Introduction

Cardiovascular disease is estimated to complicate 0.2–4% of all pregnancies in westernised countries, and is the most common cause of death during pregnancy in the UK . Anaesthetists play a vital role in the management of pregnant women with heart disease. This may be preoperatively in the antenatal clinic, peripartum to provide labour analgesia, and perioperatively to provide anaesthesia for caesarean section. Postoperatively, anaesthetists are involved in the provision of analgesia or in the delivery of high-dependency care. In addition, anaesthetists can offer expertise in areas such as fluid balance optimisation, monitoring, and vasoactive drug administration.

Guidelines exist for the management of cardiovascular disease during pregnancy; however, specific recommendations about optimal anaesthetic techniques are lacking . Evidence that is available comes from observational data, case reports, or expert opinion. As a result, many clinicians advocate individualised care guided by severity of disease and obstetric indications. This review will look at the anaesthetic management of parturients with cardiac disease during the peripartum period. Relevant physiology, the general principles of anaesthetic management, and the specific considerations of individual cardiac conditions will be covered.

Cardiovascular physiological changes in pregnancy

A clear understanding of the normal physiological changes during pregnancy as well as the pathophysiology of heart disease is crucial to optimise management and recognise complications. Changes in the cardiovascular system during pregnancy are designed to meet the increasing metabolic demands of the mother and fetus. By 24 weeks’ gestation, plasma volume has increased by 40%, with a 30–50% increase in cardiac output. In early pregnancy, the rise in cardiac output is driven by an increase in stroke volume. As pregnancy advances, an increase in heart rate becomes the main contributor, peaking around 32 weeks and remaining high up to 5 days after delivery. Initially, both systolic and diastolic blood pressure drop because of vasodilation and a fall in systemic vascular resistance (SVR). This is thought to be a result of local mediators such as nitric oxide and prostacyclin. By term, diastolic blood pressure may have reached pre-pregnancy values. Aortocaval compression develops from 20 weeks gestation onwards, and can cause complete occlusion of the inferior vena cava in the supine position towards term .

During labour, further increases in cardiac output occur. Contributing factors include an additional volume load (up to 500 ml) with each uterine contraction, as well as pain, physical exertion and anxiety. In early labour, this can lead to further increases in cardiac output of up to 15%, and up to 50% during maximum expulsive efforts in the second stage of labour . Immediately after delivery, the effect of auto-transfusion and reabsorption of leg oedema contribute to further increases in cardiac output of up to 80% from pre-labour values . After delivery, cardiac output remains high and SVR rises sharply. This elevation can last from weeks to months . As such, the risk of cardiovascular complications continues well into the postpartum period.

Cardiovascular physiological changes in pregnancy

A clear understanding of the normal physiological changes during pregnancy as well as the pathophysiology of heart disease is crucial to optimise management and recognise complications. Changes in the cardiovascular system during pregnancy are designed to meet the increasing metabolic demands of the mother and fetus. By 24 weeks’ gestation, plasma volume has increased by 40%, with a 30–50% increase in cardiac output. In early pregnancy, the rise in cardiac output is driven by an increase in stroke volume. As pregnancy advances, an increase in heart rate becomes the main contributor, peaking around 32 weeks and remaining high up to 5 days after delivery. Initially, both systolic and diastolic blood pressure drop because of vasodilation and a fall in systemic vascular resistance (SVR). This is thought to be a result of local mediators such as nitric oxide and prostacyclin. By term, diastolic blood pressure may have reached pre-pregnancy values. Aortocaval compression develops from 20 weeks gestation onwards, and can cause complete occlusion of the inferior vena cava in the supine position towards term .

During labour, further increases in cardiac output occur. Contributing factors include an additional volume load (up to 500 ml) with each uterine contraction, as well as pain, physical exertion and anxiety. In early labour, this can lead to further increases in cardiac output of up to 15%, and up to 50% during maximum expulsive efforts in the second stage of labour . Immediately after delivery, the effect of auto-transfusion and reabsorption of leg oedema contribute to further increases in cardiac output of up to 80% from pre-labour values . After delivery, cardiac output remains high and SVR rises sharply. This elevation can last from weeks to months . As such, the risk of cardiovascular complications continues well into the postpartum period.

