Thromboembolism in pregnancy is an important clinical issue. Despite identification of maternal and pregnancy-specific risk factors for development of pregnancy-associated venous thromboembolism, limited data are available to inform on optimal approaches for prevention. The relatively low overall prevalence of pregnancy-associated venous thromboembolism has prompted debate about the validity of recommendations, which are mainly based on expert opinion, and have resulted in an increased use of pharmacological thromboprophylaxis in pregnancy and postpartum. A pragmatic approach is required in the absence of more robust data. Anticoagulation management of pregnant women with mechanical prosthetic heart valves is particularly challenging. Continuation of therapeutic anticoagulation during pregnancy is essential to prevent valve thrombosis. Warfarin, the most effective anticoagulant, is associated with adverse fetal outcomes, including embryopathy and stillbirth. Fetal outcome is improved with therapeutic-dose low-molecular-weight heparin, but there may be more thromboembolic complications. More intensive anticoagulation, targeting higher trough anti-Xa levels, may reduce the risk of valve thrombosis.
Pregnancy-associated venous thromboembolism
Venous thromboembolism (VTE) is an infrequent but potentially preventable cause of maternal morbidity and mortality. Maternal mortality associated with pregnancy-associated venous thromboembolism (PA-VTE) is reported to be between 0.4 and 1.6 per 100,000 pregnancies in developed countries, and is one of the most common causes of maternal death . Deciding who would benefit from thromboprophylaxis requires an understanding of when thromboembolic events are likely to occur and which women are at risk of developing this complication. The relative infrequency of PA-VTE means that most clinicians have limited experience, and clinical trial data are insufficient to inform on best practice. Published guidelines and recommendations are ‘eminence based’ rather than ‘evidence based’.
Epidemiology and risk factors for pregnancy-associated venous thromboembolism
The reported rate of PA-VTE from 15 epidemiological studies ranges from 0.61 to 2.22 per 1000 deliveries ( Table 1 ), which represents around 5–10-fold increase compared with the rate in non-pregnant women of the same age . Across the studies , an average of 23.5% of events are diagnosed as pulmonary embolism, of which up to one in 30 are fatal.
| Country | Study period | PA-VTE N | Deliveries N | Rate of VTE per 1000 deliveries | Pulmonary embolism N (%) |
|---|---|---|---|---|---|
| Hong Kong | 1998–2000 | 32 | 16,993 | 1.88 | 2 (6.3) |
| USA | 1978–1996 | 165 | 268,525 | 0.61 | 38 (23.0) |
| USA | 1966–1995 | 100 | 50,080 | 2.0 | 24 (24.0) |
| Norway | 1990–2003 | 615 | 613,232 | 1.0 | Data not reported |
| USA | 2000–2001 | 14335 | 8,330,927 | 1.72 | 3009 (21.0) |
| Scotland | 1980–2005 | 2006 | 1,475,301 | 1.36 | 290 (14.5) |
| Denmark | 1980–2001 | 129 | 71,729 | 1.8 | 17 (13.2) |
| Sweden | 1990–1993 | 607 | 479,422 | 1.27 | Data not reported |
| Canada | 1991–2006 | 6821 | 3,852,569 | 1.77 | 2144 (31.4) |
| UK | 1997–2007 | 82 | 82,000 | 1.00 | 24 (29.3) |
| USA | 2003–2008 | 74 | 33,311 | 2.22 | 37 (50.0) |
| Australia | 1999–2006 | 8 | 6987 | 1.14 | 0 (0) |
| UK | 1988–1997 | 336 | 395,335 | 1.33 | 42 (12.5) |
| UK | 1995–2009 | 500 | 376153 a | 1.33 a | Data not reported |
| Denmark | 1995–2005 | 727 | 460464 a | 1.58 a | Data not reported |
a Pregnancy-years not deliveries; PA-VTE, pregnancy-associated venous thromboembolism; VTE – venous thromboembolism.
