32 Obstetric anaesthesia and analgesia

Analgesia for labour 738

Regional labour analgesia 739

Epidural analgesia for labour 740

The poorly functioning epidural 744

Complications of epidural analgesia 746

Dural puncture 748

Remifentanil for labour analgesia 752

Caesarean section 754

Caesarean section: epidural 756

Caesarean section: spinal 758

Caesarean section: combined spinal/epidural (CSE) 760

Inadequate anaesthesia 762

Hypotension 763

Caesarean section: general anaesthesia 764

Antacid prophylaxis 766

Postoperative analgesia 767

Retained placenta 768

Breast feeding and drug transfer 770

Placenta praevia and accreta 772

Massive obstetric haemorrhage 774

Amniotic fluid embolism 777

Cardiac disease and pregnancy 778

Maternal resuscitation 780

Pregnancy-induced hypertension, pre-eclampsia, and eclampsia 782

Eclampsia 787

HELLP syndrome 788

Surgery during pregnancy 790

Cervical cerclage 794

Controversies  796

See also:

Cardiac output increases further in labour, even with effective epidural analgesia, peaking immediately after delivery. It is in this period, when the preload and afterload of the heart are changing rapidly, that women with impaired myocardial function are at greatest risk.

Uteroplacental blood flow is not autoregulated and so is dependent on uterine blood pressure.

Aortocaval occlusion occurs when the gravid uterus rests on the aorta or the inferior vena cava. Even in the absence of maternal hypotension, placental blood supply may be compromised in the supine position. After the 20th week of gestation, a left lateral tilt should always be employed. If either mother or fetus is symptomatic, the degree of tilt should be increased.

Plasma volume increases 50% by term while the red cell mass only increases by 30%, resulting in the physiological anaemia of pregnancy.

Pregnant women become hypercoagulable early in the first trimester. Antepartum maternal deaths from pulmonary embolism occur most commonly in the first trimester. Plasma concentrations of Factors I, VII, VIII, IX, X, and XII are all increased. Antithrombin III levels are depressed.

PaCO₂ falls to ∼4.0kPa (30mmHg). Functional residual capacity is reduced by 20% resulting in airway closure in 50% of supine women at term. This, in combination with a 60% increase in oxygen consumption, renders pregnant women at term vulnerable to hypoxia when supine.

In labour painful contractions and excessive breathing of Entonox can result in further hyperventilation and marked alkalosis may occur. Arterial pH in excess of 7.5 is common.

Gastric emptying and acidity are little changed by pregnancy. However, gastric emptying is slowed in established labour and almost halted if systemic opioids are administered for analgesia. Barrier pressure (the difference in pressure between the stomach and lower oesophageal ‘sphincter’) is reduced, but the incidence of regurgitation into the upper oesophagus during anaesthesia, in otherwise asymptomatic individuals, is not significantly different in the first and second trimesters.

By 48hr postpartum, intra-abdominal pressure, gastric emptying, volume, and acidity are all similar to non-pregnant controls. Although lower oesophageal sphincter tone may take longer to recover, mask anaesthesia is acceptable 48hr after delivery in the absence of other specific indications for intubation.¹

Renal blood flow increases by 75% at term and glomerular filtration rate by 50%. Both urea and creatinine plasma concentrations fall.

Neurological tissue has a greater susceptibility to the action of local anaesthetics during pregnancy—‘MAC’ is also reduced.

The volume of distribution increases by 5 litres, affecting predominantly polar (water-soluble) agents. Lipid-soluble drugs are more affected by changes in protein binding. The fall in albumin concentration increases the free active portion of acidic agents, while basic drugs are more dominantly bound to α₁ glycoprotein. Some specific binding proteins such as thyroxin binding protein increase in pregnancy.

Although plasma cholinesterase concentration falls by about 25% in pregnancy, this is counteracted by an increase in volume of distribution, so the actual duration of action of agents such as suxamethonium is little changed.

Randomised studies show only weak evidence of analgesia for transcutaneous electrical nerve stimulation (TENS). Opioids in labour act predominantly as sedatives and amnesics—pain scores are minimally changed. Entonox is more efficacious than pethidine, but complete analgesia is never attained. When regional analgesia is contraindicated, a fentanyl or remifentanil PCA may be beneficial. Diamorphine has also been advocated.

Regional analgesia provides the most effective pain relief. If hypotension is avoided, fetal condition in the first stage of labour may be improved as maternal sympathetic stimulation and hyperventilation are reduced. However, a degree of maternal motor block is almost universal, and most randomised studies comparing regional and parenteral analgesia demonstrate an association between epidural analgesia and prolonged labour, together with a higher incidence of instrumental deliveries. Careful obstetric management of labour and appropriate anaesthetic management of analgesia may negate this effect.¹

Uterine pain is transmitted in sensory fibres which accompany sympathetic nerves and end in the dorsal horns of T10–L1. Vaginal pain is transmitted via the S2–S4 nerve roots (the pudendal nerve). Spinal, combined spinal/epidural (CSE), and epidural analgesia have largely replaced other regional techniques (paracervical, pudendal, caudal block). Neuraxial techniques can be expected to provide effective analgesia in over 85% of women.

Acceptable analgesia must be provided, but minimising the incidence of hypotension and motor blockade is important. Reducing the degree of motor block increases maternal satisfaction and may decrease the incidence of assisted delivery. Motor block can be reduced by:

Expectation of operative delivery (e.g. multiple pregnancy, malpresentation).

Maternal disease—in particular, conditions in which sympathetic stimulation may cause deterioration in maternal or fetal condition.

Specific cardiovascular disease (e.g. regurgitant valvular lesions).

Severe respiratory disease (e.g. cystic fibrosis).