Physiological changes of regional anaesthesia

Central neuraxial anaesthesia further disrupts normal cardiovascular physiology, particularly when intrathecal or spinal anaesthesia is used. Local anaesthetic blockade of the sympathetic chain is unavoidable, and leads to vasodilation in the lower body with compensatory vasoconstriction in the upper limbs. Overall, SVR reduces, and a consequent fall in blood pressure occurs. Cardiac output falls as a result of decreased venous return and, if the sympathetic block reaches the cardiac accelerator fibres at T1–T4, decreased heart rate. This is common in pregnant women for a number of reasons. Firstly, there is a shift in the autonomic equilibrium in pregnancy in favour of sympathetic compared with parasympathetic activity . Experimental studies have shown that those with a higher baseline sympathetic tone are more prone to severe hypotension during spinal anaesthesia . Secondly, pregnant women are relatively resistant to the effects of vasopressors, a phenomenon that has been attributed to increased baroreceptor sensitivity , increased endothelial nitric oxide synthase activity , and variations in the genotype of the β-2-adrenergic receptor . Severe acute under-filling of the right ventricle secondary to a combination of aortocaval compression and venodilation may precipitate a Bezold–Jarisch type phenomenon and profound bradycardia.

Management principles

Labour analgesia

The use of epidural analgesia for labour is associated with a reduction of the stress response and greater haemodynamic stability . As such, low-dose epidural labour analgesia is often recommended in women with cardiac disease, particularly as the severity of disease increases . Although the haemodynamic effects of modern low-dose epidurals are modest, caution and careful monitoring are still required with high-risk lesions.

Absolute contraindications to regional anaesthesia include patient refusal, local anaesthetic allergy, moderate or severe bleeding diathesis, and infection at the site of insertion. Relative contraindications include raised intracranial pressure, severe hypovolaemia, some skeletal abnormalities and after certain types of back surgery . As we will discuss, conditions with a ‘fixed’ cardiac output, such as aortic stenosis, are now considered only a relative contraindication.

Test doses are given to check for inadvertent intrathecal insertion of the epidural catheter. A positive test dose will result in rapid onset of sensory and motor block, accompanied with marked vasodilation. This could have deleterious consequences in a woman with heart disease, and caution is required. The addition of adrenaline to the test dose is used to check for intravascular catheter insertion. Although some suggest it is safe, most agree it is of limited benefit, with the potential to cause profound tachycardia .

Regional compared with general anaesthesia for caesarean section

Regional anaesthesia forms the foundation of modern obstetric anaesthesia, allowing parturients to be awake when the baby is delivered while minimising transfer of drugs across the placenta. There is, however, no evidence that regional or general anaesthesia is superior in reducing morbidity or mortality in either mother or baby . Historically, the reduction in SVR with regional anaesthesia limited its use in women with cardiac disease. In addition, concerns were raised about the development of right to left shunts in women with congenital heart defects. Refinement of the neuraxial technique, along with widespread use of phenylephrine infusions to maintain SVR, has meant this is no longer the case.

Successful outcomes have been reported with regional anaesthesia for virtually every type of cardiac pathology, from simple septal defects to the most severe types of congenital heart disease . A prospective review from Scandinavia looked at 9000 deliveries from 2003 to 2008, of which 113 were classified as having cardiac disease . Out of these, 29% were considered to be high risk, including conditions such as cardiomyopathy, pulmonary hypertension, and Eisenmenger’s syndrome. The caesarean section rate was high (52.2%), most of which were managed using regional anaesthesia (86%), with no maternal or neonatal mortality at 6 months.

Advantages of general anaesthesia include the ability to manipulate respiratory and cardiovascular parameters, the avoidance of sharp drops in SVR, and the ability to monitor cardiac function with transoesophageal echocardiography . Positive pressure ventilation will support a failing ventricle and facilitate management of pulmonary oedema. General anaesthesia also negates the problem of having an insufficient block, particularly pertinent in cases where surgery is prolonged or complex. Many women with cardiac disease are fully anti-coagulated, thus excluding the option of regional anaesthesia.

Disadvantages of general anaesthetic techniques include the risk of failed intubation and respiratory complications. In one Centre for Maternal and Child Enquiries report, as many as six maternal deaths were attributed to complications from general anaesthesia . Concern has also been raised about the sympathetic stimulation from direct laryngoscopy and intubation. The effect of positive pressure ventilation with a reduction in venous return, hypotension, and raised pulmonary artery pressures must also be considered .

Conduct of regional anaesthesia

The options for neuraxial techniques in cardiac parturients include single shot spinal, combined spinal epidural (CSE), spinal catheter insertion, or incremental epidural anaesthesia. Incremental epidural anaesthesia can either be de novo or by supplementation of an epidural sited for labour. Although no trial evidence supports a particular technique, survey data suggest that anaesthetists favour incremental epidural anaesthesia, for both vaginal and caesarean deliveries .

