Abstract

Venous thomboembolism (VTE) causes only about 2% of maternal deaths in the developing world but is a leading cause of direct maternal deaths in developed countries. Pregnancy increases the risk of VTE through venous stasis, changes in blood coagulability and damage to vessels. Early diagnosis of VTE depends crucially on awareness of the condition but clinical diagnosis is unreliable in pregnancy and objective testing is essential. Compression or duplex ultrasonography is used to diagnose deep venous thrombosis and a ventilation/perfusion scan for pulmonary embolism. Low molecular weight heparins are safe and effective for treatment and for thromboprophylaxis in pregnancy. All women should undergo risk assessment in early pregnancy or preferably before pregnancy. Identifying risk factors such as obesity, or a past or family history of thromboembolism, allows at-risk women to be offered thromboprophylaxis. Guidelines on thromboprophylaxis have reduced deaths after caesarean section and are now being developed for all women.

Introduction

During pregnancy and the puerperium the risk of thromboembolism is increased compared to the non-pregnant state. In most parts of the world this is overshadowed by other complications of pregnancy which account for a much greater proportion of maternal mortality and morbidity. In developed countries, however, as mortality from other causes has been reduced, thromboembolism has emerged as a leading cause of maternal death.

Thromboembolism occurs mainly in women with other risk factors in addition to pregnancy. Attention is now focused on identifying such women and targeting prophylactic measures, and this will be discussed later in this chapter. Recognition and appropriate management of early disease are also important if morbidity and mortality are to be reduced, and this chapter discusses diagnosis and treatment of thromboembolic disorders as well as their prevention.

Incidence

Venous thromboembolic disease has an estimated annual incidence in developed countries of 1 in 1000 people1. The figure increases with age from 1 in 100,000 in childhood to nearly 1% in old age2. Among non-pregnant young women the incidence is 1 in 20,000 per year (and 1 in 6600 among those using second-generation oral contraceptives)3. In pregnancy the incidence is 1 in 1200 pregnancies but it is also related to age4: in women under 35 it is about 1 in 1600 pregnancies3,5—about 12 times that in non-pregnant young women—but over the age of 35 it is about 1 in 8004,5.

The mortality is 1–2%2. Pregnancy-associated death from thromboembolism occurs once in around 70,000 pregnancies6—again, a 12-fold increase compared to the non-pregnant state, where the risk is around one in a million. By comparison, the contraceptive pill increases the risk approximately three-fold7. The risks of thrombosis with the pill are better known than those of pregnancy, and it is regrettable that inaccurate perceptions are helping to restrict the pill’s availability in countries where it could help to reduce maternal mortality by preventing unwanted pregnancy8.

Contribution to maternal mortality

Globally, thromboembolism is not among the leading causes of maternal mortality as estimated by the World Health Organization. Most of the world’s maternal deaths occur in developing countries, where other causes predominate. In a hospital-based study in rural Ghana, embolism accounted for 2.2% of direct deaths (five out of 156 cases)9 and in the second report on Confidential Enquiries into Maternal Deaths in South Africa (1999–2001), thromboembolism also caused 2% of direct deaths (48 out of 1462 cases)10. In a comparison of hospitals in Zambia and the USA, pulmonary embolism did not figure among the 45 deaths in the Zambian hospital but caused 15% of the 20 deaths in the US hospital11.

In the USA in 1991–1997 there were 3201 pregnancy-related deaths, with hypertensive disorders accounting for 509, infection for 422, haemorrhage for 401, and thrombotic embolism for 32712. Compared to the UK, thromboembolism seems under-represented, bearing in mind the importance of obesity as a risk factor in both countries. The pattern in the USA may be due to suboptimal management of the other conditions (there has been no improvement in maternal mortality in the USA since 1982)12 or to under-reporting of thromboembolic deaths, many of which occur outside hospital.

In Europe there is a variable picture of under-reporting of direct deaths13,14 but data from the UK are among the most accurate. In England and Wales thromboembolism has been one of the major causes of maternal mortality since 1952, when confidential enquiries into maternal deaths began. Table 1 shows that deaths from all the leading causes fell between 1952 and 1984, but the fall in deaths from pulmonary embolism was less than the others.