Specific neurological disease (intracranial A–V malformations, etc.).

Obstetric disease (e.g. pre-eclampsia).

Conditions in which GA may be life threatening (e.g. morbid obesity).

Allergy (true allergy to amide local anaesthetics is rare).

Local infection.

Uncorrected hypovolaemia.

Coagulopathy. (Although guidelines suggest that with platelet count >80 × 10⁹/l and INR <1.4 neuraxial procedures are safe, clinical judgement for each individual patient remains of paramount importance. The cause of the clotting abnormality and the indication for the procedure have to be considered.)

Untreated systemic infection (providing systemic infection has been treated with antibiotics, the risk of ‘seeding’ infection into the epidural space with neuraxial procedures is minimal).

Specific cardiac disease (e.g. severe valvular stenosis, Eisenmenger's syndrome, peripartum cardiomyopathy). Although regional analgesia has been used for many of these conditions, extreme care must be taken to avoid any rapid changes in blood pressure, and preload and afterload of the heart. Intrathecal opioid without local anaesthetic may be advantageous for these patients.

‘Bad backs’ and previous back surgery do not contraindicate regional analgesia/anaesthesia, but scarring of the epidural space may limit the effectiveness of epidural analgesia and increase the risk of inadvertent dural puncture. Intrathecal techniques can be expected to work normally.

Establish IV access. In the absence of previous haemorrhage or dehydration, when low-dose local anaesthetic techniques are used, large fluid preloads are unnecessary.

Position in either a full lateral or sitting position. Finding the midline in the obese may be easier in the sitting position. Accidental dural puncture may be slightly lower in the lateral position.

Fetal heart rate should be recorded before and during the establishment of analgesia. Whether regional analgesic technique requires routine continuous fetal heart rate monitoring after analgesia has been established is controversial.

Skin sterilisation with chlorhexidine is more effective than with iodine.

Locate the epidural space (loss of resistance to saline may have slight advantages in both reduced incidence of accidental dural puncture and reduced incidence of ‘missed segments’ compared with loss of resistance to air).

Introduce 4–5cm of catheter into the epidural space. (Longer has an increased incidence of unilateral block and shorter increases the chance that the catheter pulls out of the space.) Multihole catheters have a lower incidence of unsatisfactory blocks.

Check for blood/CSF.

Give appropriate test dose. An ‘appropriate’ test remains controversial. Using 0.5% bupivacaine significantly increases motor block. Using 1:200 000 adrenaline to detect IV placement of a catheter has both high false-positive and false-negative rates. Many anaesthetists will use 10–15ml 0.1% bupivacaine with a dilute opioid (2µg/ml fentanyl) as both the test and main dose. This will exclude intrathecal placement but may give false negatives if the catheter is sited intravascularly. The complete absence of a detectable block after a normal epidural loading dose for labour analgesia should therefore warn of possible IV cannulation. Remember every dose is a ‘test dose’! If required give further local anaesthetic to establish analgesia. There should be no need to use concentrations >0.25% bupivacaine.

If required give further local anaesthetic to establish analgesia. There should be no need to use concentrations >0.25% bupivacaine.

Measure maternal blood pressure every 5min for at least 20min after every bolus dose of local anaesthetic.

A continuous infusion of local anaesthetic (5–12ml/hr of 0.0625–0.1% bupivacaine with 2µg/ml fentanyl).

Intermittent top-ups of local anaesthetic administered by a PCEA (5ml boluses of 0.0625–0.1% bupivacaine with 2µg/ml fentanyl and a 10–15min lockout period).

Inject subarachnoid solution (e.g. 0.5–1.0ml 0.25% bupivacaine with 5–25µg fentanyl). Do not rotate epidural needle. Insert epidural catheter

or

Perform spinal at L3/4 with a 25–27G pencil-point needle.

Inject subarachnoid solution.

Perform epidural once analgesia has been established.

Minimal motor block associated with these techniques increases maternal satisfaction scores. Intrathecal, as opposed to epidural, techniques produce a more rapid onset of analgesia, possibly a more rapid cervical dilation, and a marginal reduction in assisted delivery.

Impaired proprioception may make walking dangerous even when leg strength has been maintained. Whilst dynamic posturography suggests that following an initial intrathecal dose of 2.5mg bupivacaine and 10µg fentanyl, proprioception is adequate for safe walking, this may no longer be true after repeated epidural top-ups.

Intrathecal opioid may cause temporary fetal bradycardia, probably by altering uterine blood flow through a change in maternal spinal reflexes.

Assessing fetal condition is difficult when the mother is mobile.

Ask the woman if she feels able to stand.

When the woman first stands, have two assistants ready to offer support if required.

Perform a knee bend.

Ask the woman if she feels safe.

Allow full mobilisation.

After each top-up the same sequence must be repeated.

Measure the blood pressure on the dependent arm.

Give an IV fluid bolus of crystalloid solution.

Give 6mg IV ephedrine and repeat as necessary.

Block is often patchy and asymmetrical. Sparing of motor fibres to the lower limbs may occur.

A total spinal may occur following a top-up dose. This is likely to be a consequence of the increased volume rupturing the arachnoid mater.

If a subdural is suspected resite the epidural catheter.

Even if consciousness is not lost, intubation may be required to protect the airway.

Careful maternal and fetal monitoring is essential and, if appropriate, delivery of the fetus. In the absence of fetal distress, Caesarean section is not an immediate requirement. Ventilation is usually necessary for 1–2hr.

Always being alert to symptoms of IV injection with every dose of local anaesthetic, even when previous doses have been uncomplicated.

Divide all large doses of local anaesthetic into aliquots.

Use appropriate local anaesthetics.

See treatment of local anaesthetic toxicity ( p. 1182).

