Article Text

Download PDFPDF
Pregnancy and heart disease
Valvular heart disease and pregnancy part II: management of prosthetic valves
  1. Catriona J Bhagra1,
  2. Rohan D'Souza2,
  3. Candice K Silversides1
  1. 1Division of Cardiology, University of Toronto Pregnancy and Heart Disease Program, Mount Sinai Hospital and University Health Network, Toronto, Ontario, Canada
  2. 2Department of Obstetrics and Gynaecology, Division of Maternal and Fetal Medicine, Mount Sinai Hospital, University of Toronto, Toronto, Ontario, Canada
  1. Correspondence to Dr Candice K Silversides, Division of Cardiology, University of Toronto Pregnancy and Heart Disease Program, Mount Sinai Hospital and University Health Network, Mount Sinai Hospital, Room 3-913, OPG Building, 700 University Avenue, Toronto, Ontario, Canada M5G 1Z5; candice.silversides{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Learning objectives

  • To recognise the maternal, fetal and neonatal risks of pregnancy in women with prosthetic heart valves.

  • To review anticoagulation strategies for women with mechanical heart valves during pregnancy.

  • To understand the principles of management in pregnant women with prosthetic heart valves.


Prosthetic heart valves (PHV) have been used to treat patients with both congenital and acquired valve lesions since the first surgical replacement in 1960.1 The two principal types of PHV are the mechanical prostheses and the tissue or bioprostheses. Advances in surgical technique, valve design and anticoagulation have improved overall outcomes since their introduction.2 Worldwide, there is an increasing prevalence of rheumatic heart disease, with more young women being considered for surgical treatment.3 ,4 There are also increasing numbers of women with congenital heart disease reaching childbearing age, some of whom have PHV.

During pregnancy, there is an increase in heart rate, stroke volume and cardiac output. These haemodynamic changes can lead to decompensation in women with PHV. Pregnancy is associated with additional risks in women with mechanical heart valves (MHV) as pregnancy is a prothrombotic state with an increased risk of thromboembolic complications (TECs), coupled with the adverse effects of oral anticoagulants on the fetus. In this article, we will review the pregnancy risks in women with PHV, strategies for anticoagulation in women with MHV and the management of complications.

Preconception counselling and risk assessment

Pregnancy in a woman with a PHV should be carefully planned. Preconception counselling is imperative, allowing for a comprehensive discussion of the risks of pregnancy for the mother and baby. The degree of maternal cardiac risk depends on many factors including the women's clinical status, her underlying cardiac condition, the type and position of the valve, the presence of other cardiac lesions and left ventricular systolic function. Maternal obstetric, fetal and neonatal risks should also be discussed during the preconception visit. Due to the complexity in risk assessment, prepregnancy counselling should be provided by cardiologists and maternal fetal medicine specialists, or high-risk obstetricians, with expertise in heart disease in pregnancy.5 If a thrombosis clinic is available, women with MHV may also benefit from consultation with a haematologist with expertise in pregnancy.

The initial assessment, prior to conception or at the time of the first antenatal visit in women who present pregnant, should include a detailed medical history, physical examination, 12-lead ECG and transthoracic echocardiogram (TTE) (box 1). The determination of valvular haemodynamics prior to or early in pregnancy help in risk stratification and act as an important baseline for ongoing antenatal and postpartum comparisons.6 Medical therapy should be reviewed to ensure drugs contraindicated in pregnancy are stopped or substituted. Preconception assessment also allows for cardiac lesions that require intervention to be addressed.

