Haemostatic disorders in pregnancy

Learning objectives.

By reading this article, you should be able to:

• •Describe normal haemostatic processes.
• •Discuss the implication of von Willebrand disease and haemophilia in the management of a pregnant patient.
• •Recall the causes of thrombocytopenia in pregnancy.
• •Consider the risks and benefits of a central neuraxial block in a patient with a disorder of haemostasis.

Key points.

• •Pregnant patients with a disorder of haemostasis require a multidisciplinary approach.
• •The risks of a central neuraxial block must be balanced against the risks of a general anaesthetic.
• •Von Willebrand disease is the commonest inherited bleeding disorder.
• •Gestational thrombocytopenia is the commonest cause of low platelets in pregnancy.
• •Pre-eclampsia may be associated with thrombocytopenia.

The management of obstetric patients with disorders of haemostasis is a common challenge for anaesthetists. The risk of haemorrhage and the safety of a central neuraxial block (CNB) may require careful consideration. To provide safe and effective care, it is important that anaesthetists have an understanding of the pathophysiology and management of common haemostatic disorders in pregnancy.

Overview of normal haemostasis

Normal haemostasis is dependent on close interactions between blood vessels, platelets, clotting factors, anticoagulants and the fibrinolytic system. When a blood vessel is injured, platelets bind to exposed collagen, leading to platelet activation and the formation of initial platelet aggregates. Circulating von Willebrand factor (VWF) also binds to the exposed collagen and aids platelet adhesion. Blood coagulation pathways centre on the generation of thrombin, which cleaves fibrinogen to form fibrin, the structural scaffold that stabilises platelet aggregates at sites of vascular injury (Fig 1).¹

Figure 1.

Cell-based model of coagulation. At sites of vessel damage, the exposure of tissue factor to circulating Factor VIIa leads to the generation of a small amount of thrombin. This thrombin mediates the activation of clotting factors on the surface of activated platelets, leading to a rapid burst in thrombin generation. a, activated; TF, tissue factor; vWF, von Willebrand factor.

The haemostatic response is confined to the site of injury by the action of inhibitory anticoagulant pathways, which help regulate thrombin generation and are critical for controlled haemostatic plug formation.

During pregnancy, there is an increase in several clotting factors, whilst fibrinolytic and anticoagulant (specifically protein S) activities are reduced (Supplementary Table S1).² Whilst this has evolved to minimise bleeding at the time of delivery, it increases thrombotic risk.³

General management

A multidisciplinary approach is crucial when managing a pregnant patient with a disorder of haemostasis. Patients should be counselled before conception about the risks of the haemostatic disorder on pregnancy, delivery and the neonate.⁴ Once pregnant, patients should be reviewed early by obstetricians and haematologists to ensure optimal treatment of the disorder.⁴ Early in the third trimester, a multidisciplinary plan for delivery should be made with the patient and relevant clinicians, including an anaesthetist. This should include monitoring required, mode of delivery, precautions needed for the fetus, method of analgesia and anaesthesia, haemostatic management, postnatal care and management of the neonate.⁴

The departmental protocol for the management of major obstetric haemorrhage should still be followed in patients with a disorder of haemostasis, with the specific management for each disease incorporated.

The decision to perform CNB in these patients can be challenging. The Third National Audit Project (NAP 3) showed the incidence of death or paraplegia secondary to CNB to be low in normal pregnancies (one in 80,000 to one in 360,000).⁵ The extent to which a disorder of haemostasis increases this risk is difficult to quantify but is likely to be small.^6,7 Therefore, haemostatic disorders should not be deemed an absolute contraindication to a CNB.⁷ It should also be remembered that risk is not a binary ‘risk and no risk,’ but is rather a continuum. The risk of CNB must be balanced against the risks of general anaesthesia in pregnancy, including failed tracheal intubation, awareness and aspiration. This decision requires the involvement of senior clinicians and a discussion with the patient so that she can make an informed choice.⁷

Von Willebrand disease

Von Willebrand factor is a large glycoprotein required for platelet adhesion at the site of tissue injury and to stabilise Factor VIII (FVIII).⁸ Plasma VWF concentrations increase by more than two to three times during normal pregnancy, with a rapid return to baseline by around 7 days post-partum.⁹

Von Willebrand disease (VWD) is the commonest inherited bleeding disorder, with an estimated prevalence of 1%, although clinically significant disease occurs in only one in 10,000.⁹ A diagnosis of VWD can be made when the VWF activity is less than 0.3 IU ml⁻¹ in a patient with a history of bleeding.⁹ There are three main types of the disease (Table 1). Type 1 VWD is the most prevalent, accounting for 75% of symptomatic patients and is defined by a partial quantitative deficiency of VWF; in most cases, VWF concentrations increase into the normal range during pregnancy.¹⁰ Type 2 VWD affects around 20% of patients and results from defects in VWF that affect its function. It is further classified according to whether the impact of the defect is on VWF binding to platelets, collagen or FVIII. In Type 3 VWD, there is a total deficiency of VWF.^9,10 Without VWF, survival of FVIII is substantially reduced.

