Thrombosis
Thrombosis in pregnancy: maternal and fetal issues
Ian A Greer
Pulmonary thromboembolism is the main cause of maternal death in the UK and current trends show an increase.
Deep-vein thrombosis underlies this disorder. Important issues include pathophysiology, diagnosis, and management
of thrombosis in pregnancy, especially the use of anticoagulants. Congenital and acquired thrombophilias contribute
to the pathophysiological processes that underlie miscarriage, intrauterine growth restriction, and pre-eclampsia, and
raises new possibilities for intervention. The high prevalence of thrombophilic defects in the population, the
association of defects with maternal and fetal disorders, and special considerations for management make it
essential for obstetricians to understand this area.
In the UK, pulmonary thromboembolism is the leading
cause of maternal death.1 Deep-vein thrombosis (DVT)
underlies this disorder and is frequently unrecognised.2
DVT in pregnancy is associated with an increased risk
of further thrombosis and deep-vein insufficiency. Risk
factors for DVT include congenital and acquired
thrombophilias, which probably play a part in the
pathogenesis of miscarriage, intrauterine growth
restriction, and pre-eclampsia. Perhaps the most
compelling evidence is the association between thrombo-
philia due to antiphospholipid antibodies and the
activation of the coagulation-system3 and placental
infarction. A randomised controlled trial4 has shown that
aspirin and heparin given to reduce coagulation activity
improved the outcome of pregnancy. The role of
anticardiolipin antibody syndrome in such complications
is explored later in this series by Michael Greaves.5 This
paper focuses on maternal and fetal issues associated with
venous thrombosis and congenital thrombophilias.
Epidemiology of maternal venous
thromboembolism
The overall incidence of fatal pulmonary thrombo-
embolism associated with pregnancy has fallen since the
early 1950s, especially among women after vaginal
delivery (figure 1). However, since the 1980s, this downward
trend has been reversed,1 which highlights the need for
thromboprophylaxis after vaginal delivery in all women at
risk of thromboembolism. The rate of non-fatal events is
more difficult to establish than that for fatal ones. The
incidence of antenatal DVT is about 0·615 per 1000
pregnancies in women aged younger than 35 years and
1·216 per 1000 in women older than 35 years.6 The rate
of postpartum DVT is about 0·304 per 1000 pregnancies
in women younger than 35 years and 0·72 per 1000 in
women older than 35 years. Antenatal DVT is more
common than postpartum DVT,6,7 for which data may
not be accurate and complete because women may
present to internal medicine rather than to obstetric
services. Almost 40% of postpartum DVT manifests after
discharge from hospital. Age and operative delivery are
major risk factors for venous thromboembolism (figure 2).
Lancet 1999; 353: 1258–65
Department of Obstetrics and Gynaecology, University of Glasgow,
Glasgow Royal Infirmary, Glasgow G31 2ER, UK (Prof I A Greer MD)
1258
Other factors that underlie venous thromboembolism
in pregnancy include weight over 80 kg, family or
personal history of thrombosis, and thrombophilia.8
Venous thromboembolism during pregnancy is associated
with an increased risk of future venous
thromboembolism. Almost two-thirds of women with
venous thromboembolism during pregnancy develop
deep-vein insufficiency in the affected leg,9 a rate
substantially higher than would be expected after DVT
outside of pregnancy. This increase may reflect the higher
incidence of iliofemoral vein thrombosis during
pregnancy.
Pathogenesis of venous thromboembolism
Virchow’s triad of underlying factors in venous
thrombosis—hypercoagulability, venous stasis, and
vascular damage—all occur in pregnancy. During
pregnancy there are increases in procoagulant factors,
such as von Willebrand factor, factor VIII, factor V,
and fibrinogen, that occur together with an acquired
resistance to the endogenous anticogulant, activated
protein C, and a reduction in protein S, the co-factor for
protein C. 10 These changes are accompanied by impaired
fibrinolysis through increases in plasminogen activator
inhibitors 1 and 2, the latter being produced by the
placenta.11 These changes represent the physiological
preparation for the haemostatic challenge of delivery.
Venous stasis occurs in pregnancy by the end of the first
trimester and reaches a nadir at 36 weeks.12 Endothelial
damage to pelvic vessels can occur during vaginal or
abdominal delivery. Thus, the scene is set for the
development of thrombosis in pregnancy.
Almost 90% of DVT affect the left side among
pregnant women compared with 55% among women
who are not pregnant.8,9 This difference may reflect
compression of the left iliac vein by the right iliac and the
ovarian arteries which cross the vein on the left side only.
Furthermore, in pregnancy most cases of DVT are
ileofemoral rather than calf vein thrombosis (72% vs 9%),
and iliofemoral DVT is more likely than calf-vein
thrombosis to lead to pulmonary thromboembolism.
DVT in pregnancy can present, or be associated with,
lower abdominal pain due to periovarian collateral
circulation or thrombosis. When coupled with the mild
pyrexia and leucocytosis of venous thromboembolism,
this pain can be mistaken for other intra-abdominal
disorders, such as urinary tract infection or appendicitis.
THE LANCET • Vol 353 • April 10, 1999

Figure 1: Number of cases of fatal pulmonary thromboembolism in pregnancy
and the puerperium in England 1955–84, UK 1985–96
Data from Confidential Enquiries into Maternal Deaths.