Box 1

Principles of counselling for women with prosthetic heart valves

Counselling checklist for women with prosthetic heart valves

Detailed clinical history including surgical/interventional history

Family history—particularly in women with congenital heart disease

Physical examination

Review of drug therapy to ensure potentially teratogenic drugs are stopped or substituted prior to conception

12-lead ECG

Transthoracic echocardiogram—assessment of prosthetic valve function, ventricular function, concomitant valvular disease and pulmonary artery systolic pressures

Review the need for further prepregnancy investigations on an individual basis; exercise stress testing/cardiopulmonary exercise testing, cardiac MRI, transoesophageal echocardiogram or cardiac catheterisation

Counselling on the maternal cardiac, maternal obstetric and perinatal risks

Detailed discussion on the risks and benefits of anticoagulation therapies for women with mechanical heart valves

Conception counselling

For those women considering pregnancy, pregnancy follow-up planning

Risk estimates for maternal cardiac complications should incorporate known risk factors from general pregnancy risk indexes, known lesion-specific risk factors and patient-specific variables. Two commonly used risk scores used to predict complications during pregnancy in women with cardiac disease are the Cardiac Disease in Pregnancy (CARPREG) and Zwangerschap bij Aangeboren Hartafwijkingen (ZAHARA) risk scores.7 ,8 These risk scores have identified prior cardiac event (heart failure, transient ischaemic attack or stroke prepregnancy, arrhythmia), baseline New York Heart Association (NYHA) functional class >II, cyanosis, left heart obstruction and systemic ventricular systolic function <40% as risk factors for maternal cardiac complications during pregnancy. The ZAHARA risk score additionally identified significant systemic or pulmonary atrioventricular valve regurgitation and MHV as important predictors of outcome.8 The use of the WHO maternal risk classification, which incorporates underlying heart lesions and general risk factors, is recommended in the 2011 European Society of Cardiology (ESC) guidelines on the management of cardiovascular diseases during pregnancy.5 In this classification, risk is graded from I to IV. Women with MHV fall into WHO III where pregnancy carries a significantly increased risk of maternal mortality or severe morbidity. Bioprosthetic valves generally fall into WHO II-III depending on individual circumstances, and carry a small increased risk of maternal mortality and moderate risk of morbidity.5 Lesions-specific risks for women with bioprosthetic and MHV are discussed in detail below.

Maternal and neonatal health is inter-related.7 Maternal variables which predict fetal and neonatal complications include baseline NYHA class II or cyanosis, maternal left heart obstruction, smoking during pregnancy, multiple gestation, use of oral anticoagulants during pregnancy and mechanical valve prosthesis.8 ,9

Bioprosthetic valves and pregnancy

The haemodynamic adaptations of pregnancy may be well tolerated in a woman with a bioprosthetic valve provided the valve is functioning normally and no other significant maternal cardiac disease is present. However, complications can occur in these women. In the subset of 134 women with bioprosthetic valves in the Registry of Pregnancy and Cardiac Disease (ROPAC) study, heart failure complicated 8.2% of pregnancies in women with bioprosthetic valves; endocarditis and thrombotic complications in <1% and haemorrhagic complications in 5.1% of pregnancies.10 Maternal deaths are rarely reported in women with bioprosthetic valves.10–12 In the ROPAC study, maternal mortality was 1.5%.10 One death was due to an unexplained out-of-hospital arrest; the other was at the time of emergency caesarean delivery undertaken for fetal indications. There was no reported fetal mortality beyond 24 weeks gestational age and miscarriage rates (<24 weeks) were 1.5%.10

Bioprosthetic valves have limited long-term durability, with structural valve deterioration (SVD) developing over a period of years.13 The mitral valve tends to be more vulnerable than the aortic and young women appear to be at particular risk. There have been conflicting reports in the literature as to whether pregnancy leads to an acceleration of this process. An early study, supporting the hypothesis of accelerated SVD due to pregnancy, was confounded by younger age among the women who became pregnant.14 A later retrospective study of 232 women demonstrated that although valve loss at 10 years was significantly higher in women with bioprosthetic (82%) compared with mechanical (29%) or homografts valves (28%), pregnancy was not related to increased bioprosthetic or homograft valve loss.15 In a prospectively study of 85 women after prosthetic valve implantation, valve degeneration was not related to pregnancy and was attributed to the natural course of the bioprosthesis.16

MHV and pregnancy

Pregnancy in a woman with a MHV is associated with high maternal and perinatal risk. Maternal mortality estimates range between 1% and 15% depending on the series and anticoagulation strategy.17 ,18 Pregnant women with MHVs are at an increased risk of TECs, heart failure and arrhythmias. Women receiving anticoagulants also have higher rates of bleeding. Among 212 women with a MHV in the ROPAC registry, 7.5% of pregnancies were complicated by heart failure and supraventricular and ventricular arrhythmias that occurred in 2.8% and 0.5% of pregnancies, respectively. The rate of TECs was 13%.10 Postpartum haemorrhage directly after delivery and up to 24 hours post partum occurred in 10.4%.10 Perinatal complications in women with MHV can include fetal loss, preterm delivery, low birth weight and teratogenicity from the use of oral anticoagulants in pregnancy.


