Evidence is limited regarding the prevalence and optimal management of pregnancy in individuals with mechanical heart valves (MHVs). Studies are scarce, often with small numbers of patients included. Mechanical valve thrombosis (MVT) occurs more frequently in pregnancy and there is a high risk of postpartum haemorrhage (PPH). There are a number of options for anticoagulation with differing risks to the pregnant individual and foetus. In the absence of high-quality data, this guideline aims to give recommendations using observational data, evidence from outside of pregnancy and expert opinion to address the key risks at different stages during pregnancy, at delivery and postpartum. Although no single strategy for anticoagulation can be recommended, key risks are identified with recommendations to optimise current management options. Prior to cardiac surgery, all individuals of childbearing age should be counselled about the impact of valve replacement choice on future pregnancy risk by an appropriately trained Cardiologist, recognising individual clinical situations. Discussion should include the risks of re-do surgery and involve a cardiac surgeon. Guidelines recommend consideration of bio-prosthetic valves or, where appropriate, a Ross operation in individuals of childbearing age in view of the high risk of complications during pregnancy in individuals with MHV.1, 2 Assisted conception should not proceed without prior involvement of the specialist team who can recommend on the suitability, preparatory investigations and management of assisted conception. Assisted reproductive techniques (ART) involving ovarian stimulation are an additional risk factor for thrombosis. For invasive procedures associated with bleeding, for example, egg collection in an individual anticoagulated with a VKA should be bridged with twice-daily therapeutic LMWH for the minimum period of time before restarting the VKA. There is insufficient evidence to support admission for UFH.3 The last dose of therapeutic LMWH should be ≥24 h before the scheduled procedure and immediate therapeutic anticoagulation should be avoided post-procedure as per bridging practices used in international studies.4 The bleeding and thrombotic risks of egg collection in individuals anticoagulated for MHV are unknown and likely to be significant. Any individual with an MHV considering ART should have a consultation with the specialist team and their anticoagulation titrated accordingly. All individuals with MHV and confirmed pregnancy should be referred to the specialist team as a matter of urgency (as soon as a pregnancy test is positive) and ideally reviewed by the MDT before 6 weeks of gestation. Individuals should have been informed how to self-refer but any healthcare professional made aware of the pregnancy can refer as a matter of urgency. The first encounter may be with the anticoagulant clinic whose staff should escalate to the specialist team immediately. In very early pregnancy, complex decisions are required regarding optimal anticoagulation regimens and balancing competing risks. If not already completed pre-pregnancy, a written care plan should be agreed and widely circulated with a copy to the individual. Risks associated with pregnancy in this population overall are outlined in Table 2. Pregnancy in individuals with MHV is considered very high risk. Data from the United Kingdom Obstetric Surveillance System (UKOSS)5 showed that maternal mortality was 9% and the risk of severe morbidity was 41%, with 16% of individuals suffering valve thrombosis and 9% having a cerebrovascular accident. Foetal risks are also high, with an increased risk of miscarriage, stillbirth, foetal haemorrhage, and warfarin-induced teratogenicity. Only 16 (28%) pregnant individuals had a good maternal and foetal outcome compared to the Registry of Pregnancy and Cardiac (ROPAC) disease European registry,6 where individuals with a MHV had a 58% chance of experiencing an uncomplicated pregnancy with a live birth. Caring for pregnant individuals with MHV is logistically challenging involving multiple specialist teams including maternal/foetal medicine, cardiology, haematology and anaesthesia. The UKOSS study estimated the incidence of MHV in pregnancy to be 3.7 (95% CI: 2.7–4.7) per 100 000 maternities, so familiarity in management of pregnant patients outside of specialist centres will be limited. In view of the high risks involved and requirement for input by an experienced multidisciplinary specialist team, we support guidance by other specialist societies and recommend that all pregnant individuals with MHV are managed in a tertiary specialist centre with the relevant expertise.1, 2, 7 All anticoagulation regimens are associated with risks for the mother and foetus and management will be planned on an individual basis through discussion between the pregnant individual and the MDT ideally in a pre-pregnancy setting. Table 2 summarises the maternal and foetal benefits and risks of different anticoagulant regimens in pregnant individuals with MHV. Warfarin is the superior anticoagulant in terms of prevention of MVT but is associated with a higher risk to the foetus. The use of LMWH anticoagulation regimens avert the foetal risks observed with VKAs as they do not traverse the placenta, but their use is associated with poorer maternal outcomes. The most commonly used VKA in the UK is warfarin and most of the published data are for this coumarin; however, the same principles apply to other VKAs. Warfarin is an established teratogen. Foetal warfarin syndrome (FWS) comprises of nasal bone hypoplasia and skeletal abnormalities and occurs following warfarin exposure between 6 and 12 weeks of gestation. The risk of FWS varies, but recent studies would suggest that it affects between 6% and 12% of foetuses exposed. Some studies suggest a dose–response relationship with lower levels of FWS at warfarin dosages ≤5 mg/day9, 10, 14; however, the data are far from conclusive.15 Running the INR target below the recommended range to keep the warfarin dose <5 mg is not recommended due to the increased risk of MVT with this approach and the limitations of the studies on which the evidence regarding <5 mg warfarin is based. Warfarin readily crosses the placenta and the foetus is more anticoagulated than the mother, attributed to the immature foetal liver enzymes with low levels of vitamin K-dependent clotting factors and relative absence of foetal vitamin K. Beyond the first trimester of pregnancy, VKA use is associated with foetal, placental and neonatal haemorrhage.16-18 Data on risk are summarised in Table 2. It should be noted that published rates of adverse events are susceptible to reporting bias. When restricted to prospective studies only, Xu et al.10 found similar foetal loss rates for VKA, LMWH/VKA and LMWH regimens (18.44% vs. 18.36% and 16.98%). However, the prospective, observational ROPAC registry6 showed that the use of VKA during pregnancy resulted in fewer live births, with a higher rate of miscarriage (28.6% vs. 9.2% in individuals receiving heparin; p < 0.001) and late foetal death (7.1% vs. 0.7%; p = 0.016). In this study, the reported rate of miscarriage and foetal loss were not significantly different in high- versus low-dose VKA (≤5 mg/day warfarin or ≤2 mg/day acenocoumarol or ≤3 mg/day phenprocoumon). It is currently considered likely that the neurodevelopmental effects of VKA are secondary to bleeding complications such as intracerebral haemorrhage. In a study of 274 school-age children with in utero exposure to coumarins,19 the vast majority showed no clinically significant difference in growth and development compared with controls; however, 18 children (7%) of the coumarin-exposed cohort and 2 children (<1%) in the control cohort had two or more adverse outcome-measures, although the authors acknowledge potential confounding by indication, because the mothers of the exposed children had a medical indication for anticoagulation. Cohort studies indicate that the use of a VKA throughout pregnancy is the best option for preventing MVT in pregnancy, although some international guidelines nuance the advice for the first trimester based on the individual's usual daily warfarin dose.7 The choice of regimen ultimately requires careful counselling and an individual's preference and likely compliance will be a key factor. However, it is reasonable to weight recommendations from the specialist team towards warfarin for pregnant individuals at higher risk of MVT (Table 1) and any decision can be reviewed if circumstances change during the pregnancy. The European Society of Cardiology (ESC) guidance7 recommends using the same INR range as outside of pregnancy with INR monitoring weekly or every 2 weeks with self-monitoring of INR in suitable patients. It is particularly important to try and avoid high INRs in view of the enhanced anticoagulation in the foetus. In the prospective multicentre study by van Hagen et al.,6 half of the MVT occurred in pregnant individuals who were undergoing transition from a VKA to a LMWH in the first trimester with three events in second trimester and two in the third trimester and none described with onset postpartum. In a review of maternal and foetal complications in 92 individuals from 5 prospective cohort studies treated with dose-adjusted LMWH throughout pregnancy,20 nine episodes of valve thrombosis were reported and were attributed to poor compliance or suboptimal LMWH doses in the majority of the cases. The majority of events occurred antenatally rather than postpartum. In a prior case series by the same author,21 five cases of MVT were associated with enoxaparin therapy; three occurred in the first or second trimester and were all associated with inadequate anticoagulation and compliance issues and both cases of postpartum thrombosis were associated with subtherapeutic anticoagulation in late pregnancy (both received UFH peri-delivery). The UKOSS survey did not give detail on the timing of onset of MVT in all cases but two of the five deaths occurred in the first half of pregnancy.5 Transition from VKA to heparin in the first trimester therefore requires particular attention as this represents a period of very high risk for MVT. As soon as a positive pregnancy test is confirmed, the individual should stop VKA and commence twice-daily LMWH injections. The INR need not be in the normal range when commencing LMWH. Many guidelines focus on the peripartum management of anticoagulation; however, evidence suggests that the risk of MVT is predominantly antenatal and poor adherence with anticoagulation is a key theme. Although there is no firm evidence that monitoring LMWH with anti-Xa levels in pregnancy improves maternal outcomes,22 there is evidence that standard fixed doses of LMWH are associated with a higher risk of MVT in pregnancy.23, 24 Fatal thrombotic events in two pregnant individuals in the HiP-CAT study occurred with low peak anti-Xa levels.15 In a small study of 11 pregnant patients25 with a starting dose of 1 mg/kg BD enoxaparin and subsequent monitoring of LMWH to achieve a peak enoxaparin anti-Xa level of 1.0–1.2 IU/mL, a mean increase in LMWH dose of 54% was required. In another retrospective study,26 an enoxaparin dose of 1.3 mg/kg BD was required to achieve a peak enoxaparin anti-Xa level of 1.0–1.2 IU/mL. Based on data from the UKOSS study, the authors suggest starting doses of LMWH of 2.5 mg/kg/day for enoxaparin, 250 IU/kg/day for dalteparin and 250 IU/kg/day for tinzaparin to ensure minimal delay in reaching a reasonable level of anticoagulation.5 It was also noted that most pregnant individuals required dose escalation between 10 and 20 weeks of gestation. Although LMWH products differ in terms of ratio of anti-Xa to antithrombin activity due to variation in processing and average molecular weight,27 there is currently insufficient data to recommend a specific product for use in pregnant individuals with MHV. Trough LMWH testing has been recommended by some groups.28, 29 A meta-analysis of six studies of mainly prophylactic LMWH dosing suggested a benefit to trough monitoring versus no monitoring but not for peak monitoring30; however, this was dominated by a study checking anti-Xa levels 12 h post a standard prophylactic dose of dalteparin in high-risk trauma patients.31 Goland et al. defined subtherapeutic anticoagulation with LMWH in individuals with MHV in pregnancy as a trough level of <0.6 IU/mL; however, this definition was accepted as arbitrary.32 They found that even with peak anti-Xa levels >1.2 IU/mL on a BD dosing regimen, 31% trough levels were <0.6 IU/mL. Peak and trough levels showed some correlation but with variability. In the UKOSS study,5 the incidence of maternal complications was similar in individuals with post-dose only LMWH monitoring versus pre- and post-dose. The American College of Cardiology and American Heart Association (ACC/AHA) guidelines recommend target LMWH levels by anti-Xa assay of 0.8–1.2 IU/mL at 4–6 h after dose.1 They also state that measurement of trough levels to maintain a trough level >0.6 IU/mL may help pregnant individuals to maintain therapeutic anticoagulation while on LMWH. The ESC and European Association for Cardio-Thoracic Surgery guidelines recommend switching to LMWH during the first trimester with strict monitoring (peak therapeutic range: 0.8–1.2 IU/mL for aortic valve prosthesis; and 1.0–1.2 IU/mL for mitral and right-sided valve prosthesis).2 In summary, there are sufficient data to suggest that standard therapeutic doses of LMWH are inadequate and that monitoring is prudent. However, there are insufficient data on trough versus peak monitoring to make firm recommendations. We consider that a trough target of 0.6 IU/mL is arbitrary. An additional consideration is that it is more time-consuming and inconvenient for pregnant individuals to have both trough and peak levels monitored and the use of more frequent LMWH dosing than BD is highly invasive and not adequately studied in terms of risk/benefit. Moreover, one of the most important causes of MVT is poor adherence with anticoagulation and the physical burden and injection site toxicity of such a regimen could easily be counterproductive. In the absence of further supportive data, we consider a peak target of 1.0–1.4 IU/mL at 3–4 h post-BD dosing to be a reasonable target in the absence of bleeding complications and gives a better opportunity for a reasonable trough with a BD dosing regimen. Peak testing was preferred by the majority of haematology/thrombosis specialists in an ISTH SSC survey, although this survey did not have many cardiology respondents due to questionnaire distribution.33 Routine use of trough LMWH levels cannot be recommended currently. Trough levels should not currently replace peak level monitoring. We consider that trough monitoring to achieve >0.6 IU/mL LMWH levels along with more than twice-daily LMWH dosing regimens should be evaluated as a matter of urgency to assess whether they offer superior protection against MVT to pregnant individuals without excessive bleeding. Monitoring requires regular access to outpatient setting with timed blood sampling and excellent communication of results and impact on dose with plans for next levels. There is no clear evidence to guide the frequency of LMWH monitoring. However, in view of the risks of valve thrombosis in the initial transition to LMWH, we recommend at least weekly monitoring until the target level is achieved or when there is a below target at any stage and then regular monitoring thereafter (e.g. every 2–4 weeks depending on stability). Historical studies outside of pregnancy show that the addition of LDA to VKAs in patients with MHV reduces the risk of MVT but at a cost of a higher risk of major bleeding.34 Aspirin 75–100 mg in addition to anticoagulation has been recommended routinely in individuals with MHV in pregnancy.14, 35, 36 In the ROPAC registry, no pregnant patients on aspirin developed MVT but more haemorrhagic events were described; however, only 6.1% of patients were given aspirin so no conclusions can be drawn.6 The ACC/AHA guidelines1 state that LDA is regarded as safe during pregnancy and can be used in individuals with MHV if needed for other indications, for example, prevention of pre-eclampsia toxaemia (PET) rather than for thromboprophylaxis per se. Although there is an absence of high-quality data on the role of aspirin in pregnant individuals with MHV, in view of the increased risk of valve thrombosis in pregnancy, it is reasonable to add LDA (75 mg daily) from early pregnancy onwards if there are no contraindications or bleeding concerns, especially in pregnant individuals with a higher risk MHV (Table 1) and this should be continued for the duration of pregnancy. In pregnant individuals with risk factors for placental failure/PET who are eligible for LDA at 12 weeks of gestation, there is indirect evidence that aspirin at a dose greater than 75 mg may be associated with the highest reduction in preterm pre-eclampsia.37 In the absence of clear evidence of increased bleeding when increasing the aspirin dose from 75 to 150 mg daily in combination with therapeutic anticoagulation, we consider that an increase in the dose of aspirin from 75 to 150 mg at 12-week gestation can be considered if indicated for prevention of pre-eclampsia in individuals with MHV. Non-surgical management of first trimester miscarriage, that is expectant management (waiting for spontaneous miscarriage) or medical management with prostaglandin has become increasingly popular in comparison to surgical management. For anticoagulated pregnant individuals, the reduced success rates in achieving complete miscarriage, higher need for unplanned surgical evacuation and increased bleeding with these management options make them less suitable in this context.38 Likewise, for termination of pregnancy, the benefits of surgical management in minimising blood loss and unplanned surgical intervention make it the recommended option. The gestation up to which this can be carried out will depend on the expertise within a unit. For later gestations, medical management will be required. For both miscarriage treatment and termination of pregnancy, the duration of interrupted anticoagulation should be minimised whilst avoiding an excessive risk of bleeding which could prolong that interruption. A risk assessment by the specialist team is required and discussion regarding management of anticoagulation should be carried out. Contraception should be discussed and administered if appropriate (in the case of an intrauterine contraceptive device or progesterone implant). This requires a low index of suspicion in view of the heightened risk in pregnancy. Signs and symptoms of potential valve thrombosis include no longer hearing the clicks from the valve closure, obstructive symptoms such as increased breathlessness, heart failure, syncope or pre-syncope, cardiogenic shock ± non-obstructive symptoms such as an embolic event, for example, stroke, renal or splenic infarct (manifesting as new onset abdominal pain). Right-sided mechanical valve replacement thrombosis obstructive symptoms may be more insidious with loss of appetite, lower limb oedema and/or ascites ± non-obstructive symptoms such as pulmonary emboli as embolic events. Examination may reveal increased heart rate, low blood pressure, quiet or absent mechanical valve sounds, a new murmur, increased respiratory rate and inspiratory crepitations. A high index of suspicion for the potential of MVT with embolism is required with urgent senior cardiology review, urgent echo and further imaging (e.g. fluoroscopy, transoesophageal echo, +/or CT) as deemed appropriate. In cases where MVT presents with a stroke, emergency stroke team review and assessment is essential. Senior MDT discussions including obstetrics, cardiology, anaesthetics, cardiac surgery and neurology/stroke medicine are recommended and location of care needs to balance the competing needs of the pregnant individual, considering MVT management, stroke and ongoing pregnancy needs and should be individualised. Initial assessment is based on clinical stability of patient. If haemodynamically unstable, then this is a clinical emergency requiring MDT input from senior clinicians including obstetrician, cardiologist with interest in obstetrics, anaesthetist and cardiothoracic surgeon to consider management options. Ensure correct location for care and clear escalation plan, usually critical care unit/cardiac intensive care unit with obstetric input.7 Opinion generally supports a trial of thrombolysis, especially for right-sided lesions but there are no comparative data of strategies.7 Obstetric management will be based on the foetal condition and gestation in conjunction with the maternal condition. A molecular weight > 1000 Da prevents most thrombolytic agents from easily crossing the placenta. Alteplase (a recombinant tissue plasminogen activator) has the highest molecular weight and does not cross the placenta. Successful thrombolysis is defined as achieving at least two out of three of the following39: 1. Resolution of elevated Doppler gradient 2. Reduction in thrombus size (area or length) of >50% 3. Improvement in symptoms If unsuccessful fibrinolysis by these criteria, consider repeating thrombolysis if haemodynamically stable. Escalation of anticoagulant management should be considered in pregnant individuals with MVT, for example switch from LMWH to IV UFH and then to a VKA with aspirin 75 mg daily when no further immediate interventions required. If the MDT opinion is that cardiac surgery is required either as the primary treatment modality or because of ongoing haemodynamic instability despite medical treatment, then an experienced MDT team should consider the optimal foetal management. If gestation supports emergent delivery at the start of the cardiac surgical procedure, then this should be performed with exquisite attention to haemostasis at the obstetric site. A meta-analysis in 2018 of cardiac surgery during pregnancy for all indications, that is not only MVT, suggested a maternal mortality rate of 11.2%, maternal complication rate of 8.1%, pregnancy loss rate of 33.1% and neonatal complications in 10.8%.40 There are considerations in the cardiac surgical bypass procedure that may reduce the risks to the foetus including minimising hypothermia, maintaining mean arterial pressure > 70 mmHg, avoiding maternal hypoglycaemia, maintaining maternal haematocrit and maintaining a left lateral tilted position to minimise compression of the great vessels. Birth represents a particularly high-risk time to pregnant individuals with MHV, due to the need to balance the risk of haemorrhage with the risk of MVT because of a prolonged period of reduced anticoagulation. The MDT should document a delivery plan that should include the anticoagulation regimen, analgesic options during labour, anaesthetic options for an operative intervention (e.g. caesarean birth), haemodynamic monitoring, the uterotonic agents to be utilised and the recommended postpartum anticoagulation regimen following childbirth. A high proportion of pregnant individuals will deliver by caesarean section (CS), but vaginal birth has advantages for selected individuals as it avoids the risk of surgical bleeding, in particular wound haematoma and is recommended by ESC guidelines.7 The UKOSS surveillance data5 reported a 53% rate of CS, only one of which was an emergency procedure and a 45% rate of vaginal birth. In one retrospective case series comparison between two large centres in the UK and the Netherlands,41 the relative increased incidence of primary/secondary PPH and wound haematoma in one centre appeared to correlate with a higher rate of CS (Table 3). Scheduled delivery has been recommended for individuals on therapeutic LMWH in the context of venous thromboembolic (VTE) disease42; citations include a single-centre retrospective observational study of pregnant individuals receiving therapeutic-dose LMWH for the management of VTE; individuals with spontaneous onset of labour had a 1.9-fold (95% CI: 0.6–5.8) increase in the risk of PPH (≥500 mL) compared with planned induction of labour.43 There is an absence of data to inform the best options for anticoagulation around the time of birth. Any data available are derived from registries/observational retrospective data. For individuals receiving VKAs, a transition to therapeutic LMWH or UFH is recommended by 36 weeks (or 2 weeks before the planned birth). If presenting in labour on VKA, the INR should be measured and corrected to decrease maternal haemorrhage risk with a four-factor prothrombin complex concentrate (PCC) at a dose of 25–50 u/kg in addition to vitamin K 10 mg IV. This is preferable to administering 15 mL/kg of fresh frozen plasma, which is only partially effective at improving the INR, time-consuming and requires higher fluid volume.44 A CS should be considered in view of the foetal bleeding risks. For individuals with MHV, the risk of prolonged interruption of LMWH during the labour induction process is a potential risk for MVT. It is possible that this risk is reduced by bridging with UFH. It should be recognised that there are no studies examining the competing risks of bleeding versus valve thrombosis to inform recommendations on the mode of delivery in individuals with MHV. It is recommended that, in individuals receiving therapeutic LMWH, the last dose should be ≥24 h prior to the planned induction. Consideration can be given to further doses, including prophylactic and intermediate doses, and this should be discussed by the MDT as it may impact choices for labour analgesia. However, there are no data on which to make clear recommendations. Alternatively, a switch to therapeutic IV UFH at least 36 h prior to scheduled induction can be considered, especially in individuals where induction may be prolonged. The UFH infusion needs to be discontinued 4–6 h prior to delivery, which may be difficult to predict. Practically, the infusion is stopped when the patient is in early labour. A meta-analysis in non-pregnant patients did not show superiority of UFH over LMWH for bridging patients with MHV and familiarity of doctors with UFH is poor and the risks of over and under anticoagulation are significant so LMWH may in practice offer less overall risk.3 Many guidelines recommend a specific activated partial thromboplastin time ratio (APTTR) target range for pregnant individuals with MHV; however, the therapeutic APTTR range for UFH should be determined by the local laboratory as the appropriate therapeutic target depends on the reagents used. In addition, monitoring UFH by APTTR is problematic as the heparin dose to achieve the equivalent APTTR in pregnancy may differ from outside of pregnancy, usually requiring higher doses to achieve the same target APTTR with a risk of overdose and bleeding. UFH can be monitored using an appropriately calibrated anti-Xa assay45 and we recommended this in preference to sole reliance on the APTTR; however, we acknowledge that this may not be available in a timely fashion in all laboratories. Protamine is a reversal agent for UFH and has been used occasionally in pregnant individuals with MHVs on IV UFH who present with life-threatening haemorrhage. Its role in reversing LMWH however is limited and is not recommended routinely. Potential risks of protamine include anaphylaxis, pulmonary oedema, bronchoconstriction, bradycardia and increased bleeding and thrombosis risk. In pregnant individuals, who are at least partially anticoagulated at the time of delivery (<24 h), we recommend a low threshold for use of tranexamic acid 1 g IV post-delivery. Otherwise, it can be used as for normal PPH criteria. Evidence is lacking regarding any increased risk of MVT with the use of tranexamic acid46 but a single dose is unlikely to be associated with a significant risk when considering the known very high risk of bleeding (Table 3) which could lead to prolonged period of pausing anticoagulation. However, the use of repeated doses of tranexamic beyond the two doses used in the WOMAN trial47 is probably best avoided. If using aspirin, it is not mandated to stop however, in view of the high risk of PPH in this group of individuals, we recommend consideration of stopping for at least 3 days prior to scheduled delivery with anticipated correction of any associated bleeding risk.48, 49 There are no high-quality studies assessing the bleeding risk versus the benefits of early therapeutic anticoagulation post-delivery in individuals with MHV; however, there are extensive data on these risks in the postoperative setting. Therapeutic anticoagulation with heparins has been associated with a fivefold increase bleeding without reducing the thrombosis risk.50 Specifically, in patients with MHV, Passaglia et al.51 concluded that ‘Early bridging therapy with LMWH appears to be associated with consistently high bleeding rates across multiple analyses’. In the PERIOP2 study52 including patients with mechanical valve replacement ± AF, all received pre-procedure therapeutic LMWH bridging but were randomised to post-procedure escalation to therapeutic LMWH after 48 h versus no post-op therapeutic LMWH. There was no difference in bleeding; however, patients considered at higher risk of bleeding only received prophylactic LMWH post-procedure. No significant benefit was found for postoperative dalteparin bridging to prevent thromboembolism. Although pregnancy is associated with a pro-thrombotic state and significant risk of valve thrombosis, the risk of thrombosis after pausing anticoagulation over a brief period is likely to be low; however, the risk of bleeding is high in the peripartum period and is likely to be exacerbated using very early po
Search from 250 M+ research papersSearch from 250 M+ research papers
Apr 1, 2024
Ellen Kupka + 5
Open Access