Table 1.

Types of VWD.

Type	Effect	Inheritance	Bleeding propensity	Responsive to desmopressin?	Other therapeutic options	CNB
1	Partial quantitative deficiency	Autosomal dominant	Mild to moderate	Most	TXA Plasma-derived VWF concentrates (not in mild Type 1)	Yes, if normal VWF levels achieved
2A	Selective qualitative deficiency with decreased VWF function as a result of loss of high-molecular-weight multimers	Usually autosomal dominant	Variable, usually moderate	Variable, usually inadequate functional response	VWF-containing concentrates TXA	No
2B	Qualitative deficiency with increased platelet adhesion; can cause thrombocytopenia	Autosomal dominant	Variable, usually moderate	May worsen thrombocytopenia—caution with use	VWF-containing concentrates Platelet transfusion	No
2M	Qualitative deficiency with decreased VWF-dependent platelet adhesion	Autosomal dominant	Variable, usually moderate	Variable, usually inadequate functional response	VWF-containing concentrates TXA	No
2N	Decreased binding to FVIII	Autosomal recessive	Variable, usually mild/moderate	Variable	VWF-containing concentrates TXA	No
3	Complete quantitative deficiency	Autosomal recessive	Severe	No	VWF-containing concentrates Platelet transfusion TXA	No

There is also a group of patients who have a bleeding history and milder reductions in VWF activity (0.3–0.5 IU ml⁻¹). Recently, these patients have been reclassified as having ‘low VWF’, a risk factor for bleeding, but additional causes of bleeding should be sought.⁹

Management of VWD in pregnancy

Women with severe Types 1, 2 and 3 VWD should be managed in high-risk obstetric units with access to clinicians who have expertise in haemophilia.⁴ Delivery of women with Type 1 VWD and ‘low VWF’ in whom VWF concentrations have been shown to increase into the normal range can often be managed in local units in conjunction with input from haematologists.⁴ Although bleeding has rarely been reported in affected neonates, the use of external cephalic version, ventouse and mid-cavity forceps should be avoided in fetuses at risk of Types 2 and 3 VWD.⁹

Pregnant women with VWD have an increased risk of bleeding. Primary postpartum haemorrhage (PPH) occurs in 15–30% of women, and 25% will have a secondary PPH.⁴ Therefore, patients with VWD should have active management of the third stage of labour.⁴ Levels of VWF antigen, a suitable VWF activity assay (such as the ristocetin cofactor assay, which assesses functional platelet-binding activity) and FVIII coagulant activity (FVIII:C) should be checked at booking, in the third trimester and before any invasive procedure.⁴ Although FVIII:C and VWF activity concentrations above 0.5 IU ml⁻¹ are considered acceptable for invasive procedures during pregnancy and management of delivery, these levels are less than normal intra- and postpartum concentrations. Recent data suggest that higher treatment thresholds may provide better protection against PPH.¹¹ After delivery, VWF decreases rapidly. Whilst the optimal dose and duration of haemostatic treatments are not known, it is advised that VWF activity and FVIII:C concentrations be maintained above 0.5 IU ml⁻¹ in the postpartum period for a minimum of 3 days after an uncomplicated vaginal delivery and 5 days after instrumental delivery or Caesarean section.⁴ Women with more severe types of VWD may need to continue VWF replacement therapy for several weeks.⁴