Patients with MHVs require lifelong anticoagulation. Recommended international normalised ratios (INR) vary depending on the type and position of the valve. Anticoagulation is particularly important during pregnancy as pregnancy is a prothrombotic state with increased circulating levels of fibrinogen, factors VII, VIII and X and plasminogen activator inhibitor, increased platelet adhesiveness, resistance to activated protein C and reduction in fibrinolysis.19 The main determinant of TECs in pregnancy is related to the type of anticoagulant used. While vitamin K antagonists (VKA) such as warfarin and acenocoumarin are preferred in non-pregnant women, these anticoagulants cross the placenta. When taken during embryogenesis they can impact upon fetal development. Heparins do not cross the placenta and alternative anticoagulation regimens using heparin in pregnancy have been proposed with the aim of reducing fetal risks. These include low molecular weight heparin (LMWH), unfractionated heparin (UFH) and combination treatment with LMWH or UFH in the first trimester and VKAs in the second and third trimesters.1 ,20–23 The direct oral anticoagulant agents currently have no role. Table 1 summarises the major maternal and fetal outcomes stratified according to the three most common anticoagulation regimens used.18

Table 1

Adverse maternal and fetal outcomes in women with mechanical heart valves according to the type of anticoagulation therapy used during pregnancy.

Vitamin K antagonists

For women with MHV, warfarin is the anticoagulant associated with the lowest maternal risk of TECs (2.9%–3.9%) and maternal mortality (1.1%–1.8%) during pregnancy.17 ,18 However, warfarin is associated with an embryopathy with features of nasal hypoplasia, stippled epiphyses or both when exposure occurs between 6 and 12 weeks gestation. Between 3.7% and 6.4% of women exposed to VKA throughout pregnancy will develop warfarin embryopathy.17 ,18 Warfarin fetopathy (central nervous system or ocular abnormalities) can result from later exposure, and fetal demise can occur at any gestational age.24 Fetal wastage due to any cause occurs in approximately 33% of pregnancies.

Vitale et al first described a dose-dependent relationship between warfarin dose and fetal risk. Their study investigated 58 pregnancies in 43 women with MHV with INR 2.5–3.5. Twenty-seven fetal complications including 22 spontaneous abortions and two cases of warfarin embryopathy were reported. Women using >5 mg of warfarin per day had significantly more fetal complications compared with those taking ≤5 mg/day (22 vs 5, p=0.0001).23 The same group later reported outcomes in 71 pregnancies in 52 women with MHV receiving warfarin throughout pregnancy and found a daily dosage >5 mg/day was a significant predictor of poor fetal outcomes.25 Not all groups have shown a warfarin dose-dependent relationship with fetal complications. The ROPAC registry described similar rates of miscarriage or fetal loss in women taking high-dose or low-dose VKA.10

Newer generations of MHVs are less thrombogenic, particularly in the aortic position, and anticoagulation thresholds may differ for women with these valves. One small study of 16 pregnancies in women with a St. Jude MHV in the aortic position reported no adverse maternal TECs in women taking 4.1 mg/day of warfarin with a mean INR of 1.9 throughout pregnancy.26 However, the number of women in the study was small, the results are not supported by other series using low-dose warfarin,12 ,27 and further work is necessary to determine whether these outcomes can be reproduced in a larger patient cohorts.


