand by should there be a need for
at home therapy with consequent delay in hospital urgent tracheostomy. The technique of infiltration an-
presentation) is unfortunately typical of eclampsia in aesthesia is one we were reasonably familiar with, but
developing countries.9 The challenge for anaesthetists we usually use it in combination with neuroleptanalge-
working in this environment is daunting as they quite sia. Clearly, we wanted to avoid all types of sedation in
often have to manage patients presenting in extremis this patient.
with the most basic of resources. Epidural anaesthesia is generally preferred by most
As this patient would have to be delivered by emer- anaesthetists for pain relief in labour as well as for op-
gency caesarean section, a number of adverse factors erative delivery in patients with preeclampsia/eclamp-
were of concern to us: sia.3 It is generally associated with milder maternal
haemodynamic and neuroendocrine perturbation than
• Clear evidence of upper airway compromise (shown
general anaesthesia during elective caesarean section. In
by dyspnoea and stridor)
a retrospective study, Moodley et al.5 found that ma-
• Obviously difficult airway (Figs. 1 & 2) with poten-
ternal and neonatal outcome were not affected adversely
tial for total loss of control
by epidural compared with general anaesthesia in se-
• Recurrent seizures and therefore a need to expedite
lected cases of eclampsia. However, in the setting of
delivery
mild thrombocytopenia, we reasoned that a 16-gauge
• Mild thrombocytopenia in a patient with poorly in-
epidural needle would produce greater tissue trauma
vestigated coagulation status
(with potential for epidural haematoma) than a 26-gauge
• Full stomach: home administration of “native medi-
spinal needle. Additionally, we only had facilities for
cations”
single-shot epidural block in our unit at the time and we
• Poorly controlled hypertension.
considered this dangerous for the following reasons:
Our dilemma was how best to provide safe anaes- possibility of inadvertent dural puncture with attendant
thesia without further compromising her airway and headache, total spinal and loss of airway control and,
maintaining cardiovascular stability. We considered the perhaps more worrying, the potential for coning fol-
following anaesthetic options and weighed the risks and lowing dural tap with a 16-gauge needle in a patient with
benefits of each: possibly elevated intracranial pressure.6
We decided on spinal anaesthesia with 2.5 mL of
1. Epidural anaesthesia
hyperbaric 0.5% bupivacaine because this was a familiar
2. Spinal anaesthesia
technique and also we considered the risk of high spinal
3. Field block for caesarean section
block quite small with this technique. Many obstetric
4. Awake oral or nasal fibreoptic intubation followed by
anaesthetists avoid spinal anaesthesia fearing that sud-
general anaesthesia
den, extensive sympatholysis may cause severe hypo-
5. Awake tracheostomy under local anaesthesia fol-
tension. This may result in the need for rapid infusion of
lowed by general anaesthesia.
large volumes of intravenous fluid with potential for
6. General anaesthesia with rapid sequence induction.
pulmonary oedema.10 The use of vasopressors to correct
None of these options was entirely without risk to this hypotension in these patients may also be hazardous
patient, but we were able to eliminate the ones we felt because of possible enhanced sensitivity to these
were downright dangerous. We immediately rejected agents.11 We are not aware of any study comparing the
option 6 as it was clear to us that we would not be able efficacy of fluids vs. vasopressors in patients with severe
to maintain the patient’s airway with the induction of preeclampsia/eclampsia although we feel that judicious
general anaesthesia. Option 4 was considered purely for use of both fluids and pressors would be advisable in
academic reasons as we had neither the expertise nor the these patients. Our patient had a single episode of hy-
equipment for fibreoptic intubation in our department. potension which responded promptly to intravenous
We wondered, however, about the wisdom of instru- fluid and ephedrine 3 mg with no cardiopulmonary
menting a potentially haemorrhagic airway. Also we did sequelae.
