全文

Abstract

In the United Kingdom, the national guidance from Royal College of Obstetricians and Gynaecologists (RCOG) and National Institute for Health and Clinical Excellence (NICE) encourages the use of low molecular weight heparin thromboprophylaxis in high risk pregnancies. The recommendation, however, is based largely on expert opinion with almost no evidence from randomised controlled trials or meta-analyses. Here we examine the evidence for and against use of thromboprophylaxis and suggest that careful consideration is needed in implementing change in practice with follow-up of complications due to a real risk of unintended consequences. Therefore, large-scale and well-designed studies are urgently needed. We find that health economic assessments, which should be central to any major health policy change, appear entirely absent in this context.

Keywords: Obstetrics, Thromboprophylaxis, National policy.

Introduction

Venous thromboembolism (VTE) is the third commonest cause of direct maternal death in the United Kingdom (UK) and Europe [1] and [2]. Recent guidance from Royal College of Obstetricians and Gynaecologists (RCOG) and National Institute for Health and Clinical Excellence (NICE) encourages the use of low molecular weight heparin (LMWH) thromboprophylaxis in high risk pregnancies [3] and [4]. The recommendation, however, is largely based on expert opinion and pays insufficient attention to the potential haematological, neurological and medical side effects, nor has a formal health economic assessment been carried out. Here we examine the evidence for and against thromboprophylaxis guidelines that are now applied to approximately 700,000 women per year in the UK [5].

 

Incidence of VTE in pregnancy and puerperium

The Eighth Report on Confidential Enquiries into Maternal Deaths in the UK (2006–2008) estimated 0.79 VTE deaths/100,000 maternities; 16 from pulmonary embolism (PE) and 2 from cerebral vein thrombosis [1]. The UK's Obstetric Surveillance System (UKOSS) estimates an incidence of 1.3 PE/10,000 maternities, with 143 cases reported from February 2005 to August 2006 [6].

During pregnancy, the risk of VTE is 5–10 fold higher than that outside pregnancy [7]; its incidence in pregnancy and the puerperium is 1–2/1000 [7] and [8]. There is no consensus as to whether VTE is more likely antenatally or postpartum [7] and [8].

An RCOG ‘green-top’ guideline estimated 700–1400 pregnancy-related VTE episodes/year nationally, in addition to those related to miscarriage and termination of pregnancy. This translates into an overall case fatality for VTE in pregnancy of approximately 1% [3].

 

Evidence for thromboprophylaxis

The evidence underlying the importance of risk factors of VTE in pregnancy is drawn mainly from case control or population studies and derived risk assessments and decision models [9].

Thromboprophylaxis was deemed to be safe in a prospective cohort study with risk stratification which recruited 810 pregnant women [10]. The medium and high risk groups received both antenatal and postnatal dalteparin at routine and higher doses respectively, whilst the low risk group received postnatal dalteparin only. A clinically relevant bleed was found in 4.6% with 1.1% thought to be dalteparin-related: 0.1% developed osteoporosis and 2.2% had thrombocytopenia but without features of heparin-induced thrombocytopenia (HIT). In another prospective study of 286 high risk patients with a history of VTE and/or thrombophilia, serious bleeding occurred in 0.35% of cases but no instances of HIT or osteoporosis were seen [11]. All patients received postnatal LMWH prophylaxis for at least 6 weeks while 61.8% were given antenatal prophylaxis: 0.35% suffered a DVT antepartum and 0.7% postpartum.

A decision model was constructed to evaluate the risks and benefits associated with a seven-day regimen of prophylactic LMWH following caesarean section (CS), compared with no treatment [9]. LMWH prophylaxis led to the highest quality-adjusted life expectancy, with a net gain of 1.5–2.8 days. The model showed the incidence of VTE after CS to be 0.15–0.22% with the haemorrhagic risk due to LMWH 0.23–0.35%. This compares favorably with 0.5% incidence of VTE following caesarean delivery without LMWH prophylaxis.

 

Evidence against routine thromboprophylaxis

The 2010 Cochrane review on prophylaxis for VTE in pregnancy and puerperium examined 13 randomised trials involving 1774 women [12]. There was insufficient evidence on which to base recommendations for thromboprophylaxis during pregnancy and the early postnatal period. For antenatal prophylaxis, LMWH was associated with fewer bleeding episodes compared with unfractionated heparin (UFH). New to the 2010 version, there was some evidence of side effects associated with heparin (UFH or LMWH) thromboprophylaxis in women undergoing CS (n = 796): with more bleeding or bruising episodes (risk ratio 5.15) [12]. Current RCOG and NICE guidelines reference the 2002 version of the Cochrane review which included only 8 trials involving 649 women [13].