Both UFH and LMWH are associated with higher rates of maternal cardiac complications during pregnancy when compared with VKA. The high rates of TECs are likely due to a combination of factors including the prothrombotic state of pregnancy, changing dose requirements, the difficulties in monitoring heparin levels during pregnancy and patient compliance. There are additional concerns over associated thrombocytopenia and osteoporosis if UFH is used throughout pregnancy.28 While maternal risk is high with UFH and LMWH, these agents do not cross the placenta, eliminating the risk of embryopathy and fetopathy seen with VKA.

Unfractionated heparin

In a systematic review of 1234 pregnancies in women with MHV (49.7% ball and cage type, 36% with single tilting discs and 6% bileaflet), the use of heparin throughout pregnancy, either in low or adjusted dose, was associated with very high rates of TECs (33.3%) and maternal death (15%).17 These findings may be partially attributed to inadequate heparin dosing. A more recent review of the literature spanning 2000–2009 reported on 1343 pregnancies in women with MHV including a higher proportion of the newer and less thrombogenic valves (12% ball and cage, 34% tilting disc and 31% bileaflet) showed that the use of heparin throughout pregnancy was accompanied by lower rates of TECs (13.4%) and maternal mortality (4.7%).18 Fetal wastage due to any cause was 38.8%. The use of subcutaneous UFH is not recommended in the 2014 American Heart Association/American College of Cardiology (AHA/ACC) guidelines.6

Low molecular weight heparins

Outside of pregnancy, LMWH is associated with a more predictable pharmacokinetic profile.29 It was introduced as an alternative to UFH, and experience with its use in pregnancy is growing.20 ,21 ,30–35 In 2009, four groups reported, in detail, on their experience with LMWH during pregnancy. All groups monitored anti-Xa levels throughout pregnancy. Variable rates of TECs were reported ranging between 4% and 17% of pregnancies.21 ,32–34 Across the four studies, the live birth rate was 91% and major postpartum bleeding occurred in 22%. Our group reported on 23 pregnancies in 17 women treated with LMWH and low-dose aspirin through pregnancy; one thromboembolic event occurred (4%). Other cardiac complications occurred in five (22%) pregnancies including two cases of pulmonary oedema (9%), three cases of arrhythmia (13%) and one case of endocarditis (4%).21

In 1999, the Heparin In Pregnancy-Cardiac Valve Thrombosis (HIP-CAT) study, a prospective randomised controlled study comparing enoxaparin (1 mg/kg body weight with twice daily dosing) to warfarin with UFH, was stopped prematurely after two deaths from valve thrombosis in women treated with enoxaparin.36 Thromboembolic complications are more likely to occur if anti-Xa levels are subtherapeutic37 and some speculate that the TECs in HIP-CAT were related to failure to adjust LMWH doses to levels as is currently recommended.38 ,39 However, anti-Xa monitoring is not simple; a small retrospective study of four pregnancies demonstrated that achieving target peak anti-Xa levels did not guarantee the maintenance of adequate trough levels.40 Some experts recommend that LMWH dosing be guided by both peak and trough Xa levels.

Recommendations for anticoagulation during pregnancy

There are a number of guidelines that address anticoagulation options in this population including the ESC guidelines on the management of cardiovascular diseases in pregnancy,5 the AHA/ACC guidelines on the management of patients with valvular heart disease6 and the American College of Chest Physicians evidence-based practice guidelines on antithrombotic therapy and prevention of thrombosis.41 Recommendations within the guidelines are based on ‘level C’ evidence as no large randomised trials are available in this field. Figure 1 illustrates the current AHA/ACC algorithm for the management of anticoagulation in women with MHV.6 The AHA/ACC recommendations are similar to the ESC guidelines.5 It is imperative that women be involved at all stages of decision making as compliance has a significant impact on outcomes.

Figure 1

The American Heart Association/American College of Cardiology algorithm for the management of anticoagulation in pregnant women with a mechanical heart valve.6 LMWH, low molecular weight heparin; UFH, unfractionated heparin; aPTT, activated partial thromboplastin time; ASA, aspirin; INR, international normalised ratio.