Congenital thrombophilia and maternal venous
thromboembolism
The thrombophilias, which were discussed by Fritz
Rosendaal13 earlier in this series, underlie many
thrombotic disorders in pregnancy. 14 The main congenital
thrombophilias are deficiencies of antithrombin, protein
C, and protein S, and the presence of factor V Leiden,
the prothrombin gene variant, and homozygosity for
the thermolabile variant of methylenetetrahydrofolate
reductase (MTHFR). Factor V Leiden manifests as
resistance to activated protein C and is the most common
defect underlying venous thrombembolism, but such 16
resistance can be caused by disorders other than factor V
Leiden, including antiphospholipid antibody syndrome
and other genetic defects in the factor V molecule. The
resistance can also be acquired in pregnancy as a result of
increases in factor V and factor VIII.10,16
17
The genetic variation in the prothrombin gene has
been linked to venous thromboembolism in pregnancy.18
However, more information is needed about this defect
in pregnancy, particularly because it may be found with
other inherited thrombophilias such as factor V Leiden.
Hyperhomocysteinaemia is an established risk factor for
venous and arterial thrombosis, and is most commonly
caused by homozygosity for the thermolabile variant
of MTHFR.19 Pregnancy is associated with decreased
concentrations of homocysteine and folic acid
supplements will also lower homocysteine concentrations.
However, the contribution of homocysteine to venous
thromboembolism in pregnancy is unclear.
The risk of maternal venous thromboembolism with
underlying thrombophilia will depend on the underlying
thrombophilic defect(s), history of thrombotic events,
and additional risk factors. As Rosendaal13 has pointed
out, clinical thrombosis is now considered a multicausal
disease, resulting from interaction between congenital
and acquired risk factors. The risk of thromboembolism
THE LANCET • Vol 353 • April 10, 1999
during pregnancy among women with
thrombophilia in the absence of anti-
coagulant therapy was initially judged to be
about 60% among pregnant women
with antithrombin deficiency.14,20 However,
retrospective studies20,23 of symptomatic
kindred showed the incidence of venous
thromboembolism among these women to
be 32% to 44%. For pregnant women with
abnormalities of the protein C and protein
S system the risk is substantially lower than
for antithrombin-deficient women, and
postpartum thromboses are more common
than antepartum thrombosis. The risk of
thrombosis in pregnancy is 3–10% for
protein C deficiency and 0–6% for protein
S deficiency. In postpartum women the
risk is 7–19% for protein C deficiency and
7–22% for protein S deficiency.20–23
Activated protein C resistance, this defect
was found in up to 78% of women
investigated for venous thrombosis in
pregnancy,24 whereas the factor V Leiden
genotype was found in up to 46%.25 These
studies, however, investigated women with
venous thromboembolism and provide
limited information about the risk of
thrombosis in previously symptom-free
women with the mutation. In a study of 43 women with
the factor V Leiden mutation from symptomatic families,
the overall incidence of pregnancy-associated thrombosis
was 14%.22 This risk may be higher postpartum.26
Thus, pregnancy seems to be a significant factor that can
precipitate venous thromboembolic complications in
women with this mutation. However, these studies
assess symptomatic kindred and so may overestimate the
risk.
In McColl and colleagues’ 1997 retrospective study8 of
about 72 000 pregnancies, pregnant women with venous
thromboembolism were screened for thrombophilia. The
underlying prevalence of congenital thrombophilia in
this population had already been established and the
investigators estimated the risk of venous thrombo-
embolism in pregnancy to be 1 in 437 for factor V
Leiden, 1 in 113 for protein C deficiency, 1 in 2·8 for
type I (quantitative) antithrombin deficiency, and 1 in 42
Figure 2: Incidence of postpartum deep-vein thrombosis
and pulmonary thromboembolism by maternal age and
method of delivery
1259
Study and study design Study population Prevalence of factor V Leiden, activated protein C resistance, or both
protein C resistance, or both
Cases Controls
Rai et al 276 Cases: women with >3 miscarriages (a) 4/70 (5·7%) 3/70 (4·3%)
Case-control to assess prevalence (a) confined to first trimester (b) 10/50 (20%)*
of activated protein C resistance (b) including at least one second
trimester miscarriage
Controls: parous women with no history
of miscarriage
Grandone et al28 Cases: women with >2 miscarriages 7/42 (16·3%)† 5/118 (4·2%)
Case-control to assess prevalence Controls: parous women with no history
of factor V Leiden of miscarriage
Dizon-Townson et al30 Cases: women with spontaneous 12/176 (6·8%) 17/403 (4·2%)
Case-control to assess prevalence miscarriage
of factor V Leiden Controls: unselected pregnancies
Dizon-Townson et al32 Cases: women with >3 miscarriages 0/40 0/25
Case-control to assess prevalence Controls: parous women
of factor V Leiden
Belasch et al31 Cases: women with >2 consecutive 1/50 (2%) 1/50 (2%)
Case-control to assess prevalence first-trimester miscarriages
of factor V Leiden Controls: parous women with no history
of miscarriage
Brenner et al29 Cases: women with >3 consecutive first 19/39 (48%)‡ Not available
Cohort to assess activated protein or second-trimester miscarriages
C resistance and factor V Leiden
*p<0·02 versus controls and first-trimester miscarriage group.
†p<0·02 versus controls; cases were also more likely to have second-trimester loss (p<0·05).
‡An additional 5/39 had activated protein C resistance without factor V Leiden.