Table 1

Main causes of direct maternal death in England and Wales, 1952–1984

 Abortion Pulmonary embolism Haemorrhage Hypertensive disease All other causes Total 
1952–54 153 138 220 246 369 1094 
1961–63 139 129 92 104 228 692 
1970–72 81 61 27 47 139 355 
1973–75 29 35 21 39 111 235 
1976–78 19 45 26 29 108 227 
1979–81 14 23 14 36 89 176 
1982–84 11 25 25 68 138 
 Abortion Pulmonary embolism Haemorrhage Hypertensive disease All other causes Total 
1952–54 153 138 220 246 369 1094 
1961–63 139 129 92 104 228 692 
1970–72 81 61 27 47 139 355 
1973–75 29 35 21 39 111 235 
1976–78 19 45 26 29 108 227 
1979–81 14 23 14 36 89 176 
1982–84 11 25 25 68 138 

Adapted from CEMD reports15,16.

Since 1973 in England and Wales there has been no reduction in the number of deaths from pulmonary embolism, and from 1991 onwards it has been the leading cause of maternal death in the UK, with 30 deaths in 1991–1993 and 46 in 1994–1996. In 1997–1999 there were 31 deaths, equating to a rate of 14 per million maternities. In addition, however, there were 14 ‘late’ deaths from this cause, so the true rate is probably 50% higher.

In 1997–1999 more than half of the antepartum deaths were in the first trimester, before the woman had booked for antenatal care. Table 2 shows the striking reduction in deaths after caesarean section (despite an increasing national caesarean section rate), which is discussed in the final section of this chapter. Of the 10 deaths after vaginal delivery, most occurred after the woman had been discharged from hospital (Table 3).

Table 2

Deaths from pulmonary embolism (excluding late deaths), UK, 1985–99

 Antepartum deaths Deaths after caesarean section Deaths after vaginal delivery Total (including deaths after miscarriage, ectopic and in labour) 
1985–87 16 30 
1988–90 10 24 
1991–93 12 13 30 
1994–96 15 15 10 46 
1997–99 13 4a 10 31 
 Antepartum deaths Deaths after caesarean section Deaths after vaginal delivery Total (including deaths after miscarriage, ectopic and in labour) 
1985–87 16 30 
1988–90 10 24 
1991–93 12 13 30 
1994–96 15 15 10 46 
1997–99 13 4a 10 31 

From ‘Why Mothers Die’6.

a

Another three deaths followed caesarean section which was carried out after antepartum collapse.

Table 3

Interval between delivery and death from pulmonary embolism, UK, 1997–99

 Days post-partum
 
    Total 
 0–7 8–14 15–28 29–42 43–365  
Vaginal delivery 18 
Caesarean sectiona – – 
Total 23 
 Days post-partum
 
    Total 
 0–7 8–14 15–28 29–42 43–365  
Vaginal delivery 18 
Caesarean sectiona – – 
Total 23 

From ‘Why Mothers Die’6.

a

Another three deaths followed caesarean section carried out after antepartum collapse.

Of the 18 deaths after vaginal delivery (including eight late deaths), only one followed instrumental delivery. Most of the women who died in 1997–1999 had risk factors in addition to pregnancy and in particular many were overweight.

Pathology

Pathogenesis of thromboembolism

Virchow’s triad, described almost 150 years ago2, consists of venous stasis, changes in the vessel wall and changes in the composition of the blood. Pregnancy affects all three factors. Firstly, progestogen-mediated changes to the blood vessels cause venous stasis, which begins by the end of the first trimester and is greatest at 36 weeks. Compression of the pelvic veins adds to this in later pregnancy and affects particularly the left side: almost 90% of deep venous thrombosis (DVT) is on the left side in pregnancy, compared to 55% in the non-pregnant state. More worryingly, 72% of DVT in pregnancy is ileofemoral (compared to 9% in the non-pregnant) and therefore more likely to embolize5.

Secondly, the endothelium of the walls of the pelvic veins may be damaged during vaginal delivery or at caesarean section. Thirdly, clotting factors alter in pregnancy. There are increases in factors V and VIII and fibrinogen, acquired resistance to activated protein C (an endogenous anticoagulant) and a reduction in its co-factor, protein S. There are also increases in inhibitors of plasminogen activator, resulting in impaired fibrinolysis5.

Congenital thrombophilia

Since the first description in 1965 of a family with hereditary thrombosis2, a number of thrombophilias have been discovered. The most common abnormality underlying venous thromboembolism is factor V Leiden, and the other main thrombophilias include deficiencies of protein C and protein S. Antithrombin III deficiency is the rarest but carries the highest thrombogenic risk17.