The ESC and the AHA/ACC guidelines advise continuing warfarin in the first trimester if the dose is ≤5 mg/day, after patient information and consent. In women whose warfarin dose is >5 mg/day, and for those who do not wish to take warfarin in the first trimester, it may be discontinued between weeks 6 and 12 and replaced with either weight-adjusted twice-daily LMWH or an intravenous infusion of UFH. The first trimester, when women are converted from VKA to LMWH, is a vulnerable period with at least half the valve thromboses occurring around this time.10 The optimal time interval for monitoring and additional value of trough anti-Xa levels is not standardised, although both sets of guidelines are unanimous in stating that a body weight method should not be used in pregnancy.5 ,6

VKAs should be used during the second and third trimester, until approximately 36 weeks gestation. After discussion, some women may continue LMWH during the second and third trimesters if otherwise deemed at low risk (normal functioning bileaflet valve, no other thrombosis risk factors) in place of warfarin. The ESC guidelines suggest that changes in anticoagulation should be implemented in hospital, although practically this can be difficult. The increased maternal and fetal bleeding associated with administration of VKA in the late third trimester is addressed by changing to LMWH or UFH approximately 2 weeks before the anticipated date of delivery.

Principles of pregnancy management

Antepartum management

Follow-up through pregnancy should be individualised. The frequency of visits is determined by clinical status, the type of PHV and anticoagulation regimen. In women with good baseline bioprosthetic valve function, cardiology assessment with echocardiographic evaluation of the valve can be performed in each trimester. For all women with MHV, the effectiveness of the chosen anticoagulation regimen should be monitored closely and clinical follow-up including TTE performed frequently. With the increase in circulating blood volume and cardiac output, the mean PHV gradients measured by TTE will increase through pregnancy, but the calculated valve area typically remains stable and can be used for serial evaluation.

Labour and delivery

A structured labour and delivery plan should be formalised with the intent to optimise maternal and neonatal outcomes. The plan should specifically address pregnancy risk, the need for induction, timing and mode of delivery, monitoring at the time of labour and delivery, analgesia and anaesthesia plans and postpartum monitoring. From a cardiac perspective, vaginal delivery is preferred in most women. For women with MHV, a detailed management plan for peripartum anticoagulation should be included.

Delivery of a woman with a MHV who is on warfarin must be planned to allow warfarin to be stopped and transitioned to either LMWH or UFH at least 2 weeks prior to delivery. The timing of this transition must be individualised as some women with prosthetic valves are at risk of preterm delivery. If a woman taking warfarin presents in active labour, a caesarean delivery is indicated due to the high risk of fetal intracranial haemorrhage during vaginal delivery. Fresh frozen plasma (FFP) can be administered to the mother prior to caesarean delivery often along with vitamin K. The fetus may also require FFP and vitamin K and this should be addressed by the neonatology and haematology teams.5 For women taking LMWH, the LMWH should be stopped prior to delivery and replaced with an UFH infusion keeping the activated partial thromboplastin time >2 times control. If emergent delivery is required in a woman still on LMWH or UFH, protamine can be considered. Protamine will only partially reverse the anticoagulant effect of LMWH.5 When in active labour, UFH can be stopped to allow safe epidural anaesthesia and minimise the risks of bleeding. UFH can usually be re-commenced 4–6 hours post partum provided there are no bleeding concerns.5 Timing of the re-introduction of warfarin differs by centre. The haemodynamic and the haematologic changes of pregnancy take weeks to months to return to normal and continued surveillance post partum is necessary.

Breast feeding

Nursing mothers may safely breast feed their babies while taking VKAs, LMWH or UFH.42 Diuretics, digoxin and β-blockers with high protein-binding (such as metoprolol, labetalol and propranolol that are excreted in small quantities in breast milk) are also safe for use during breast feeding.37 β-Blockers with low protein-binding such as atenolol, nadolol and sotolol are usually not advised if alternatives are available.43 The ESC guidelines on the management of cardiovascular diseases during pregnancy provide a comprehensive guide to the safety of individual drugs during pregnancy and in breast feeding.5

