A search for clarity in an uncertain world.

Venous thromboembolism (VTE) is a relatively uncommon complication of pregnancy and the postpartum, with an incidence of around one in 1000–1500 pregnancies.1 It remains an important cause of maternal death in developed countries and was the second most-common cause of direct maternal death in the Australian maternal mortality report from 2003–052 and third most common in the UK.3 In New Zealand, the Maternal Mortality Working Group of the Perinatal and Maternal Mortality Review Committee (PMMRC) reported only two deaths from VTE from a total of the 21 direct maternal deaths between 2006 and 2011.4 Most women who develop pregnancy-associated VTE (PA-VTE) do not die, but studies suggest that around one in 40 of pulmonary emboli (PE) are fatal.5-7 One-third to one-half of PA-VTE occur postpartum5,8,9, most within the first six weeks, with the daily risk of VTE in the six-week postpartum period greater than during the antenatal period.6, 9,10 At least half of the antepartum events occur in the first two trimesters7,11-13, emphasising the importance of pre-conception counselling in high-risk women as well as an assessment of risk factors early in pregnancy.

Risk factors for PA-VTE

Pregnancy predisposes to thrombosis, as all three components of Virchow’s triad are present in pregnancy: venous stasis, induced by venous dilation and obstruction to venous return; increases in procoagulant factors and reduction in natural anticoagulants; and vessel wall injury during labour and following caesarean section (CS). Pregnancy is also a pro-inflammatory state, with activation of endothelial cells. Additional maternal factors further increase the risk of VTE and recognisable risk factors were identified in as many as 75 per cent of women who died from PE in the UK.14 Risk factors for PA-VTE are summarised in Table 1. Some risk factors, such as obesity, immobilisation and VTE, are common to many patient populations, whereas others, such as CS and pre-eclampsia, are specific to pregnancy. Increased BMI is an important and consistent risk factor for PA-VTE.15,16 In the UK, two-thirds of women who died from PE had a BMI >25 and almost 20 per cent were morbidly obese (BMI >40).14 Increased BMI combined with immobilisation significantly increases the risk of PA-VTE.16

Prior history of VTE

Previous VTE is one of the most important risk factors for PA-VTE. The risk of recurrence is higher following previous unprovoked (no identified risk factors) than provoked (associated with a risk factor) events.17 In pregnant women with a prior VTE associated with a major (surgical or traumatic) provoking factor, the risk of antepartum recurrence appears to be low, with no recurrent events seen in this patient group in two studies.17,18 The risk of recurrence in women with either a prior unprovoked or hormonally provoked VTE (pregnancy or oral contraceptive associated) is higher, with antepartum recurrence risk ranging from two to five per cent and two to ten per cent, respectively.18-20 Some women have risk factors that place them at an increased risk of VTE throughout pregnancy and the postpartum period and warrant extended thromboprophylaxis. Others will develop complications during pregnancy or will require thromboprophylaxis only if they are hospitalised.

Inherited thrombophilias

The inherited thrombophilias have gained much attention in the last couple of decades as risk factors for VTE. Factor V Leiden (FVL) and the prothrombin mutation are relatively weak in terms of their prothrombotic risk, but are found in five per cent and two to three per cent of Caucasians, respectively, whereas deficiencies of antithrombin, protein C and protein S are more thrombogenic, but are rare each with a prevalence of around 0.1 per cent.21 While these inherited clotting tendencies are associated with an increased risk of development of an initial VTE, they are less important risk factors for recurrent VTE.

Table 1. Clinical risk factors for PA-VTE

Standard risk
Adjusted OR
Previous VTE 24.8
Age >35 1.4-1.7
Obesity (BMI > 30kgm)* 1.7-5.3
Active medical illness 2.1-8.7
Smoking 1.7-3.4
Family history VTE 2.9-4.1
Immobility 7.7-10.1
Varicose veins 2.4
risk factors
Multiparity (>2) 1.6-2.9
Multiple pregnancy 1.6-4.2
Preeclampsia 3.0-5.8
Assisted reproduction technology 2.6-4.3
Hyperemesis 2.5
Postpartum risk
Planned caesarean section 1.3-2.7
Emergency caesarean section 2.7-4.0
Placental abruption 2.5-16.6
Postpartum infection 4.1-20.2
Postpartum haemorrhage 1.3-12.0

Adapted from McLintock et al.1

Thromboprophylaxis – who should get it?