Clinical thrombosis is now recognized to be a multicausal disease resulting from interaction between congenital and acquired risk factors2,5. Thrombophilias interact with pregnancy. Among pregnant women with antithrombin deficiency the incidence of thromboembolism has been estimated at 32–44%5. With protein C and protein S deficiency the risks of thrombosis in pregnancy have been estimated at 3–10% and 0–6%, respectively.

Consequences of thromboembolism

As well as fatal pulmonary embolism (PE), venous thrombosis results in a disabling post-thrombotic syndrome in at least 20% of patients2. Chronic venous hypertension causes limb pain, swelling, hyperpigmentation, dermatitis, ulcers, venous gangrene and lipodermatosclerosis1. In one survey of patients 9–144 months (mean 41 months) after DVT, 49% had pain, 26% pigmentation, 21% oedema and 3% ulceration18. Post-thrombotic syndrome has a significant impact on quality of life, which worsens with increasing severity of the syndrome19.

Diagnosis

The diagnosis of thromboembolism depends first and foremost on awareness of the condition. Successive reports from the UK confidential enquiries have included the histories of women whose doctors—in general practice, accident and emergency departments and specialist units—failed to recognize classic symptoms, even when risk factors such as obesity or a family history were present.

Symptoms and signs

With DVT, pain or swelling of the leg (particularly the left leg) are the common presenting complaints. Other features include tenderness, increased temperature, lower abdominal pain or a raised white cell count. Clinical diagnosis is unreliable, however, and less than half the cases suspected clinically are confirmed on objective testing. Even in the non-pregnant woman, individual signs and symptoms are of little value and ‘Homan’s sign’ is of no value. In pregnancy, pitting oedema may be physiological and structured clinical models for diagnosis may not be valid1,20. Nevertheless, features such as a history of immobilization, asymmetrical calf swelling (>3 cm difference in calf circumference measured 10 cm below the tibial tuberosity) or swelling of the entire leg are still important in pregnancy1,21.

PE normally produces symptoms and signs and rarely causes sudden death without warning. Dyspnoea is the main clinical feature, usually associated with chest pain. Other features are faintness or collapse, haemoptysis, focal signs in the chest, raised jugular venous pulse and signs or symptoms of DVT.

The Royal College of Obstetricians and Gynaecologists (RCOG) has produced a detailed guideline on the management of thromboembolic disease in pregnancy and the puerperium21. It points out that in pregnancy the vast majority of DVTs are ileofemoral and will therefore require treatment, and advises that when the condition is suspected clinically, anticoagulant treatment should be started until an objective diagnosis is made. It states: ‘Any woman with signs and symptoms suggestive of venous thromboembolism should have objective testing carried out expeditiously to avoid the risks, inconvenience and costs of inappropriate anticoagulation. Individual hospitals should have an agreed protocol for the objective diagnosis of suspected venous thromboembolism during pregnancy.’

Screening tests

D-dimers are derivatives of fibrin, produced by plasmin degradation, and several assays are available to measure circulating levels. In the non-pregnant, D-dimer testing is useful because of the high negative predictive value if the blood level is low, and this is also true in pregnancy. D-dimer levels can be elevated in pregnancy, however, particularly if there is also pre-eclampsia, so a positive test is unhelpful.

Definitive investigations

For DVT, the diagnostic gold standard has been contrast venography but this is invasive and ultrasonography is now preferred. Compression ultrasound involves gently pressing the vascular lumen with the ultrasound probe: a fully compressible vein indicates the absence of thrombosis. Duplex ultrasonography involves additionally evaluating the blood flow characteristics with pulsed Doppler: normal flow is phasic with respiration, and absence of this phasic pattern indicates venous outflow obstruction1.

If ultrasound is negative and there is a low index of clinical suspicion, anticoagulant treatment can be discontinued but if there is a high level of clinical suspicion, anticoagulant treatment should be continued and the ultrasound examination repeated in 1 week. X-ray venography may be considered. If the repeat examination is also negative, treatment may be discontinued then20,21.