Complications during pregnancy

Prosthetic valve thrombosis

Prosthetic valve thrombosis (PVT) can be life threatening and is the most feared complication in a pregnant woman with a MHV. Accurate diagnosis and timely management are critical to maternal and fetal survival. The mother's presenting complaint may be dyspnoea, heart failure, arrhythmias, a thromboembolic event or muffling of the normal MHV click.4 Full assessment including urgent echocardiography is required. Transoesophageal echocardiography (TOE) should be performed if there is a clinical suspicion of valve thrombosis. Thrombus may be visible on the valve and there may be increased Doppler valve gradients or regurgitation. Cine-fluoroscopy with abdominal shielding should be performed if ongoing suspicion persists despite negative TOE.44

The management of PVT is dependent on the patient's clinical status, gestational age of the fetus and degree of valve obstruction. In general, management should be similar to that of non-pregnant patients.5 ,45 Figure 2 shows an adapted algorithm for the management of obstructive left-sided prosthetic thrombosis during pregnancy. If the mother is clinically stable or in the setting of a small thrombus with a history of subtherapeutic anticoagulation, it may be reasonable to treat medically, with intravenous UFH. Emergent surgery is generally reserved for patients who present critically unwell with obstructive thrombi. Fetal loss associated with cardiopulmonary bypass surgery has been reported between 20% and 30%.37 ,46 Fibrinolysis can be considered. Thrombolytics do not cross the placenta, although there is a risk of subplacental bleeding and placental abruption. The largest prospective, non-randomised single centre series reported on 25 pregnancies with 28 PVT events. All PVT occurred in the mitral position. Low dose (25 mg), slow infusion (over 6 hours) of tissue-type plasminogen activator (tPA) was used. TOE guidance was used to determine the success of tPA. A 100% rate of complete thrombolysis was reported using this protocol with no maternal deaths. No women had stroke or transient ischaemic attack. One woman had a placental haemorrhage with preterm live birth at 30 weeks and another epistaxis.47 Irrespective of the chosen treatment, the mother and fetus will be placed at high risk. Thoughtful consideration must be given to the anticoagulation strategy for the remainder of the pregnancy.

Figure 2

Algorithm for the management of mechanical valve thrombosis in pregnant women. Modified from the 2012 European Society of Cardiology guidelines on the management of valvular heart disease.45 MHV, mechanical heart valve; TTE, transthoracic echocardiogram; TOE, transoesophageal echocardiogram; UFH, unfractionated heparin.

Infective endocarditis

Infective endocarditis in pregnancy is rare (estimated overall incidence of 1 per 100 000 pregnancies).48 ,49 PHV endocarditis in pregnancy is even more unusual with published case reports only.50 ,51 Management should involve the cardiologist, infectious disease specialists, cardiothoracic surgeons and maternal fetal medicine specialists. Care must be taken to tailor antibiotic therapy to blood culture results and avoid antibiotics contraindicated in pregnancy.

Choice of prosthetic valves in women of childbearing age

Mechanical, bioprosthetic and native valvular disease each pose specific risks during pregnancy. The selection of an appropriate prosthetic valve is challenging in young women who may desire a future pregnancy. As no ideal is currently available, care must be individualised, with detailed preoperative discussions informing the patient of the risks and benefits of each option.

MHVs are well recognised for their long-term durability; re-operation rates are low but they necessitate lifelong anticoagulation.52 Tissue valves (homografts, heterografts or autografts) do not require anticoagulation, but are associated with high rates of valve deterioration, with redo surgery inevitable. Surgery itself may carry up to 5% mortality depending on the valve position and degree of urgency,14 ,45 ,53 and rates of redo surgery are between 50% and 80% at 10 years.13 While there continues to be debate about the best valve for young women, the 2012 ESC guidelines for the management of valvular heart disease recommend that a bioprosthetic valve should be considered in women who desire a pregnancy (class IIa, level of evidence C) due to the high risk of TECs in pregnancy with a MHV.45

The Ross operation is an alternative surgical approach for native aortic valve disease. Surgery involves the transfer of the pulmonary autograft to the aortic position and insertion of a homograft in the pulmonary position. This surgery is principally carried out in children and young adults at centres with this surgical expertise. There is no need for anticoagulation after the Ross operation. Re-operation rates vary by centre, and less information is available on long-term (>10 years) outcomes. Freedom from re-operation at 15 years has been reported as high as 92% for the pulmonary autograft, 97% for the pulmonary valve and from any re-operation 85%.54 Pregnancies in women with transcutaneous aortic valve replacements have not been reported.