Table 1: Miscarriage, activated protein C resistance, and factor V Leiden

for type II (qualitative) antithrombin deficiency. There is
a pressing need for more information on the natural
history of these disorders in pregnancy so that
obstetricians can assess risk and provide appropriate
thromboprophylaxis. Because the risk in symptom-free
women with a familial thrombophilic trait, excluding
antithrombin deficiency, is uncertain in pregnancy, a
randomised placebo-controlled trial of thrombopro-
phylaxis is justified to guide practice.
Congenital thrombophilia and fetal loss
The association between acquired resistance to activated
protein C and disorders such as anticardiolipin antibody
syndrome (an underlying cause of recurrent miscarriage),
prompted investigation of the effect of congenital
activated protein C resistance on miscarriage Rai and
colleagues27 found a significantly higher prevalence of
such resistance in women with recurrent miscarriages
(including at least one second-trimester loss) than in
women with recurrent fetal loss in the first trimester or
in a parous control group (table 1). This link to fetal loss
during the second trimester was supported by Grandone
and colleagues’ case-control study28 which showed a
significantly higher prevalence of factor V Leiden in
recurrent miscarriage, particularly in the second
trimester. Brenner and colleagues’ cohort study29 of
women with recurrent fetal loss showed a 48% prevalence
of the factor V Leiden mutation; in women with factor V
Leiden, only 19% of 128 pregnancies ended in livebirths.
Other studies30–32 have shown no relation between
recurrent miscarriage and factor V Leiden (table 2).
These data from case-control and cohort studies suggest
that recurrent first-trimester miscarriage is not associated
with factor V Leiden, although second-trimester miscarriage
may be. A possible explanation for this discrepancy is that
first-trimester miscarriages reflect a failure in implantation
and second-trimester miscarriages reflect thrombotic event
in the placenta. The cohort study by Dizon-Townson and
colleagues30 showed a link with fetal genotype: 8·6% of
aborted fetuses carried the genotype for factor V. This
finding accords with the higher frequency of placental
infarction in pregnancies associated with maternal and
fetal carriage of factor V Leiden.30,33
These case-control and cohort studies focused on
women with a history of miscarriage. Taking another
approach, Preston and colleagues34 investigated
pregnancy outcome in a cohort of women with
thrombophilia. Compared with data from a healthy
control population, there was an increased risk of fetal
loss (odds ratio 1·35 [95% CI 1·01–1·82]) and stillbirth
(3·6 [1·4–9·4]), however, the risk of miscarriage was not
significantly increased (1·27 [0·94–1·71]) (table 2). The
investigators also found no difference in the number of
pregnancies or fetal losses between women with and
those without thrombophilic partners.34 Sanson and
colleagues,35 found an increased incidence of pregnancy
loss in women with a deficiency of antithrombin, protein
C, and protein S, with an odds ratio of 2·0 (1·2–3·3) for all
deficiencies combined. 35 Other thrombophilic defects such
as dysfibrinogenaemia have also been linked to fetal loss.36
These data come from case-control or retrospective
cohort studies. Prospective longitudinal studies are
needed to define the risk of pregnancy-related
complications in women with thrombophilic abnormalities.
Nonetheless, available data are consistent with the concept
that maternal thrombophilia carries an increased risk of
fetal loss due to placental vascular disorders.
Thrombophilia, intrauterine growth restriction,
and pre-eclampsia
Since pre-eclampsia is associated with vascular injury,
which in turn is associated with a disturbance in
coagulation, the congenital thrombophilias could have
a role in this disorder. Dekker and colleagues’ cohort
study37 of 101 patients with a history of pre-eclampsia
found that 15% of the women had underlying activated
protein C resistance, although genotyping for factor V
Leiden was not done. These investigators also found that

1260 THE LANCET • Vol 353 • April 10, 1999

Type of thrombophilia Number of cases
(pregnancies)
Antithrombin deficiency 108 (250)
Protein C deficiency 162 (430)
Protein S deficiency 145 (378)
Factor V Leiden 141 (410)
Combined defects 15 (46)
Data are from refs 30–34.
Control group n=395 with 1019 pregnancies.
*Adjusted for number of pregnancies and centre (data from Preston and colleagues34).
Table 2: Odds ratios for fetal loss in women with thrombophilia compared with a control group by type of thrombophilia
25% of the women had a prevalence of protein S
deficiency, and that 18% of patients had a positive result
to a methionine load test, which assesses the possibility
of hyperhomocysteinaemia, Furthermore, 30% of
patients had anticardiolipin antibodies.37 Several patients
had combined disorders, but the only combination that
recurred was protein S deficiency and a moderate
increase of anticardiolipin antibodies. De Vries and
colleagues38 assessed 62 women who had had placental
abruption, intrauterine death, or a small-for-gestational
age infant in the absence of antihypertensive disorders,
and found that 65%, 56%, and 85%, respectively, had
underlying thrombophilic disorders. Overall, 26% had
protein S deficiency, 24% of women had hyperhomo-
cysteinaemia, and 6% had protein C deficiency; no
women had a deficiency of antithrombin.38 Several case
reports have also described activated protein C resistance
in pregnancies complicated by HELLP syndrome
(haemolysis elevated liver enzymes and low platelets), a
variant of pre-eclampsia, severe pre-eclampsia,
intrauterine growth restriction, and fetal loss.29,39 Although
these studies point to a possible link between
thrombophilia, pre-eclampsia, abruption, and intrauterine
growth restriction, interpretation is limited by the
observational design. However, some case-control studies
have examined activated protein C resistance and factor
V Leiden.
Dizon-Townson and colleagues40 case-control study
showed that 8·9% of women with severe pre-eclampsia
were heterozygous for the factor V Leiden mutation,
compared with 4·2% of controls. In another case-control
study, Grandone and colleagues41 reported a prevalence
of 10·5% for factor V Leiden in women with pre-
eclampsia versus 2·3% in controls, which gave an odds
ratio of 4·9 (1·3–18·3). These researchers also assessed
underlying homozygosity for the thermolabile variant of
MTHFR: 29·8% of the pre-eclamptic patients had this
variant, compared with 18·6% of controls, with an odds
ratio of 1·8 (1·0–3·5). This finding suggests that
hyperhomocysteinaemia may be a risk factor for pre-
eclampsia, presumably because of the toxic effects that
homocysteine has on the endothelium. If so, it may be
useful to measure folic acid, and vitamin B6 and B12
in women with adverse pregnancy outcome, since a
deficiency of these vitamins can lead to acquired
hyperhomocysteinaemia, which is treatable with folic acid
and vitamin B6 supplements.42
In a case-control study by Lindoff and colleagues43
22% of the 50 women with previous pre-eclampsia had
activated protein C resistance, compared with 10% of the
50 controls. There was no difference in concentrations of
protein C or antithrombin concentrations between the
groups and no association between intrauterine growth
restriction and activated protein C resistance. However,
Lindqvist and colleagues 44 found that 18% of 122 women
THE LANCET • Vol 353 • April 10, 1999
Odds ratio (95% CI)*
All spontaneous fetal losses Miscarriages Stillbirths
2·1 (1·2–3·6) 1·7 (1·0–2·8) 5·2 (1·5–18·1)
1·4 (0·9–2·2) 1·4 (0·9–2·2) 2·3 (0·6–8·3)
1·3 (0·8–2·1) 1·2 (0·7–1·9) 3·3 (1·0–11·3)
1·0 (0·6–1·7) 0·9 (0·5–1·5) 2·0 (0·5–7·7)
2·0 (0·5–8·1) 0·8 (0·2–3·6) 14·3 (2·4–86·0)
with pre-eclampsia, intrauterine growth restriction, or
both had factor V Leiden, compared with 10% of 465
healthy pregnant controls. In their case-control study of
women with pregnancy complications (pre-eclampsia,
abruption, growth restriction, and stillbirth), Kupferminc
and colleagues45 reported odds ratios of 3·7 (1·5–9·0), 3·1
(1·4–7·1), and 3·9 (1·1–14·6) in women with factor V
Leiden, homozygosity for the thermolabile variant of
MTHFR, and the prothrombin gene variant, respectively.
These data suggest that factor V Leiden and
homozygosity for the thermolabile variant of MTHFR
predispose women to pre-eclampsia. Of interest is the
fairly high rate (2–10%) of the factor V Leiden gene in
the populations studied. Why such a genotype persists in
the population warrants further investigation. Analogies
have been made with sickle-cell disease, in which the
presence of sickle-cell trait affords protection against
malaria. Lindqvist and colleagues44 reported a fall in the
number of women with a blood loss at vaginal delivery of
greater than 600 mL and the development of postpartum
anaemia in the women with factor V Leiden.
Activated protein C resistance can be acquired during
pregnancy, probably because of high concentrations of
factor VIII and factor V.10 However, acquired resistance
does not account for all of the association of activated
protein C resistance with pre-eclampsia. The disturbance
of the coagulation system in pre-eclampsia means that
vascular damage may result from the combination of
an underlying thrombophilic disorder, the physiological
changes in activated protein C resistance in pregnancy,
and the pathological changes of pre-eclampsia.
Screening for congenital thrombophilia
There is no evidence to support routine screening of
all pregnant women for congenital thrombophilia. The
natural history of many of the congenital thrombophilias
in individuals is not known, nor is therapy for them.
Moreover, the cost of screening is likely to be high. There
is no doubt that screening is indicated for women with
venous thromboembolism in pregnancy or who have a
personal or family history of this disorder. About 50% of
such women will be found to have a thrombophilic
defect. Although the need for antenatal thrombo-
prophylaxis may be uncertain for some disorders for
example, symptomless factor V Leiden, the consensus is
that women with a personal history of venous thrombo-
embolism and an underlying thrombophilic trait should
receive thromboprophylaxis during pregnancy (see
below).
There are no data to suggest that screening for
congenital thrombophilia in women with recurrent first-
trimester fetal loss is justified, but patients with recurrent
pregnancy loss, including a second-trimester miscarriage
or a history of an intrauterine death, should be screened.
Screening is also justified for women with severe or
1261
recurrent intrauterine growth restriction or pre-eclampsia.
However, in a patient with recurrent pregnancy loss
and an underlying thrombophilic trait management is
uncertain. Some extrapolation can be made from the
benefits of aspirin plus heparin in women with
antiphospholipid antibody syndrome, but controlled trials
are needed to assess the effectiveness of such intervention
for women who screen positive.
Anticoagulants and antithrombotic techniques
Unfractionated and low-molecular-weight heparins do
not cross the placenta and, so, there is no risk of
teratogenesis or fetal haemorrhage. These drugs are not
secreted in breastmilk and can be used during lactation.
The disadvantages are osteoporosis, heparin-induced
thrombocytopenia, and allergy.46,47
Warfarin is not secreted in breastmilk in clinically
significant amounts and is safe to use during lactation.
However, this drug crosses the placenta, and warfarin
embryopathy may occur in 4–5% of fetuses exposed to
the drug between 6 and 9 weeks of gestation.48 Warfarin
embryopathy can be prevented by substitution of heparin
for warfarin during the first trimester. Central-nervous-
system abnormalities reported with warfarin exposure in
the fetus may be due to spontaneous intracerebral
bleeding. Since the liver is immature and concentrations
of vitamin-K-dependent coagulation factors are low in
the fetus, maternal warfarin therapy maintained in the
therapeutic range is associated with excessive anti-
coagulation and potential bleeding in the fetus. Warfarin
should, particularly, be avoided beyond 36 weeks of
gestation,48,50 because of the excessive risk of haemorrhage
to both mother and fetus in the peripartum period.
Dextran has been used for peripartum thrombo-
prophylaxis, particularly during caesarean section, and
carries a substantial risk of anaphylactic and anaphylactoid
reactions. Of greater concern in pregnancy is the risk of
maternal anaphylactoid reactions associated with uterine
hypertonus, profound fetal distress, and a high incidence
of fetal death or profound neurological damage.51 Thus,
dextran should be avoided in pregnancy.
Since thromboembolic-deterrent stockings are effective
in women who are not pregnant, they might be useful
in pregnancy. These stockings stop overdistenstion
of veins and so prevent endothelial damage and exposure
of subendothelial collagen.52 Other mechanical
techniques such as intermittent pneumatic compression
are of value during caesarean section and immediately
postpartum.
Aspirin can prevent DVT,53 and its effectiveness in
pregnancy, compared with heparin, remains to be
established. The effectiveness of aspirin is likely to be less
than that of heparin and low-molecular-weight heparin.
In women who are unable to take heparin or in whom the
balance of risk is not deemed sufficient to merit heparin,
aspirin may be useful. Low doses (60–75 mg daily) of
aspirin are not associated with adverse pregnancy
outcome. Hirudin is used in women who are not
pregnant for treatment of thrombocytopenia induced by
heparin, but is also used for postoperative prophylaxis.
Since hirudin crosses the placenta it should not be used
in pregnancy; its use after delivery has not been assessed.
The best option seems to be the established
thromboprophylactic agents.
1262
Management of venous thromboembolism
The objective diagnosis of venous thromboembolism
during pregnancy is crucial. The diagnosis has serious
implications not only for management of the pregnancy,
but also for other features of the woman’s life, such
as contraception and management of future pregnancies.
All women with symptoms or signs compatible with
venous thromboembolism must have appropriate
investigation.
Most events occur outside of hospital and carers of
pregnant women in the community must be aware of this
possibility and the underlying risk factors for venous
thromboembolism. Real time or duplex ultrasonography
is the first-line diagnostic tool,54 particularly because
most cases of DVT are iliofemoral. If the initial
investigation is negative and there is continuing concern
about possible DVT, then the women should undergo
repeat ultrasonography. If pulmonary thromboembolism
is suspected then ventilation-perfusion lung scan should
be done; in the first instance, perfusion scanning can
be used alone. Lung scans should be combined with
ultrasound venography. The radiation dose to the
fetus from limited venography with abdominal
shielding, ventilation-perfusion lung scan, and chest
radiography, even taken together, is small and deemed to
pose a negligible risk. If the diagnosis of pulmonary
thromboembolism remains uncertain after lung scan,
the results of ultrasonography are useful. If positive,
anticoagulant therapy should be started. Pulmonary
angiography is the gold standard for diagnosis of
pulmonary thromboembolism but is rarely necessary,
although it should be considered in cases with a high
clinical probability of the disorder and inconclusive test
reports. If there is a persistent clinical suspicion of venous
thromboembolism in the face of a negative or low-
probability test result, treatment should be started and
tests repeated within 7 days and therapy discontinued if
tests remain negative.
Therapeutic doses of unfractionated or low-molecular-
weight heparin are used, 55 with thromboembolic deterrent
stockings, and continued throughout the pregnancy.
Twice daily subcutaneous administration is appropriate
and the patient should be taught how to self-inject.
The activated partial thromboplastin time is unreliable for
monitoring doses of unfractionated heparin in pregnancy
because of the pregnancy-associated increase in factor
VIII,51,55 so antifactor Xa concentrations may offer a
better alternative way to measure heparin concentrations
in pregnancy. The dose of low-molecular-weight heparin
depends on maternal weight, and although monitoring
is not required in the women who are not pregnant,
measurement of peak antifactor Xa concentrations
(target range 0·4–1·0 m/mL) may be of value until more
experience is gained of low-molecular-weight heparin
in pregnancy.56 The dose of unfractionated and low-
molecular-weight heparin is reduced during delivery
to prophylactic concentrations and the timing of
administration is adjusted to allow epidural or spinal
anaesthesia. Warfarin can be used postpartum,
particularly to avoid the risk of osteoporosis associated
with heparin, although many women prefer to stay on
low-molecular-weight heparin. Treatment should
continue for at least 6 weeks and 3 months is often more
appropriate.
THE LANCET • Vol 353 • April 10, 1999
Thromboprophylaxis in pregnancy and the
puerperium
All patients with a personal or family history of venous
thromboembolism should be considered for antenatal
prophylaxis and be screened for a thrombophilia.
Thromboprophylaxis is also indicated when there are
additional risk factors, such as hyperemesis or surgery or
when patients are obese or immobile, particularly if they
also have pre-eclampsia or concurrent medical conditions
associated with thrombosis, such as nephrotic syndrome,
inflammatory bowel disease, or infection. If previous
DVT is the only risk factor, antenatal anticoagulation is
controversial, because of the potential hazards of heparin,
although use of low-molecular-weight heparin reduces
the hazards. However, controlled trials are needed to
guide management.
Anticoagulant therapy for thromboprophylaxis in
pregnancy usually consists of unfractionated heparin
in a dose of 10 000 IU twice daily during the latter half of
pregnancy, or for low-molecular-weight heparin, 40 mg
enoxaparin daily or 5000 IU dalteparin daily. There is no
need to adjust the dose of low-molecular-weight heparin
during pregnancy.45,57 Women with a bodyweight of
less than 50 kg are likely to have satisfactory heparin
concentrations on low doses (20 mg enoxaparin, 2500 IU
dalteparin),58 whereas obese women (>80 kg) may require
higher doses.
In women with multiple thrombotic events during
previous pregnancies, antenatal prophylaxis should start
at least 4–6 weeks in advance of the gestation at which
the previous events occurred. If previous thrombosis was
not associated with pregnancy, then prophylaxis should
start by about the mid point (20 weeks) of pregnancy or
sooner, particularly if additional risk factors are present.
Indeed, with the low incidence of side-effects with
low-molecular-weight heparin there is a case to start
prophylaxis from much earlier in pregnancy. Since the
risk of venous thromboembolism is likely to be high in
patients with past thromboses, low-molecular-weight
or unfractionated heparin should be combined with
graduated elastic compression stockings. If heparin is
contraindicated, graduated elastic compression stockings
and low-dose aspirin may be used.
After delivery, thromboprophylaxis should be
continued for a minimum of 6 weeks, but in patients with
severe thrombocytic problems, for 3 months. Patients on
heparin must be told of the risk of osteoporosis with
heparin. If the patient wants to change to warfarin it can
be started 48 h after delivery with the heparin continued
until the international normalised ratio is between 2
and 2·5. Unfractionated heparin is prescribed in a dose of
7500–10 000 IU twice daily, and low-molecular-weight-
heparin such as enoxaparin or dalteparin in doses of
40 mg or 5000 IU daily, respectively. Gibson and
colleagues59 reported that low-molecular-weight heparin
provided better antifactor Xa concentrations than did
unfractionated heparin in conventional doses after
caesarean section.59
Many patients with underlying congenital or acquired
thrombophilia will require antenatal prophylaxis, the
timing of which will depend on the patient’s history and
thrombophilic disorder.14 In patients with symptomatic
protein C deficiency, protein S deficiency, and factor V
Leiden, standard treatment is unfractionated or low-
molecular-weight heparin antenatally in the doses
THE LANCET • Vol 353 • April 10, 1999
discussed above. In women with antithrombin deficiency,
the risk may be substantially greater and the dose of
heparin needs to be adjusted throughout the pregnancy.
Antithrombin preparations can be added at times of
particularly high risk, such as delivery. Symptom-free
carriers of thrombophilia require special consideration. Since
the risk varies with the type of thrombophilia, all patients
with thrombophilia should be referred to a unit that
specialises in the management of thromophilia in pregnancy.
During labour thromboprophylaxis should be
continued, but because of the risk of epidural haematoma
associated with heparin, the timing of the insertion (and
removal) of the epidural catheter, caesarean section, or
heparin dose may need to be adjusted to avoid peak
heparin concentration at the time of placement of the
catheter.50 There have been postmarketing reports in the
US Food and Drug Administration of low-molecular-
weight heparin and epidural haematoma. Most of these
events have occurred in elderly women (median age 75
years) undergoing orthopaedic surgery. Additional risk
factors have included concomitant use of non-steroidal
anti-inflammatory agents or multiple puncture attempts
at spinal or epidural. The true incidence of epidural
haematoma is impossible to determine because of lack of
denominator data. In addition, the dose of enoxaparin
given in Europe is 40 mg daily, compared with 30 mg
twice daily in North America. However, caution is
needed in giving epidural anaesthesia in patients on low-
molecular-weight heparin. There must be vigilence for
signs of cord compression. Epidural anaesthesia must be
avoided in patients on therapeutic anticoagulation. In
women on prophylactic doses of unfractionated and low-
molecular-weight heparin, epidural anaesthesia should be
avoided around the time of peak-heparin concentrations.
Thus placement of the epidural catheter must be delayed
for 4 h with unfractionated heparin and longer with low-
molecular-weight heparin the delay is arbitrary, since it is
broadly extrapolated from the time of peak
concentrations and pharmacokinetics of the drugs. In my
unit the delay is at least 6 h, but others suggest a delay of
12 h. 60 Unfractionated and low-molecular-weight heparin
can be restarted 2 h after catheter removal.
Knowledge of thromboprophylaxis for women with no
personal or family history of venous thromboembolism,
but with congenital thrombophilia and a history of fetal
loss or pre-eclampsia is limited. Aspirin is not beneficial
in the prevention of pre-eclampsia. The combination of
low-dose aspirin and heparin is effective in preventing
recurrent fetal loss in women with anticardiolipin
antibodies,4 and could be considered for women with
congenital thrombophilia and fetal loss, pre-eclampsia, or
both.
Caesarean section, particularly emergency caesarean
section, carries substantial increased risk of
thromboembolic disease and guidelines have been drawn
up for management according to risk.61 Patients at low
risk are those who undergo elective caesarean section
with an uncomplicated pregnancy and no additional risk
factors. Women at moderate risk are those who undergo
caesarean section and have another risk factor such as age
older than 35 years or obesity, or if the caesarean section
is an emergency procedure. Women at high risk are those
with multiple risk factors or severe disorders such as
thrombophilia or paralysis of the lower limbs. These
women should receive unfractionated or low-molecular-
weight heparin thromboprophylactically; duration of
1263
therapy depends on the magnitude of the risk. In view of
the increase in death after vaginal delivery, guidelines
should also be developed for the management of women
with risk factors after vaginal delivery.
Conclusion
Thrombosis in pregnancy and underlying thrombophilia
have serious implications for mother and fetus. Evidence
is accumulating to link congenital thrombophilia to fetal
loss, intrauterine growth restriction, pre-eclampsia and
maternal venous thromboembolism. However, the
knowledge is based mostly on case-control and cohort
studies, so the natural history of these conditions in
symptom-free individuals is uncertain and large-scale,
prospective, longitudinal studies are needed. Only when
the implications of thrombophilic defects are known,
can appropriate intervention be assessed, ideally in large
randomised trials. In the meantime, women with a
personal or family history of venous thromboembolism,
second-trimester fetal loss, intrauterine death, and severe
or recurrent intrauterine growth restriction and pre-
eclampsia, should be screened for thrombophilia and
consideration given to thromboprophylaxis.
References
1 Department of Health, Welsh Office, Scottish Home and Health
Department and Department of Health and Social Services Northern
Ireland. Confidential Enquiries into Maternal Deaths in the United
Kingdom 1994–97. London: TSO, 1998.
2 Greer IA, de Swiet M. Thrombosis prophylaxis in obstetrics and
gynaecology. Br J Ostet Gynaecol 1993; 100: 37–40.
3 Vincent T, Rai R, Regan L, Cohen H. Increased thrombin generation
in women with recurrent miscarriage. Lancet 1998; 353: 116.
4 Rai R, Cohen H, Dave M, Regan L. Randomised controlled trial of
aspirin and aspirin plus heparin in pregnant women with recurrent
miscarriage associated with phospholipid antibodies (or
antiphospholipid antibodies). BMJ 1997; 314: 253–57.
5 Greaves M. Antiphospholipid antibodies and thrombosis. Lancet
(in press).
6 Macklon NS, Greer IA. Venous thromboembolic disease in obstetrics
and gynaecology: the Scottish experience. Scot Med J 1996; 41: 83–86.
7 Rutherford S, Montoro M, McGhee W, Strong T. Thromboembolic
disease associated with pregnancy: an 11 year review. Am Obstet
Gynecol 1991; 164 (suppl): 286.
8 McColl M, Ramsay JE, Tait RC, et al. Risk factors for pregnancy
associated venous thromboembolism. Thromb Haemost 1997; 78:
1183–88.
9 Lindhagen A, Bergqvist A, Bergqvist D, Hallbook T. Late venous
function in the leg after deep venous thrombosis occurring in relation
to pregnancy. Br J Obstet Gynaecol 1986; 93: 348–52.
10 Clark P, Brennand J, Conkie JA, McCall F, Greer IA, Walker ID.
Activated protein C sensitivity, protein C, protein S and coagulation
in normal pregnancy. Thromb Haemost 1998; 79: 1166–70.
11 Greer IA. Haemostasis and thrombosis in pregnancy. In: Bloom AL,
Forbes CD, Thomas DP, Tuddenham EGD, eds. Haemostasis and
thrombosis. Edinburgh: Churchill Livingstone, 1994: 987–1015.
12 Macklon NS, Greer IA, Bowman AW. An ultrasound study of
gestational and postural changes in the deep venous system of the leg
in pregnancy. Br J Obstet Gynaecol 1997; 104: 191–97.
13 Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet
1999; 353: 1167–73.
14 Walker ID. Congenital thrombophilia. In: Greer IA, ed. Bailliere’s
clinical obstetrics and gynaecology—thromboembolic disease in
obstetrics and gynaecology. London: Bailliere Tindall, 1997: 431–45.
15 Zoller B, Holm J, Dahlback B. Resistance to activated protein C due
to factor V gene mutation: the most common inherited risk factor of
thrombosis. Trends Cardiovasc Med 1996; 6: 45.
16 Bokarewa MI, Wramsby N, Bremme K, Blomback M. Variability of
the response to activated protein C during normal pregnancy. Blood
Coag Fibrinol 1997; 8: 239.
17 Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common
genetic variation in the three untranslated region of the prothrombin
gene is associated with elevated plasma prothrombin levels and an
increase in venous thrombosis. Blood 1996; 88: 36–38.
18 McColl MD, Walker ID, Greer IA. A mutation in the prothrombin
1264
gene contributing to venous thrombosis in pregnancy. Br J Obstet
Gynaecol 1998; 105: 923–25.
19 Den Heijer M, Koster T, Blom HJ, et al. Hyperhomocysteinemia as a
risk factor for deep vein thrombosis. N Engl J Med 1998; 334: 759–62.