For PE, the cornerstone of diagnosis is the ventilation/perfusion (V/Q) scan, though in many places spiral computed tomography is now the first-line imaging technique in the non-pregnant22. When PE is suspected in pregnancy, ultrasound examination for DVT should be carried out20,21 and a chest X-ray should be done if another diagnosis is considered possible20. If these examinations are negative a V/Q scan should be carried out, though a perfusion scan alone may be sufficient. If the V/Q scan reports a low probability of PE, treatment should nevertheless be continued if either the ultrasound investigation is positive or the clinical suspicion is high, in which case further investigation by angiography, magnetic resonance imaging or spiral computed tomography (if postpartum) should be considered.

A pregnant woman who is advised to undergo radiological investigation will be concerned about the risk to her baby. She can be reassured that the radiation dose from either a chest X-ray or venography with abdominal shielding is extremely low20. The reports of the UK Confidential Enquiries into Maternal Deaths have repeatedly stressed that chest X-rays are not contraindicated in pregnancy, and the most recent report advises that ‘pregnancy is not a reason for withholding plain X-ray films of the abdomen, chest X-rays, some CT scans or MRI from sick women’6.

Treatment

Patients with suspected thromboembolism should be treated with heparin until the diagnosis has been confirmed or excluded, unless treatment is strongly contraindicated. A thrombophilia screen, full blood count and coagulation screen should be checked before treatment, bearing in mind that pregnancy may affect the thrombophilia screen (for example, protein S levels fall in normal pregnancy).

Initial treatment

Standard initial treatment has been unfractionated heparin, which reduces the risk of extension of thromboembolism compared to oral anticoagulation only. The most common mistake when starting treatment is failure to achieve adequate anticoagulation23 but achieving the target activated partial thromboplastin time (APTT) may be difficult in late pregnancy, when pregnancy-induced changes produce apparent heparin resistance, and monitoring may require measurement of factor Xa levels. The need for APTT monitoring can be avoided by the use of low molecular weight heparins.

Intravenous unfractionated heparin remains the treatment of choice in massive PE, however, because of its rapid action. With DVT, subcutaneous unfractionated heparin is an effective alternative. Detailed dosage regimens are given in the RCOG guideline on thromboembolic disease in pregnancy and the puerperium21, which is also available at www.rcog.org.uk.

Low molecular weigh heparins (LMWH) have been shown in non-pregnant patients to be more effective than unfractionated heparin and warfarin, and to carry a lower risk of haemorrhagic complications21,24. They are safe in pregnancy25,26, have a lower risk of heparin-induced thrombocytopaenia27 and appear to carry a lower risk of osteoporosis than unfractionated heparin28,29. The RCOG guideline recommends a twice-daily dosage regimen, with the exact dose being related to the patient’s body weight21, but a recent systematic review concluded that once daily treatment is as effective as twice-daily treatment30. The peak anti-Xa activity should be checked 3 h after injection but frequent monitoring does not appear to be necessary and need not be done until the next routine working day after starting treatment. The platelet count should be checked 7–9 days after starting treatment.

With DVT, the affected leg should be elevated and a graduated compression stocking applied. If the viability of the leg is threatened by DVT or if there is massive life-threatening PE, embolectomy or thrombolytic therapy may be needed. Recurrent PE may indicate insertion of a filter in the inferior vena cava.

Maintenance treatment of DVT or PE

Oral anticoagulants are generally avoided in pregnancy because they cross the placenta. They can cause embryopathy in the first trimester and central nervous system abnormalities at any time during pregnancy, and their anticoagulant effect on the fetus may cause haemorrhagic problems.

Subcutaneous heparin, in the form of either unfractionated heparin or LMWH, can be given for the remainder of the pregnancy, in an adjusted-dose regimen. LMWH are preferable because they are simpler to administer and women can be taught to self-inject. Treatment can be monitored on an outpatient basis, and the platelet count should be checked monthly to detect heparin-induced thrombocytopaenia. Treatment should be continued for at least 6 months and for at least 6–12 weeks after delivery.

Warfarin may be used after delivery. Women taking warfarin may breastfeed without risk to the baby, and the same is probably true for LMWH. Graduated elastic compression stockings should be worn on the affected leg for 2 years after the event, as this substantially reduces the risk of post-thrombotic syndrome. Some authorities have suggested that they should be worn for 5 years after the acute event31.

Anticoagulant therapy during labour and delivery

During labour, heparin should be given in a thromboprophylactic dose, returning to a therapeutic dose after delivery. Epidural anaesthesia should not be used until at least 12 h after the last prophylactic dose of LMWH32, or 24 h after the last therapeutic dose. Afterwards, the epidural cannula should not be removed until 10–12 h after the last injection and at least 4 h should then elapse before the next LMWH injection.