Levels of VWF may be increased using desmopressin, or VWF-containing concentrates.¹⁰ Desmopressin stimulates the release of stored VWF from endothelial cells and can increase VWF three-to five-fold. Response is variable and should be checked before pregnancy.¹⁰ As VWF concentrations normalise during pregnancy in the majority of patients with Type 1 VWD and low VWF concentrations, desmopressin is less useful in this setting. Furthermore, repeated doses should be avoided because of the risk of hyponatraemia, to which the fetus may be particularly sensitive. The need to fluid restrict to 1 L in 24 h can also be problematic around delivery.⁹ Maximal response is achieved with doses of 0.3 μg kg⁻¹ given i.v. or s.c. with time to peak effect of 30–60 and 90 min, respectively.¹² Desmopressin is of limited value in most patients with Type 2 VWD, and the exacerbation of thrombocytopenia in Type 2B VWD is a relative contraindication. There is a theoretical risk of placental insufficiency caused by vasoconstriction, and desmopressin should be avoided in pre-eclampsia.⁹

Plasma-derived VWF containing concentrates is used in most patients with Type 2 and severe Type 1, and all patients with Type 3 VWD. If treatment is deemed necessary, peak plasma VWF activity concentrations of more than 1.0 IU ml⁻¹ should be targeted.⁴

Tranexamic acid (TXA) is an antifibrinolytic that reduces clot lysis by plasmin. It can be used as a sole agent in patients with VWF activity greater than 0.5 IU ml⁻¹, or in combination with other treatments when VWF activity is less than this.⁴ Oral TXA should be started at the onset of labour, or administered intravenously before Caesarean section and continued in the postpartum period.⁹ It may be required for several weeks.⁴ A limited amount of TXA (∼1%) may be secreted in breast milk, but this is not considered likely to produce an antifibrinolytic effect in the infant.⁹

Platelets contain 10% of total blood VWF, and platelet transfusions are used as adjunctive therapy in some patients.¹⁰

I.M. injections should be avoided where FVIII:C concentrations are low, and NSAIDs should generally be avoided when VWF activity is less than 0.5 IU ml⁻¹. Postpartum thromboprophylaxis with low-molecular-weight heparin can be used in patients with VWD who have had adequate correction of VWF activity and FVIII.⁴

Neuraxial anaesthesia and VWD

Central neuraxial blocks can be offered in Type 1 VWD, in which concentrations of VWF activity and FVIII:C are above 0.5 IU ml⁻¹ with a normal coagulation screen and platelet count.⁴ In Types 2 and 3 VWD, UK guidance advises against the use of CNB, even after VWF replacement therapy.⁴ This is because the achievement of normal laboratory measurements of VWF activity does not usually equate to restoration of normal haemostasis in these patients.⁹ As VWF concentrations decrease early in the postpartum period, an epidural catheter should be removed soon after delivery or repeat treatment considered before removal.³

Haemophilia A and B

Haemophilia is an X-linked congenital bleeding disorder caused by a coagulation factor FVIII (haemophilia A) or Factor IX (haemophilia B) deficiency. Typically, the disorder affects males, but female carriers may also have factor concentrations below the normal range. Haemophilia occurs in 1:5,000 male births, 80% of whom have haemophilia A. Severity of the disease is determined by baseline factor concentrations (Table 2).^4,12

Table 2.

Severity of haemophilia.

Severity	Clotting factor level (IU ml−1)	Bleeding episodes
Mild	0.05–0.5	Bleeding with major trauma or surgery Spontaneous bleeding is rare
Moderate	0.01–0.05	Prolonged bleeding with minor trauma or surgery Some spontaneous bleeding
Severe	<0.01	Spontaneous bleeding into joints and muscles

Management of haemophilia A and B carriers in pregnancy

The antenatal care of potential (those women with a positive family history) or known carriers of haemophilia A and B should be undertaken in an obstetric unit with the ability for laboratory monitoring 24 h a day, and in close liaison with a haemophilia centre.¹³ The majority of haemophilia carriers have normal factor concentrations before pregnancy. Those women with low FVIII plasma concentrations would be expected to be in the normal range by the third trimester given the increase in FVIII during pregnancy. Factor IX concentrations do not change during pregnancy.¹² Ideally, baseline factor concentrations should be checked before conception and then at booking, in the third trimester, before any invasive procedure and if there is any bleeding.⁴ Factor concentrations of at least 0.5 IU ml⁻¹ are required, and treatment should be given when below this.⁴

Desmopressin can be used to increase FVIII concentrations but not factor IX.^3,4 Recombinant factor concentrates are the treatment of choice when desmopressin cannot be used; factor concentrations should be closely monitored after replacement therapy.⁴ Tranexamic acid should also be considered, either as a sole agent when factor concentrations are >0.5 IU ml⁻¹, or as an adjunct when below this. It should be continued postpartum until there is normal lochia.⁴