Label the catheter clearly as an intrathecal catheter and only allow the anaesthetist to perform top-ups.

Give intermittent top-ups through the catheter. (1ml 0.25% bupivacaine ±5–25µg fentanyl. Tachyphylaxis may occur with prolonged labour.)

Advantages:

Disadvantages:

Reinsert the epidural at a different interspace—usually one interspace higher. If the reason for tap was difficult anatomy, a senior colleague should take over.

Run the epidural as normal, but beware of intrathecal spread of local anaesthetic. All top-ups should be given by an anaesthetist.

Characteristically worse on standing. Headache is often absent after overnight bed rest but returns after mobilising.

Usually in the fronto-occipital regions and may be associated with neck stiffness.

Photophobia and difficulty in accommodation are common. Hearing loss, tinnitus, and VIth nerve palsy with diplopia are possible.

Nausea occurs in up to 60% of cases.

Bed rest alleviates symptoms, but the incidence of post dural puncture headache after 48hr is the same for those cases that mobilised throughout. Because of the risk of thromboembolism, bed rest should not be routinely encouraged in asymptomatic women.

Epidural infusion of saline (1 litre/24hr) will compress the dural sac and can alleviate symptoms. It may also reduce the flow of CSF through a dural tear. After 24hr of continuous infusion the incidence of post dural puncture headache is marginally reduced. However, radicular pain in the lower limbs may occur and patients are immobilised.

Adequate fluid intake should be encouraged although there is no evidence that hydration reduces the incidence of post dural puncture headache.

Caffeine/theophyllines have become controversial. Concern has been expressed that the incidence of seizures following dural puncture may be increased in the presence of caffeine. These drugs act by reducing intracranial vasodilation, which is partially responsible for the headache. However, while symptoms may be improved, they are not cured. Therefore it is probably sensible to avoid using these agents.

Sumatriptan. Although this cerebral vasoconstrictor is of benefit, it is expensive, is given subcutaneously (6mg), and may cause coronary artery spasm. Its use is best reserved for those in whom blood patching is contraindicated.

Epidural blood patch—see below.

The injected blood forms a clot over the site of the dural tear and this seals the CSF leak.

Two operators are required. One should be an experienced ‘epiduralist’, the other is required to take blood in a sterile manner.

The patient should have a period of bed rest before performing the patch to reduce the CSF volume in the epidural space.

Aseptic technique must be meticulous both at the epidural site and the site of blood letting (usually the antecubital fossa).

An epidural should be performed at the same or a lower vertebral interspace as the dural puncture with the woman in the lateral position to minimise CSF pressure in the lumbar dural sac.

Once the epidural space has been identified, 20ml of blood is obtained.

Inject the blood slowly through the epidural needle until either a maximum of 20ml has been given or pain develops (commonly in the back or legs). If pain occurs, pause and if the pain resolves, try continuing with slow injection. If the pain does not resolve or recurs, then stop.

To allow the clot to form, maintain bed rest for at least 2hr and then allow slow mobilisation.

As far as possible the patient should avoid straining, lifting, or excessive bending for 48hr, although there are obvious limitations when a woman has a newborn infant to care for.

Follow-up is still required. Every woman should have clear instructions to contact the anaesthetists again if symptoms recur even after discharge home.

Serious complications of blood patching are rare. However, backache is common, with 35% of women experiencing some discomfort 48hr post epidural blood patch and 16% of women having prolonged backache (mean duration 27d). Other reported complications include repeated dural puncture, neurological deficits, epileptiform fits, and cranial nerve damage.

No opioid should have been used in the previous 4hr.

Establish dedicated IV access.

PCA with bolus dose of 40µg and lockout of 2min.

Monitor with continuous pulse-oximetry.

Give O₂ if SpO₂ <94% on air.

30min observations of respiratory rate, sedation score, and pain scores.

Always flush the cannula when PCA is discontinued.

Respiratory rate <8 breaths/min.

SpO₂ <94% despite O₂ supplementation.

The neonate is more alert, promoting early bonding and breastfeeding.

Fewer drugs are administered, with less ‘hangover’ than after general anaesthesia.

Better postoperative analgesia and earlier mobilisation.

Both mother and partner can be present at the delivery.

Ultrasound reports to establish the position of the placenta. A low-lying anterior placenta puts a woman at risk of major haemorrhage, particularly if associated with a scar from a previous Caesarean section.

Using an intrathecal catheter and incrementally topping up the block.

Using a combined spinal epidural approach and injecting a small dose of intrathecal local anaesthetic. The block can be subsequently extended using the epidural catheter.

Specific maternal disease (e.g. cardiac disease) where rapid changes in systemic vascular resistance might be problematic.

Ensure that antacid prophylaxis has been given.

Establish 16G or larger IV access. Start 10–15ml/kg crystalloid co-load.

Insert epidural catheter at the L2/3 or L3/4 vertebral interspace.

Test dose, then incrementally top-up the epidural with local anaesthetic and opioid:

Opioid (e.g. 100µg fentanyl or 2.5mg diamorphine) improves the quality of the analgesia and a lower level of block may be effective if opioid has been given.

Establish an S4–T4 block (nipple level) measured by loss of light touch sensation. Always check the sacral dermatomes, as epidural local anaesthetic occasionally does not spread caudally. Anaesthesia to light touch is more reliable at predicting adequacy of block than loss of cold sensation.¹ Document the level of block obtained and the adequacy of perioperative analgesia.

Position patient in the supine position with a left lateral tilt or wedge. Give supplemental oxygen by facemask if SpO₂ <96% on air. (This is very important in obese patients who may become hypoxic when supine, and may have benefit for a compromised fetus.)

