Use of anticoagulants during pregnancy and postpartum

INTRODUCTION — Anticoagulation is sometimes needed during pregnancy and/or the postpartum period in individuals at high risk of deep vein thrombosis, a history of venous thromboembolism, with prosthetic heart valves, atrial fibrillation, left ventricular dysfunction, or a history of fetal loss.

Use of anticoagulants during pregnancy is challenging due to the potential teratogenic effects and dosing complexities of the various agents, and the management of anticoagulation around the time of labor. In addition, individuals receiving chronic anticoagulation who are contemplating pregnancy need counseling regarding how to avoid the potential teratogenic effects of warfarin and the passage of some of the anticoagulant to the fetus.

This topic review describes anticoagulant use during attempted conception, pregnancy, and postpartum. Our practices are largely consistent with guidelines from societies such as the American Society of Hematology (ASH), American College of Chest Physicians (ACCP), American College of Obstetricians and Gynecologists (ACOG), and European Society of Cardiology (ESC), with the acknowledgment that most of the data that inform the recommendations are not from randomized trials. (See 'Society guideline links' below.)

Indications for anticoagulation during pregnancy are discussed in disease-specific topic reviews:

●DVT and PE treatment – (See "Venous thromboembolism in pregnancy and postpartum: Treatment".)

●DVT and PE prevention – (See "Venous thromboembolism in pregnancy: Prevention".)

●Inherited thrombophilia – (See "Inherited thrombophilias in pregnancy".)

●Sickle cell disease – (See "Sickle cell disease: Obstetric considerations", section on 'VTE prophylaxis'.)

●Atrial fibrillation – (See "Supraventricular arrhythmias during pregnancy", section on 'Anticoagulation'.)

●Prosthetic heart valve – (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

●APS – (See "Antiphospholipid syndrome: Obstetric implications and management in pregnancy".)

General issues related to anticoagulants are discussed separately. (See "Heparin and LMW heparin: Dosing and adverse effects" and "Warfarin and other VKAs: Dosing and adverse effects" and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects".)

CHOICE OF ANTICOAGULANT

Overview of anticoagulant choice — In contrast to anticoagulation of nonpregnant individuals, the choice of anticoagulant during pregnancy needs to take into account fetal safety and maternal peripartum issues including the unpredictable onset of labor and the use of neuraxial anesthesia for management of labor pain.

●Heparins – Heparins are used most often because they do not cross the placenta and do not result in fetal anticoagulation.

Heparins have not been associated with an overall increase in risk of congenital anomalies or any specific anomaly; however, the latter is challenging to ascertain due to the presence of underlying conditions that are independently associated with specific anomalies, the low number of cases, and the limitations of adjusting for multiple comparisons [1].

•Low molecular weight (LMW) heparins – We recommend LMW heparin rather than unfractionated heparin for all but the final weeks of pregnancy because LMW heparins are effective and easier to administer than unfractionated heparin [2,3]. LMW heparins produce a more predictable anticoagulant response than unfractionated heparin and do not require routine monitoring when used at therapeutic doses [4-6]. The risk of heparin-induced thrombocytopenia (HIT) is also less with LMW heparin than unfractionated heparin. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia", section on 'Incidence and risk factors'.)

•Unfractionated heparin – Unfractionated heparin is a reasonable alternative to an LMW heparin when need for rapid reversal is important (eg, for vaginal or cesarean birth, placement of neuraxial anesthesia, or surgery). Unfractionated heparin is preferred over LMW heparin in individuals with severely reduced kidney function (creatinine clearance <30 mL/min) because LMW heparin clearance is almost exclusively by the kidney, while elimination of unfractionated heparin is by the kidney and liver.

●Fondaparinux, argatroban, danaparoid – There is less information on the fetal effects of these agents, but available evidence suggests they are reasonable options. Fondaparinux is preferred for individuals with a history of HIT (or active HIT). Argatroban is most likely to cross the placenta. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia", section on 'Terminology and HIT variants'.)

●Avoidance of warfarin – Warfarin is generally avoided during pregnancy because it crosses the placenta, is a teratogen, and causes fetal anticoagulation. Exposure during early pregnancy can result in embryopathy, while exposure later in pregnancy can cause fetal bleeding, including intracranial hemorrhage.

An exception is an individual at especially high risk for thrombosis or thromboembolism (eg, mechanical heart valve), as discussed in more detail separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

●Avoidance of direct oral anticoagulants – Dabigatran, rivaroxaban, apixaban, and edoxaban are not used during pregnancy due to concern that information on efficacy and fetal safety is lacking.

Individuals receiving chronic anticoagulation who are contemplating pregnancy need counseling regarding how to avoid the adverse fetal effects of oral anticoagulants. (See 'Already taking warfarin' below and 'Already taking a DOAC' below.)

LMW heparin — An LMW heparin is generally the preferred anticoagulant [7]. This is largely because available evidence has shown LMW heparins to be effective and safe for the fetus. LMW heparins do not cross the placenta and do not cause fetal anticoagulation [2,3].

A systematic review of studies of LMW heparin for prevention or treatment of venous thromboembolism (VTE) in pregnancy concluded that LMW heparin was both safe and effective (64 studies, 2777 pregnancies) [8]. Rates of venous and arterial thrombosis were 0.8 and 0.5 percent, respectively, and rates were acceptably low for significant bleeding (2 percent), skin reactions (1.8 percent), and osteoporotic fractures (0.04 percent). There were no maternal deaths and no cases of HIT.

Unfractionated heparin — Unfractionated heparin is an acceptable alternative to LMW heparin. It may be more appropriate than LMW heparin when rapid temporal control of anticoagulation is required (approaching the time of birth, if surgery is required). (See 'Timing of heparin discontinuation' below.)

Unfractionated heparin is also preferred over LMW heparin in patients with severely reduced kidney function because LMW heparin clearance is exclusively by the kidney, while elimination of unfractionated heparin is by the kidney and liver.

Unfractionated heparin does not cross the placenta, and available evidence has not indicated any harmful effects on the fetus [9-11].

Alternatives to heparin — Non-heparin anticoagulants are generally not used during pregnancy unless there is a contraindication to heparins (such as HIT) or an inability to use injections.

●Danaparoid – Danaparoid is a low molecular weight heparinoid (heparan derivative) that is available in many countries (eg, Canada, Japan, Europe, Australia) but not the United States. It does not cross the placenta. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Danaparoid'.)

High-quality data regarding the use of danaparoid in pregnancy are lacking; this agent generally is reserved for pregnancies complicated by HIT [12-15]. (See 'HIT during or immediately preceding pregnancy' below.)

●Fondaparinux – Fondaparinux is a synthetic pentasaccharide structurally similar to the active moiety of heparin. Experience with fondaparinux during pregnancy is extremely limited, and data regarding placental passage are mixed. A series of 13 individuals treated with fondaparinux during 15 pregnancies reported 10 uncomplicated deliveries of healthy babies [16]. Of the remaining five pregnancies, three ended in miscarriage, one had a pregnancy termination due to fetal anomalies, and one was associated with recurrent VTE, possibly due to underdosing. One of the adverse outcomes may have been related to a delay in birth due to lack of experience with fondaparinux management, which emphasizes the importance of having a clear birth plan. Another series reported the use of fondaparinux during 12 pregnancies in 10 individuals who had hypersensitivity reactions to LMW heparin [17]. Outcomes were good; there was no major bleeding or fetal abnormalities. The American College of Chest Physicians (ACCP) suggests limiting the use of fondaparinux during pregnancy to individuals with severe reactions to heparin (eg, HIT) who are unable to receive danaparoid [15]. (See "Fondaparinux: Dosing and adverse effects" and 'HIT during or immediately preceding pregnancy' below.)

●Argatroban – Argatroban is a parenteral direct thrombin inhibitor reserved for those with severe reactions to heparins (eg, HIT) who cannot receive danaparoid or fondaparinux due to lack of availability or allergic reactions [18-20]. Argatroban requires continuous intravenous administration and is monitored by the activated partial thromboplastin time (aPTT). Argatroban is likely to cross the placenta due to its small size, although this has not been well studied [20]. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Argatroban'.)

Anticoagulants that are generally avoided during pregnancy

●Warfarin – Warfarin is generally avoided in pregnancy, except in those considered especially high risk such as having a mechanical heart valve. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

Warfarin and other vitamin K antagonists freely cross the placenta and are teratogenic, with the highest risk occurring with administration between weeks 6 and 12 of gestation; these agents can also cause fetal bleeding at any stage of pregnancy [9,10,21,22].

Warfarin has also been reported to be associated with early miscarriage [23], although the incidence is unknown, and it is unclear whether the increased rate of early miscarriages is due to the use of warfarin or to the underlying conditions for which warfarin was administered. Warfarin has been used during the second trimester of pregnancy when organogenesis has been mostly completed, but the risk of fetal bleeding is a disadvantage compared with other agents. (See 'Warfarin teratogenicity' below and "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

●DOACs – Direct oral anticoagulants (DOACs) include the oral direct thrombin inhibitor dabigatran and factor Xa inhibitors (rivaroxaban, apixaban, edoxaban). DOACs should not be used during pregnancy due to increased reproductive risks in animal studies and insufficient human safety and efficacy data [15,24,25]. DOACs also should not be used during breastfeeding [24]. (See 'Breastfeeding' below.)

A 2020 cohort study reported on 614 unique pregnancies of which 83 percent had DOAC exposure restricted to the first trimester (apixaban 50, dabigatran 36, edoxaban 23, and rivaroxaban 505); the median duration of exposure was 5.3 weeks [26]. Of the 336 pregnancies with known outcomes, 188 resulted in live births, 74 in miscarriages, and 74 in pregnancy terminations. Fetal abnormalities occurred in 21 (6 percent), 15 of which were classified as major; adjudication determined that 12 (4 percent) were possibly related to DOAC exposure. The incidence of fetal bleeding was not reported. The authors concluded that major birth defects were less frequent with DOACs than what was previously reported with warfarin (7.4 to 10.8 percent); no specific pattern of birth defects was observed with DOACs, and overall miscarriage rates were similar to the general population. These results should not be interpreted as providing evidence for the safety of DOACS in pregnancy, especially given the relatively brief exposures; however, they provide some reassurance to recommend that a pregnancy not be terminated solely based on DOAC exposure early in pregnancy.

In a 2016 series of 137 DOAC-exposed pregnancies from case reports and pharmacovigilance sources for which outcomes were available, there were 67 live births and 31 miscarriages (49 and 23 percent, respectively), which is comparable to the general population [27]. There were no instances of fetal bleeding; however, there were three anatomical abnormalities that might be interpreted as embryopathy (facial dimorphism, limb deformity, cardiac defect). This frequency is similar to population estimates, but conclusions are difficult to draw because outcomes were unavailable for an additional 96 DOAC-exposed pregnancies.

TIMING FOR STARTING LMW HEPARIN

Switching from oral anticoagulants to LMW heparin — Individuals receiving chronic oral anticoagulation who are contemplating pregnancy need counseling regarding avoidance of the potential teratogenic effects of warfarin. (See 'Already taking warfarin' below.)

Those receiving dabigatran, apixaban, edoxaban, or rivaroxaban should plan to switch to low molecular weight (LMW) heparin when they become pregnant. (See 'Already taking a DOAC' below.)

The management of anticoagulation around the time of conception is discussed below and in disease-specific topics. (See 'Specific scenarios' below and "Supraventricular arrhythmias during pregnancy", section on 'Anticoagulation' and "Management of risks of prosthetic valves during pregnancy", section on 'Preconception evaluation and counseling'.)

Initiating LMW heparin — For individuals who are not receiving chronic anticoagulation but require thromboprophylaxis during pregnancy, anticoagulation is initiated when pregnancy is established rather than before conception. LMW heparin is generally started in the first trimester after pregnancy is confirmed by a positive pregnancy test, as long as there is no vaginal bleeding.

ADMINISTRATION DURING PREGNANCY — Low molecular weight (LMW) and unfractionated heparin are usually administered subcutaneously, although unfractionated heparin can also be administered intravenously to achieve a more constant level of anticoagulation or during times when rapid discontinuation is advantageous (eg, birth, surgery). (See 'Timing of heparin discontinuation' below.)

Injections are usually well tolerated. Ice applied to the proposed injection site for 20 minutes prior to the injection can help to minimize bruising, although this generally is not necessary.

Baseline laboratory testing — We obtain a baseline platelet count in all individuals.

We do not routinely monitor the platelet count because the risk of heparin-induced thrombocytopenia (HIT) is very low in pregnancy [28]. However, it is reasonable to check the hemoglobin, hematocrit, and platelet count after approximately three to four weeks of therapy.

If the platelet count at three to four weeks is stable, it does not need to be repeated because subsequent development of HIT is unlikely, although there may be opportunities for additional monitoring if other laboratory testing is obtained. A platelet count should always be checked if thrombosis occurs during treatment with any form of heparin. (See 'Suspected HIT' below.)

We also obtain a baseline serum creatinine level in all patients. Diminished kidney function may indicate a need for more intensive monitoring or use of unfractionated heparin rather than LMW heparin. (See 'Choice of anticoagulant' above.)

Preservative-free vials — The multi-dose vials of LMW and unfractionated heparin contain benzyl alcohol and/or other preservatives. Alcohols can have adverse fetal effects and are contraindicated in pregnancy, although the concentration of alcohol in maternal blood from the preservative and a correlation of specific levels of benzyl alcohol with fetal harm have not been well characterized.

Prefilled, single-dose syringes of LMW heparin are generally preservative free. Confirmation of the absence of preservatives from the product label is advised. A disadvantage of unfractionated heparin is that the preservative-free formulation may be too dilute to enable injection of a small volume of medication.

Dosing and laboratory monitoring — Anticoagulants such as LMW heparin can be administered at different doses depending upon the risk of thromboembolism and desired degree of anticoagulation. The appropriate dosing level is discussed in topic reviews that describe the indications for anticoagulation. (See "Inherited thrombophilias in pregnancy" and "Venous thromboembolism in pregnancy: Prevention".)

We use the following terminology to describe heparin anticoagulation dosing:

●Low-dose anticoagulation refers to the use of low doses that aim to reduce the risk of venous thromboembolism (VTE) while minimizing bleeding complications.

●Intermediate-dose anticoagulation refers to the adjustment of prophylactic-dose anticoagulation with weight gain during pregnancy. (See 'Low and intermediate dose (used for VTE prophylaxis)' below.)

●Therapeutic-dose anticoagulation refers to full therapeutic doses typically reserved for treatment of thromboembolic disease.

Despite the nomenclature, therapeutic dosing may be used prophylactically (to prevent thromboembolism in some individuals at very high risk of thromboembolism when lower dosing is thought to be insufficient for thromboembolism prophylaxis). Examples include severe thrombophilias, mechanical heart valves and other high-risk situations.

For LMW heparin, therapeutic dosing is based on weight and, in rare cases, using anti-factor Xa levels. For unfractionated heparin, therapeutic dosing is titrated to keep the activated partial thromboplastin time in the therapeutic range (1.5 to 2.5 times baseline). (See 'Therapeutic dose' below.)

Dosing regimens for prophylactic and therapeutic LMW and unfractionated heparin are summarized in the table (table 1) and below. (See 'Low and intermediate dose (used for VTE prophylaxis)' below and 'Therapeutic dose' below.)

There is no exact formula for converting between LMW and unfractionated heparin. A reasonable approach is as follows:

●Converting from prophylactic-dose LMW to unfractionated heparin – Use the doses of unfractionated heparin listed below by trimester. (See 'Unfractionated heparin' below.)

●Converting from therapeutic-dose LMW to unfractionated heparin (eg, acute pulmonary embolism in the third trimester) – A rough calculation for converting from enoxaparin 1 mg/kg every 12 hours would be to use 250 units of unfractionated heparin per kg of body weight, given subcutaneously every 12 hours (sample calculation for a 100 kg patient: 250 units/kg x 100 kg = 25,000 units subcutaneously every 12 hours) [29].

For treatment of acute VTE, clinicians should be aware that the unfractionated heparin dose of 250 units/kg every 12 hours has not been validated in pregnancy, was used without laboratory monitoring, and was overlapped with three months of warfarin therapy in a comparative trial of unfractionated versus LMW heparin [29]. Therefore, clinical judgment should be used, and we generally start with somewhat lower doses of unfractionated heparin and monitor the aPTT six hours after administration, as described below. (See 'Unfractionated heparin' below.)

Dosing of heparins in pregnancy is altered because pregnancy-associated weight gain and metabolism affect the pharmacokinetics of these agents; higher doses are necessary compared with nonpregnant individuals due to alterations in a variety of factors including increased plasma volume and increased clearance by the kidney [25]. In a study of LMW heparin pharmacokinetics in 24 individuals at 12, 24, and 36 weeks of gestation and six weeks postpartum, peak anti-factor Xa activity levels during pregnancy were lower than postpartum levels and occurred later after injection (four versus two hours) [30].

Dosing postpartum is discussed below and in separate topic reviews. (See 'Postpartum and breastfeeding' below and "Cesarean birth: Preoperative planning and patient preparation", section on 'Thromboembolism prophylaxis' and "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'During pregnancy, labor and delivery, and postpartum'.)

Low and intermediate dose (used for VTE prophylaxis)

LMW heparin — Low-dose or intermediate-dose LMW heparin is used for VTE prophylaxis in individuals at increased risk for VTE. Indications are discussed separately. (See "Inherited thrombophilias in pregnancy", section on 'Prevention of VTE' and "Venous thromboembolism in pregnancy: Prevention", section on 'Indications'.)

●Low dose – Low-dose LMW heparin uses a fixed dose with adjustment for extremes of body weight (table 1) [31].

●Intermediate dose – Intermediate-dose LMW heparin is another prophylactic strategy that uses an increasing dose as the pregnancy progresses and the patient's weight increases, up to a maximum dose of enoxaparin 1 mg/kg once daily (table 1).

The frequency of dose increases is individualized depending on the interval between office visits. Intermediate dosing provides a higher dose than standard prophylactic regimens of LMW heparin, but with only a single daily injection, which is better tolerated than twice daily injections.

For individuals who require VTE prophylaxis, the 2018 American Society of Hematology (ASH) suggested low-dose rather than intermediate-dose LMW heparin for VTE prophylaxis during pregnancy and either low- or intermediate-dose postpartum [7].

The relative efficacy of low-dose versus intermediate-dose LMW heparin for VTE prophylaxis was compared in the Highlow trial, which randomly assigned 1110 individuals who were ≤14 weeks of gestation and had a prior history of VTE to receive weight-adjusted intermediate-dose or fixed low-dose LMW heparin (eg, enoxaparin 40 mg for weight <100 kg or 60 mg for weight ≥100 kg) once daily until six weeks postpartum [31]. The risk of VTE was similar between groups (2 percent with intermediate dose versus 3 percent with low dose; relative risk [RR] 0.69, 95% CI 0.32-1.47). The trend towards increased risk in the low-dose group was due to a slightly higher number of postpartum VTE events that did not meet statistical significance. The risk of bleeding was also similar (4 percent in each group; RR 1.16, 95% CI 0.65-2.09). Editorialists commented on methodologic concerns, including that it took seven years to accrue 1110 patients and that there were numerous protocol deviations and a low event rate [32].

Prior to this trial, intermediate dosing was thought to have an efficacy advantage based on pharmacokinetic studies that suggested a greater dose requirement after 20 weeks of gestation [33]. A small retrospective study (126 pregnancies) had found a high incidence of VTE in high-risk individuals treated with low-dose LMW heparin antepartum and postpartum, suggesting that low-dose LMW heparin prophylaxis may have been insufficient in high-risk individuals [34].

Though the best available evidence favors low-dose LMW heparin for antepartum thromboprophylaxis for most individuals, there remains a role for individualized risk assessment based on history of VTE, patient weight, and personal preference. Monitoring is not required for low-dose or intermediate-dose LMW heparin.

Unfractionated heparin — Unfractionated heparin is given subcutaneously every 12 hours, with increasing doses as the pregnancy progresses, from 5000 to 7500 units in the first trimester, to 7500 to 10,000 units in the second trimester, to 10,000 units in the third trimester (reduce if the activated partial thromboplastin time [aPTT] is elevated) [25,35]. Some clinicians use 5000 units subcutaneously every 12 hours throughout the pregnancy, but studies have suggested that this dose is probably insufficient in some patients based on plasma heparin levels [36-40].

Monitoring of prophylactic dose unfractionated heparin is generally not performed, although the aPTT can be measured if there are concerns about bleeding or thrombosis.

Therapeutic dose

LMW heparin — Therapeutic-dose LMW heparin is based on body weight and usually administered every 12 hours by subcutaneous injection (table 1). Our preference is enoxaparin 1 mg/kg every 12 hours or dalteparin 100 units/kg every 12 hours; we do not employ once daily regimens (enoxaparin 1.5 mg/kg or dalteparin 200 units/kg once daily) because we wish to avoid higher peak and lower trough levels.

We do not routinely monitor anti-factor Xa activity levels, consistent with a 2018 ASH guideline [7]; however, some experts do recommend such monitoring. (See "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'Most patients: Anticoagulation'.)

An exception is extremes of body weight. As an example, we occasionally monitor anti-factor Xa activity for individuals >100 kg and especially for those >150 kg and use the results to make dose adjustments if needed.

If monitoring is done, peak anti-factor Xa activity levels are measured four to six hours after dosing, and the dose is titrated to maintain a target peak anti-factor Xa activity of approximately 0.6 to 1.2 units/mL [25]. Data supporting the need for laboratory monitoring of therapeutic-dose LMW heparin are sparse [30,41,42]. (See "Clinical use of coagulation tests", section on 'Monitoring heparins'.)

Unfractionated heparin — Therapeutic-dose unfractionated heparin is administered every 12 hours by subcutaneous injection and titrated based on the aPTT. In some patients, the volume of medication in the injection may become unacceptably high; this can be avoided by using more concentrated heparin solutions (eg, 20,000 units/mL).

Heparin can be administered intravenously if the need for rapid discontinuation is likely (eg, imminent labor/birth, surgery).

Unfractionated heparin is monitored using the aPTT, measured six hours after injection. The dose should be adjusted to maintain the aPTT at 1.5 to 2.5 times the mean of the control value or the patient's baseline aPTT value. We monitor the aPTT daily until the proper dose is achieved and every one to two weeks once it is in the therapeutic range.

Anti-factor Xa activity calibrated for unfractionated heparin can also be used to monitor and titrate unfractionated heparin dosing. One test (aPTT or anti-factor Xa activity) should be used, rather than a combination of both tests. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Laboratory monitoring and dose titration (unfractionated heparin)'.)

Intravenous heparin dosing and heparin dose monitoring in individuals with a baseline prolonged aPTT (such as due to a lupus anticoagulant phenomenon in antiphospholipid syndrome [APS]) are discussed separately. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Prolonged baseline aPTT'.)

PREVENTION OF THROMBOSIS AND BLEEDING — Anticoagulants are administered to patients at increased risk of thromboembolic disease, and this risk, although mitigated by therapy, persists and may increase during pregnancy.

Individuals treated with anticoagulation should be aware of the signs and symptoms of thromboembolism (leg swelling, pleuritic chest pain, central nervous system symptoms) and bleeding and have instructions to contact their clinician should these occur. (See "Deep vein thrombosis in pregnancy: Epidemiology, pathogenesis, and diagnosis" and "Cerebrovascular disorders complicating pregnancy".)

Concerns about excess bleeding in a patient receiving low molecular weight (LMW) or unfractionated heparin do not apply to the fetus, since heparins do not cross the placenta.

Other than appropriate dosing and monitoring, strategies to decrease the risk of bleeding include the following:

●Convert from subcutaneous LMW heparin to unfractionated heparin prior to birth and from subcutaneous unfractionated heparin to intravenous unfractionated heparin prior to anticipated birth in those who require more continuous anticoagulation. Discontinue heparin at the onset of labor. Maximum control of anticoagulation can be achieved if the timing of birth is planned (scheduled cesarean or induction of labor). (See 'Timing of heparin discontinuation' below.)

●If preterm labor develops in a patient receiving heparin, protamine sulfate has been used to reverse maternal heparinization. However, it is best to avoid administration of protamine antepartum unless hemorrhage cannot be controlled using routine supportive measures. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal' and "Heparin and LMW heparin: Dosing and adverse effects", section on 'Bleeding'.)

●Place or remove a neuraxial needle or catheter only after the patient is no longer anticoagulated. (See 'Neuraxial anesthesia' below.)

MANAGEMENT OF COMPLICATIONS

Treatment of bleeding on heparin — Protamine sulfate can be used to rapidly reverse the effects of unfractionated heparin in patients with serious or severe bleeding unrelated to pregnancy; patients at risk of severe bleeding because of imminent vaginal or cesarean birth; or patients at risk of serious or severe bleeding due to antepartum complications (eg, placental abruption, placenta previa, expanding subchorionic hematoma). Repeated small doses of protamine may be required because of ongoing absorption of heparin from subcutaneous tissue. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal'.)

Low molecular weight (LMW) heparin may not be completely reversed by protamine, but protamine may reduce bleeding and should be used if bleeding is severe. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal'.)

Minor bleeding that is easily visible (eg, spotting) does not require pharmacologic reversal of anticoagulation. If bleeding persists, one option is to withhold anticoagulation, if possible, until the bleeding stops and then resume the anticoagulant. Decisions depend on the degree and site of bleeding and the indication for anticoagulation and are made on a case-by-case basis by the hematologist and obstetrician.

Patients with bleeding who require VTE prophylaxis can be managed with mechanical means or vena cava filters if needed; these interventions are discussed separately. (See "Prevention of venous thromboembolic disease in acutely ill hospitalized medical adults", section on 'Mechanical methods of thromboprophylaxis' and "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'Patients at high bleeding risk or contraindications to anticoagulation: IVC filter'.)

Suspected HIT — Heparin-induced thrombocytopenia (HIT) is an adverse reaction to heparin in which antiplatelet antibodies are induced by the drug. Unlike other drug-induced thrombocytopenias, HIT antibodies cause platelet activation that can result in life-threatening arterial and venous thrombosis. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia" and 'HIT during or immediately preceding pregnancy' below.)

HIT can occur in any patient receiving any amount of heparin; however, the incidence in pregnancy is very low. This was illustrated in a meta-analysis of 2777 pregnancies during which LMW heparin was administered; no instances of HIT were reported [8]. LMW heparin appears less likely to precipitate HIT compared with unfractionated heparin in some populations (eg, surgical patients); this has not been studied in pregnancy [4].

Any patient who develops thrombocytopenia while receiving heparin should have a clinical evaluation for HIT that includes platelet count and clinical assessment for thrombosis and/or skin changes. Those with a high clinical suspicion for HIT should have immediate discontinuation of heparin, institution of a non-heparin anticoagulant, and laboratory testing for HIT (heparin-PF4 antibody testing). (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia", section on 'Evaluation' and 'HIT during or immediately preceding pregnancy' below.)

A modest decrease in platelet count is expected during pregnancy and should not prompt laboratory testing for HIT. (See "Thrombocytopenia in pregnancy", section on 'Gestational thrombocytopenia (GT)'.)

In addition to evaluation for HIT, pregnant patients with a drop in platelet count that is greater than expected should be evaluated for other potential causes of thrombocytopenia. (See "Thrombocytopenia in pregnancy".)

Bone loss — Prolonged use of unfractionated heparin (more than a few weeks) is associated with decreases in bone mineral density (BMD). Individuals taking unfractionated heparin for more than a few weeks should have adequate intake of calcium and vitamin D, and regular weight bearing exercise (eg, walking) should be encouraged. (See "Overview of the management of low bone mass and osteoporosis in postmenopausal women", section on 'Lifestyle measures to reduce bone loss'.)

It is unclear whether bone loss may be reduced or prevented by using LMW heparin instead of unfractionated heparin. A study that randomly assigned 44 pregnant women to receive either dalteparin (target anti-Xa greater than 0.20 international units/mL three hours after injection) or unfractionated heparin (mean dose 17,250 units/day) found that mean BMD in the lumbosacral spine was significantly lower with unfractionated heparin at one week to three years postpartum; those who received LMW heparin had similar BMD to postpartum women who were not exposed to any form of heparin [43]. Other studies have found no difference in the effect of LMW versus unfractionated heparin on BMD [44]. (See "Drugs that affect bone metabolism", section on 'Anticoagulants'.)

LABOR AND DELIVERY — Anticoagulation during labor should be avoided except in the highest risk settings (eg, reduced cardiopulmonary reserve and recent pulmonary emboli). Normal coagulation is desirable intrapartum to allow safe administration of a neuraxial anesthetic and reduce the risk of excessive bleeding at delivery [45,46]. (See 'Neuraxial anesthesia' below.)

Timing of heparin discontinuation

●Patients not at very high thrombotic risk – For patients in whom a period of up to 24 hours without anticoagulation is reasonably safe, the following approaches to prepare for labor and delivery are acceptable:

•Continue LMW heparin and plan induction or cesarean birth at 39 weeks of gestation – Planned induction at 39 weeks is a reasonable alternative to expectant management in any well-dated pregnancy in which the patient desires a trial of labor and vaginal birth (see "Induction of labor with oxytocin", section on 'At 39 weeks'). For patients planning a cesarean birth (eg, repeat cesarean, breech presentation), the procedure is typically scheduled for 39 weeks.

With this approach, patients on low-dose LMW heparin take their last dose 12 hours before the procedure; patients on intermediate-dose or therapeutic-dose LMH take their last dose 24 hours before the procedure [7,25]. If the patient develops signs or symptoms of spontaneous labor before 39 weeks, LMW is discontinued at that time. This approach works best for patients at low risk for preterm labor and no pregnancy complications who are likely to require prompt early birth.

Extending LMW heparin to 38 to 39 weeks or beyond in selected cases is supported by an analysis from the Highlow trial, which compared low-dose versus intermediate-dose LMW heparin [47]. Patients were not eligible for a neuraxial procedure when the time interval since the last LMW heparin injection was <12 hours for the low-dose arm and <24 hours for the intermediate-dose arm. Based on these time intervals, 82 percent of patients on low-dose and 61 percent of patients on intermediate-dose LMW heparin were eligible for a neuraxial procedure at the unplanned onset of labor; for planned labor, 81 percent on intermediate-dose and 93 percent on low-dose were eligible. Thus, most patients were eligible despite use of LMW heparin.

•Switch from LMH heparin to unfractionated heparin – Another approach is to switch from LMW heparin to unfractionated heparin before term, typically at approximately 36 weeks. If induction or cesarean birth is planned, patients on low-dose unfractionated heparin take their last dose 12 hours before the procedure; patients on intermediate-dose or therapeutic-dose unfractionated heparin take their last dose 24 hours before the procedure. If induction or cesarean birth is not planned, then unfractionated heparin is discontinued when the patient first develops signs and symptoms of spontaneous labor.

In patients on low-dose unfractionated heparin, the heparin effect may dissipate within four to five hours following the last dose. In patients on therapeutic doses of unfractionated heparin, the activated partial thromboplastin time (aPTT) can take 24 hours to normalize after stopping subcutaneous administration.

The advantages of unfractionated heparin over LMW heparin are that the anticoagulant effect probably diminishes faster after discontinuation because of a shorter half-life, the anticoagulant effect is readily assessed by the activated partial thromboplastin time (aPTT), and anticoagulation can be reversed with protamine sulfate when required.

●Patients at very high thrombotic risk – Some patients have a strong need for ongoing anticoagulation (eg, prosthetic heart valve, atrial fibrillation with thrombus, antepartum deep vein thrombosis or pulmonary embolism), and in these individuals, a period of 24 hours without anticoagulation may be undesirable. Details of management in these settings are presented separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy", section on 'Peripartum management' and "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'During pregnancy, labor and delivery, and postpartum'.)

Neuraxial anesthesia — Neuraxial anesthesia techniques (spinal, epidural, or combined spinal epidural) should not be performed if a patient is anticoagulated, due to the risk of spinal or epidural hematoma. Anticoagulation is discontinued before labor to allow neuraxial analgesia or anesthesia, which is performed for >95 percent of cesarean deliveries and >60 percent vaginal deliveries in the United States [48].

Examples of the interval between anticoagulant discontinuation and neuraxial anesthesia/analgesia include the following (table 2):

●Prophylactic-dose LMW heparin – After at least 12 hours have elapsed since the last dose.

●Intermediate- and therapeutic-dose LMW heparin – After at least 24 hours have elapsed since the last dose.

●Prophylactic- and therapeutic-dose unfractionated heparin – Once the aPTT has normalized following discontinuation (timing of aPTT testing depends on dose and route of administration). In patients on therapeutic doses of unfractionated heparin, the aPTT is usually normal 6 hours after stopping intravenous administration but can take 24 hours to normalize after stopping subcutaneous administration.

It is possible that in the future, thromboelastography (TEG) may be used to determine when neuraxial anesthesia can be initiated following discontinuation of LMW heparin, but more data are needed before this approach can be used [49].

We do not administer a reversal agent (eg, protamine) unless there is excessive or unexpected bleeding due to the anticoagulant.

Further details, and timing of anticoagulant resumption after catheter removal, are discussed in detail separately. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

POSTPARTUM AND BREASTFEEDING

Resuming or initiating anticoagulation postpartum — Anticoagulation is reinstituted following birth in most patients who were receiving an anticoagulant during pregnancy; a possible exception is patients who received anticoagulation to prevent miscarriage or fetal loss.

●Acute VTE (therapeutic dosing) – For acute venous thromboembolism (VTE) still in the active treatment period (first three months) or other indications requiring continuous anticoagulation (eg, high-risk thrombophilia), either unfractionated heparin or low molecular weight (LMW) heparin can be given alone at therapeutic doses, or one of these anticoagulants should be administered during the initiation of warfarin and continued for a minimum of five days and an additional one to two days after an appropriate International Normalized Ratio (INR) has been obtained. (See "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'During pregnancy, labor and delivery, and postpartum'.)

Acute VTE confers the highest risk for VTE recurrence, progression, or complications, and the risk-benefit calculation favors initiation of heparin as early as safely possible.

The ideal time to start anticoagulation after giving birth is based on clinical judgment, as comparative data are very limited. Unfractionated heparin or LMW heparin are generally resumed 4 to 6 hours after vaginal birth or 6 to 12 hours after cesarean birth, unless there was significant postpartum bleeding or traumatic neuraxial catheter placement [25].

A retrospective study reported that bleeding rates were lower when more time elapsed following delivery (≥9.25 hours for vaginal and ≥15.1 hours for cesarean) [50]. However, bleeding rates in a matched control group not receiving anticoagulation were not provided, co-interventions may have differed among groups, and the confidence intervals were wide and overlapping. Major factors to consider in deciding when to restart anticoagulation include the underlying indication for anticoagulation (such as acute treatment of VTE versus prophylaxis) and other risk factors for bleeding. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

Heparin can be transitioned to warfarin, fondaparinux, or danaparoid (not available in the United States-), or, if not breastfeeding, to a direct oral anticoagulant if appropriate [25]. (See 'Breastfeeding' below and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Transitioning between anticoagulants' and "Warfarin and other VKAs: Dosing and adverse effects", section on 'Transitioning between anticoagulants/bridging'.)

Resumption of anticoagulation for individuals mechanical heart valves is discussed separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy", section on 'Postpartum management'.)

●VTE prophylaxis – Additional patients who were not receiving anticoagulation during pregnancy may initiate it postpartum for VTE prevention (eg, inherited or acquired thrombophilia or prior VTE). These indications are discussed separately. (See "Inherited thrombophilias in pregnancy", section on 'Prevention of VTE' and "Venous thromboembolism in pregnancy: Prevention".)

The urgency of starting heparin is less than for acute VTE (or VTE within the prior three months). We initiate unfractionated heparin or LMW heparin at a low or intermediate dose starting 4 to 6 hours after vaginal birth and 6 to 12 hours after cesarean birth (in patients who received epidural anesthesia) [25]. LMW heparin is not started sooner than four hours after epidural catheter removal [51,52]. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication", section on 'Low molecular weight heparin (LMWH)'.)

In a review of 95 individuals treated with postpartum anticoagulation, mostly low dose, we found very few cases of bleeding, with the exception of two incisional hematomas following cesarean birth in individuals who received low dose enoxaparin within 24 hours following delivery [53]. If oral anticoagulation is chosen for continued postpartum thromboprophylaxis, warfarin can be started immediately after delivery; unfractionated or LMW heparin can be discontinued when a therapeutic INR has been obtained if the patient is not at very high risk for thrombotic complications.

Prophylactic anticoagulation after cesarean birth is discussed separately. (See "Cesarean birth: Preoperative planning and patient preparation", section on 'Thromboembolism prophylaxis'.)

Duration of postpartum anticoagulation — The duration of postpartum anticoagulation depends upon the underlying reason for anticoagulation.

●Anticoagulation for VTE prophylaxis should be continued for at least six weeks postpartum, due to the continued high risk of VTE in the early postpartum period. (See "Inherited thrombophilias in pregnancy" and "Venous thromboembolism in pregnancy: Prevention", section on 'Duration'.)

●The duration of anticoagulation for VTE treatment during pregnancy is discussed separately. (See "Venous thromboembolism in pregnancy and postpartum: Treatment".)

Breastfeeding — Many anticoagulants can be used during breastfeeding because they do not accumulate in breast milk. The 2018 American Society of Hematology (ASH) guidelines recommend continuation of the following anticoagulants (if indicated) during breastfeeding [7]:

●LMW heparin

●Unfractionated heparin

●Fondaparinux

●Warfarin or other vitamin K antagonists

●Danaparoid (not available in the United States)

Continuation of low-dose aspirin for vascular indications is also appropriate [15].

The 2012 American College of Chest Physicians (ACCP) guidelines agreed with the above list with the exception of fondaparinux, for which there was considered to be insufficient evidence at the time the 2012 guideline was published [15].

In contrast to the agents listed above, we do not use the following during breastfeeding, consistent with the 2018 ASH and 2012 ACCP guidelines [7,14]:

●Oral direct thrombin inhibitors (dabigatran)

●Oral direct factor Xa inhibitors (rivaroxaban, apixaban, edoxaban)

A review of observational data and a case report with testing of drug levels, however, suggest that dabigatran and rivaroxaban levels in breast milk are quite low, whereas apixaban levels are higher [54,55].

If anticoagulation is required during breastfeeding, another agent should be selected.

Contraception — Contraceptive options for individuals with inherited thrombophilias are discussed separately. (See "Contraception: Counseling for women with inherited thrombophilias".)

SPECIFIC SCENARIOS

HIT during or immediately preceding pregnancy — A prior episode of heparin-induced thrombocytopenia (HIT) that has been adequately treated is not an indication for anticoagulation during pregnancy. However, if a pregnant patient with a history of HIT requires anticoagulation for another reason, or if a patient develops HIT immediately prior to or during pregnancy, an anticoagulant other than heparin should be used. All sources of heparin (including heparin flushes) should be avoided.

The 2012 American College of Chest Physicians (ACCP) guidelines recommend danaparoid as the preferred alternative to heparin for pregnant patients; this agent is not available in the United States [15]. For patients who cannot receive danaparoid, argatroban or fondaparinux can be used. We generally use fondaparinux because of the ease of administration during pregnancy. Other aspects of the management of HIT are discussed separately. (See "Management of heparin-induced thrombocytopenia".)

The use of danaparoid during pregnancy complicated by HIT is supported by the outcomes of 91 pregnancies in 83 individuals with HIT; some of the patients also had antiphospholipid syndrome [56]. Maternal adverse events included two post-cesarean deaths in the setting of placenta previa and placental abruption (one from pulmonary embolus and one from bleeding in which the patient refused transfusion); three major bleeds; three thromboembolic events; and 10 rashes. There were seven early miscarriages, one therapeutic termination, and one neonatal death associated with a lupus anticoagulant and placental infarction [57]. HIT is associated with a high baseline mortality, even with treatment. The live birth rate in this review was 90 percent, which compares favorably with use of aspirin or heparin during pregnancy [56].

In the rare pregnant patient with a prior history of HIT who requires anticoagulation for another reason and cannot be treated with an alternative to unfractionated or low molecular weight (LMW) heparin, we screen for HIT prior to administering unfractionated or LMW heparin by checking for HIT antibodies; platelet counts are checked before initiating heparin and monitored periodically thereafter. (See 'Suspected HIT' above.)

Baseline platelet count testing and platelet count monitoring is discussed above. (See 'Baseline laboratory testing' above.)

Thromboembolism during pregnancy — If a new thromboembolic event is suspected during pregnancy, this should be evaluated immediately.

Diagnosis and management of venous thromboembolism (VTE) during pregnancy is discussed separately. (See "Deep vein thrombosis in pregnancy: Epidemiology, pathogenesis, and diagnosis" and "Venous thromboembolism in pregnancy and postpartum: Treatment".)

Cesarean birth — Cesarean birth increases the risk of VTE, especially when performed urgently or in an emergency. Prevention of VTE after cesarean birth is reviewed separately. (See "Cesarean birth: Preoperative planning and patient preparation", section on 'Thromboembolism prophylaxis'.)

Already taking warfarin — Warfarin and other vitamin K antagonists cross the placenta; these agents can be teratogenic and can cause fetal bleeding, including hemorrhagic fetal death. Warfarin is generally avoided during pregnancy, or, rarely, restricted to the second and early third trimester. If thromboembolic risk is exceptionally high (eg, mechanical heart valve) the benefits and risks of warfarin, including fetal risks, are weighed during each trimester and an alternative anticoagulant is used approaching birth, as discussed separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

Warfarin-associated bleeding — The management of excessive anticoagulation with or without bleeding depends on the original indication for warfarin therapy, the degree and site of bleeding, and the International Normalized Ratio (INR). Management decisions are made on a case-by-case basis in consultation with the hematologist and other specialists (cardiologist, neurologist, obstetrician). Available agents for reversal (Fresh Frozen Plasma [FFP], prothrombin complex concentrates [PCCs], vitamin K) and their use in specific situations are discussed separately. (See "Management of warfarin-associated bleeding or supratherapeutic INR" and "Reversal of anticoagulation in intracranial hemorrhage".)

For preterm labor and/or unplanned delivery in an individual taking warfarin, it is important to remember that the fetus is anticoagulated. (See 'Birth' below.)

Warfarin teratogenicity — Warfarin is a teratogen. The precise incidence of warfarin embryopathy is unknown, with different series reporting widely ranging numbers. The best overall estimate of the risk is <10 percent. The teratogenic effect appears to be dose-related rather than correlating with maternal INR; doses less than 5 mg per day appear to provide the highest margin of safety, but teratogenicity at these doses has been reported [58-60].

The risk of teratogenicity is greatest for fetuses exposed to warfarin between the sixth and 12^th weeks of gestation [36]. However, toxicity before or after this period is still possible [61-64]. A series of 72 pregnancies in individuals taking warfarin for a heart valve reported the following [21]:

●Virtually no embryopathic events occurred in the 23 pregnancies in which warfarin was discontinued by the sixth week of gestation and not restarted until after the 12^th week.

●Warfarin embryopathy occurred in 25 percent of the 12 pregnancies in which warfarin was not stopped until after the seventh week.

●Embryopathy occurred in 30 percent of the 37 pregnancies in which warfarin was continued throughout the entire pregnancy.

The most common developmental abnormalities affect bone and cartilage; these simulate chondromalacia punctata, with stippled epiphyses and nasal and limb hypoplasia [36]. The mechanism of this type of warfarin teratogenicity has not been established; it may be related to the drug's interference with the post-translational modification of calcium-binding proteins that are important for the normal growth and development of bony structures [22]. (See "Vitamin K-dependent clotting factors: Gamma carboxylation and functions of Gla".)

Less well-documented are reports of central nervous system abnormalities (optic atrophy, microcephaly, intellectual disability, spasticity, and hypotonia) associated with warfarin use at any stage during pregnancy [9,65-68]. This complication may be related to fetal anticoagulation leading to cerebral hemorrhage.

Attempted conception/first trimester — Individuals taking warfarin prior to conception should have a clear plan to switch to another anticoagulant (typically LMW heparin) during attempted conception or immediately upon becoming pregnant, to avoid teratogenic effects of the drug during the first trimester. Though the direct oral anticoagulants (DOACs) have not been shown to be teratogenic, individuals taking these anticoagulants should be switched to LMW heparin prior to conception; if pregnancy occurs unintentionally, the DOAC should be discontinued immediately and LMW heparin commenced [24].

The optimal timing for changing to LMW heparin is unclear, and it is necessary to balance the importance of avoiding warfarin or a DOAC during early pregnancy/organogenesis with the inconvenience of using LMW heparin for an undefined period until conception occurs. The two major options are to change to LMW heparin during attempted conception, or to wait until pregnancy is confirmed before changing from warfarin to LMW heparin, as long as the switch can be made before six weeks of pregnancy (≤14 days after the missed first day of expected menses).

For warfarin, the 2012 ACCP guidelines made a weak suggestion in favor of performing frequent pregnancy tests and substituting treatment with LMW heparin as soon as pregnancy is documented, rather than changing to LMW heparin while attempting conception [15]. (See 'Society guideline links' below.)

We believe that this is a reasonable option for an individual who meets all of the following criteria:

●Regular monthly menstrual cycles.

●Willingness to have a blood pregnancy test within the first seven days of the missed first day of expected menses. This can be facilitated by having a standing order at a laboratory or providing laboratory requisitions in advance.

●Ability to switch to LMW heparin preparation promptly if the pregnancy test is positive, with a second blood pregnancy test if the first test is negative and menses have not begun within 10 days of the missed first day of expected menses. This can be facilitated by having a prescription for LMW heparin readily available or filled in advance.

●Understanding of the increased risk and types of embryopathy if warfarin is continued during or after the sixth week of pregnancy (≥14 days after the missed first day of expected menses). (See 'Warfarin teratogenicity' above.)

Individuals who do not meet all of the above criteria; those who prefer to minimize the possible increased risk of early miscarriage associated with warfarin therapy or any possible risk of a DOAC; and those who place less value on avoiding risks, inconveniences, and costs of LMW heparin of uncertain duration should be switched to LMW heparin during attempted conception. An exception is an individual with a mechanical heart valve, for whom warfarin generally is continued during attempted conception, as discussed separately. (See "Management of risks of prosthetic valves during pregnancy", section on 'Counseling regarding anticoagulant therapy'.)

If pregnancy occurs while an individual is taking warfarin, LMW heparin should be substituted as soon as possible. Prompt obstetric consultation should be obtained to establish the gestational age of the fetus and to provide appropriate counseling regarding any potential teratogenic risk.

Birth — The risk of fetal hemorrhage related to warfarin use is thought to be greatest during and immediately after birth [21,58,67-71]. If warfarin is used during pregnancy, it should be discontinued after 34 to 36 weeks of gestation and an alternative anticoagulant substituted if appropriate. (See 'Timing of heparin discontinuation' above.)

If preterm birth occurs in a patient receiving warfarin, cesarean birth should be considered to reduce the risk of fetal bleeding [36]. Vitamin K and fresh frozen plasma should be administered to the neonate if preterm birth occurs in a patient receiving warfarin.

The immaturity of fetal enzyme systems and the relatively low concentration of vitamin K-dependent clotting factors render the fetus more sensitive to the anticoagulant effects of warfarin [9,65,66]. Importantly, fetal levels of coagulation factors do not correlate with maternal levels, and infusion of FFP to the mother does not reliably reverse fetal anticoagulation.

Already taking a DOAC — Direct thrombin inhibitors and factor Xa inhibitors are not used during pregnancy. An individual taking one of these agents who plans to become pregnant should be monitored very closely, as described for warfarin. (See 'Attempted conception/first trimester' above.)

If an individual taking one of these agents becomes pregnant, she should switch immediately to LMW heparin.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Anticoagulation in pregnancy".)

SUMMARY AND RECOMMENDATIONS

●Choice of agent – For most patients who require anticoagulation during pregnancy (except those with mechanical heart valves), heparins are considered safer than other anticoagulants. We recommend low molecular weight (LMW) heparin rather than unfractionated heparin for all but the final weeks of pregnancy (Grade 1B). Unfractionated heparin can be used when cost or need for rapid reversal is important and is the preferred agent for patients with creatinine clearance ≤30 mL/min and in preparation for labor and delivery. (See 'Choice of anticoagulant' above.)

The choice of anticoagulant with mechanical heart valves is discussed separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy".)

●Dosing – LMW heparin dosing depends on thromboembolism risk and the agent used (table 1). We obtain a baseline creatinine and platelet count and monitor platelet counts periodically. (See 'Administration during pregnancy' above and "Inherited thrombophilias in pregnancy", section on 'Prevention of VTE' and "Venous thromboembolism in pregnancy: Prevention", section on 'Indications'.)

●Risks

•Bleeding – Anticoagulation increases bleeding risk. Decisions regarding discontinuation and reversal of anticoagulants are made on a case-by-case basis in consultation with a hematologist and other specialists. (See 'Prevention of thrombosis and bleeding' above and 'Treatment of bleeding on heparin' above.)

•HIT – Heparin-induced thrombocytopenia (HIT) is very rare in pregnancy Patients with suspected HIT should have an immediate assessment of clinical parameters (platelet count, presence of thrombosis or skin changes). If suspicion for HIT is high, a non-heparin anticoagulant should be substituted and laboratory testing for HIT obtained. (See 'Suspected HIT' above and 'HIT during or immediately preceding pregnancy' above.)

●Labor and delivery – In patients at low risk of giving birth before 39 weeks of gestation, it is reasonable to plan induction or cesarean birth at 39 weeks and discontinue low-dose LMW heparin 12 hours before the procedure and discontinue intermediate- or therapeutic-dose LMW heparin 24 hours before the procedure. If labor occurs before 39 weeks, heparin is discontinued at the first signs of labor. Alternatively, patients on LMW heparin may switch to unfractionated heparin before term, usually at 36 weeks, and discontinue it as described for LMW heparin. The advantage of switching is that unfractionated heparin has a shorter half-life so its effect dissipates faster after discontinuation, the anticoagulant effect is readily assessed by the activated partial thromboplastin time (aPTT), and anticoagulation can be reversed with protamine sulfate when required. (See 'Timing of heparin discontinuation' above.)

Some patients have a strong need for ongoing therapeutic anticoagulation (eg, prosthetic heart valve, atrial fibrillation with thrombus, antepartum deep vein thrombosis or pulmonary embolism), and in these individuals, a period of 24 hours without anticoagulation may be undesirable. Details of management in these settings are presented separately. (See "Management of antithrombotic therapy for a prosthetic heart valve during pregnancy", section on 'Peripartum management' and "Venous thromboembolism in pregnancy and postpartum: Treatment", section on 'During pregnancy, labor and delivery, and postpartum'.)

●Neuraxial anesthesia – A catheter for neuraxial anesthesia cannot be inserted if a patient is anticoagulated. The timing of discontinuation depends on the agent and dose used. (See 'Neuraxial anesthesia' above.)

●Postpartum – Anticoagulation can be reinstituted or started following birth. Low-dose- or intermediate-dose heparin for VTE prophylaxis can typically be resumed as early as 4 to 6 hours after vaginal birth and 6 to 12 hours after cesarean birth in the absence of bleeding. Many anticoagulants (heparins, warfarin) can be used during breastfeeding. (See 'Postpartum and breastfeeding' above.)

●Chronic anticoagulation – Warfarin and direct oral anticoagulants (DOACs) should not be continued during the first trimester. A switch to LMW heparin should occur before attempted conception or immediately upon confirmation of pregnancy. (See 'Already taking warfarin' above and 'Already taking a DOAC' above and 'Timing for starting LMW heparin' above.)