Determining the sex and haemophilia status of the fetus can help inform options for delivery. Neonatal intracranial bleeding rates are lowest when delivery is by elective Caesarean section, and this should be considered in male fetuses known to be affected or where antenatal diagnosis has not been performed.⁴ When vaginal delivery is planned, ventouse extraction and midcavity forceps should be avoided. No special measures are generally required for delivery of a female fetus to a haemophilia carrier, and delivery may be managed in a local unit.⁴

Factor VIII or IX concentrations should be maintained above 0.5 IU ml⁻¹ for a minimum of 3 days after vaginal delivery and 5 days after a Caesarean section.⁴ Factor VIII concentrations decrease rapidly after delivery, and the risk of primary and secondary PPH in a haemophilia carrier is high, even in those women who have received treatment or have factor concentrations >0.5 IU ml⁻¹.^4,14 I.M. injections and pharmacological thromboprophylaxis can be used when factor concentrations are in the normal range.⁴

CNB and carriers of haemophilia A and B

Factor VIII and IX concentrations should be above 0.5 IU ml⁻¹ before a CNB, and should also be checked and normalised before the epidural catheter is removed.⁴

Thrombocytopenia

Thrombocytopenia is defined as less than 150×10⁹ platelets L⁻¹ of blood. It affects 7–11% of all pregnancies and after anaemia is the commonest haematological disorder in pregnancy.¹⁵ The cause of thrombocytopenia may be pregnancy specific, independent of but worsened by pregnancy, or relate to a preexisting disease (Table 3).¹⁶

Table 3.

Causes of thrombocytopenia in pregnancy. AFLP, acute fatty liver of pregnancy; TTP, thrombotic thrombocytopenic purpura.

Pregnancy specific	Pregnancy associated	Non-pregnancy associated
•GT •Hypertensive syndromes •Pre-eclampsia •HELLP •AFLP	•TTP •Disseminated intravascular coagulation •Type 2B VWD	•ITP •Hereditary •Marrow infiltration •Viral infection •Hypersplenism •Drugs (e.g. heparin)

Platelet counts above 50×10⁹ L⁻¹ are targeted before delivery, as surgical bleeding is less likely to occur above this level.¹⁵ This threshold assumes platelets to be functionally normal. Some forms of hereditary thrombocytopenia also affect platelet function, highlighting the importance of personal bleeding and family history.

There are no studies that define the lowest platelet count at which it is safe to perform a CNB, although guidance published by the Association of Anaesthetists provides expert opinion.⁷ A large study by Lee and colleagues estimated the risk of an epidural haematoma to be 0.2% with platelet counts between 70 and 100×10⁹ L⁻¹. Estimates of risk with platelet counts below this (3% with counts between 50×10⁹ and 69×10⁹ L⁻¹, and 11% with counts between 0 and 49×10⁹ L⁻¹) should be used with caution.¹⁷

The risk of an epidural or spinal haematoma must be weighed against the benefits of superior analgesia and the avoidance of general anaesthesia. The cause of the thrombocytopenia, the rate at which the platelet count is decreasing, the bleeding history and the risks of a general anaesthetic for an individual parturient must be considered (Table 4).^7,18

Table 4.

Selected causes of thrombocytopenia in pregnancy. AFLP, acute fatty liver of pregnancy; DBP, diastolic blood pressure; LDH, lactate dehydrogenase; LFT, liver function test; SBP, systolic blood pressure.

Disease	Incidence in pregnancy (%)	Platelets	Features	CNB	Other
GT	5–9	Typically >70×109 L−1 Majority above 100×109 L−1 If less than 70×109 L−1 should be investigated for other causes Normal platelet function	Onset often in third trimester Spontaneously resolves Asymptomatic No neonatal effects	Safe if platelets >75×109 L−1 and no other adverse features Can be considered at platelets >50×109 L−1 if stable/no adverse features	May deliver in a midwife-led unit if platelets stable and above 100×109 L−1
ITP	<1	Typically <100×109 L−1 Platelets appear large on a blood film Platelets are hyper-functional	Can be a preexisting disease Can be asymptomatic, minor bruising, petechiae or significant bleeds Onset any trimester Can cause neonatal thrombocytopenia	As for GT	Deliver in an obstetric-led unit with a neonatal unit Avoid ventouse Use mid-cavity forceps with caution
Pre-eclampsia	5–8	Platelet count can decrease rapidly Reduction in platelet count may precede other symptoms	Onset after 20 weeks' gestation SBP >140 mmHg or DBP >90 mmHg with proteinuria May have headache, visual disturbance, subcostal pain, and oedema	Safe if platelets >100×109 L−1 Can be considered if platelets >75×109 L−1 and stable with normal clotting studies	Deliver in an obstetric-led unit with a neonatal unit
HELLP	<1	Platelet count can decrease rapidly Reduction in platelet count may precede other symptoms	Raised LFTs and LDH Micro-angiopathic haemolytic anaemia Features of pre-eclampsia 15% have no hypertension or proteinuria	Could be considered if platelets >75×109 L−1, and stable platelet count and clotting studies should be checked immediately before the procedure Treat as severe pre-eclampsia	Deliver in an obstetric-led unit with a neonatal unit
AFLP	<0.01	Typically >50×109 L−1	Onset third trimester Abdominal pain and jaundice Hepatic encephalopathy Raised LFTs, ammonia, creatinine Deranged clotting studies Hypoglycaemia	Case reports of use in a well patient with no bleeding history and normal clotting studies	Referral to an ICU or a specialist liver unit may be required