Treat hypotension with (see also p. 763):

At delivery give 5IU oxytocin IV bolus. If tachycardia must be avoided then an IV infusion of 30–50IU oxytocin in 500ml crystalloid given over 4hr is an acceptable alternative.

At the end of the procedure give NSAID unless contraindicated (100mg diclofenac PR).¹

Epidural diamorphine given at the time of surgery improves postoperative analgesia, while epidural fentanyl has little postoperative analgesic benefit.

Ensure that antacid prophylaxis has been given.

Establish 16G or larger IV access. Start 10–15ml/kg crystalloid co-load.

Perform spinal anaesthetic at L3/4 interspace using a 25G or smaller pencil-point needle. With the orifice pointing cephalad, inject the anaesthetic solution, e.g. 2.5ml 0.5% hyperbaric bupivacaine with 300µg diamorphine or 15µg fentanyl. Intrathecal diamorphine improves postoperative analgesia, while intrathecal fentanyl has little postoperative analgesic benefit. (Preservative-free morphine 100µg is also used but has little intraoperative benefit. Morphine provides prolonged postoperative analgesia but with a higher incidence of postoperative nausea and vomiting, plus an increased risk of late respiratory depression.)

During the insertion of a spinal anaesthetic, some anaesthetists place patients in a sitting position, while others lie patients on their side. The sitting position usually makes the midline easier to find (important in obese patients), and may be associated with a faster onset, although the height of block is less predictable. A lateral position is associated with a slower onset of block, particularly if a full lateral position is maintained until the block has fully developed. This position also avoids aortocaval compression. With both positions, when hyperbaric local anaesthetic solutions are used, it is important that the cervical spine is kept elevated (pillow) to prevent local anaesthetic spreading to the cervical dermatomes.

Hypotension is more common with spinal anaesthesia than epidural anaesthesia. Patients may benefit from a continuous infusion of pressor agent initiated at the time of insertion of spinal block—see prevention and treatment of hypotension, p. 763.

Continue as for epidural anaesthesia for Caesarean section ( p. 756).

Using the epidural catheter for postoperative analgesia.

Limiting the speed of onset of a block. A small initial intrathecal dose of local anaesthetic can be supplemented through the epidural catheter as required.

Ensure that antacid prophylaxis has been given.

Establish 16G or larger IV access. Give 10–15ml/kg crystalloid co-load.

Insert epidural catheter. Aspirate the catheter carefully for CSF. Testing the catheter with local anaesthetic before the intrathecal dose has receded may be unreliable. However, using the catheter intraoperatively is reasonable, as the anaesthetist is continuously present to deal with the consequences of an intrathecal injection. This may not be true if opioids are given through the catheter for postoperative analgesia at the end of the procedure before the block has receded.

If the spinal block is inadequate, inject local anaesthetic or 10ml 0.9% sodium chloride through the epidural catheter; 0.9% sodium chloride works by compressing the dural sac, causing cephalad spread of intrathecal local anaesthetic.

Continue as for epidural anaesthesia for Caesarean section ( p. 756).

If a partial but inadequate block has developed, the epidural may be resited or withdrawn slightly. Should the toxic limit for the local anaesthetic agent have been reached, elective procedures can be abandoned, but for urgent procedures a general anaesthetic or a spinal anaesthetic will be required. If a spinal is chosen, exceptional care with positioning and observation of the block level is required, as high or total spinal can occur. Use a normal spinal dose of hyperbaric local anaesthetic, as this should ensure adequate anaesthesia, but control the spread with careful positioning.

If a partial but inadequate block develops, an epidural may be inserted and slowly topped up.

Use a GA if required.

To reduce the incidence of hypotension associated with regional anaesthesia.

Limited to 10–15ml/kg crystalloid. Larger volumes should be avoided as they offer little advantage and may be harmful.

More fluid should only be given as clinically indicated.

Colloids are preferred by some anaesthetists if excessive fluid load is likely to be harmful.

Emergency Caesarean section should not be delayed to allow a preload to be administered.

Urgent surgery. (In experienced hands and with a team that is familiar with rapid regional anaesthesia, a spinal or epidural top-up can be performed as rapidly as a general anaesthetic.)

Regional anaesthesia contraindicated (e.g. coagulopathy, maternal hypovolaemia).

Failed regional anaesthesia.

Additional surgery planned at the same time as Caesarean section.

Antacid prophylaxis (see p. 766).

Start appropriate monitoring (see below).

Position supine with left lateral tilt or wedge.

Preoxygenate for 3–5min or, in an emergency, with 4–8 vital capacity breaths with a high flow through the circuit. Ensure a seal with the facemask. At term, women have a reduced FRC and a higher respiratory rate and oxygen consumption. This reduces the time required for denitrogenation, but also reduces the time from apnoea to arterial oxygen desaturation.

Perform rapid sequence induction with an adequate dose of induction agent (e.g. 5–7mg/kg thiopental). Isolated forearm techniques suggest that awareness without recall may be common when the dose of induction agent is reduced. A 7.0mm endotracheal tube is adequate for ventilation and may make intubation easier.

Propofol has also been used for Caesarean section without any major reported complications, although at present thiopental probably is still the most commonly used agent in the UK.

Ventilate with 50% oxygen in nitrous oxide. If severe fetal distress is suspected then 75% oxygen or higher may be appropriate. Maintain ETCO₂ at 4.0–4.5kPa (30–34 mmHg).

Use ‘overpressure’ of inhalational agent to rapidly increase the end tidal concentration of anaesthetic agent to at least 0.75 of MAC (e.g. 2% isoflurane for 5min, then reduce to 1.5% for a further 5min).

At delivery:

At end of procedure give an NSAID (e.g. 100mg diclofenac PR). Bilateral ilioinguinal nerve blocks or TAP blocks (see p. 1153 and p. 1155) are also effective for postoperative analgesia.