 Thrombocytopenia and an impossible airway 113

We were concerned about performing a neuraxial rient. We believe we chose the best technique for this
block in the setting of mild thrombocytopenia and patient given our limited resources and the paucity of
incompletely screened coagulation status. The lowest options presented by the complex nature of the case.
acceptable platelet count for safe central neuraxial block
is unclear. Also there appears to be no consensus on the
REFERENCES
reliability of platelet count in predicting the risk of
epidural haematoma. The recommendation that neurax- 1. Shnider S M, Levinson G. Anesthesia for cesarean section. In:
Shnider S M, Levinson G (eds). Anesthesia for obstetrics.
ial anaesthesia is contraindicated in the presence of a Baltimore: Williams & Wilkins, 1989: 159–178.
platelet count less than 100  109 /L has no data to 2. Gofton E N, Capewell V, Natale R, Gratton R J. Obstetrical
support it and has being challenged.12 The reviewer intervention rates and maternal and neonatal outcomes in women
with gestational hypertension. Am J Obstet Gynecol 2001; 185:
recommends that a healthy parturient with a platelet 798–803.
count >75  109 /L can receive regional anaesthesia. 3. Howell P. Spinal anaesthesia in severe preeclampsia: time for
However, the patient with eclampsia is not a healthy reappraisal or time for caution? International Journal of Obstetric
Anesthesia 1998; 7: 217–219.
customer and platelet dysfunction is a well recognised 4. Wallace D H, Leveno K J, Cunningham F G, Giesecke A H,
feature of severe preeclampsia/eclampsia. We agree Shearer V E, Sidawi J E. Randomized comparison of general and
with the recommendation that the risk/benefit ratio of regional anesthesia for Cesarean delivery in pregnancies
complicated by severe preeclampsia. Obstet Gynecol 1995; 86:
regional anaesthesia should be carefully assessed in 193–199.
every patient with thrombocytopenia and where appli- 5. Moodley J, Jjuuko G, Rout C. Epidural compared with general
cable the most experienced anaesthetist should perform anaesthesia for caesarean delivery in women with eclampsia.
the block. Br J Obstet Gynaecol 2001; 108: 378–382.
6. Richards A M, Moodley J, Graham D I, Bullock M R R. Active
We also considered the thorny issue of “informed management of the unconscious eclamptic patient. Br J Obstet
consent” in a drowsy, acutely ill patient, but concluded Gynaecol 1986; 93: 554–562.
that we could only rely on implied consent given by the 7. Ramanathan J, Coleman P, Sibai B M. Anesthetic modification of
hemodynamic and neuroendocrine responses to cesarean delivery
husband and other family members. In an ideal world in women with severe preeclampsia. Anesth Analg 1991; 73:
anaesthetists would provide care with clinical decisions 772–779.
based on controlled randomised clinical trials and pa- 8. Yuen T S T, Kua J S W, Tan I K S. Spinal haematoma following
epidural anaesthesia in a patient with eclampsia. Anaesthesia 1999;
tients would be able to give prospective consent for such 54: 350–371.
therapy. However, we do not live in an ideal world; we 9. Jamelle R N. Eclampsia — a taxing situation in the third world. Int
often provide care in emergency settings when treatment J Gynecol Obstet 1997; 58: 311–312.
10. Sibai B M, Mabie B C, Harvey C J, Gonzalez A R. Pulmonary
must be started before any meaningful discussion with edema in severe preeclampsia-eclampsia: analysis of 37
the patient or the next of kin is possible. We agree with consecutive cases. Am J Obstet Gynecol 1987; 156: 1174–1179.
Balser et al.13 that in some emergency situations, the 11. Talledo O, Chesley L C, Zuspan F P. Renin-angiotensin system in
concept of “deferred consent” may be the only reason- normal and toxemic pregnancies: Differential sensitivity to
angiotensin II and norepinephrine in toxemia of pregnancy.
able approach. To have waited until she was able to give Am J Obstet Gynecol 1968; 100: 218–221.
consent for regional anaesthesia in this setting we felt 12. Douglas M J. Platelets, the parturient and regional anesthesia.
was not in her best interest. International Journal of Obstetric Anesthesia 2001; 10: 113–120.
13. Balser J R, Martinez E A, Winters B D et al. Beta-adrenergic
Each case is unique and it is impossible to prescribe blockade accelerates conversion of postoperative tachyarrhythmias.
the best anaesthetic method for every eclamptic partu- Anesthesiology 1998; 89: 1052–1059.