Timing and presentation of venous thromboembolism in pregnancy and the postpartum period
The timing of presentation with PA-VTE in the antenatal and postpartum period shows some variation , ( Table 2 ). The risk per day is higher in the shorter postpartum period. Over 95% (369 out of 383) of women with postpartum VTE presented in the first 6 weeks, with the remainder in weeks 7–12 . Pulmonary embolism is more frequent in the postpartum period compared with the antenatal period: 0.22 out of 1000 v 0.06 out of 1000 deliveries, respectively . A 15-fold increase in the observed incidence of pulmonary embolism was reported in the first 3 months postpartum compared with pregnancy . Up to 50% of women who develop VTE during pregnancy present in the third trimester ( Table 2 ), but a significant number occur in the first and second trimesters. Most women present with deep vein thrombosis (DVT) in the left leg ( Table 3 ) , particularly proximal and iliofemoral DVT, although distal DVT present with equal frequency in either leg .
| Study | All PA-VTE N | Antenatal VTE N (%) | First trimester N (%) | Second trimester N (%) | Third trimester N (%) | Postpartum VTE (N) |
|---|---|---|---|---|---|---|
| Chan et al. | 32 | 8 (25.0) | 3 (37.5) | 3 (37.5) | 2 (25.0) | 24 (75.0) |
| Chan et al. | 60 | 31 (51.7) | 14 (45.2) | 7 (22.6) | 10 (32.3) | 29 (48.3) |
| Heit et al. | 100 | 36 (36.0) | 4 (11.1) | 14 (38.9) | 18 (50.0) | 64 (64.0) |
| Jacobsen et al. | 615 | 301 (48.9) | 62 (20.6) | 64 (21.3) | 175 (58.1) | 314 (51.1) |
| Lyall and Myers | 82 | 56 (68.3) | 13 (23.2) | 21 (37.5) | 22 (39.5) | 26 (31.7) |
| James et al. | 53 | 34 a (64.2) | 15 (44.1) | 8 (23.5) | 9 (26.5) | 19 (35.8) |
| Blanco-Molina et al. | 136 | 72 (52.9) | 29 (40.3) | 13 (18.1) | 30 (41.7) | 64 (47.1) |
| Virkus et al. | 709 | 491 (69.3) | 61 (12.4) | 75 (15.3) | 355 (72.3) | 218 (30.7) |
| Sharma and Monga | 8 | 7 (87.5) | 5 (71.4) | 0 (0) | 2 (28.6) | 1 (12.5) |
| Voke et al. b | 126 | 126 | 31 (24.6) | 37 (29.4) | 58 (46.0) | – |
| Gherman et al. | 165 | 109 (66.1) | – | – | – | 56 (33.9) |
| James et al. | 14335 | 7177 (50.1) | – | – | – | 7158 (49.9) |
| Larsen et al. | 129 | 61 (47.3) | – | – | – | 68 (52.7) |
| Lindqvist et al. | 608 | 308 (50.7) | – | – | – | 300 (49.3) |
| Morris et al. c | 375 | 145 (38.7) | – | – | – | 230 (61.3) |
| O’Connor et al. | 74 | 36 (48.6) | – | – | – | 38 (51.4) |
| Simpson et al. | 336 | 109 (32.4) | – | – | – | 256 (76.2) |
| Sultan et al. | 500 | 215 (43.0) | – | – | – | 285 (57.0) |
| Trimester at presentation | 237 (20.4) | 242 (20.9) | 683 (58.7) | |||
| Antenatal v postpartum | 9196 d (50.1) | – | – | – | 9150 (49.9) | |
a Trimester at presentation unknown in two cases.