A systematic review including 2777 pregnancies showed that despite LMWH use, VTE still occurred in 0.86% of pregnancies. Bleeding, generally associated with primary obstetric causes, occurred in 1.98%, allergic skin reactions in 1.80%, HIT in 0%, thrombocytopenia in 0.11%, and osteoporotic fractures in 0.04% of pregnancies [14].

 

Anaesthetic implications of anticoagulation in pregnancy

The current RCOG and NICE guidelines for prevention of VTE in pregnancy and the puerperium have the potential to impact on the use of regional analgesia and anaesthesia in pregnant patients [3] and [4]. Concurrent use of anticoagulation with central neuraxial blocks (CNB) increases the risk of vertebral canal haematoma (both spinal and epidural haematoma) [15]. Vertebral canal haematoma, defined as symptomatic bleeding within the spinal neuraxis, is a rare and potentially catastrophic complication of spinal or epidural anaesthesia. It requires neurosurgical intervention (ideally within 8 h) to achieve the best chance of complete neurological recovery.

In the UK, it is recommended that 10–12 h elapse between administering LMWH and performing CNB [16]. This equates to twice the elimination half-life of LMWH and it is therefore considered safe to perform CNB after this time [17]. For this reason it is usual to prescribe pre-operative LMWH for the evening before surgery so that CNB may occur safely on the day of elective surgery. The nature of obstetric interventions, however, means that they are unpredictable and the recommended 10–12 h may not have elapsed before CNB is required for labour analgesia or CS [18]. Women with a raised body mass index (>30 kg/m2), in association with two other risk factors, should be anticoagulated antenatally [3]. This means that the very parturients in whom an anaesthetist is likely to recommend early regional analgesia for labour or regional anaesthesia for CS may be prevented from this and require general anaesthesia for urgent CS, with the associated risks of failed intubation and/or pulmonary aspiration. Additionally, there is the risk of an increased incidence of spinal/epidural haematoma if CNB is inadvertently performed less than 10–12 h since the administration of LMWH [16].

The incidence of neurologic dysfunction resulting from haemorrhagic complications associated with CNB is unknown. As 13 cases of spinal haematoma were identified after 850,000 epidural anaesthetics and 7 cases among 650,000 spinal techniques, the calculated incidence is less than 1 in 150,000 epidural and less than 1 in 220,000 spinal anesthetics [19]. Recent case series and epidemiologic surveys suggest the risk of vertebral canal haematoma is 1 in 168,000 to 1 in 200,000 for CNB due to obstetric indications [20]. The risk is substantially higher with epidural than spinal anaesthesia. Bleeding complications occur more rarely after epidural anaesthesia in obstetrics than in female orthopaedic patients (1 in 3600) [20], so major complications of CNB are probably less frequent in the obstetric population than the general population. These risks, however, were calculated before the new guidelines for prophylaxis had been introduced.

Additional caution is required in groups at high risk of bleeding with LMWH. Renal impairment, concomitant administration of an antiplatelet agent or other anticoagulant may also increase the risk of haematoma [16] and [17]. This may occur during pregnancy since the NICE guidelines for the treatment of hypertension in pregnancy encourage the use of aspirin daily from 12 weeks until delivery in women at high risk of pre-eclampsia [21].

 

Other side effects of LMWH

The reported rate of allergic skin reactions to LMWH varies between 0.6% and 2% [14]. Enoxaparin has been associated with local reactions, such as erythema and ecchymosis [22]. Up to 6% of patients in trials receiving enoxaparin for other indications had reversible elevations in ALT or AST levels to over three times the normal level. This has also been reported with UFH and other LMWH [22].

A 1.3–1.98% risk of bleeding due to LMWH use, such as antenatal bleeding, postpartum haemorrhage >500 mL and wound haematoma, has been reported in meta-analysis, prospective and retrospective studies [14] and [23]. Maternal thrombocytopenia arising from enoxaparin use in pregnancy has been shown to occur in 1.6% out of 624 pregnancies in a retrospective case-note review [23]. The low rate of HIT (0.08%) [14] support the United States (US) recommendation that monitoring of platelet counts is unnecessary in pregnant women treated exclusively with LMWH as HIT is often induced by previous exposure to UFH.

Although osteoporosis is associated with long-term heparin use, the incidence of osteoporosis is thought to be lower with LMWH than UFH. One case of osteoporosis occurred out of 486 pregnancies treated with LMWH [24]. In a prospective observational study of 123 pregnant women with antiphospholipid syndrome, there was no difference in bone mineral density between UFH and enoxaparin treatment [25].