Extubate awake in the head-down left lateral position.

Give additional IV analgesia as required.

10mg metoclopramide orally 2hr before surgery.

30ml 0.3M sodium citrate immediately before surgery. (Gastric pH >2.5 is maintained for only 30min after 30ml 0.3M sodium citrate. If a GA is required after this, a further dose of citrate is required.)

10mg metoclopramide IV injection immediately before surgery.

30ml 0.3M sodium citrate orally immediately before surgery.

Intrathecal/epidural opioid:

Assess total amount and rate of blood loss and cardiovascular stability. Blood loss may be difficult to accurately assess. If rapid blood loss is continuing then urgent crossmatch and evacuation of placenta under general anaesthetic is required.

Regional anaesthesia is safe provided estimated blood loss is less than 1000ml, but if there are any signs suggesting hypovolaemia, a general anaesthetic may be required.

Remember antacid prophylaxis.

For general anaesthesia use a rapid sequence induction technique with cuffed ETT.

Regional anaesthesia can be obtained either by topping up an existing epidural or by performing a spinal (e.g. 2ml 0.5% hyperbaric bupivacaine intrathecally). A T7 block reliably ensures analgesia.¹

Occasionally uterine relaxation is required. Under general anaesthesia this can be produced by increasing the halogenated vapour concentration. Under regional anaesthesia SL GTN spray or IV aliquots of 100µg glyceryl trinitrate are effective (dilute 1mg in 10ml 0.9% sodium chloride and give 1ml bolus repeated as required). With either technique expect transient hypotension.

On delivery of the placenta give 5IU oxytocin ± an infusion of oxytocin (e.g. 30–50IU in 500ml crystalloid over 4hr).

At the end of the procedure give an NSAID unless contraindicated.

Lipophilicity or, with hydrophilic drugs, a molecular weight of <200Da.

Weak bases (which increase the proportion of ionised drug in the weakly acidic breast milk, leading to ‘trapping’).

If not, does the placenta cover the anterior lower segment of the uterus. (If it does, the obstetrician will have to divide the placenta to deliver the fetus and blood loss can be expected to be large.)

Is there a uterine scar from previous surgery? (Placenta accreta is more common if the placenta overlies a uterine scar.)

Lower uterine segment does not contract as effectively as the body of the uterus so the placental bed may continue to bleed following delivery.

All patients admitted with placenta praevia should be seen and assessed by an anaesthetist.

Interventional radiology should be considered (see p. 724 and p. 775).

When Caesarean section is to be performed 2–8U of blood should be crossmatched, depending on the anticipated risk of haemorrhage.

Cell salvage should be used if available. In obstetric practice, to reduce the fetal tissue that is reinfused, efforts should be made to minimise the amount of amniotic fluid that is collected and, after processing, the red cells should be reinfused through a leucocyte depletion filter.

Obstetric staff experienced in Caesarean hysterectomy should be immediately available.

Two 14G cannulae should be inserted and equipment for massive haemorrhage must be present.

If regional anaesthesia is used, a CSE may offer advantages as the surgery may be prolonged.

For bleeding patients a general anaesthetic is the preferred choice.

Have a selection of uterotonics to hand (see p. 776). Even if massive haemorrhage is not encountered, an infusion is advantageous (e.g. oxytocin 30–50IU in 500ml crystalloid over 1–2hr).

If massive bleeding does occur, hysterectomy may be the only method of controlling bleeding. Excessive delay in making this decision may jeopardise maternal life (see also p. 774).

Do not forget surgical methods of controlling haemorrhage—bimanual compression of the uterus, ligation of internal iliac arteries, temporary compression of the aorta.

Even if no significant bleeding occurred intraoperatively, continue to observe closely in the postnatal period as haemorrhage may still occur.

Placenta praevia/accreta. Usually a small painless bleed. May be catastrophic.

Uterine rupture. Fetal distress is almost universal. Classically uterine rupture is said to be painful, but painless dehiscence of a previous uterine scar is not uncommon.

Retained placenta. Haemorrhage may be massive but is usually less than 1 litre and occasionally minimal.

Retained products of conception. This is the leading cause of late haemorrhage but is rarely massive.

Genital tract trauma. Vaginal and vulval haematomas are usually self-limiting, but retroperitoneal haematomas may be extensive and life threatening.

Uterine inversion. This is a rare complication in the Western world. It is associated with uterine atony and further relaxation may be required to enable replacement. After replacement uterotonics should be administered.

Give supplemental oxygen. If laryngeal reflexes are obtunded, intubate and ventilate. In antenatal patients avoid aortocaval compression.

Insert two 14G cannulae and take blood for crossmatching. Request type-specific blood (this can be retrospectively crossmatched).

Fluid resuscitate with crystalloid and/or colloid.

If required give Group O Rhesus negative blood (i.e. blood loss of 2–3 litres and ongoing without the imminent prospect of crossmatched blood being available and/or the presence of ECG abnormalities).

Start appropriate monitoring of mother and fetus. Urine output and invasive monitoring of central venous and arterial pressures may be indicated depending on rate of blood loss and maternal condition. However, early monitoring of CVP is not essential as hypotension is almost always due to hypovolaemia.

Treat the cause of haemorrhage. If surgery is required:

With continuing haemorrhage further equipment including warming devices and rapid transfusion devices should be available.

Correct coagulopathy with platelets, fresh frozen plasma, and cryoprecipitate as indicated.

Once blood loss has been controlled, continue care on HDU or ICU.

Most postnatal haemorrhage is due to uterine atony and can be temporarily controlled with firm bimanual pressure while waiting for definitive treatment.