| Study | Deep vein thrombosis (N) | Left leg N (%) | Right leg N (%) | Bilateral N (%) |
|---|---|---|---|---|
| Chan et al. | 32 | 15 (47) | 13 (41) | 4 (13) |
| Chan et al. | 44 | 29 (66) | 12 (27) | 3 (7) |
| Gherman et al. | 127 | 104 (82) | 22 (17) | 1 (1) |
| Lyall and Myers | 54 | 36 (67) | 18 (33) | Not reported |
| James et al. | 53 | 35 (66) | 11 (21) | 6 (11) |
| O’Connor et al. | 40 | 23 (58) | 14 (35) | 3 (8) |
| Blanco-Molina et al. | 111 | 76 (68) | 35 (32) | Not reported |
| Pomp et al. | 74 | 55 (74) | 19 (26) | Not reported |
| Voke et al. | 76 | 54 (71) | 22 (29) | Not reported |
| Sharma and Monga | 8 | 2 (25) | 6 (75) | Not reported |
| Overall | 618 | 429 (69) | 172 (28) | 17 (3) |
Risk factors for pregnancy-associated venous thromboembolism
All components of ‘Virchow’s triad’ are present in pregnancy. Venous stasis is induced by venous dilation and obstruction to venous return , pro-coagulant factors are increased, natural anticoagulants reduced, and vessel wall injury occurs during labour and after caesarean section. Pregnancy is also a pro-inflammatory state with activation of endothelial cells . Inherent maternal factors further increase the risk of VTE in pregnancy. A number of epidemiological studies have explored risk factors for thrombosis ( Table 4 ). The reported contribution of risk factors for developing PA-VTE, particularly in more recent studies, will be influenced by whether women received thromboprophylaxis. Information relating to use of thromboprophylaxis is not generally available in population studies, but can be assumed to have been given to women with clinical factors perceived to place them at higher risk of thrombosis, introducing a source of bias, with the effect of minimising the effect of that risk factor in the cohort studied.
| Risk factor | Adjusted odds ratio for PA-VTE |
|---|---|
| Maternal characteristic | |
| Previous VTE | 4.2–24.8 |
| Age > 35 years | 1.0–2.7 |
| BMI > 30 | 1.5–5.3 |
| Smoking | 1.0–3.4 |
| Antiphospholipid syndrome | 5.1–15.8 |
| Cardiac disease | 3.2–7.1 |
| Sickle cell disease | 1.3–6.7 |
| SLE | 2.3–8.7 |
| Varicose veins | 2.7–3.8 |
| Inflammatory bowel disease | 3.5–4.6 |
| Diabetes | 1.4–2.0 |
| Hypertension | 0.9–1.8 |
| Cancer (any past diagnosis) | 1.2–2.0 |
| Pregnancy-specific factors | |
| Assisted reproductive technology | 2.2–4.4 |
| Multiple gestation | 0.8–2.7 |
| Parity ≥ 3 | 0.8–2.8 |
| Gestational diabetes | 1.7–4.1 |
| Caesarean section | |
| All | 1.8–11.2 |
| Planned | 1.4–3.1 |
| Emergency | 2.2–4.0 |
| Hypertensive disorders of pregnancy | 0.5–5.8 |
| Preterm delivery < 37 weeks | 1.8–4.5 |
| Placenta previa | 3.6 |
| Placental abruption | 2.5 |
| Stillbirth | 6.0–6.2 |
Intrinsic maternal risk factors
Maternal age
The rate of VTE was 38% higher in women over the age of 35 years (2.27 out of 1000 deliveries) compared with younger women (1.64 out of 1000 deliveries) . Heit et al. reported a three-fold (95% CI 1.4 to 6.5) increased risk of pulmonary embolism for every 10 years of maternal age.
Body mass index
The contribution of increased body mass index (BMI) has been recently reviewed . Increased (BMI ≥ 25 kg m −2) , combined with immobility, was associated with a major increase in the risk of antenatal and postpartum VTE, aOR 62.3 (95% CI 11.5 to 337.6) and aOR 40.1 (95% CI 8.0 to 201.5), respectively .
Racial factors
Background rates of VTE in a general population are lower in individuals of Asian descent , but published studies in pregnancy are limited. A study of ethnic Chinese women from Hong Kong in a tertiary obstetric unit reported a rate of 18.8 out of 10,000 deliveries between 1998 and 2000, with only one DVT recorded between 1988 and 1992. The investigators postulate that a more westernised diet and affluent lifestyle might account for the increasing rate. Lower rates of PA-VTE were reported in Asian women (10.7 out of 10,000 deliveries) and Hispanic women (12.5 out of 10,000 deliveries) in a large US study compared with white women (17.5 out of 10,000 deliveries), with higher rates in black women (26.4 out of 10,000 deliveries).