Other points to consider are that LMWHs are costly and the subcutaneous route of administration makes long-term treatment painful and inconvenient for the patient. According to pricing in the British National Formulary, the cost of a 1-week course of enoxaparin, tinzaparin or dalteparin prophylaxis is £28.28, £24.92 or £19.71 (35.51, 31.29 or 24.75 Euros), respectively for a woman weighing between 50 and 90 kg.

An Irish study reviewed 100 consecutive deliveries in 2010 and found that 51% would have been deemed intermediate or high VTE risk at some point in pregnancy and required LMWH treatment if the RCOG guideline were adopted. They calculated the cost would have risen by 17-fold to €2973.60 compared with the actual expenditure of €173.46 [26]. This cost estimation does not include that of staff education, time taken for risk stratification and patient teaching of self-administration, needle care and disposal. A separate English series of 97 deliveries suggested risk scores necessitating antenatal or postnatal LMWH treatment were present in 2.1% or 40% of the cohort [27]. They also calculated the annual cost of TEDS, medication and equipment to be £44,847 per 1000 deliveries, and £2.6 million for each life saved. In this study, 10% of normal-weight postnatal women who achieved a vaginal birth had a risk score requiring thromboprophylaxis of at least 1 week.

In terms of compliance to local thromboprophylaxis guidelines, a Scottish audit showed that 31% (4 out of 13) spontaneous vaginal births and 22% (4 out of 18) of assisted instrumental deliveries in which thromboprophylaxis was indicated actually received the required treatment [28]. In a prospective study from Ireland, there was 100% adherence to the recommendation of thromboprophylaxis following a caesarean section (n = 60) [29], but only 6% of these women received the optimum dose of LMWH. A more recent study indicated compliance with the hospital guidelines in 60% (3 out of 5) antenatal versus 68% (166 out of 244) postnatal patients [30].

 

International perspective

Thromboprophylaxis recommendations in other countries have incorporated most or parts of the RCOG guideline (Table 1 and Table 2). Within the UK, NICE and RCOG guidelines differ. For obstetric inpatients, only one additional risk factor is required before considering anticoagulation according to NICE [4], whereas the RCOG suggests two or more risk factors [3]. The main differences between international and the RCOG guidelines are outlined below.

In the US, for women with thrombophilia (other than homozygous factor V Leiden and prothrombin 20210A), no prophylaxis is recommended. In those with homozygous factor V Leiden and prothrombin 20210A, antepartum vigilance and 6-week postpartum anticoagulaition are recommended. Patients with severe ovarian hyperstimulation syndrome (OHSS) are advised to have thromboprophylaxis for 3 months after OHSS resolution.

The Australia and New Zealand Working Party guideline in 2007 suggests that women undergoing CS should receive thromboprophylaxis provided there are no contraindications for anticoagulant therapy [31]. A primary postpartum haemorrhage (PPH) of >1000 mL is considered a contraindication for anticoagulation in a separate guideline from the Australian National Health and Medical Research Council in 2009. This contradicts the RCOG guideline, in which PPH >1000 mL is a risk factor for VTE and prophylaxis is advocated. The recommendations endorsed by Councils of the Society of Obstetric Medicine of Australia and New Zealand and the Australasian Society of Thrombosis and Haemostasis are comparable to those of the RCOG.

The French recommendations are similar to those from RCOG, except patients with >3 risk factors would not be considered for antenatal thromboprophylaxis, but for postnatal prophylaxis only. In Spain, the postnatal intermediate risk group is advised to have 3–5 days of anticoagulation rather than 7 days as in the UK. The Italian guidelines recommend antenatal and postnatal prophylaxis in those with previous VTE regardless of thrombophilia or family history status except for those related to temporary factors [32]. In all women with thrombophilia, they recommend postnatal prophylaxis for 6 weeks.

The Swedish guideline uses a weighted risk score based on major factors associated with a 5-fold increased risk or its multiples [33]. For example, women with risk score 1 are at 5-fold increased risk, risk score 2 (2 variables with 1 point or 1 variable with 2 points) are at a 25-fold increased risk, and risk score 3 have a 125-fold increased risk, and so on. The score also determines the timing and duration of anticoagulation.

The German guideline and European international consensus statement are both similar to those from RCOG.

 

Conclusion

As VTE in pregnancy remains a major but potentially preventable cause of maternal death and morbidity, it is entirely understandable and appropriate that VTE prevention should be a focus of positive national policy. RCOG and NICE guidelines are largely based on case-control studies and expert opinions with limited or no evidence from randomised controlled trials or meta-analyses. It is worth noting that the RCOG guideline highlighted the low grade of evidence for many of its recommendations. There is also a statement recognising the need for individualised treatment in some women, particularly in discussion with the woman concerned and with input from a local expert. Though there has been a decline in the number of deaths for VTE 2006–2008 coinciding with the 2004 publication of RCOG guidelines on LMWH prophylaxis, the reason for this remains unknown: whether it represents cyclical variation; improved diagnosis and management of VTE or prophylaxis itself.