Another option that has gained popularity for people with cardiac disease is low-dose sequential CSE . In this technique, lower doses of intrathecal bupivacaine (e.g. 5 mg v 10–15 mg) are followed by cautious epidural supplementation until surgical anaesthesia is achieved. Reports have suggested this provides better anaesthesia with less hypotension, and reduces the risk of inadequate anaesthesia . A study in healthy parturients, however, suggested no difference in the incidence of hypotension between 5 mg and 10 mg of bupivacaine . Moreover, the onset of block was significantly faster in the 10 mg group. Of course, it is difficult to extrapolate these findings, particularly to women with cardiac disease who may be more dependent on sympathetic tone.

Incremental spinal anaesthesia has also been used for cardiac parturients with good success . The risk of hypotension, however, is still present even with careful titration, with one study showing an 18% incidence of symptomatic hypotension . Furthermore, the technique is unfamiliar to many anaesthetists in the UK, and risk of post-dural puncture headache is increased .

Conduct of general anaesthesia

Little evidence supports a particular general anaesthetic technique, and much depends on the type and severity of disease. General principles that can be applied to most cardiac conditions include obtunding the sympathetic response to laryngoscopy, avoidance of tachycardia, and maintaining normotension.

Pregnant women are considered at risk of regurgitation and aspiration of gastric contents. Standard practice in obstetrics is to carry out a rapid sequence induction (RSI) with thiopentone and a depolarising muscle relaxant such as suxamethonium. In women with cardiac pathologies, most anaesthetists ‘modify’ the standard RSI with the addition of an opioid, or avoid RSI altogether . The addition of an opioid obtunds the sympathetic response to laryngoscopy, but risks respiratory depression in the neonate. The successful use of ultra-short-acting opioids such as remifentanil and alfentanil, with minimal neonatal respiratory depression, has been described .

Standard induction agents such as propofol and thiopentone have been used in cardiac parturients, despite an accompanying reduction in SVR . In severe cardiac disease, etomidate may be used as it maintains SVR . Maintenance of anaesthesia is generally with inhalational agents, such as isoflurane and sevoflurane. Inhalational agents are negatively inotropic, reduce SVR, and impair the pulmonary vascular response to hypotension, so must be used with care. Nitrous oxide, a vapour used as a carrier gas, is cardio-depressant, and may cause a rise in pulmonary vascular resistance (PVR). It would usually be avoided, particularly in women with pulmonary hypertension . Some would also advocate avoiding depolarising muscle relaxants owing to their propensity to cause bradycardia and histamine release.

Use of vasopressors

As mentioned previously, many of the unwanted side-effects from neuraxial anaesthesia can be counteracted using an infusion of vasopressor. Traditionally, ephedrine, a combined α and β-agonist, was considered the drug of choice for treating hypotension during caesarean section. Several studies, however, have since found that ephedrine can cause fetal acidaemia, particularly in large doses . In addition, it is difficult to titrate, frequently ineffective, and causes maternal tachycardia. Phenylephrine, a pure α-agonist, has shown to be highly effective, easily titratable, and safe for the baby. It is now widely considered the vasopressor of choice during routine caesarean section, and for most people with cardiac disease .

Phenylephrine does, however, have significant side-effects, the pure α-agonist activity causing reflex bradycardia and reduced cardiac output. This can be dramatic, particularly if the woman is already on medications such as beta-blockers, and has led to concern about its use in cardiac parturients. Concern has also been raised from animal studies about its use if fetal compromise is already present . Careful monitoring and titration are essential.

Blood loss and fluid management

The risk of peripartum haemorrhage is theoretically greater in women with cardiac disease given the prevalence of anti-coagulation use, potentially raised venous pressures, and the avoidance of certain uterotonics. The consequence of blood loss can be profound, as many with heart disease lack the mechanisms to compensate acutely. Many women may also be on medications such as beta-blockers, which prevent compensatory rises in heart rate and disguise the severity of blood loss. Another concern is the response to fluid administration and potential for developing pulmonary oedema.