Interventional radiology is especially useful when major haemorrhage is anticipated. It may not reduce the incidence of Caesarean hysterectomy but probably reduces blood loss. Balloon catherters can be prophylactically placed in the internal illiac vessels before delivery (but only inflated at the moment of delivery!). However, interventional radiology can also be used in response to a major bleed, but be very cautious of moving a patient to a radiology suite if they are cardiovascularly unstable. Intrauterine tamponade with Bakri or Sengstaken catheters left in situ for 24hr can also be useful.

Incidence: ∼1:12 000 live births.²

Effects are probably due to an anaphylactic response to fetal tissue.

Within the first 30min after amniotic fluid embolism, intense pulmonary vasoconstriction occurs and is associated with right heart failure, hypoxia, hypercarbia, and acidosis.

This is followed by left heart failure and pulmonary oedema.

Expect a coagulopathy.

Incidence of AFE is increased with:

AFE is often a diagnosis of exclusion, but clinical features include:

Senior staff should be present (obstetric, anaesthetic, paediatric, and midwifery).

Haematology services should be alerted, as large quantities of blood products may be required.

Early delivery of the fetus is vital for both maternal and fetal survival.

Measure coagulation profile regularly. Platelets fresh frozen plasma, and cryoprecipitate may all be required.

Intensive care will be required for those who survive the initial insult.

Plan the delivery. There is little evidence that vaginal delivery is associated with a worse outcome for many cardiac patients. However, on occasions, an elective Caesarean section may offer advantage as appropriate personnel can be guaranteed to be present.

Investigations should be performed as indicated. The risk to the fetus from procedures such as chest radiographs is minimal.

Make sure that the woman is in an appropriate place for delivery (this may be the normal delivery suite or it may be the cardiac theatres in a tertiary centre).

With each condition, in consultation with the cardiologists, consider the effects of vasodilation, vasoconstriction, and positive and negative inotropic and chronotropic agents. This will help with the planning of the delivery, and can allow written guidance on the likely acceptability of regional analgesia and regional or general anaesthesia, as well as the use of oxytocin (potent vasodilator) and ergometrine. It can also allow planning of the appropriate agent to use in the event of hypotension.

Some centres use a form where all these elements of the anaesthetic management of complex patients are specified. The form remains with the mother throughout her pregnancy.

In most situations rapid changes in pre- or afterload should be avoided, so always use oxytocin with extreme caution and preferably only as an infusion.

Expect the period of highest risk to be in the 1–2hr post delivery (cardiac output peaks and autotransfusion plus blood loss leads to a variable effect on pre- and afterload).

Continue management on ICU if appropriate.

Extreme caution is required to avoid sudden changes in afterload for patients with fixed cardiac output.

Cyanotic heart lesions (i.e. right-to-left shunts) will not tolerate reductions in systemic vascular resistance. Nevertheless epidural analgesia is sometimes used to minimise the stress of labour, but onset of analgesia must be slow, and use phenylephrine to maintain afterload. General anaesthesia is probably the technique of choice for Caesarean section.

Aortic stenosis may become symptomatic during pregnancy. Serial echocardiography is often used. Tachycardia and reduction in afterload should be avoided. Loss of sinus rhythm should be treated promptly. General or slow-onset regional anaesthesia have both been advocated for Caesarean section. The technique is probably less important than the skill with which it is applied.

Valvular insufficiencies are usually well tolerated during pregnancy.

Women with symptomatic Marfan's disease (see p. 64 and p. 317), particularly if the aortic root is dilated, have a high risk of aortic dissection. They are usually maintained on β-blockers. Unexplained severe chest pain is an indication for a chest X-ray and an echocardiogram.

Myocardial infarction during pregnancy has a 20% mortality. Infarction occurs most commonly in the third trimester. If possible, delivery should be delayed at least 3wk after infarction. Both elective Caesarean section and vaginal delivery have been advocated. In either case, cardiac stress should be minimised with effective analgesia.

After 20wk gestation the fetus should be delivered as soon as possible. This improves the chance of maternal survival (and that of a term fetus) as aortocaval compression severely limits the effectiveness of chest compressions. The fetus is likely to be severely acidotic and hypoxic at delivery.

Remember that pregnant women have reduced oesophageal sphincter tone and that cricoid pressure and intubation should both be performed as rapidly as possible.

Normal resuscitation drugs should be used. Adrenaline is the drug of choice despite its effect on uterine circulation.

Adrenaline is also the drug of choice in major anaphylactic and anaphylactoid reactions. Severe hypotension associated with anaphylaxis results in very poor fetal outcome. Early delivery is vital for the fetus.

Consideration should be given to the diagnosis and treatment of obstetric causes of maternal arrest.

Pulmonary embolism.

Amniotic fluid embolus.

Intracranial haemorrhage.

Myocardial infarction.

Iatrogenic events:

Chronic hypertension: hypertension that existed before pregnancy.

Pregnancy-induced hypertension: hypertension that develops in pregnancy. In the absence of other signs of pre-eclampsia, this has minimal effect on pregnancy but may be indicative of a tendency to hypertension in later life.

Pre-eclampsia: pregnancy-induced hypertension in association with renal involvement causing proteinuria (>300mg/24hr or 2+ on urine dipstick). Incidence 6–8% of all gestations.

Severe pre-eclampsia: pre-eclampsia in association with any of the following: a sustained BP >160/110; proteinuria >5g/24hr or 3+ on urine dipstick; urine output <400ml/24hr; pulmonary oedema or evidence of respiratory compromise; epigastric or right upper quadrant pain; hepatic rupture, platelet count <100 × 10⁹/l; evidence of cerebral complications. Incidence is 0.25–0.5% of all gestations.

Eclampsia: convulsions occurring in pregnancy or puerperium in the absence of other causes. Almost always occurs in the presence of pre-eclampsia, although signs of pre-eclampsia may not be manifest until after a fit.