The LMWH prophylaxis policy has been implemented with no apparent health economic or clinical risk benefit analysis. Even notwithstanding this, according to generally accepted principles for introducing a therapeutic intervention, there should be some form of organised national follow-up of complications and to determine unintended consequences of such a policy. To our knowledge, no such surveillance is planned.

Whether the incidence of complications will actually increase in this population with the introduction of more stringent LMWH prophylaxis remains to be seen. Nevertheless, there is a risk that the widespread use of LMWH in the pregnant population will result in rising pharmacotherapy costs and increased demand on resources due to more haematological, neurological and medical complications.

There is an unarguable urgent need for large-scale, well-designed studies with adequate power in order to establish the clinical effectiveness of thromboprophylaxis in pregnancy and puerperium and direct true evidence-based practice. Until the policy for routine thromboprophylaxis is aligned with the basic principles of introducing new treatments to a population, we urge that the policy is reconsidered for all women except those with the highest level of prior risk.

 

References

[1] Centre for Maternal Child Enquiries (CMACE). Saving mothers’ lives: reviewing maternal deaths to make motherhood safer: 2006–2008. The eighth report on confidential enquiries into maternal deaths in the United Kingdom. BJOG: An International Journal of Obstetrics & Gynaecology. 2011;118(Supp1):1-203

[2] K. Wildman, M.H. Bouvier-Colle. Maternal mortality as an indicator of obstetric care in Europe. BJOG: An International Journal of Obstetrics & Gynaecology. 2004;111:164-169

[3] Green-top guideline no. 37: reducing the risk of thrombosis and embolism during pregnancy and the puerperium. (RCOG, London, 2009)

[4] J. Hill, T. Treasure. Reducing the risk of venous thromboembolism in patients admitted to hospital: summary of NICE guidance. British Medical Journal. 2010;340:c95

[5] Live births in England and Wales by characteristics of birth, 2010. Office for National Statistics, November 2011.

[6] M. Knight. Antenatal pulmonary embolism: risk factors, management and outcomes. BJOG: An International Journal of Obstetrics & Gynaecology. 2008;115:453-461

[7] R.B. Gherman, T.M. Goodwin, B. Leung, J.D. Byrne, R. Hethumumi, M. Montoro. Incidence, clinical characteristics, and timing of objectively diagnosed venous thromboembolism during pregnancy. Obstetrics and Gynecology. 1999;94:730-734

[8] A.F. Jacobsen, F.E. Skjeldestad, P.M. Sandset. Incidence and risk patterns of venous thromboembolism in pregnancy and puerperium – a register-based case-control study. American Journal of Obstetrics and Gynecology. 2008;198:233e1-2337e

[9] M. Blondon. Thromboprophylaxis after cesarean section: decision analysis. Thrombosis Research. 2011;127(Suppl 3):S9-S12

[10] R.M. Bauersachs, J. Dudenhausen, A. Faridi, et al. Risk stratification and heparin prophylaxis to prevent venous thromboembolism in pregnant women. Thrombosis and Haemostasis. 2007;98:1237-1245

[11] Y. Dargaud, L. Rugeri, M.C. Vergnes, et al. A risk score for the management of pregnant women with increased risk of venous thromboembolism: a multicentre prospective study. British Journal of Haematology. 2009;145:825-835

[12] R. Tooher, S. Gates, T. Dowswell, L.J. Davis. Prophylaxis for venous thromboembolic disease in pregnancy and the early postnatal period. Cochrane Database of Systematic Reviews. 2010;(5):CD001689

[13] S. Gates, P. Brocklehurst, L.J. Davis. Prophylaxis for venous thromboembolic disease in pregnancy and the early postnatal period. Cochrane Database of Systematic Reviews. 2002;(2):CD001689

[14] I.A. Greer, C. Nelson-Piercy. Low-molecular-weight heparins for thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a systematic review of safety and efficacy. Blood. 2005;106:401-407

[15] T.T. Horlocker, D.J. Wedel, J.C. Rowlingson, et al. Regional anesthesia in the patient receiving antithrombotic or thrombolytic therapy: American Society of Regional Anesthesia and Pain Medicine Evidence-Based Guidelines (Third Edition). Regional Anesthesia and Pain Medicine. 2010;35:64-101