Decisions regarding the removal of an epidural catheter can also be challenging. A platelet count that is satisfactory at insertion may decrease subsequently. Minimum platelet counts for catheter removal are also not defined. A proportion of the vertebral haematomas seen in the NAP 3 study occurred at the time of catheter removal, and the consensus document notes that catheter removal is not a benign process.^5,7 It is therefore prudent to follow the same criteria for epidural removal as for epidural insertion.

Gestational thrombocytopenia

Gestational thrombocytopenia (GT) accounts for 75% of patients presenting with low platelets in pregnancy. The majority have platelet counts above 100×10⁹ L⁻¹, but 10% are below this.¹⁶ It usually presents in the third trimester, resolves spontaneously after delivery, causes no adverse fetal effects and patients are asymptomatic.¹⁹ It tends to recur in subsequent pregnancies, but in between the platelet count remains normal.¹⁵ The pathophysiology remains unknown, but theories include insufficient thrombopoietin and increased consumption and turnover.¹⁶ A recent study showed that platelet counts decrease by an average of 17% during pregnancy, so those women with a low-to-normal starting platelet count often fall into the thrombocytopenic range.²⁰

Immune thrombocytopenia

Immune thrombocytopenia (ITP) affects one in 1,000 pregnancies, and is the commonest cause of low platelets in the first and second trimesters.¹⁶ The platelet count may decrease in the third trimester alone, making it difficult to distinguish from GT. It is an autoimmune disease, in which immunoglobulin G (IgG) antibodies form against glycoprotein complexes on the platelet membrane. When the antibody binds to the platelet, it is sequestered and destroyed by the spleen and reticuloendothelial system.¹⁶ In around 60–70% of patients, it is a preexisting disease, but it may present in pregnancy and is characterised by an isolated thrombocytopenia with platelet counts less than 100×10⁹ L⁻¹.²¹ The majority of patients have no symptoms or just minor bruising, and there are no specific tests. Some do experience serious bleeding events, such as intracranial haemorrhage or gastrointestinal bleeding, but platelet counts in these patients have usually decreased to less than 30×10⁹ L⁻¹.²² The antibodies cross the placenta and may cause destruction of platelets in the newborn, resulting in neonatal thrombocytopenia. This is rarely clinically significant, and mode of delivery should be determined by obstetric indications alone.²¹ However, ventouse extraction should be avoided and mid-cavity forceps used with caution in those women delivering vaginally.²³

Management of GT and ITP

The isolated thrombocytopenia seen in GT is rarely of concern, especially if counts remain above 100×10⁹ L⁻¹. Management is observational and requires no specific treatment. Platelet counts less than 70×10⁹ L⁻¹, not associated with pre-eclampsia, warrant further investigation as this rarely occurs in GT.¹⁹

In ITP, platelets should be monitored regularly; around 30% of patients will need treatment during pregnancy.¹⁶ In the first and second trimesters, the aim is to prevent major bleeds, and treatment is given if the platelet count decreases below 30×10⁹ L⁻¹ or if there is any bleeding.²¹ First-line therapy is usually corticosteroids (e.g. prednisolone 10–20 mg daily).²¹ Two-thirds of patients will show improvement in their platelet counts within 5–7 days. Patients who do not respond to steroids or who require a more rapid response are given i.v. IgG (IVIg). Although IVIg causes a rapid increase in the platelet count, the duration of effect is variable and often short-lived. A single dose of 0.5 g kg⁻¹ is used, but it causes malaise, headache and there is a risk of haemolysis.^16,21 Platelet transfusion is not generally recommended except in the emergency setting and should be given with IVIg, as transfused platelets are rapidly consumed and do not raise the platelet count for long. Dexamethasone or high-dose methylprednisolone with IVIg can also be used for uncontrolled bleeding.²¹