For elective caesarean section, the heparin injection should be omitted on the morning of the operation. A prophylactic dose can be given 3–4 h after the operation and a therapeutic dose in the evening. Wound drains should be considered at operation, and the skin closed with interrupted sutures or staples.

Women at high risk of haemorrhage should be treated with intravenous unfractionated heparin, which has a shorter half-life than LMWH and can be more completely reversed with protamine sulphate.

Prevention

The first advance in preventing deaths from VTE was in the 1960s when the importance of early mobilization after normal delivery was recognized. In the 1950s most deaths from VTE occurred after vaginal delivery and the numbers fell when the practice of ‘lying in’—enforced bed rest after normal delivery—was abolished.

The next advance was the reduction in deaths from VTE after caesarean section in the UK in the late 1990s6. This followed the publication in 1995 of RCOG guidelines on thromboprophylaxis for operative procedures including caesarean section33. The number of deaths from postoperative embolism fell from 15 in 1994–6 to four in 1997–9 (Table 2), despite a rise in the caesarean section rate34. Table 3 shows that there was no increase in the number of late deaths, indicating that VTE deaths after caesarean section were not merely being delayed.

This dramatic improvement suggested that deaths from thromboembolism can be prevented if women at risk are identified and guidelines on prophylaxis are followed. Guidelines for thromboprophylaxis in vaginal deliveries were included in Why Mothers Die 1997–19996, and in June 2003, the RCOG published draft guidelines for thromboprophylaxis during pregnancy and after normal delivery35.

The key to prevention, as with diagnosis, is awareness of the condition. Recent reports of the UK Confidential Enquiries into Maternal Deaths have raised awareness of thromboembolism and especially of risk factors. Of the 31 women whose deaths were reported in 1997–1999, obvious risk factors were present in 25 cases36. Thirteen women were overweight, five had had a period of bed rest, four had a family history, three had previous thromboembolism, two had undertaken long-haul flights during pregnancy and one had varicose veins. Some had multiple risk factors. In addition, 18 of the women were aged over 30.

Of the six women with no recorded risk factors, five died before 25 weeks’ gestation. Thrombophilia may have been a factor in these deaths. Five of the late deaths were associated with the oral contraceptive pill: all five women were overweight and in two their obesity should have contraindicated the combined pill36.

Risk assessment and thromboprophylaxis

All women should undergo an assessment of their risk for VTE, ideally before pregnancy, or failing that, in early pregnancy. Risk is high in the first trimester and death may occur before the antenatal booking visit. The assessment should be repeated if the woman is admitted to hospital or develops other problems.

LMWH is effective for thromboprophylaxis in pregnancy37. The draft RCOG guideline suggests that a woman with three or more of the risk factors in Table 4 should receive prophylactic LMWH antenatally and for at least 3–5 days postpartum. A woman with two or more risk factors should receive prophylactic LMWH for 3–5 days after normal delivery.

Table 4

Risk factors for venous thromboembolism in pregnancy and the puerperium

Pre-existing New onset/transient 
Previous VTE Surgical procedure in pregnancy or puerperium 
Family history of VTE Immobility (>4 days bed rest) 
Obesity (BMI > 30 kg/m2Hyperemesis 
Thrombophilia Dehydration 
    Congenital Ovarian hyperstimulation syndrome 
    Antithrombin deficiency Infection (e.g. UTI) 
    Protein C deficiency Pre-eclampsia 
    Protein S deficiency Excessive blood loss 
    Factor V Leiden Long haul air travel 
    Prothrombin gene variant Prolonged labour 
    Acquired (antiphospholipid syndrome)  
    Lupus anticoagulant  
    Anticardiolipin antibodies  
Age over 35 years  
Parity greater than 4  
Gross varicose veins  
Paraplegia  
Sickle cell disease  
Inflammatory disorders, e.g. inflammatory bowel disease  
Some medical conditions, e.g. nephrotic syndrome, some cardiac diseases  
Pre-existing New onset/transient 
Previous VTE Surgical procedure in pregnancy or puerperium 
Family history of VTE Immobility (>4 days bed rest) 
Obesity (BMI > 30 kg/m2Hyperemesis 
Thrombophilia Dehydration 
    Congenital Ovarian hyperstimulation syndrome 
    Antithrombin deficiency Infection (e.g. UTI) 
    Protein C deficiency Pre-eclampsia 
    Protein S deficiency Excessive blood loss 
    Factor V Leiden Long haul air travel 
    Prothrombin gene variant Prolonged labour 
    Acquired (antiphospholipid syndrome)  
    Lupus anticoagulant  
    Anticardiolipin antibodies  
Age over 35 years  
Parity greater than 4  
Gross varicose veins  
Paraplegia  
Sickle cell disease  
Inflammatory disorders, e.g. inflammatory bowel disease  
Some medical conditions, e.g. nephrotic syndrome, some cardiac diseases  