Prevention of PA-VTE by offering thromboprophylaxis to women at highest risk of an event is clearly a sensible clinical approach. However, just how to identify these high-risk women is hampered by the paucity of evidence from clinical trials with wide estimates of risk that are likely to be subject to various sources of bias. This means that decision-making in this clinical setting and recommendations for thromboprophylaxis are ‘eminence based’ rather than evidence based. An individual’s appreciation of risk is also personal and for these reasons it is difficult to base recommendations for prophylaxis on arbitrary quantitative estimates of risk. Decisions should therefore be made after explaining the available evidence to the patient and taking into account her perception of the balance of risk and benefit in thromboprophylaxis.

An assessment of risk of thromboembolism should be carried out in all pregnant women. Limited evidence suggests that risk factors are synergistic16 therefore a patient with multiple risk factors warrants discussion with regard to thromboprophylaxis, even in the absence of a personal or family history of VTE. There is a higher daily risk of PA-VTE in the postpartum period so the threshold for recommending thromboprophylaxis is lower during this period.

Guidelines have been published by the Royal College of Obstetricians and Gynaecologists in the UK22 and more recently Australian and New Zealand (ANZ) specialists in haematology, obstetric medicine, anesthesia, obstetrics and maternal fetal medicine who are actively involved in this field have published ANZ recommendations for thromboprophylaxis1 and also for diagnosis and management of acute VTE in pregnancy.23 The ANZ guidelines were endorsed by the Australasian Society of Thrombosis and Haemostasis and the Society of Obstetric Medicine of Australia and New Zealand. The National Health and Medical Research Council of Australia has also developed guidelines for thromboprophylaxis for inpatients, including obstetric patients.24 One key difference between the RCOG guidelines and the ANZ recommendations for thromboprophylaxis relates to the importance of the inherited thrombophilias in risk assessment. In the absence of a personal or family history of VTE, the ANZ recommendations do not consider the presence of an inherited thrombophilias, such as heterozygosity for FVL or the prothrombin gene mutation, protein C or protein S deficiency, to be an indication for antenatal thromboprophylaxis if a woman has no other risk factors. The ANZ recommendations also take the prothrombotic potential of the inherited thrombophilias into account, recognising that higher rates of VTE are reported in patients with deficiencies of antithrombin, protein C, protein S or who are homozygous for FVL or the prothrombin gene mutation or compound heterozygotes for these mutations. The RCOG also stipulated that higher doses of LMWH for thromboprophylaxis should be used in woman who weigh more than 90kg. In contrast, the ANZ group did not consider that there was sufficient evidence to make this recommendation.

In the absence of evidence, what should we do?

Development of guidelines or recommendations in the absence of high-level evidence from clinical trials opens the authors to the criticism that thromboprophylaxis will be given to large numbers of women whose risk of developing PA-VTE is relatively low.25 Some groups26 have attempted to develop estimates of the number-needed-to-treat to prevent an thromboembolic event. However, in the absence of randomised clinical trials in pregnancy cohorts, accurate statistics are not available and extrapolating efficacy rates with thromboprophylaxis from other patient groups is problematic.

So where does that leave clinicians? Very few guidelines or recommendations would be published in medicine if we had to wait for incontrovertible truths to be available to support them. In our imperfect world, most clinicians are happy to have pragmatic, carefully considered recommendations that provide guidance for clinical decision-making, while clearly identifying the limitations of the data that support these recommendations and acknowledging the lack of certainty. Development of eminence-based recommendations should not close the door to continuing to conduct more robust clinical research. To answer the important clinical questions relating to prevention of PA-VTE will require international multicentre trials – no easy feat, but hopefully attainable.