Reduced circulating volume but increased total body water.

In severe pre-eclampsia systemic vascular resistance is increased and cardiac output reduced. However, some women have elevated cardiac output with normal or only marginally increased systemic vascular resistance. Fetal prognosis is improved in this group.

Poor correlation between central venous and pulmonary capillary wedge pressure.

Increased capillary permeability which may result in:

Increased haematocrit resulting from decreased circulating volume.

Increased permeability to large molecules resulting in proteinuria.

Decreased urate clearance with rising serum uric acid level.

Oliguria in severe disease.

Cerebrovascular haemorrhage.

Eclampsia (resulting from cerebral oedema or cerebrovascular vasoconstriction).

Poor placental perfusion and increased sensitivity to changes in maternal BP.

Reduction of umbilical arterial diastolic blood flow and particularly reverse diastolic flow are indicative of poor fetal outcome without early intervention.

In established pre-eclampsia the only definitive treatment is the delivery of the placenta. Symptoms will usually start to resolve within 24–48hr.

When pre-eclampsia develops at term there is no advantage to delaying delivery. However, if pre-eclampsia develops before term, a compromise has to be made between maternal and fetal health. Maternal blood pressure is controlled for as long as possible to allow fetal growth to be optimised. If fetal or maternal condition deteriorates, delivery must be expedited.

Antihypertensive therapy. Blood pressure should be controlled to below 160/110 to prevent maternal morbidity particularly from intracranial haemorrhage, encephalopathy, and myocardial ischaemia/failure.

Magnesium prophylaxis in pre-eclampsia effectively reduces the incidence of eclampsia, but increases the frequency with which the side effects of magnesium therapy are seen. Severe pre-eclamptic patients are usually treated with magnesium.

Fluid management in severe pre-eclampsia is critical. Intravascular volume is depleted, but total body water is increased. Excessive fluid load may result in pulmonary oedema, but underfilling may compromise fetal circulation and renal function. General principles are:

Invasive arterial pressure monitoring is indicated in severe pre-eclampsia for:

Central venous pressure monitoring is rarely indicated even in severe pre-eclampsia as central venous pressure may not correlate well with pulmonary arterial wedge pressure. It should be considered:

Check platelet count before performing an epidural. If maternal condition is rapidly deteriorating or if the platelet numbers are falling then a count must be performed immediately before placement. The ‘acceptable’ level of platelet count is debatable and based on little evidence. However, common general guidelines are:

Preload before regional analgesia is not required, but monitor BP and fetus carefully and treat changes in BP promptly with cautious doses of ephedrine or phenylephrine (IV infusions may be preferable).

Obtund the hypertensive response to laryngoscopy, e.g. alfentanil 1–2mg (inform paediatrician that opioids have been used) or labetalol 10–20mg before induction. A remifentanil infusion may be useful if the anaesthetist is familiar with its use. In very severe pre-eclampsia, intra-arterial pressure monitoring is required before induction.

If magnesium has been used, expect prolongation of action of non-depolarising muscle relaxants. Use a reduced dose and assess muscle function with a nerve stimulator.

Ensure adequate analgesia before extubation. The hypertensive response to extubation may also need to be controlled with antihypertensive agents (e.g. labetalol 10–20mg).

Effective postoperative analgesia is required, but avoid NSAIDs as these patients are prone to renal impairment and may have impaired platelet count or function. When the proteinuria has resolved, which is often within 48hr, NSAIDs may be introduced.

Continue care in HDU or ICU.

A reduced volume of preload should be used (possibly with colloid).

Expect ephedrine to have an increased effect. The role of phenylephrine in pre-eclamptic patients has still to be established.

A slow-onset block may be beneficial.

Effective postoperative analgesia is required, but avoid NSAIDs as these patients are prone to renal impairment and may have impaired platelet count or function.

Care should be continued on HDU or ICU.

Most fits occur in the third trimester, and nearly one–third occur postpartum, usually within 24hr of delivery.

Eclampsia is a life-threatening event.

Management is aimed at immediate control of the fit and secondary prevention of further fits.

Control of fits with 4g magnesium given IV over 5–10min.

Patients on calcium channel antagonists are at particular risk of toxicity.

Toxicity can be treated with IV calcium (e.g. 10ml 10% calcium chloride).

Elevated liver function tests—AST (serum aspartate transaminase), ALT (serum alanine aminotransferase), alkaline phosphatase, and γ-glutamyl transferase. Epigastric or right upper quadrant abdominal pain are present in 90% of women with HELLP. Liver failure and hepatic rupture may occur. Extreme elevation in AST is associated with poor maternal prognosis. Most women with right upper quadrant pain and a platelet count of <20 × 10⁹/l will have an intrahepatic or subcapsular bleed.

A falling platelet count. Counts of less than 100 × 10⁹/l are of concern, while a count of less than 50 × 10⁹/l is indicative of severe disease.

Hypertension and proteinuria are present in 80% of women with HELLP and 50% suffer nausea and vomiting. Convulsions and gastrointestinal haemorrhage are occasional presenting features.

The risk of epidural haematoma may preclude the use of regional analgesia/anaesthesia. Consideration must be given to both the absolute platelet number as well as rate of fall in platelet count. All patients require a clotting screen.

Be prepared for major haemorrhage.

Further management is supportive, with appropriate replacement of blood products as required.

Invasive monitoring is dictated entirely by the clinical condition of the patient.

ARDS, renal failure, and disseminated intravascular coagulation may develop.

After delivery of the placenta, recovery can be expected to start within 24–48hr. These patients should all be on an HDU or ICU.

Make sure that the obstetric team are aware that surgery is planned.

Remember gastric acid prophylaxis.