CNB in patients with GT and ITP

Platelet function is normal in GT and can be considered to be hyperfunctional in ITP.²² Stable platelet counts above 75×10⁹ L⁻¹, in the absence of bruising or history of bleeding and a normal coagulation screen, are deemed to carry no increased risk from CNB.^7,19 An experienced anaesthetist may consider a CNB with stable platelet counts above 50×10⁹ L⁻¹ after a careful risk–benefit assesment.⁷ A smaller needle is less likely to cause vascular injury, and therefore, a spinal may be safer than an epidural/combined spinal–epidural.¹⁹

Pre-eclampsia and haemolysis, elevated liver enzymes and low platelet count syndrome

Pre-eclampsia and haemolysis, elevated liver enzymes and low platelet count (HELLP) syndrome are on the spectrum of the same disease process.¹⁹ These are multisystem disorders, and abnormal haemostasis is just one part of their physiological derangement.

Pre-eclampsia is a hypertensive disorder of pregnancy with associated proteinuria and occurs after 20 weeks' gestation.²⁴ It is the second most common cause of thrombocytopenia in pregnancy.¹⁵ Patients may complain of severe headache, visual disturbance, subcostal pain, swelling of their hands and feet and there may be a family or personal history of the disease.²⁵ HELLP occurs as a complication in 10–20% of patients with severe pre-eclampsia.²⁶ Endothelial damage, which is thought to be the cause of the disease process, activates the coagulation cascade and causes a consumption of platelets and clotting factors. The severity of thrombocytopenia correlates with the severity of the disease, and platelet counts can decrease rapidly.²⁴

Management of pre-eclampsia and HELLP

Pre-eclampsia has wide-ranging implications for the provision of anaesthesia and analgesia. It is a large topic that has been discussed previously in this journal.²⁵ In general, management is supportive with arterial blood pressure control, seizure prophylaxis and careful fluid balance, but the disease process will not reverse until the fetus is delivered. In pre-eclampsia, platelet counts above 100×10⁹ L⁻¹ are not likely to carry an increased risk of vertebral canal haematoma with CNB.⁷ More caution is required for platelet counts between 75 and 100×10⁹ L⁻¹.⁷ When platelet counts are less than 100×10⁹ L⁻¹, a coagulation screen (prothrombin time, activated partial thromboplastin time and fibrinogen concentration) should be performed. As platelet counts can decrease rapidly in pre-eclampsia, a platelet count and coagulation screen within the previous 6 h should be available before performing a CNB.⁷ In HELLP, platelet counts should be checked immediately before siting a CNB.⁷

Conclusion

Disorders of haemostasis are not uncommon in pregnancy, and they may have a significant impact on the analgesia and anaesthesia that can be offered to the parturient. An understanding of the common causes, their pathophysiology and treatment will inform the management of major haemorrhage, and will help to determine the risks and optimal timing of CNB.

Funding

CMM is supported by the National Institute for Health Research Imperial Biomedical Research Centre.

Declaration of interests

CMM has received research support from Baxter, CSL Behring and Grifols; speaker's fees from CSL Behring, LFB, Octapharma and Takeda; consultancy fees from CSL Behring, LFB, Novo Nordisk and Takeda; and educational support from Novo Nordisk, Octapharma and SOBI. RA and JC declare no conflict of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Rita Agarwala FRCA is a specialty registrar in anaesthesia at Queen Charlotte's and Chelsea Hospital.

Carolyn Millar MD FRCP FRCPath is a senior lecturer at the Centre for Haematology, Imperial College London and a consultant haematologist at Imperial College Healthcare NHS Trust. Her clinical and research interests include von Willebrand factor and von Willebrand disease, and she leads the obstetric haematology service at Queen Charlotte's and Chelsea Hospital.

Jeremy Campbell MRCSEd FRCA is a consultant anaesthetist at Queen Charlotte's and Chelsea Hospital whose interests include obstetric anaesthesia, anaesthesia for renal transplantation and major gynaecological cancer surgery.

Matrix codes: 1A01, 2B01, 3B00

Appendix A. Supplementary data

The following is the Supplementary data to this article: