SeeAppendix I for patient data contribution.
The management of antenatal venous thromboembolism in the UK and Ireland: a prospective multicentre observational survey*
Version of Record online: 3 OCT 2007
British Journal of Haematology
Volume 139, Issue 4, pages 545–558, November 2007
How to Cite
Voke, J., Keidan, J., Pavord, S., Spencer, N. H., Hunt, B. J. and British Society for Haematology Obstetric Haematology Group (2007), The management of antenatal venous thromboembolism in the UK and Ireland: a prospective multicentre observational survey. British Journal of Haematology, 139: 545–558. doi: 10.1111/j.1365-2141.2007.06826.x
- Issue online: 8 OCT 2007
- Version of Record online: 3 OCT 2007
- Received 12 May 2007; accepted for publication 19 June 2007
- antenatal venous thromboembolism;
This prospective observational study reports on 126 women from 25 UK centres with image-proven antenatal venous thromboembolism (VTE), 62% deep vein thrombosis and 38% pulmonary embolism. Thrombophilia screening was of limited benefit except to identify antithrombin deficiency. Sixteen (13%) patients had previous VTE, all but one was related to previous pregnancy or combined oral contraceptive and 12 received no thromboprophylaxis in the index pregnancy, the other four thus received inadequate low molecular weight heparin (LMWH) doses. Treatment was with dalteparin in 25%, enoxaparin in 47%, tinzaparin in 25% and unfractionated heparin alone in 3%. 66% of patients received once-daily LMWH. Anti-activated factor X (anti-Xa) monitoring was performed at 90% of centres, with a wide range of target values. Thus current management of antenatal VTE, despite widely diverse clinical practice, appeared effective and safe, for there were no recurrent events and postpartum haemorrhage was not increased when compared to known rates. Larger studies are required to confirm this. The need for twice as opposed to once daily LMWH and for anti-Xa monitoring is questioned by this study. The importance of clinical risk assessment and adherence to the Royal College of Obstetricians and Gynaecologists guidelines on antenatal thromboprophylaxis, with adequate LMWH dosing is confirmed.
Venous thromboembolism (VTE) remains a leading direct cause of maternal death in the UK and yet data on current practice is poor. The relatively low incidence of antenatal venous thrombosis has so far made randomized research studies impossible to organize and evidence for best practice is lacking, as highlighted in the 2007 guidelines from the Royal College of Obstetricians and Gynaecologists (RCOG, 2007). Multi-centre surveys are therefore an important aid to clinical decisions. To this purpose, we conducted a multicentre, prospective, observational study of antenatal VTE, to assess diversity in practice and determine whether this had a bearing on outcome and complication rate and to collect data on risk factors, timing and presentation of the condition.
This observational study was conducted by the British Society for Haematology Obstetric Haematology Group, following approval from the Multicentre Research Ethics Committee (MREC). Twenty-five centres, from the UK and Ireland, entered data on up to 10 consecutive patients presenting with antenatal VTE.
The questionnaire (seen on http://www.ukivtepsurvey.co.uk) included age and gestation at diagnosis, personal and family history of VTE, previous pregnancy loss, any known or subsequently diagnosed thrombophilia, presenting clinical features, diagnostic methods, treatment and maternal and foetal outcome. Data were entered directly or by post, the latter entered on-line by the study co-ordinator. Patients were registered at presentation and data accrued as the pregnancy progressed. Patient and obstetrician consent forms, patient information sheets, and example letters for Trust Research and Development Departments and Local Research Ethical Committees were provided. Anonymity was ensured by coding, to comply with the Data Protection Act 1998.
Inclusion and exclusion criteria
Patients were excluded if VTE was not objectively confirmed by imaging or if treatment details, delivery and outcome data were not submitted. The initial objective to limit numbers to five consecutive patients from each centre was modified in the second year, as some centres were unable to provide data on five cases. Six to 10 cases were therefore accepted from some of the larger obstetric centres.
Frequency tables were used to look at individual variables and cross-tabulations to explore relationships between variables. Where these investigations yielded patterns worthy of further examination, appropriate statistical hypothesis tests were carried out. Values of P < 0·05 were considered significant.
Consultant Haematologists from 40 centres initially expressed an interest, 32 registered for the survey and 25 returned patient data, with a median of five (range 1–10) patients per centre. Twenty-two (88%) of the 25 centres had local protocols for antenatal VTE management with the remaining three submitting protocols for approval at the time of the study. The study was closed after 2 years, with a total of 130 patients recruited. Four patients were excluded after data entry, two because of lack of positive imaging and two because of lack of outcome data. Before data entry, three requests to enter patients without imaging confirmation of VTE were rejected. Thus 126 patients were analysed. Data were missing in some of the eligible patients because of inability to find the required facts in the notes or patients lost to follow-up.
Age, weight and parity. The median age at diagnosis of VTE was 32 (range 16–42 years). Details of age at diagnosis of VTE plotted against weeks gestation is shown in Fig 1. Data on either early pregnancy weight or body mass index (BMI) were provided for 123/126 (98%) women. The median early pregnancy weight was 68 (range 45–119) kg and BMI 26 (range 19–43). Correlations between the weeks gestation at presentation of VTE and the age, early pregnancy weight and BMI were −0·240, −0·166 and −0·280 respectively. Of these, only the negative correlation between weeks of gestation at time of VTE with age was statistically significant (P < 0·05). Forty of 124 (32%) women were primiparous (no data on two). Fifty-three of 121 (44%) women had suffered previous miscarriages (no data on four) of whom 14 had two to three losses and four had more than three losses. Further details were not requested for this study.
Deep venous thromboses were diagnosed by Doppler ultrasound scan and pulmonary embolism (PE) by ventilation-perfusion (V/Q) or computed tomography (CT) scan (Table I). Image-proven DVT was present in 79 (63%) of 126 patients and PE in 49 (39%). Two patients had both DVT and PE positive imaging: one with clinical features of DVT alone, therefore, analysed in that group, the other with symptoms only of PE therefore included in the PE group. There was a delay in performing imaging in some patients; data were available for 119 patients. Sixty-one (51%) patients were scanned the same day, 28 (24%) the next day, 13 (11%) had a 2 d wait for scanning, 14 (12%) waited 3–4 d, and 3 (2%) had to wait >4 d.
|Positive imaging result||N (% of those with PE)|
|High probability V/Q||17 (35)|
|Intermediate probability V/Q||10 (21)|
|Q scan positive, no V scan||4 (8)|
|CT lung scan||15 (31)|
Time and site of presentation
Figure 2 shows the time of presentation of the VTE events according to trimester. Thirty-eight (49%) of all 78 DVTs, 28 (60%) of 47 proximal DVTs and 14 (67%) of the 21 iliofemoral DVTs occurred in the third trimester. There was a significant increase in the third trimester for iliofemoral DVTs (P = 0·006 for comparison with first trimester and P = 0·015 compared with second trimester) and near significance for other proximal DVTs (P = 0·058 for comparison with first trimester and the same compared with second trimester) but no difference for popliteal or calf DVTs. More PEs occurred in the third trimester than in the first and second but the differences were not significant.
Figure 3 shows the site of DVTs. The more proximal the DVT, the greater the proportion on the left side, 71% overall, including 39 (83%) of all 47 proximal DVTs, (P ≤ 0·001 for comparison with right side), 14 (78%) of 18 third trimester proximal DVTs (P ≤ 0·001), which included 19 (90%) of 21 iliofemoral DVTs (P < 0·001), 12 (86%) of 14 third trimester iliofemoral DVTs (P = 0·006), and 20 (77%) of 26 proximal DVTs excluding iliofemoral (P < 0·001). There was no significant difference between the left and right legs regarding the incidence of the 29 calf and popliteal DVTs.
Data were available on clinical features in 108 (86%) patients (Table II). Clinical features of both DVT and PE were recorded in six patients, four of whom had scan confirmation of PE and two of DVT. Fifty (64%) of 78 women with DVT had pain and swelling in the leg, with discolouration also reported in five. Eight had a swollen leg without pain and six had pain without swelling. A clinical diagnosis of superficial phlebitis had been made in two patients who then had positive scans for DVT. Forty-three (90%) of 48 women with PE had recorded clinical features of PE: pleuritic chest pain with dyspnoea in 25, haemoptysis in two of these patients and 12 women had only one clinical feature of PE (pleuritic pain or dyspnoea). One patient presented with loss of consciousness, vomiting and palpitations. Another patient had image-proven PE but clinical features only of DVT and another with confirmed DVT only had clinical features of PE.
|Scan diagnosis||Clinical features of DVT only||Clinical features of PE only||Clinical features of both DVT and PE||No clinical features recorded||Total|
|Both PE and DVT||1||1||–||–||2|
Risk factors for VTE
Previous VTE was reported in 16 (13%) cases (Tables III and IV). Of these, 15 were either on a combined oral contraceptive pill (COC), pregnant or post-Caesarean section at the time of the first event. Four had a family history of VTE. One patient was not tested for thrombophilia but seven of the remaining 15 (47%) had inherited thrombophilia; five had been confirmed before the second VTE and two were diagnosed after the second event. Twelve of the 16 women (75%) were not prescribed thromboprophylaxis in the index pregnancy prior to their recurrent event.
|Risk factors prior to first VTE/weeks gestation||Timing of second event (weeks gestation)||Family history of VTE||Thrombophilia identified||Antenatal prophylactic LMWH before second event|
|Before second VTE||After second VTE|
|COC and long flight||24||+||Antithrombin deficiency||confirmed||yes|
|COC and leg injury||12||−||NA||−||−|
|32 weeks||22||−||FVL heterozygote||−|
|36 weeks||30||−||NA||FVL heterozygote||−|
|37 weeks||15||+||FVL heterozygote||−|
|Total 16 patients||Mean 25 weeks||Four with family history of VTE||Seven of 15 women tested had a recognized inherited thrombophilia||Four failed prophylaxis|
|Target anti-Xa range recorded||No. patients|
Failed prophylactic low molecular weight heparin (LMWH) after a previous VTE. Four women were receiving LMWH at the time of the second event. All had thrombophilia screening. One had antithrombin deficiency and was receiving enoxaparin 40 mg daily when the second DVT occurred at 24 weeks. Two were heterozygous for factor V Leiden (FVL), one of whom was on enoxaparin 40 mg s.c. daily prior to VTE at 36 weeks, and the other was on dalteparin 2500 IU daily prior to the second VTE at 34 weeks The fourth patient had a negative thrombophilia screen and no family history of VTE but was 38 years old, para 4 + 3, with a history of stillbirth at 39 weeks gestation in her first pregnancy and toxaemia in her third pregnancy, after which she developed the first PE following Caesarean section. The recurrent PE occurred at 38 weeks gestation in her eighth (index) pregnancy while receiving enoxaparin 40 mg daily from 12 weeks gestation. Factor VIII and antiglycoprotein antibody results were not available on these four patients. Fasting homocysteine levels were checked in one patient and were normal.
Previous VTE in pregnancy. Six (37%) of the 16 with previous VTE had the first event in a previous pregnancy and none were receiving thromboprophylaxis in the index pregnancy (Table III). In four of these six patients, the second antenatal VTE occurred earlier in gestation than the first. One had her first PE earlier in the index pregnancy, stopped LMWH against medical advice and entered the study after a second PE. Three of the six patients were heterozygous for FVL.
Previous VTE after Caesarean section. Three women had the first VTE after Caesarean section. One was heterozygous for FVL and was given prophylactic dalteparin 2500 IU s.c. daily in the subsequent pregnancy prior to the second event, The other two women had no thrombophilia identified and received no prophylactic heparin prior to the second VTE.
Previous VTE when taking a COC. Six of the 16 patients were taking a COC at the time of the first event. Of these, two had a family history of VTE and they both had inherited thrombophilia – one had antithrombin deficiency and the other had heterozygous FVL. Both were on prophylactic antenatal LMWH prior to the second VTE. One woman with no family history of VTE, had no thrombophilia screen after her first VTE on COC, but was found to have antithrombin deficiency after the second VTE at 9 weeks gestation. She had not received any antenatal prophylactic LMWH.
Family history and thrombophilia
A family history of VTE was reported in 28 (23%) of 123 patients (Tables III and IV). A previous VTE had occurred in 4 (14%) of these 28 women and in 12 (13%) of the 95 women with no family history of VTE. Data on thrombophilia were available for 105 patients and were positive in 23 (22%), 22 of these being inherited abnormalities. The timing of thrombophilia screening varied, with most tests performed after postpartum anticoagulation was completed. The majority of centres tested antithrombin, proteins C and S, FVL (with preliminary APC resistance in 12 patients), prothrombin 20210 screening, and antiphospholipid antibody assays – IgG and IgM anticardiolipin antibody (ACA) and lupus anticoagulant (LAC), the latter was available for 83 patients. There were no results for antibeta-2-glycoprotein antibody assays. Fasting homocysteine and FVIII levels were included for nine and 11 patients, respectively, with normal results. Thrombophilia was present in significantly more women with a family history of VTE than in those without (P = 0·009), with abnormalities in 10 (43%) of 23 women with a family history (no results available on five patients) and in 13 (16%) of 79 women without a family history of VTE (no results on 16 patients). In total, 34 of 40 women who had either a personal or family history of VTE were screened, with 13 (38%) having inherited thrombophilia. Four women had both a personal and family history of VTE. One was antithrombin-deficient and three heterozygous for FVL. No women had persistently raised ACA but one patient had persistent LAC confirmed after the pregnancy. Seven (27%) of 26 women tested for thrombophilia before the surveyed event had inherited thrombophilia. Reasons for previous screening were stated in 20 women: previous VTE (seven patients, of whom four were heterozygous for FVL and one, antithrombin deficient), family history of VTE (six women: five heterozygous for FVL and one protein S deficient).
Management of VTE
Admission. Sixteen of 110 patients (13%) were managed entirely as outpatients, 41 (33%) were in hospital for 1–3 d and 23 (18%) were admitted for more than a week.
Attending physician. At the time of diagnosis, 90 (71%) of the patients were seen by an obstetrician, 108 (85%) by a haematologist and 33 (26%) by a general physician, with 96 (76%) being seen by more than one specialist. An anticoagulant nurse practitioner or specialist midwife also saw 55 (43%) of the patients initially. Haematologists were not aware of 10 (8%) of the patients at the start of treatment but were involved later or after delivery.
Pharmacological treatment. Type of heparin: treatment with LMWH or unfractionated heparin (UFH) was given to all from clinical diagnosis: enoxaparin in 59 (47%) patients, dalteparin in 32 (25%) and tinzaparin in 31 (25%) (Table V). UFH was used in 16 (13%) patients. Four (3%), with VTE diagnosis in late gestation, received only intravenous UFH. Seven were transferred to LMWH after initial UFH and in four others, initial LMWH was replaced with UFH late in the pregnancy. One patient had UFH before and after LMWH.
|No. patients treated (%)||Initial twice daily LMWH||Transfer to once daily||Initial once daily LMWH||Transfer to twice daily|
|Intravenous unfractionated heparin* alone||4 (3)|
|Total (%)||126||39 (31)||13 (10)||83 (66)||19 (15)|
Frequency and administration of injections. Initial subcutaneous LMWH doses were given once daily to 83 (66%) and twice daily to 39 patients (31%). Nineteen (15%) transferred from once daily to twice daily LMWH, ten were within 4 weeks of delivery at the time of the change. Thirteen (10%) patients transferred from twice to once daily doses antenatally.
LMWH was self-administered by 100 (82%) of 122 patients, with 11(9%) receiving injections from a relative and 9% from a nurse. Eight women (7%) reported needle phobia. Four were switched to a different LMWH, three because of skin rash and one because of painful subcutaneous swellings at the injection sites.
Dosage of LMWH. Initial doses for each LMWH plotted against early pregnancy weight are shown in Figs 4–6. Doses were unchanged antenatally in 24 (51%) patients, increased from once to twice daily in 16% and decreased in 22%. Reasons given for dose reductions (over 24 h) were clinical improvement several weeks after diagnosis, transfer from twice to once daily injections, high antiactivated factor X (anti-Xa) levels and minor vaginal bleeding. Dose increases were because of transfer from once to twice daily doses.
Monitoring of LMWH. Anti-Xa assays were performed in 90% of centres but results were available for only 73/122 (60%) women (Table VI). Target ranges varied widely. Peak anti-Xa levels 2–5 h postdose were reported in 41 (34%), 5–6 h postdose levels in 2 (2%), 12 h postdose levels in 5 (4%) and 12–24 h postdose levels in 3 (2%) of the 122 women. One centre measured both 2- and 12-h postdose levels. Fifty-two (43%) patients had further anti-Xa assays at varying intervals throughout the pregnancy. Monthly anti-Xa assays were planned for 31 (25%) patients and 7 (9%) were tested weekly or at other intervals, depending on results.
Presenting symptoms and signs improved on treatment in 96% of cases. Three patients were re-scanned for possible recurrent PE or extension of DVT, all with negative results. One patient had a temporary inferior vena cava filter inserted before Caesarean section, following PE at 38 weeks gestation. One patient had pre-eclampsia and one had intra-uterine growth restriction. No women had heparin-induced thrombocytopenia or clinical features suggestive of osteoporosis.
A written delivery plan was available in 80%. The mean gestational age at delivery was 38·5 weeks (range 23–42), with 57 (46%) spontaneous vaginal deliveries (mean 39·3 weeks), 34 (27%) induced vaginal deliveries (mean 38·5 weeks) and 33 (27%) Caesarean sections (mean 36·9 weeks). The method of delivery was as planned in 72%. Anticoagulation for VTE was specifically mentioned as the reason for induction of labour in two patients and the reason for Caesarean sections was given as obstetric in 32 cases, although medical reasons were also listed in four. Problems at delivery were reported in 14 (11%) patients and included delivery of two intrauterine deaths, failure to progress in labour (n = 4), fetal distress (n = 3) and a breech twin.
Obstetric anaesthesia. Timing and adjustments of LMWH dose in the 24 h before and after delivery were poorly documented and very variable. Anaesthesia was regional in 33% of deliveries and general in 10%. No epidural haemorrhage was reported. Regional anaesthesia was used for 59% of Caesarean sections, for a third of induced vaginal deliveries and 17% of spontaneous vaginal deliveries.
Pregnancy outcome. There were no maternal deaths and 124 live births (96% of pregnancies) including two sets of twins. Apgar scores at 1 min were reported for 99 babies with a score of 7–10 in 83 (84%), and after 5 min in 93 babies with a score of 7–10 in 91 (98%). The median weight of the babies was 3·5 kg (range 1·5–5·0). No weight was entered for seven healthy babies. Five babies needed initial special care but were discharged in good health. One baby had hypospadias and talipes, and one had Down Syndrome. There were four pregnancy losses after starting treatment for VTE: a miscarriage at 12 weeks gestation after two weeks treatment for an iliofemoral DVT, a termination at 12 weeks for non-medical reasons and two intrauterine deaths, of which one was associated with cord compression at 38 weeks and the other, at 23 weeks, with severe oligohydramnios, uterine haematoma and chorionitis. Placental histology did not show thrombosis or infarction. The patient had had three previous losses but declined thrombophilia testing. There was one neonatal death. The 37-year-old mother, weight 63·5 kg, had a previous early miscarriage but no prior or family history of VTE. A left iliofemoral DVT at 28 weeks in her second pregnancy was treated with tinzaparin 13 000 IU s.c. once daily. No dose adjustments were made until spontaneous labour began at 37 weeks gestation, when the dose was reduced. Regional anaesthesia was given. The 2·5 kg baby had an Apgar score of 0–1 at delivery and could not be resuscitated. At postmortem, adrenal and pulmonary petechiae were present, thought to be perinatal in origin. The mother was later found to have compound heterozygosity for FVL and prothrombin 20210.
Major bleeding was not seen antenatally. Seven women (6%) had minor antenatal bleeding, including haemoptysis, bleeding gums, rectal bleeding from endometriosis, a subconjunctival haemorrhage and three with minor vaginal bleeding. The latter included the patient with uterine haematoma on scanning at 19 weeks gestation and fetal death before 23 weeks. Seven (6%) recorded bruising at injection sites.
Six women (5%) had blood loss of 500 ml or more during or within 24 h of delivery; one required an emergency Caesarean section and a three unit blood transfusion following placental abruption at 28 weeks. Two had Caesarean sections, at term, for breech presentation. One declined blood transfusion for religious reasons.
Major secondary postpartum haemorrhage (PPH) occurred in 3 (2%) patients on days 3, 5 and 9 postpartum, one with retained products of conception and two associated with early transfer from LMWH to warfarin. Less severe PPH was recorded in a further five patients, with over-anticoagulation reported as a contributory cause in two.
The mean time in hospital after delivery was 3·5 d [standard deviation (SD) 2·7, range 0–18 d], with a mean of 5 d (SD 2·9) for post-Caesarean patients and 3 d (SD 2·4) after vaginal delivery. Obstetric complications were given as the cause of delayed discharge in 60% patients, but anticoagulant adjustment, diabetic control, post-partum bleeding, baby in the Special Care Baby Unit, social factors, psychological support, drug withdrawal and patient’s wishes were recorded as other reasons for delayed discharge. Anticoagulation was continued with LMWH (30%) or by switching to warfarin (70%). The duration of postpartum anticoagulation was available for 114 patients: 5–12 weeks in 66 (58%) patients, 13–23 weeks in 21(18%) and >23 weeks in 27 (24%), with three remaining on long-term anticoagulation because of underlying thrombophilia. Clinical features of post-thrombotic syndrome (PTS) were recorded postpartum in 9 (13%) of the 68 patients who had clinical features of DVT at diagnosis, but routine clinical examination for post-thrombotic syndrome was not performed in most centres. Of the nine patients with evidence of PTS, the left side was affected in eight and seven following proximal DVTs.
This study shows that multi-centre data collection on the management of antenatal VTE is possible, and can provide valuable information to inform clinical decision-making and planning of future randomized trials. Secondly, although there was wide variation in LMWH dosage and monitoring, there were no recurrent thromboses, thus questioning the need for twice daily doses of LMWH and anti-Xa monitoring, both requiring further studies. Lastly, the data on previous thrombosis and failed thromboprophylaxis highlight the need for adequate prophylactic LMWH.
The Confidential Enquiry into Maternal and Child Health (CEMACH 2004) shows that VTE remains a major cause of maternal mortality in the UK, with about 10 cases of fatal VTE per year and a higher risk of maternal death in older women, with increased BMI and post partum. Improved awareness, along with national guidelines on management (RCOG, 2007) and prevention of antenatal VTE (RCOG, 2004) may be responsible for the fact that only four of the 25 maternal deaths because of PE recorded in the 2004 triennial report occurred in the antenatal period.
Non-fatal antenatal VTE occurs in less than one in 1000 pregnancies (Macklon & Greer, 1996; Andersen et al, 1998; Simpson et al, 2001; Heit et al, 2005). This low incidence, plus the difficulties inherent in any antenatal research, make randomized studies of treatment difficult to perform. Most recommendations on antenatal management of VTE have therefore been extrapolated from randomized studies in the non-pregnant population (Buller et al, 2005; Prandoni et al, 2006) supported by non-randomized studies of antenatal treatment with dalteparin (Ulander et al, 2002; Jacobsen et al, 2003; Barbour et al, 2004), enoxaparin (Ellison et al, 2000; Lepercq et al, 2001; Rodie et al, 2002; Rowan et al, 2003) and tinzaparin (Smith et al, 2004; Jorgensen & Nielsen, 2004). These and other studies have been well reviewed (Greer & Nelson-Piercy, 2005; Kher et al, 2007).
Although not demonstrated in our study, the increased risk of VTE with obesity has been emphasized both for pregnancy-related fatal PE (CEMACH, 2004) and non-fatal VTE (Larsen et al, 2007). Like others (Lindqvist et al, 1999) we found no correlation between VTE and advancing age, although the rate of fatal PE per million maternities rose with age in the triennial report (CEMACH, 2004). We found a weak correlation between age and gestation at diagnosis of VTE suggesting earlier presentation in older women. The incidence of VTE in different trimesters of pregnancy was as previously described, the risk starting in the first trimester and continuing throughout pregnancy (McColl et al, 1997; Ray & Chan, 1999; Blanco-Molina et al, 2007).
Over 80% of patients presented with classical clinical features of DVT and/or PE. During pregnancy a higher proportion of patients are likely to present with frank clinical signs of VTE than in the non-pregnant population because of the higher incidence of large vessel DVT – 63% proximal including 27% iliofemoral thrombi in our patients. Failure to appreciate the significance of classical signs and symptoms of VTE, even in the presence of background risk factors, has previously been identified (CEMACH, 2004).
All participating centres had, or were developing, local protocols for diagnosis of VTE in pregnancy. Because of the poor specificity of clinical diagnosis of VTE in pregnancy, objective confirmation should be obtained in all women with suggestive features (RCOG, 2007) to avoid the short- and long-term consequences of inaccurate or missed diagnosis. In this study, all DVTs were diagnosed using Doppler ultrasound with 65% of PE diagnosed by V/Q scanning. The use of CT scanning for diagnosis of antenatal PE has increased in many centres (Nijkeuter et al, 2006) since this study, but this may decline again because of a suggested increased lifetime risk of breast cancer. Each breast is exposed to a minimum radiation dose of 2 rads (20 mGy) during CT screening for PE (Parker et al, 2005) and this may be more damaging to the developing breast tissue in pregnancy and lactation than at other times. Certainly in suspected antenatal PE, Doppler ultrasound of the legs should be the first investigation as a positive scan obviates the need for lung scanning. We confirmed a high proportion of left proximal DVTs in pregnancy with 90% of iliofemoral DVTs on the left side, and two thirds of the proximal DVTs occurring in the last trimester. However, our data differed from that of earlier reports in which 85–90% of all DVTs in pregnancy were found on the left side (Lindhagen et al, 1986; Ginsberg et al, 1992), presumably because imaging techniques now detect smaller distal DVTs, which the present study found with equal frequency on left and right, and in each trimester.
Women with previous VTE have an estimated increased relative risk of recurrence in pregnancy of 3·5 compared with their risk outside pregnancy (Pabinger et al, 2002) and a 6·2% probability of antenatal recurrence, the risk being constant throughout pregnancy (Pabinger et al, 2005). However, this risk depends on the circumstances surrounding the original event, with studies showing no recurrences in those with transient risk factors and without thrombophilia (Brill-Edwards et al, 2000). An exception is oestrogen-related VTE and in our study all but one of the 16 previous events were related to pregnancy or COC. If guidelines had been followed (RCOG 2004) all these women would have been given antenatal thromboprophylaxis. Only four received LMWH, but doses were inadequate.
Thrombophilia (22%) was found less frequently than in previous studies (Gerhardt et al, 2000; McColl et al, 1997, 2000; Pabinger et al, 2005) partly because of the fact that FVIII coagulant activity and homocysteine levels were not tested in most centres, and anti-beta2 glycoprotein assays were not considered part of the defining criteria for antiphospholipid antibodies until 2006. Sixteen women with no past history of VTE were identified as having thrombophilia, two having been tested prior to the pregnancy. None of these 16 women were given antenatal thromboprophylaxis. Had they been tested before the second event, three would have been offered thromboprophylaxis as two were double heterozygotes for FVL and prothrombin 20210 and one had moderate protein S deficiency, with no personal history of VTE but a family history of protein S deficiency and VTE. Eight of these patients were heterozygous for FVL and three for prothrombin 20210 but these defects are not considered sufficient to justify antenatal thromboprophylaxis (RCOG, 2004). Some of the obstetric histories suggested antiphospholipid syndrome (APS) but there was only one patient with repeatedly positive antiphospholipid assays. The absence of women with APS in this survey suggests that once the diagnosis of APS is made, these patients receive adequate thromboprophylaxis to prevent recurrent VTE events, or that the risk of VTE in pregnancy in women with antiphospholipid antibodies is overstated. The value of thrombophilia screening is much debated (Walker et al, 2001; Baglin, 2003; Cosmi et al, 2003; Samama et al, 2003; Heit et al, 2005), and although the increased risk of VTE in pregnancy with individual thrombophilic abnormalities has been confirmed in a systematic review (Robertson et al, 2006), the absolute risk of VTE remains low in women without clinical risk factors. However, a previous venous thrombotic event associated with underlying thrombophilia is an indication for antenatal and post-partum prophylaxis (RCOG, 2004).
There is currently lack of agreement on thromboprophylactic LMWH dosage in pregnancy. Two study patients who were heterozygous for FVL and had previous oestrogen-related VTE developed recurrent thrombosis in the third trimester despite enoxaparin 40 mg daily in one and dalteparin 2500 U daily in the other, suggesting that doses greater than these are required in the third trimester i.e. ‘higher thromboprophylactic doses’ as recommended by the RCOG (2004), such as enoxaparin 40 mg twice daily. One woman with antithrombin deficiency and a previous VTE was only receiving enoxaparin 40 mg once daily in the second trimester prior to the recurrence. As heparin acts through potentiating the action of antithrombin, monitoring of anti-Xa activity is required to ensure that the combined effects of heparin and reduced antithrombin produce adequate anti-Xa activity. Current experience of the authors (unpublished observations) would suggest that, in the second and third trimesters, women with antithrombin deficiency require at least enoxaparin 40 mg twice daily. to obtain trough anti-Xa levels of 0·10 U/ml.
A family history of VTE was reported in 23% of our cohort, similar to the level of 25% reported in 189 Italian women who developed VTE on COC (Cosmi et al, 2003) in contrast to 10·2% in the Italian control group, confirming the importance of family history. Genetic thrombophilia was present in 46% of those with a family history of VTE in Cosmi’s report and in 42% of those without. In our survey, four of the 16 women with previous VTE also had a family history of VTE and seven had identifiable thrombophilia including all four of those with a positive family history. More data were needed on the risk of VTE in women with both a family history of VTE and a positive result for FVL before firm recommendations can be made. The patient with asymptomatic protein S deficiency and a family history of VTE and protein S deficiency was not offered antenatal thromboprophylaxis. Overall, our results suggest preconception thrombophilia testing in those with a family history may be of benefit.
Use and monitoring of LMWH
Prior to the 2007 update (RCOG, 2007), recommendations for the treatment of VTE in pregnancy were for either UFH initially or LMWH. In our survey, 13% of mothers received UFH at some point in their treatment, with four women having only UFH after late third trimester diagnosis. The reasons for UFH use were unclear, but seven women were started on UFH and transferred to LMWH. UFH is ideally used in the management of antenatal VTE at times of high-risk when the short half-life provides the needed flexibility to respond rapidly to clinical events. The three LMWH preparations licensed in the UK for use outside pregnancy are all used in pregnancy. There is debate as to whether they are interchangeable in view of their differing anti-Xa:antithrombin ratios (Prandoni, 2003; Nenci, 2003; White & Ginsberg, 2003). In our study, nearly half the patients were treated with enoxaparin, with the other half were equally divided between dalteparin and tinzaparin (Table V). Most reports used twice daily LMWH (Greer & Nelson-Piercy, 2005) although our results showed that many patients were managed on once daily dosing. Two thirds of those receiving LMWH in this survey were given once daily LMWH with 15% increasing to twice daily at some point. There is no data comparing once or twice daily LMWH treatment of VTE in pregnancy but twice daily dosing is recommended by the RCOG (2007) based on anti-Xa activity and reports on safety and efficacy. Anti-Xa levels have been studied after treatment doses of antenatal LMWH (Rey & Rivard, 2000; Ellison et al, 2000; Rodie et al, 2002; Smith et al, 2004) and after prophylactic LMWH in pregnancy (Blomback et al, 1998; Hunt et al, 1997; Casele et al, 1999; Sephton et al, 2003; O’Connell et al, 2004; Norris et al, 2004). The pharmacokinetics has shown a shorter heparin half-life and lower anti-Xa levels for a given dose in pregnancy. However, it is not known whether detectable heparin activity is required throughout the 24 h or whether continuing the initial LMWH treatment dose throughout pregnancy is necessary.
Our survey suggests that the standard management of VTE outwith pregnancy, with once daily dosing, may be efficacious in pregnancy and is worthy of further research. We have shown significant inconsistencies in LMWH monitoring practices, dose adjustments and target anti-Xa ranges, reflecting a lack of clinical trials and varied guidance in this area (RCOG, 2007; Kher et al, 2007), and questioning the need for routine antenatal anti-Xa monitoring. Despite the wide variation in target anti-Xa activities, there were no recurrent thrombotic events. These findings suggest inappropriate reliance on an assay which, in a past NEQAS scheme, gave a wide spread of results (Kitchen, 2000). Some centres no longer monitor anti-Xa and use a fixed dose 12-hourly regime without monitoring (Hunt et al, 1997) in accordance with RCOG guidelines. Laboratories should, however, offer anti-Xa levels when there are complications such as bleeding, possible incorrect dosage or recurrent thomboembolism.
The 5% incidence of primary post-partum haemorrhage seen in this study was similar to that quoted for the developed countries (Combs et al, 1991; Stones et al, 1993; Magann et al, 2005). A large case control study showed that bleeding complications in women receiving LMWH were not increased when compared with normal controls matched for delivery route (Kominiarek et al, 2007). It is important to optimize management of LMWH prior to delivery and develop planned strategies to minimise blood loss when higher doses are necessary (Akkad et al, 2003). Because of the risk of secondary PPH, an extended period on LMWH should be considered rather than early transfer to warfarin.
There was no increased incidence of pre-eclampsia or intrauterine growth restriction in our study. Reassuringly, the Caesarean section rate of 27% was only a little over the national rate of 23% at the time of the survey (http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsStatistics/DH_4107060) and the majority of Caesarean sections were performed for obstetric reasons. Almost half of the women had a normal spontaneous vaginal delivery, again confirming that antenatal VTE were managed successfully. A third of the women had a regional anaesthetic for delivery but, overall, anaesthesia and anticoagulant administration around delivery were poorly recorded. It would appear that most centres were aware of the reported risk of spinal and epidural haematoma (Wysowski et al, 1998) and no complications of regional anaesthesia occurred. The pregnancy losses were too few to allow comment, but are similar to reports of 3–4% adverse foetal outcome for women on LMWH (Kher et al, 2007) and in the general antenatal population. A 23-week loss following oligohydramnios and retroplacental bleeding occurred with once daily enoxaparin with therapeutic peak ant-Xa levels, reduced to 40 mg daily for 3 weeks prior to confirmation of intrauterine death. The 38-week stillbirth was related to significant maternal medical problems unrelated to anticoagulation, with foetal growth restriction and cord asphyxia. Once daily tinzaparin had been given with a dose increase following a low anti-Xa level and there was no report of bleeding. The neonatal death at 37 weeks gestation followed a fixed once daily dose of tinzaparin (200 IU/kg), which appeared to have been calculated on the patient’s weight in mid-trimester rather than on the recommended 175 IU/kg on early pregnancy weight. No bleeding was reported and this dose was continued without monitoring until the start of labour when dose reduction allowed epidural anaesthesia. The post-mortem reported petechial haemorrhages in congested lungs and adrenals but the cause of death could not be determined. LMWH does not cross the placenta and thus could not be implicated. In summary, there was no evidence that these pregnancy losses were related to LMWH. Large randomized studies would be required to detect a difference in pregnancy outcome with different LMWH protocols.
The mean time in hospital after vaginal delivery was 3 d and after Caesarean section 5 d, similar to the national averages (http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsStatistics/DH_4107060; Simon Jones, Dr Foster Research Ltd, London, UK, personal communication regarding data for 2005–2006) suggesting that anticoagulation did not delay discharge. PTS is reported to be clinically detectable though generally mild in about 50–70% of women after DVT in pregnancy (McColl et al, 2000; Rosfors et al, 2001) and severe PTS is rare with treatment (Kahn & Ginsberg, 2002). Few centres reported clinical examination of the leg post-partum and only 13% of 68 patients with recorded clinical features of DVT were reported to have postpartum features that suggested PTS.
To avoid bias, offers of funding from pharmaceutical companies were not accepted but lack of funding meant that the survey progressed slowly. No attempts were made to alter clinical practice in the participating centres, the data reflecting current multidisciplinary practice, The report does not claim to include every antenatal patient with VTE in pregnancy in the 25 centres, as the survey only used haematologists to enter patients, thus, patients managed entirely by non-haematologists may have been missed. The large obstetric centres could have entered >10 cases over 2 years. Restricting the number of patients from each centre to obtain a wide view of clinical practice across the country meant that we did not attempt to give accurate figures on the incidence of antenatal VTE. Other possible sources of selection bias were the exclusion of patients with missing outcome data, late data entry or those without image-proven VTE. Double data entry was not used routinely, although participants were contacted by the trial co-ordinator to check discrepancies or supply important missing data.
This is the largest survey of antenatal VTE to date, and suggests that the treatments used in the UK and Ireland, albeit with much variation of regimens, are safe and efficacious. Evidence for the best practice in the management of low incidence conditions in pregnancy is always difficult to obtain because of the major organization required for a multi-centre randomized trial. Such controlled international research trials were proposed for antenatal thromboprophylaxis – the APPLE study (Gates et al, 2002) and for thromboprophylaxis after Caesarean section – the PEACH study (Brocklehurst & Gates, 2002) but were abandoned because of the complexity and magnitude of the projects (Gates et al, 2004a,b). Similar problems would confront plans for a randomized national study of antenatal treatment of VTE. Multi-centre studies similar to ours are therefore important in ensuring that we give the safest and most efficacious treatment to mothers with VTE. The UK Obstetric Surveillance System (UKOSS; Knight et al, 2005) has conducted a similar, as yet unpublished study, of VTE reported by obstetricians.
In conclusion, our results confirm the timing of first and recurrent venous thromboembolic events in pregnancy, thus emphasizing the importance of early risk assessment for thromboprophylaxis in women with previous VTE or other risk factors. Our data highlight the importance of testing for antithrombin deficiency in those with previous VTE, and emphasize the need for adequate doses of prophylactic LMWH in women with previous VTE. In this survey, although not designed to provide significant outcome data, enoxaparin, dalteparin and tinzaparin used in treatment doses twice daily and possibly once daily, appeared safe and efficacious in preventing recurrent VTE in pregnancy. More research is required to confirm this.
Thanks to haematology nurses D. Elliott, S. Stuart-Smith, S. Jones and K. Willis who helped with the data collection, to Simon Jones who provided national data from Dr Foster database, to all participating haematologists and to web manager Daniel Pearce of mediamurph.com.
- 2003) Intrapartum care for women on full anticoagulation. International Journal of Obstetric Anaesthesia, 12, 188–192. , , & (
- 1998) The cumulative incidence of venous thromboembolism during pregnancy and puerperium – an 11 year Danish population-based study of 63,300 pregnancies. Acta Obstetricia et Gynecologica Scandinavica, 77, 170–173. , , , & (
- 2003) Management of thrombophilia: who to screen? Pathophysiology of Haemostasis and Thrombosis, 33, 401–404. (
- 2004) A prospective trial that demonstrates that dalteparin requirements increase in pregnancy to maintain therapeutic levels of anticoagulation. American Journal of Obstetrics & Gynecology, 191, 1024–1029. , & (
- 2007) Venous thromboembolism during pregnancy or postpartum: findings from the RIETE Registry. Thrombosis & Haemostasis, 97, 186–190. , , , , , , & (
- 1998) Thromboprophylaxis with low molecular mass heparin, Fragmin (dalteparin), during pregnancy – a longitudinal safety study. Blood Coagulation & Fibrinolysis, 9, 1–9. , , , & (
- Recurrence of Clot in This Pregnancy Study Group (2000) Safety of withholding heparin in pregnant women with a history of venous thromboembolism Recurrence of Clot in This Pregnancy Study Group. New England Journal of Medicine, 343, 1439–1444. , , , , , , , , , , & &
- 2002) Randomized controlled trial (PEACH) and meta-analysis of thromboprophylaxis using low molecular weight heparin (enoxaparin) after caesarean section. Journal of Obstetrics & Gynaecology, 22, S52. & (
- 2005) Treatment of venous thromboembolism. Journal of Thrombosis & Haemostasis, 3, 1554–1560. , & (
- 1999) Changes in the pharmacokinetics of the low-molecular-weight heparin enoxaparin sodium during pregnancy. American Journal Obstetrics & Gynecology, 181, 1113–1117. , , & (
- CEMACH (2004) Why Mothers Die 2000–2002. The Sixth Report of the Confidential Enquiries into Maternal Deaths in the United Kingdom. RCOG Press, London.
- 1991) Factors associated with postpartum hemorrhage with vaginal birth. Obstetrics and Gynecology, 77, 69–76. , & (
- 2003) Role of family history in identifying women with thrombophilia and higher risk of venous thromboembolism during oral contraception. Archives of Internal Medicine, 163, 1105–1109. , , , & (
- 2000) Antenatal use of enoxaparin for prevention and treatment of thromboembolism in pregnancy. British Journal of Obstetrics & Gynaecology, 107, 1116–1121. , & (
- 2002) Antenatal thromboprophylaxis using low molecular weight heparin (enoxaparin) for women at risk of thrombembolic disease: multicentre placebo controlled randomized trial (APPLE) and systematic review. Journal Obstetrics & Gynaecology, 22, S44. , & (
- 2004a) Recruitment to multicentre trials. BJOG: International Journal of Obstetrics and Gynaecology, 111, 3–5. , , & (
- Thromboprophylaxis in Pregnancy Advisory Group(2004b) Thromboprophylaxis and pregnancy: two randomized controlled pilot trials that used low-molecular-weight heparin. American Journal of Obstetrics & Gynecology, 191, 1296–1303. , , & &
- 2000) Prothrombin and factor V mutations in women with a history of thrombosis during pregnancy and the puerperium. New England Journal of Medicinie, 342, 374–380. , , , , , , & (
- 1992) Venous thrombosis during pregnancy: leg and trimester of presentation. Thrombosis & Haemostasis, 67, 519–520. , , , , , & (
- 2005) Low-molecular-weight heparins for thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a systematic review of safety and efficacy. Blood, 106, 401–407. & (
- 2005) Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Annals of Internal Medicine, 143, 697–706. , , , , & (
- 1997) Thromboprophylaxis with low molecular weight heparin (Fragmin) in high risk pregnancies. Thrombosis & Haemostasis, 77, 39–43. , , , , , , & (
- 2003) Low molecular weight heparin dalteparin for the treatment of venous thromboembolism in pregnancy. BJOG: An International Journal of Obstetrics and Gynaecology, 110, 139–144. , & (
- 2004) The low molecular weight heparin, tinzaparin, is effective and safe in the treatment and prophylaxis of venous thromboembolic disease during pregnancy. Blood, 104, 1774S. & (
- 2002) The post-thrombotic syndrome: current knowledge, controversies and directions for future research. Blood Reviews, 16, 155–65. & (
- 2007) The management of thrombosis in pregnancy: Role of low-molecular-weight heparin. Thrombosis & Haemostasis, 97, 505–513. , & (
- 2000) Problems in laboratory monitoring of heparin dosage. British Journal of Haematololgy, 111, 397–406. (
- 2005) The UK Obstetric Surveillance System for rare disorders of pregnancy. BJOG: An International Journal of Obstetrics and Gynaecology, 112, 263–265. , , & (
- 2007) Low-molecular-weight heparin in pregnancy: peripartum bleeding complications. Journal of Perinatology : Official Journal of the California Perinatal Association, 27, 329–334. , , , , & (
- 2007) Maternal smoking, obesity, and risk of venous thromboembolism during pregnancy and the puerperium: a population-based nested case–control study. Thrombosis Research 120, 505–509. , , & (
- 2001) Venous thromboembolism during pregnancy: a retrospective study of enoxaparin safety in 624 pregnancies. BJOG: An International Journal of Obstetrics and Gynaecology, 108, 1134–1140. , , , , , , , , , , & (
- 1986) Late venous function in the leg after deep venous thrombosis occurring in relation to pregnancy. British Journal of Obstetrics & Gynaecology, 93, 348–352. , , & . (
- 1999) Thrombotic risk during pregnancy: a population study. Obstetrics and Gynecology, 94, 595–599. , & (
- 1996) Venous thromboembolic disease in obstetrics and gynaecology: the Scottish experience. Scottish Medical Journal, 41, 83–86. & (
- 2005) Postpartum hemorrhage after vaginal birth: an analysis of risk factors. Southern Medical Journal, 98, 419–422. , , , , & (
- 1997) Risk factors for pregnancy associated venous thromboembolism. Thrombosis & Haemostasis, 78, 1183–1188. , , , , , , & (
- 2000) Prothrombin 20210 GA, MTHFR C677T mutations in women with venous thromboembolism associated with pregnancy. BJOG: An International Journal of Obstetrics and Gynaecology, 107, 565–569. , , , , & (
- 2003) Low molecular weight heparins: are they interchangeable? NO. Journal of Thrombosis & Haemostasis: JTH, 1, 12–13. (
- 2006) Diagnosis of deep vein thrombosis and pulmonary embolism in pregnancy: a systematic review. Journal of Thrombosis & Haemostasis, 4, 496–500. , & (
- 2004) Low molecular weight heparin (tinzaparin) therapy for moderate risk thromboprophylaxis during pregnancy. A pharmacological study. Thrombosis & Haemostasis, 92, 791–6. , , , & (
- 2004) Is the monitoring of anti-Xa activity necessary in pregnant women undergoing thromboprophylaxis? European Journal of Obstetrics & Gynecology, 114, 12–14. , , , & (
- 2002) Temporary increase in the risk for recurrence during pregnancy in women with a history of venous thromboembolism. Blood, 100, 1060–1062. , , , , , & (
- 2005) Risk of pregnancy-associated recurrent venous thromboembolism in women with a history of venous thrombosis. Journal of Thrombosis and Haemostasis: JTH, 3, 949–954. , , , , , & (
- 2005) Female breast radiation exposure during CT pulmonary angiography. American Journal of Roentgenology, 185, 1228–1233. , , , , & (
- 2003) Low molecular weight heparins: are they interchangeable? Yes. Journal of Thrombosis and Haemostasis: JTH, 1, 10–11. (
- 2006) New strategies for the treatment of acute venous thromboembolism. Seminars in Thrombosis and Hemostasis, 32, 787–792. , & (
- 1999) Deep vein thrombosis during pregnancy and the puerperium: a meta-analysis of the period of risk and the leg of presentation. Obstetrical & Gynecological Survey, 54, 265–271. & (
- RCOG(2004) Thromboprophylaxis during Pregnancy, Labour & After Vaginal Delivery. Guideline no.37. Royal College of Obstetricians and Gynaecologists Press,London. Available at: http://www.rcog.org.uk/resources/Public/pdf/Thromboprophylaxis_no037.pdf (accessed on 31 August 2007).
- RCOG(2007) Thromboembolic Disease in Pregnancy and the Puerperium: Acute Management. RCOG Green-top Guideline No.28 Second Edition.Royal College of Obstetricians and Gynaecologists Press, London. Available at: http://www.rcog.org.uk/resources/Public/pdf/green_top_28_thromboembolic_minorrevision.pdf (accessed on 31 August 2007).
- 2000) Prophylaxis and treatment of thromboembolic diseases during pregnancy with dalteparin. International Journal of Gynaecology and Obstetrics: the Official Organ of the International Federation of Gynaecology and Obstetrics, 71, 19–24. & (
- The Thrombosis: Risk and Economic Assessment of Thrombophilia Screening, S. (2006) Thrombophilia in pregnancy: a systematic review. British Journal of Haematology, 132, 171–196. , , , , , , , , , & &
- 2002) Low molecular weight heparin for the treatment of venous thromboembolism in pregnancy: a case series. BJOG : An International Journal of Obstetrics and Gynaecology, 109, 1020–1024. , , , , & (
- 2001) A 16-year haemodynamic follow-up of women with pregnancy-related medically treated iliofemoral deep venous thrombosis. European Journal of Vascular and Endovascular Surgery: the Official Journal of The European Society For Vascular Surgery, 22, 448–455. , , , , & (
- 2003) Prophylactic and therapeutic enoxaparin during pregnancy: indications, outcomes and monitoring. The Australian & New Zealand Journal of Obstetrics & Gynaecology, 43, 123–128. , , & (
- 2003) Pregnancy-associated venous thromboembolism VTE in combined heterozygous factor V Leiden FVL and prothrombin FII 20210 A mutation and in heterozygous FII single gene mutation alone. British Journal of Haematology, 123, 327–334. , , , , , , & (
- 2003) A longitudinal study of maternal dose response to low molecular weight heparin in pregnancy. Obstetrics & Gynecology, 101, 1307–1311. , , , , , & (
- 2001) Venous thromboembolism in pregnancy and the puerperium: incidence and additional risk factors from a London perinatal database. BJOG : An International Journal of Obstetrics and Gynaecology, 108, 56–60. , , & (
- 2004) Tinzaparin sodium for thrombosis treatment and prevention during pregnancy. American Journal of Obstetrics & Gynecology, 190, 495–501. , , , & (
- 1993) Risk factors for major obstetric haemorrhage. European Journal of Obsterics Gynecology and Reproductive Biology, 48, 15–18. , & (
- 2002) Treatment of deep venous thrombosis with low-molecular-weight heparin during pregnancy. Thrombosis Research, 106, 13–17. , & (
- 2001) Investigation and management of heritable thrombophilia. British Journal of Haematology, 114, 512–528. , & (
- 2003) Low molecular weight heparins: are they all the same? British Journal of Haematology, 121, 12–20. & (
- 1998) Spinal and epidural hematoma and low-molecular-weight heparin. The New England Journal of Medicine, 338, 1774–1775. , , & (
F. Ni Ainle, St Vincent’s University Hospital, Dublin, Ireland; J. Arnold, Barnet and Chase Farm Hospitals, London, UK; M. Bowers, Ulster Community and Hospitals Trust, Belfast, UK; V. Clough, Countess of Chester Hospital, Chester, UK; A. Copplestone, Plymouth Hospital, Plymouth, UK; D. Creagh, Royal Cornwall Hospital, Truro, UK; E. Gaminara, West Herts Hospitals NHS Trust, UK; S. Green, Swindon and Marlborough Hospitals, Swindon, UK; C. Hoggarth, Hinchingbrooke Hospital, Hinchingbrooke, UK; B. Hunt, Guys and St Thomas’s NHS Trust, London, UK; S. Janes, St Richards Hospital, Chichester, UK; J. Keidan, Queen Elizabeth Hospital, Kings Lynn, UK; H. Kelsey, Kettering General Hospital, Kettering, UK; M. Leahy, Mid-Western Regional Hospital, Limerick, Ireland; J. Mainwaring, Basingstoke General Hospital, Basingstoke, UK; K. Murphy, National Maternity Hospital, Dublin, Ireland; J. Pattinson, Buckinghamshire Hospitals NHS Trust, UK; S. Pavord, University Hospitals of Leicester, UK; K. Rice, Epsom and St Helier University Hospitals NHS Trust, UK; C. Shiach, Manchester Royal Infirmary, Manchester, UK; J. Shirley, Royal Surrey County NHS Trust, UK; S. Swart, Northampton General Hospital, Northampton, UK; D. Thompson, Luton and Dunstable Hospital, Luton, UK; B. Vadher, Bromley Hospital, UK; J. Voke, East and North Herts NHS Trust, UK; A. Watson, Stoke Mandeville Hospital, Aylesbury, UK.