Prior history of venous thromboembolism
Previous VTE seems to be one of the more important risk factors for developing PA-VTE. The risk of recurrent PA-VTE is similar (2–10%) in women who have had a previous unprovoked VTE, and those who have had a previous hormonally provoked VTE (i.e. associated with pregnancy or oral contraception) . Women who have had a previous VTE associated with a major (surgical or traumatic) provoking factor are at low risk of recurrence in pregnancy, with three separate studies reporting no recurrent events among this group of patients .
Hereditary thrombophilia
As in non-pregnant populations, laboratory markers of thrombophilia are found in 40–50% of cases compared with 6–15% of controls . Racial differences in thrombophilias exist, with factor V Leiden (FVL) and the prothrombin gene mutation (PTM) found almost exclusively in white people, with deficiencies of antithrombin, protein C and S being relatively rare ; deficiencies of these natural anticoagulants are more common in individuals of Asian descent . Studies have shown that the rate of VTE in relatives who share the thrombophilia is three to five-fold higher than in relatives without the thrombophilia . Pregnancy is a critical risk period for development of VTE in affected relatives, and administration of thromboprophylaxis at periods of risk seems to reduce the risk of VTE . Of note, family cohort studies also show increased rates of VTE even in unaffected relatives compared with the background population . Screening asymptomatic women for thrombophilias is not recommended .
Family history with no thrombophilia
Venous thromboembolism is increasingly being recognised as a multigenic disease, and a positive family history alone (one or more affected first-degree relative) has been shown to increase the risk of VTE two-fold , with the strength of the association increased when younger relatives are affected (OR 2.7, 95% CI 2.2 to 3.4) and if more than one relative is affected (OR 3.9, 95% CI 2.7 to 5.7).
Pregnancy-specific factors
Mode of delivery
Caesarean section, especially emergency caesarean section, increases the risk of postpartum VTE ( Table 4 ). It seems likely that increasing awareness of the risk of thrombosis after caesarean section has led to an increasing use of thromboprophylaxis, thus modifying the development of thrombosis. In the study by Jacobsen et al. , it was usual practice in all but one hospital to give pharmacological thromboprophylaxis to all women after caesarean section. Separate thromboprophylaxis guidelines published in 1995 in Scotland and by the UK Royal College of Obstetricians and Gynaecologists highlighted pregnancy as a risk factor for VTE, and recommended pharmacological thromboprophylaxis for women at risk (i.e. after emergency caesarean section) . In the years after publication, the Scottish population study reported a 66% reduction in the rate of VTE after emergency caesarean section (12.3 out of 10,000 deliveries between 1980 and 1985, and 3.9 out of 10,000 deliveries between 2001 and 2005), whereas no change in the incidence of VTE after vaginal delivery or elective caesarean section was reported .
Assisted reproductive technology
Nearly 50% of thrombotic events occurring in women after assisted reproductive technology were in women who developed first-trimester ovarian hyperstimulation syndrome .
Obstetric haemorrhage
Women who required surgery after postpartum haemorrhage were three times more likely to develop thrombosis than those who did not require surgery .
Infection
The use of pharmacological thromboprophylaxis in most women delivering by caesarean section may explain the observed greater risk of VTE with infection after vaginal delivery (aOR 20.2, 95% CI 6.4 to 63.5) compared with after caesarean section (aOR 6.2, 95% CI 2.4 to 16.2).
Pre-eclampsia and intrauterine growth restriction
Pre-eclampsia has been found to be a risk factor for VTE in some but not all studies , and is particularly associated with postpartum events . One study reported that postpartum VTE were more frequent in women who had both pre-eclampsia and intrauterine growth restriction than either complication alone .