Adapted from draft RCOG guideline35.

Careful enquiry should be made about a past history or a family history of VTE. A woman with a previous VTE should undergo screening for thrombophilia, ideally before pregnancy. If thrombophilia is excluded, the question of antenatal thromboprophylaxis is controversial. It is generally felt that if the previous VTE was provoked by a temporary risk factor (such as immobilization or trauma) which has now resolved, thromboprophylaxis is not necessary antenatally, though it should be given postpartum.

A woman with a previous VTE and a family history in a first-degree relative should receive antenatal thromboprophylaxis with LMWH. The same applies to a woman with recurrent VTE. A family history suggests thrombophilia. Women with a previous VTE who have thrombophilia should receive antenatal and postpartum thromboprophylaxis.

Antenatal screening of all women for thrombophilia has been suggested but has not proven effective38. A large number of women will screen ‘positive’ and it is unrealistic to recommend thromboprophylaxis to a high proportion of healthy pregnant women. Nevertheless, women may present in pregnancy with no personal history of VTE but with a known thrombophilia. Again, the risk of VTE depends on the specific thrombophilia. Antithrombin deficiency carries a 30% risk of VTE in pregnancy. Protein C or protein S deficiency carries an eight-fold increase in risk—mainly of postpartum VTE. Heterozygotes for factor V Leiden are at considerably lower risk. Antenatal thromboprophylaxis is not usually necessary except in those with combined defects, those homozygous for defects, or those with antithrombin deficiency. All women with thrombophilia should, however, receive LMWH or warfarin for 6 weeks after delivery.

In antiphospholipid syndrome (lupus anticoagulant or anticardiolipin antibodies associated with thrombosis or adverse pregnancy outcome), the risk of recurrent thromboses may be 70% or even higher. Pregnant women with antiphospholipid syndrome and previous thrombosis should receive antenatal and postnatal thromboprophylaxis with LMWH. This is not required for women with anticardiolipin antibodies and no history of thrombosis or pregnancy loss.

Timing and duration of thromboprophylaxis

Thromboprophylaxis should begin as early as practicable in pregnancy and should continue until delivery. LMWH should be continued during labour and delivery because of the high risk of initiation of VTE immediately after delivery. For epidural or caesarean section, however, and for women at high risk of haemorrhage, the guidance given in the section on treatment should be followed.

Postpartum thromboprophylaxis should be started as soon as possible after delivery, provided there is no postpartum haemorrhage. LMWH should be started 3 h after delivery (or 4 h after removal of an epidural catheter). Thromboprophylaxis is normally continued for 6 weeks in high risk women, as it takes several weeks for the prothrombotic changes of pregnancy to revert to normal. For women at lower risk (e.g. those with obesity or age >35), prophylaxis for 3–5 days after delivery is usually recommended.

Puerperal women undergoing surgery or travelling by air are at increased risk of VTE. The RCOG has issued advice for pregnant women travelling by air: for those already at increased risk and undertaking long-haul flights LMWH is recommended but low-dose aspirin is an acceptable alternative if LMWH is impractical to administer36,39.

Agents for prophylaxis

LMWH are preferred to unfractionated heparin for thromboprophylaxis. Dosages are given in the RCOG guideline (available at www.rcog.org.uk). If renal function is normal, monitoring of anti-Xa levels is not necessary, except in antithrombin deficiency. Low dose aspirin may be appropriate when the risk of VTE is not high enough to warrant antenatal LMWH, for example in women without thrombophilia but with previous VTE due to a specific cause. Warfarin is suitable for prolonged postpartum thromboprophylaxis. Dextran carries a risk of anaphylaxis and should not be used. Danaparoid has been used in case of heparin-induced thrombocytopaenia40.

All women with previous VTE or thrombophilia should be encouraged to wear graduated elastic compression stockings throughout pregnancy and for 6–12 weeks after delivery, though strong evidence for this recommendation is lacking. The use of such stockings is also recommended for pregnant women travelling by air.

Conclusion

For many years, maternal death from thromboembolism was thought to be sudden and unavoidable. Better understanding of the clinical presentation and risk factors, along with the development of more convenient methods of diagnosis and treatment, and in particular the publication of detailed guidelines for thromboprophylaxis for women at risk, open up the possibility of making a real reduction in this cause of maternal mortality and morbidity.

References

1
Tovey
C
, Wyatt S. Diagnosis, investigation, and management of deep vein thrombosis.
BMJ
 
2003
;
326
:
1180
–4
2
Rosendaal
FR
. Venous thrombosis: a multicausal disease.
Lancet
 
1999
;
353
:
1167
–73
3
British National Formulary 39. London: British Medical Association and Royal Pharmaceutical Society of Great Britian,
2001
4
Simpson
EL
, Lawrenson RA, Nightingale AL, Farmer RD. Venous thromboembolism in pregnancy and the puerperium: incidence and additional risk factors from a London perinatal database.
Br J Obstet Gynaecol
 
2001
;
108
:
56
–60
5
Greer
IA
. Thrombosis in pregnancy: maternal and fetal issues.
Lancet
 
1999
;
353
:
1258
–65
6
Lewis
G
(ed)
Why Mothers Die 1997–1999. Report of the Confidential Enquiry into Maternal Deaths in the UK
 . London: RCOG Press,
2001
7
Drife
JO
. The third generation pill controversy (‘continued’).
BMJ
 
2001
;
323
:
119
–20
8
Drife
J
. Oral contraception and the risk of thromboembolism. What does it mean to clinicians and their patients?
Drug Safety
 
2002
;
25
:
893
–902
9
Geelhoed
DW
, Visser LE, Asare K, Schagen van Leeuwen JH, van Roosmalen J. Trends in maternal mortality: a 13-year study in rural Ghana.
Eur J Obstet Gynecol Reprod Biol
 
2003
;
107
:
135
–9
10
Pattinson
R
(ed)
Saving Mothers 1999–2001. Second Report on Confidential Enquiries into Maternal Deaths in South Africa.
 www.doh.gov.za
11
Kilpatrick
SJ
, Crabtree KE, Kemp A, Geller S. Preventability of maternal deaths: comparison between Zambian and American referral hospitals.
Obstet Gynecol
 
2002
;
100
:
321
–6
12
Berg
CJ
, Chang J, Callaghan WM, Whitehead SJ. Pregnancy-related mortality in the United States,
1991
–1997.
Obstet Gynecol
 
2003
;
101
:
289
–96
13
Salanave
B
, Bouvier-Colle M-H, Varnoux N, Alexander S, Macfarlane A and the MOMS Group. Classification differences and maternal mortality: a European study.
Int J Epidemiol
 
1999
;
28
:
64
–9
14
Schuitemaker
N
, Van Roosmalen J, Dekker G, Van Dongen P, Van Geijn H, Gravenhorst JB. Underreporting of maternal mortality in the Netherlands.
Obstet Gynecol
 
1997
;
90
:
78
–82
15
Department of Health and Social Security.
Report on Confidential Enquiries into Maternal Deaths in England and Wales 1976–1978
 . London: Her Majesty’s Stationery Office,
1982
16
Department of Health.
Report on Confidential Enquiries into Maternal Deaths in England and Wales 1982–84
 . London: Her Majesty’s Stationery Office,
1989
17
Jilma
B
, Kamath S, Lip GYH. Antithrombotic therapy in special circumstances. II. In children, thrombophilia, and miscellaneous conditions.
BMJ
 
2003
;
326
:
93
–6
18
Killewich
LA
, Martin R, Cramer M, Beach KW, Strandness DE. An objective assessment of the physiological changes in the postthrombotic syndrome.
Arch Surg
 
1985
;
120
:
424
–6
19
Kahn
SR
, Hirsch A, Shrier I. Effect of postthrombotic syndrome on health-related quality of life after deep venous thrombosis.
Arch Intern Med
 
2002
;
162
:
1144
–8
20
Chan
WS
, Ginsberg JS. Diagnosis of deep vein thrombosis and pulmonary embolism in pregnancy.
Thromb Res
 