The risk of pulmonary oedema must be considered when managing the parturient with cardiac disease . Pregnant women at term are prone to fluid retention, especially if there is co-existing pre-eclampsia. Significant fluid shifts take place as a result of auto-transfusion from the uterus as it contracts after delivery. The volume of intravenous fluid administration can accumulate because of the administration of multiple medications. Consideration must be given to using higher concentrations and lower volumes of drugs . Close monitoring of hourly fluid inputs and outputs is crucial to optimise fluid management. Some advocate a small dose of furosemide at delivery if the risk from autotransfusion is deemed significant .

Use of uterotonics

The use of oxytocin in pregnancy is well established, and has been shown to significantly reduce the incidence of postpartum haemorrhage . In the UK, standard practice involves administration of a five unit bolus immediately after delivery of the baby at caesarean section followed by an infusion of 40 units over 4 h. This is in contrast to practice in the USA and other countries, where the bolus dose is frequently omitted . Oxytocin administered as a bolus causes a marked reduction in SVR (up to 50%) with a compensatory increase in cardiac output and heart rate. This has raised concern about the use of oxytocin boluses in women with heart disease . Indeed, a 10-unit bolus of oxytocin has been implicated in the death of a woman with hypovolemia in a previous Centre for Maternal and Child Enquiries report .

As a result, many investigators suggest either giving only an infusion of oxytocin, or avoiding it altogether . Others suggest that this may lead to a greater risk of bleeding and more cardiovascular instability. Moreover, it has been shown that co-administration of phenylephrine effectively negates the unwanted side-effects of oxytocin . Langesaeter et al. have suggested giving small incremental bolus doses of oxytocin (e.g. 0.1 IU) until uterine contraction is achieved . They highlight the significantly lower dosage of oxytocin required to achieve uterine contraction in elective caesarean section (ED 95% ED is effective dose 0.35 IU) than is traditionally given as a bolus . In their study, this technique caused only minor changes in cardiac output in parturients with significant cardiac disease . In emergency deliveries, where parturients have already been on oxytocin infusions, the dose required to achieved uterine contraction is significantly greater (ED 95% three IU), making this technique impractical .

Other concerns have been raised about the propensity of oxytocin to cause coronary vasoconstriction. In a recent study comparing oxytocin (10 IU) with ergometrine (0.2 mg), oxytocin caused significantly greater ST-segment changes on vector-cardiography during caesarean section . Another study confirmed these findings, and found this effect to be more pronounced when using a 10 IU compared with a 5 IU bolus . Although this may not be clinically significant in healthy women, in parturients with heart disease it may be highly significant.

The use of other uterotonics such as ergometrine, an α-agonist, and carboprost, a prostaglandin-F _2α agonist, is also controversial in parturients with heart disease. Ergometrine, in particular, is known to cause coronary vasospasm, and several reports have highlighted the risk of myocardial infarction in patients without cardiac disease . It should, therefore, be avoided in women with severe heart disease or, if absolutely necessary, given intramuscularly. It is also known to cause pulmonary vasoconstriction and hypertension and should be avoided in severe pulmonary hypertension . Carboprost, a smooth muscle contractor, can cause hypotension, bronchospasm, pulmonary oedema, and cardiovascular collapse, and should be avoided in all forms of cardiac disease . Misoprostol, a prostaglandin E1 analogue, has gained popularity as a uterotonic, and has minimal cardiovascular side-effects.

Given the side-effects of uterotonics, surgical techniques to prevent postpartum bleeding take on greater prominence. Prophylactic use of techniques such as the B-lynch suture and intra-uterine balloon has been advocated .

Aortocaval compression

Aortocaval compression can have significant consequences in the parturient with cardiac disease. Avoidance of the supine position whenever possible is therefore crucial, particularly if neuraxial anaesthesia has been carried out. The benefits of performing left lateral tilt in improving umbilical venous oxygenation are well documented , and for this reason alone can be justified. Debate continues about the degree of tilt required during caesarean section to prevent changes in maternal cardiac output and blood pressure . Studies looking at maternal blood pressure and fetal heart rate have shown no significant differences between varying degrees of tilt .

More recently, Bamber and Dresner showed that, in healthy parturients, the degree of tilt does have a significant effect on cardiac output as measured by thoracic bioimpedance . They showed that lying on the left side significantly improves cardiac output compared with lying supine with a right lateral tilt. No significant difference between varying degrees of tilt (up to 12.5°) was reported in the supine position. The investigators suggest that if aortocaval compression is suspected, and the situation allows, a full roll onto the woman’s left side is preferable. Of course, this evidence comes from healthy mothers, and the effect of aortocaval compression in mothers with heart disease is likely to be accentuated.