Thromboprophylaxis
The overall low prevalence of PA-VTE means that universal prescription of thromboprophylaxis is unwarranted. A recurring theme in maternal mortality and morbidity reports is the apparent lack of appreciation of risk factors for thrombosis in women who develop PA-VTE . Attempts to identify women at risk for PA-VTE who would merit antenatal, postpartum thromboprophylaxis, or both, is challenged by the lack of clinical data supporting benefit. A recent Cochrane review confirmed what clinicians working in the field have long recognised, that there is ‘insufficient evidence available from randomised-controlled trials to guide clinical decision-making’ and ‘practitioners must rely on consensus derived clinical practice guidelines’. An assessment of a woman’s risk factors is key, but what determines the threshold for recommendation for thromboprophylaxis is unclear. Decisions should be made after a discussion of the available evidence with the woman, taking into consideration her perception of the balance of risk and benefit with thromboprophylaxis, an approach endorsed by the most recent American College of Chest Physicians guidelines .
Risk-factor assessment
Most guidelines recognise that the threshold for recommending thromboprophylaxis in the postpartum period should be lower given a higher daily risk of thrombosis in the postpartum period. All pregnant women should have an assessment of risk of VTE at the earliest opportunity and, if the threshold for antenatal thromboprophylaxis is reached, it should be commenced as soon as pregnancy is recognised. Re-assessment should be carried out if any change occurs to a woman’s health during pregnancy, such as admission to hospital and also after delivery.
Contribution of multiple risk factors
Some thromboprophylaxis guidelines focus mainly on personal and family history of VTE and the presence of thrombophilias, whereas others emphasise the importance of additional risk factors for thrombosis ( Table 4 ). In the absence of data from clinical trials, decisions about how to implement risk assessment strategies is based only on expert opinion. A summary of recommendations from different guidelines is presented in Table 5 . The overall low incidence of PA-VTE means that many women will need to be given thromboprophylaxis to prevent an event, and considerable resources are required to provide thromboprophylaxis as per current guidelines. Clinical studies are urgently required to inform clinical practice.
| ACCP | ACOG | RCOG | ASTH and SOMANZ | |||||
|---|---|---|---|---|---|---|---|---|
| Antenatal | Postpartum | Antenatal | Postpartum | Antenatal | Postpartum | Antenatal | Postpartum | |
| Personal history VTE | ||||||||
| Idiopathic | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ |
| Pregnancy or related to combined oral contraceptives | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ |
| Temporary risk factor | ✗ | ✓✓✓ | ✗ | ✓✓✓ | ✗ | ✓✓✓ | ✗ | ✓✓✓ |
| Multiple VTE | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ |
| Family history VTE plus thrombophilia* | ||||||||
| Antithrombin deficiency | ✗ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓✓✓ | ✓✓✓ |
| Protein C or S deficiency | ✗ | ✓✓✓ | ✗ or ✓ | ✗ or ✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Homozygous FVL or PTM | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Heterozygous FVL or PTM | ✗ | ✓✓✓ | ✗ or ✓ | ✗ or ✓ | ✓ | ✓✓ | ✗ | ✓ |
| FVL/PTM comp heterozygous | ✗ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| No thrombophilia | NC | NC | NC | NC | NC | NC | ✗ | ✓ |
| Thrombophilia without personal or family history VTE* | ||||||||
| Antithrombin deficiency | ✗ | ✗ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Protein C or S deficiency | ✗ | ✗ | ✗ or ✓ | ✗ or ✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Homozygous FVL or PTM | ✗ | ✓✓✓ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Heterozygous FVL or PTM | ✗ | ✗ | ✗ or ✓ | ✗ or ✓ | ✓ | ✓✓ | ✗ | ✗ |
| FVL/PTM compound heterozygous | ✗ | ✗ | ✓✓✓ | ✓✓✓ | ✓ | ✓✓ | ✓ | ✓✓✓ |
| Family history VTE no thrombophilia | NC | NC | NC | NC | ✗ | ✗ | ✗ | ✓ |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