Remember DVT prophylaxis. Pregnant women are hypercoagulable.

Consider regional anaesthesia. The combination of a mother maintaining her own airway together with a minimal fetal drug exposure is desirable. However, data demonstrating that regional anaesthesia is safer than general anaesthesia are lacking.

Airway management in the first and early second trimester is controversial. In asymptomatic women with no other indication for intubation, it is acceptable not to perform a rapid sequence induction up to 18wk of gestation. However, be aware that lower oesophageal sphincter tone is reduced within the first few weeks of pregnancy and intra-abdominal pressure rises in the second trimester. If patients have additional risk factors for regurgitation (e.g. symptomatic reflux, obesity), use a rapid sequence induction.²

Every effort must be made to maintain normal maternal physiological parameters for the gestational age of the fetus throughout the perioperative period.

Treat haemorrhage aggressively. Avoid hypovolaemia and anaemia as both impact on fetal oxygenation.

From the 20th week of gestation use left lateral tilt to reduce aortocaval compression. Remember that although upper limb BP may be normal, uterine blood flow may still be compromised in the supine position.

If general anaesthesia is employed, use adequate doses of inhalational agents. Light anaesthesia is associated with increased catecholamine release, which reduces placental blood flow. The tocolytic effect of inhalational agents is advantageous.

Fetal monitoring may be beneficial although its value remains unproven. If fetal distress is detected, maternal physiology can be manipulated to optimise uterine blood flow.

The primary risk to the fetus is premature labour in the postoperative period. Detection and suppression of premature labour is vital. Women should be told to report sensations of uterine contractions so that appropriate tocolytic therapy can be instituted.

Effective postoperative analgesia is required to reduce maternal catecholamine secretion. Although opioids can be used, they may result in maternal hypercarbia. Regional analgesia with local anaesthetic agents may be preferential. If this would prevent the mother from detecting uterine contractions, consider external uterine pressure transduction—‘tocodynamometry’. For minor surgery local anaesthesia and simple analgesics such as paracetamol and codeine may be used. Chronic dosage with NSAIDs should be avoided—in the first trimester because of increased fetal loss and in the third trimester because of the possibility of premature closure of the fetal ductus arteriosus.

Ranitidine and cimetidine are not known to be harmful, but caution is advised with chronic exposure to cimetidine because of known androgenic effects in adults.

Propofol is not teratogenic in animal studies. Its use in early human pregnancy has not been formally investigated. Propofol is safe to use during Caesarean section at term.

Etomidate is also not teratogenic in animal studies. It is a potent inhibitor of cortisol synthesis, and when used for Caesarean section, neonates have reduced cortisol concentrations.

Ketamine should be avoided in early pregnancy as it increases intrauterine pressure, resulting in fetal asphyxia. This increase in intrauterine pressure is not apparent in the third trimester.¹

Despite early concerns, recent epidemiological studies suggest that nitrous oxide is safe. However, nitrous oxide is consistently teratogenic in Sprague Dawley rats if they are exposed to 50–75% concentrations for 24hr during their peak organogenic period. Given that anaesthesia can be safely delivered without nitrous oxide it is sensible to avoid this agent.

Muscle relaxants: because these agents are not lipophilic, only very small quantities cross the placenta and so fetal exposure is limited. These agents are safe to use.

Anticholinesterase inhibitors: these agents are highly ionised and so, like muscle relaxants, do not readily cross the placenta and are safe to use. Chronic use of pyridostigmine to treat myasthenia gravis may cause premature labour.

Chronic exposure to NSAIDs in early pregnancy may be associated with increased fetal loss and in the third trimester may cause premature closure of the ductus arteriosus and persistent pulmonary hypertension of the newborn. Single doses are unlikely to be harmful. These agents are also used to suppress labour, particularly in the second trimester.

Bupivacaine and lidocaine are safe. When used near delivery, bupivacaine has no significant neonatal neurobehavioural effects, while lidocaine may have a mild effect. Cocaine abuse during pregnancy increases fetal loss and may increase the incidence of abnormalities in the genitourinary tract.

Careful assessment of airway, gestation, symptoms of reflux, and supine hypotension.

Remember antacid prophylaxis.

Explain risks of teratogenicity/spontaneous miscarriage (see p. 791).

If general anaesthesia is used and uterine relaxation is required to allow bulging membranes to be reduced, the halogenated vapour concentration can be increased.

For regional anaesthesia, a T8–T10 level is required for intraoperative comfort. If uterine relaxation is required, 2–3 puffs of sublingual glyceryl trinitrate spray may be used and repeated as necessary, although transient hypotension is to be expected.

Vaginal cervical cerclage sutures are usually removed at the 38th week of gestation.

The transabdominal procedure requires two operations—one for insertion and another for a Caesarean section for delivery and removal of the suture. It also carries greater risk of ureteric involvement.

Transvaginal procedures are much more common. Shirodkar and McDonald procedures are the two commonest methods. They both require anaesthesia for insertion but can be removed without anaesthetic.

Stop all intake of solids if any opioid, epidural, or oxytocic is used.

Allow only ‘low residue’ foods that rapidly empty from the stomach (cereals, toast, low fat cheese, and semi-sweet biscuits). Very cold foods such as ice creams are known to delay gastric emptying.

As particulate matter is known to be especially problematic in the event of aspiration, isotonic sports drinks may offer the ideal solution. These effectively prevent ketosis without increasing gastric volume.

Change in the way the force of contraction is transferred to the pelvic floor when relaxed by an epidural.

Change in uterine force of contraction (this is not substantiated by recent studies).

Interference with the Ferguson reflex (increased oxytocic production between pressure on the pelvic floor in the second stage of labour). This is controversial, as the Ferguson reflex has not been clearly demonstrated in humans.