2002
;
107
:
85
–91
21
Royal College of Obstetricians and Gynaecologists.
Thromboembolic Disease in Pregnancy and the Puerperium: Acute Management
 . Clinical Green Top Guideline No. 28. London: RCOG Press,
2001
22
Janata
K
. Managing pulmonary embolism.
BMJ
 
2003
;
326
:
1341
–2
23
Turpie
AGG
, Chin BSP, Lip GYH. Venous thromboembolism: treatment strategies.
BMJ
 
2002
;
325
:
948
–50
24
Working Group on behalf of the Obstetric Medicine Group of Australasia. Anticoagulation in pregnancy and the puerperium.
Med J Aust
 
2001
;
175
:
258
–63
25
Eldor
A
. The use of low-molecular-weight heparin for the management of venous thromboembolism in pregnancy.
Eur J Obstet Gynecol Reprod Biol
 
2002
;
104
:
3
–13
26
Lepercq
J
, Conard J, Borel-Derlon A et al. Venous thromboembolism during pregnancy: a retrospective study of enoxaparin safety in 624 pregnancies.
Br J Obstet Gynaecol
 
2001
;
108
:
1134
–40
27
Blann
AD
, Landray MJ, Lip GYH. An overview of antithrombotic therapy.
BMJ
 
2002
;
325
:
762
–5
28
Jilma
B
, Kamath S, Lip GYH. Antithrombotic therapy in special circumstances. I—pregnancy and cancer.
BMJ
 
2003
;
326
:
37
–40
29
Rodie
VA
, Thomson AJ, Stewart FM, Quinn AJ, Walker LD, Greer IA. Low molecular weight heparin for the treatment of venous thromboembolism in pregnancy: a case series.
Br J Obstet Gynaecol
 
2002
;
109
:
1020
–4
30
van Dongen
CJ
, MacGillavry MR, Prins MH. Once versus twice daily LMWH for the initial treatment of venous thromboembolism (Cochrane Review). In:
The Cochrane Library, Issue 2, 2003
 . Oxford: Update Software
31
Franzeck
UK
, Schalch I, Bollinger A. On the relationship between changes in the deep veins evaluated by duplex sonography and the postthrombotic syndrome 12 years after deep vein thrombosis.
Thromb Haemost
 
1997
;
77
:
1109
–12
32
Crossley
D
, Zych Z. Timing of thromboprophylaxis for general surgery should be discussed with anaesthetists.
BMJ
 
2000
;
321
: 1018
33
Royal College of Obstetricians and Gynaecologists.
Report of a Working Party on Prophylaxis against Thromboembolism in Gynaecology and Obstetrics.
  London: RCOG Press,
1995
34
The National Sentinel Caesarean Section Audit Report. London: RCOG Clinical Effectiveness Support Unit,
2001
35
Royal College of Obstetricians and Gynaecologists.
Guideline: Thromboprophylaxis During Pregnancy and After Normal Vaginal Delivery.
  London: RCOG Press,
2003
36
Drife
J
on behalf of the Editorial Board. Thrombosis and thromboembolism. In: Lewis G (ed)
Why Mothers Die 1997–1999. Report of the Confidential Enquiry into Maternal Deaths in the UK
 . London: RCOG Press,
2001
;
49
–75
37
Hunt
BJ
, Doughty HA, Majumdar G, Copplestone A, Kerslake S, Buchanan N, Hughes G, Khamashta M. Thromboprophylaxis with low molecular weight heparin (Fragmin) in high risk pregnancies.
Thromb Haemost
 
1997
;
77
:
39
–43
38
Lindqvist
PG
, Olofsson P, Dahlback B. Use of selective Factor V Leiden screening in pregnancy to identify candidates for anticoagulants.
Obstet Gynecol
 
2002
;
100
:
332
–6
39
Royal College of Obstetricians and Gynaecologists.
Advice on Preventing Deep Vein Thrombosis for Pregnant Women Travelling by Air
 . Scientific Advisory Committee Opinion Paper 1. London: RCOG Press,
2001
40
Woo
YL
, Allard S, Cohen H, Letsky E, de Swiet M. Danaparoid thromboprophylaxis in pregnant women with heparin-induced thrombocytopenia.
Br J Obstet Gynaecol
 
2002
;
109
:
466
–8