[16] T.T. Horlocker. Regional anaesthesia in the patient receiving antithrombotic and antiplatelet therapy. British Journal of Anaesthesia. 2011;107(Suppl. 1):i96-i106

[17] W. Gogarten, E. Vandermeulen, H. Van Aken, S. Kozek, J.V. Llau, C.M. Samama. Regional anaesthesia and antithrombotic agents: recommendations of the European Society of Anaesthesiology. European Journal of Anaesthesiology. 2010;27:999-1015

[18] A.J. Butwick, B. Carvalho. Neuraxial anesthesia in obstetric patients receiving anticoagulant and antithrombotic drugs. International Journal of Obstetric Anesthesia. 2010;19:193-201

[19] T.T. Horlocker, D.J. Wedel. Anticoagulation and neuraxial block: historical perspective, anesthetic implications, and risk management. Regional Anesthesia and Pain Medicine. 1998;23(6 Suppl 2):129-134

[20] V. Moen, N. Dahlgren, L. Irestedt. Severe neurological complications after central neuraxial blockades in Sweden 1990–1999. Anesthesiology. 2004;101:950-959

[21] National Institute for Health and Clinical Excellence. Hypertension in pregnancy. NICE clinical guideline 107. August 2010.

[22] N.J. Carter, P.L. McCormack, G.L. Plosker. Enoxaparin: a review of its use in ST-segment elevation myocardial infarction. Drugs. 2008;68:691-710

[23] J. Lepercq, J. Conard, A. Borel-Derlon, et al. Venous thromboembolism during pregnancy: a retrospective study of enoxaparin safety in 624 pregnancies. BJOG: An International Journal of Obstetrics & Gynaecology. 2001;108:1134-1140

[24] B.J. Sanson, A.W. Lensing, M.H. Prins, et al. Safety of low-molecular-weight heparin in pregnancy: a systematic review. Thrombosis and Haemostasis. 1999;81:668-672

[25] M. Backos, R. Rai, E. Thomas, M. Murphy, C. Dore, L. Regan. Bone density changes in pregnant women treated with heparin: a prospective, longitudinal study. Human Reproduction. 1999;14:2876-2880

[26] D. Hayers-Ryah, B.M. Bryne. Prevention of thrombosis in pregnancy: how practical are consensus derived clinical practice guidelines?. Journal of Obstetrics and Gynaecology. 2012;32:740-742

[27] C. Bond, K. O’Brien, T. Draycott, et al. Financial implications and maternal impact of national recommendations for thromboprophylaxis: a retrospective cross-sectional analysis. Obstetric Medicine. 2011;6:270-272

[28] C.K. Tan, S.J. Wisdom. Thromboprophylaxis post vaginal delivery: are we forgetting it?. Journal of Obstetrics and Gynaecology. 2006;26:27-29

[29] C. Kennedy, V. O’Dwyer, S. O’Kelly, et al. Thromboprophylaxis for women undergoing caesarean section. Irish Medical Journal. 2012;105:56-57

[30] D. Touqmatchi, C. Cotzias, J. Girling. Venous thromboprophylaxis in pregnancy: the implications of changing to the 2010 RCOG guidelines. Journal of Obstetrics and Gynaecology. 2012;32:743-746

[31] The Australia & New Zealand Working Party on the management and prevention of venous thromboembolism. Best practice guidelines for Australia and New Zealand: prevention of venous thromboembolism. 4th ed.; Health Education & Management Innovations, 2007.

[32] Grandone E, Abbate R, De Stefano D, Faioni EM, Martinelli I. Prevenzione del tromboembolismo venoso associato alla gravidanza. Statement Condiviso tra ie Societa’ Italiana per lo Studio dell’ emostasi e della tromboli (SISET) e la Societa’ Italiana di ginecologia e ostetricia (SIGO).

[33] P.G. Lindqvist, M. Hellgren. Obstetric thromboprophylaxis: the Swedish guidelines. Advances in Hematology. 2011;2011:157483

 

Footnotes

a The Ian Donald Fetal Medicine Unit, Southern General Hospital, Greater Glasgow and Clyde NHS Trust, Glasgow G51 4TF, United Kingdom

b Department of Anaesthesia, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge CB2 0QQ, United Kingdom

c Heart of England NHS Foundation Trust, Birmingham B9 5SS, United Kingdom

d Department of Development and Regeneration, University Hospitals Leuven, Campus Gasthuisberg, Leuven, Belgium, B-3000

e Department of Fetal-Maternal Medicine, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge CB2 0QQ, United Kingdom

  Corresponding author. Tel.: +44 1223 217972; fax: +44 1223 216185.