Contraception options should be discussed in all women with PHV. The safety and efficacy of the various methods should be considered. Barrier methods such as condoms are not associated with cardiac risks, but are unreliable with high failure rates. Oestrogen-containing contraceptives are associated with venous and arterial thrombosis and are therefore not recommended in women with MHV. Progesterone-only contraceptives including pills, dermal implants, intramuscular progesterone injections and the levonorgestrel Intrauterine device (IUD) (Mirena) are safe for women with MHV. However, progesterone-only pills (‘mini-pills’) have high failure rates and should not be used in women in whom pregnancy carries a significant risk. In women who have completed their families or in those in whom pregnancy is contraindicated, permanent methods of contraception such as tubal ligation may be considered.5

Unanswered questions

There remain a number of unanswered questions. Currently, there is no optimal method of anticoagulation for pregnant women with MHVs. Anticoagulation therapy is restricted by competing maternal and fetal risks and the lack of high-quality data from prospective studies. A randomised controlled trial, although difficult for many reasons including patient preference, may help to answer this question. Until then, further prospective registry data remain the best approach to understand risk. For women with MHV using LMWH, larger studies are needed to assess risk, the utility of anti-Xa peak and trough levels to guide anticoagulation and the optimal way in which to deliver care to this complex population. The mechanisms contributing to late fetal death are poorly understood in women with MHV and further study in this area is needed. Finally, there has been growth in the field of the new direct oral anticoagulants, understanding the safety of these and their possible role in this population requires investigation.


Women with a normally functioning bioprosthetic valve and preserved ventricular function often do well during pregnancy. Women with MHV are at significant risk of maternal and perinatal complications. Pregnancy risks must be discussed with young women with valvular heart disease considering valve replacement and those considering pregnancy. During pregnancy, unique challenges present for women with MHV when deciding upon the safest and most appropriate anticoagulation. Currently, there is no ideal method of anticoagulation for both the mother and the baby. A multidisciplinary team including cardiologists, haematologists, obstetricians and anaesthesiologists with expertise in the management of pregnant women with cardiac disease is required to optimise outcomes.

Key messages

  • All women with prosthetic heart valves (PHV) should be well informed of the maternal and perinatal risks of pregnancy.

  • A multidisciplinary team approach to the management of pregnant women with PHV is necessary to ensure optimal outcomes.

  • Women with normally functioning bioprosthetic valves tend to have good pregnancy outcomes.

  • Pregnancy in women with mechanical heart valves is high risk.

  • There is no ideal anticoagulation therapy for pregnant women with mechanical heart valves. All anticoagulation strategies have associated maternal and perinatal risks.

    • Vitamin K antagonists (VKAs) are recommended for women in the second and third trimester in view of the lower maternal risk of thromboembolic complications.

    • Women using ≤5 mg/day may consider using VKA in the first trimester after discussion of fetal risks.

    • Women using >5 mg/day may consider using VKA or switching to low molecular weight heparin (LMWH) to avoid embryopathy.

    • If LMWH is chosen, the dose must be weight based and adjusted with frequent monitoring of peak anti-Xa levels.

    • Strict patient compliance with the chosen anticoagulation regime for the duration of pregnancy is essential.

    • A structured plan for labour and delivery, with specific guidance on the management of anticoagulation, should be established prior to delivery.

You can get CPD/CME credits for Education in Heart

Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to:

  • RCP credits: Log your activity in your CPD diary online (—pass mark is 80%.

  • EBAC credits: Print out and retain the BMJ Learning certificate once you have completed the MCQs—pass mark is 60%. EBAC/ EACCME Credits can now be converted to AMA PRA Category 1 CME Credits and are recognised by all National Accreditation Authorities in Europe (

Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal's username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.


View Abstract


  • Contributors All three authors contributed to the literature review and reporting of this review article.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Linked Articles