20 Conard J, Horellou MH, van Dreden P, Le Compte T, Samama M.
Thrombosis in pregnancy and congenital deficiencies in AT III,
protein C or protein S: study of 78 women. Thromb Haemost 1990;
63: 319–20.
21 de Stefano V, Leone G, Masterangela S, et al. Thrombosis during
pregnancy and surgery in patients with congenital deficiency of anti-
thrombin III, protein C–protein S. Thromb Haemost 1994; 71:
799–800.
22 Hough RE, Makris N, Preston FR. Pregnancy in women with
thrombophilia: incidence of thrombosis and pregnancy outcome.
Br J Haematol 1996; 93 (suppl 2): 136 (abstr).
23 Pabinger I and the Study Group on Natural Inhibitors. Thrombotic
risk in hereditary anti-thrombin III, protein C or protein S deficiency.
Arter Thromb Vasc Biol 1996; 16: 742–48.
24 Hallak M, Senderowica J, Cassel A, et al. Activated protein C
resistance (factor V Leiden) associated ith thrombosis in pregnancy.
Am J Obstet Gynecol 1997; 176: 889–93.
25 Bokarewa MI, Bremme K, Blomback M. Arg 506-Gln mutation in
factor V and risk of thrombosis during pregnancy. Br J Haematol 1996;
92: 473–78.
26 de Stefano V, Leone G, Masterangelo S, et al. Thrombotic risk during
pregnancy and puerperium in women with APC resistance: effective
subcutaneous heparin prophylaxis in a pregnant patient. Thromb
Haemost 1995; 74: 793–94.
27 Rai RS, Regan L, Hadley E, Dave M, Cohen H. Second trimester
pregnancy loss is associated with activated Protein C resistance. Br J
Haematol 1996; 92: 489–90.
28 Grandone E, Margaglione M, Colaizzo D, et al. Factor V Leiden is
associated with repeated and recurrent explained fetal losses. Thromb
Haemostas 1997; 77: 822–24.
29 Brenner B, Mandel H, Lanir N, Younis J, Rothbart H. Activated
protein C resistance can be associated with recurrent fetal loss. Br J
Haematol 1997; 97: 551–54.
30 Dizon-Townson DS, Meline L, Nelson LM, Barner M, Ward K. Fetal
carriers of the factor V Leiden mutation are prone to miscarriage and
placental infarction. Am J Obstet Gynecol 1997; 177: 402–05.
31 Dizon-Townson DS, Kinney S, Branch DW, Ward K. The factor V
Leiden mutation is not a common cause of recurrent miscarriage.
J Reprod Immunol 1997; 34: 217–23.
32 Belasch J, Riverter JC, Fabregues F, et al. First trimester repeated
abortion is not associated wtih activated protein C resistance. Hum
Reprod 1997; 12: 1094–97.
33 Rai RS, Regan L, Chitolie A, Donald JG, Cohen H. Placental
thrombosis in second trimester miscarriage in association with activated
Protein C resistance. Br J Obstet Gynaecol 1996; 103: 842–44.
34 Preston FE, Rosendaal FR, Walker ID, et al. Increased fetal loss in
women with heritable thrombophilia. Lancet 1996; 348: 913–16.
35 Sanson BJ, Friederich PW, Simioni P, et al. The risk of abortion and
stillbirth in antithrombin, Protein S and Protein S deficient women.
Thromb Haemost 1996; 75: 387–88.
36 Havercate F, Samama M. Familial dysfibrinogenaemia and
thrombophilia—report on a study of the SSC sub-committee on
fibrinogen. Thromb Haemostas 1995; 73: 151–61.
37 Dekker GA, de Vries JIP, Doelitzsch PM, et al. Underlying disorders
associated with severe early onset of pre-eclampsia. Am J Obstet
Gynecol 1996; 173: 104–45.
38 De Vries JIP, Dekker J, Huijgens PC, Jakobs C, Blomberg BME, van
Geijn HP. Hyperhomocysteinaemia and Protein S deficiency in
complicated pregnancies. Br J Obstet Gynaecol 1997; 104: 1248–54.
39 Rotmensch S, Liberati M, Mittlemann M, Ben-Rafael Z. Activated
protein C resistance and adverse pregnancy outcome. Am J Obstet
Gynecol 1997; 177: 170–73.
40 Dizon-Townson DS, Nelson LM, Easton K, Ward K. The factor V
Leiden mutation may predispose women to severe pre-eclampsia.
Am J Obstet Gynaecol 1996; 175: 902–05.
41 Grandone E, Margagilione M, Colazzo D, et al. Factor V Leiden.
C>TMTHFR polymorphism and genetic susceptibility to pre-
eclampsia. Thromb Haemostas 1997; 77: 1052–54.
42 Leeda M, Riyazi N, de Vries JIP, Jakobs C, van Geijn HP,
Dekker GA. Effects of folic acid and vitamin B6 supplementation on
women with hyperhomocysteinemia and a history of pre-eclampsia or
fetal growth restriction. Am J Obstet Gynecol 1998; 179: 136–39.
43 Lindoff C, Ingemarsson I, Martinsson G, Segelmark M, Thysell H,
Astedt B. Pre-eclampsia is associated with a reduced response to
activated protein C. Am J Obstet Gynecol 1997; 176: 457–60.
44 Lindqvist PG, Svensson PJ, Dahlback B, Masal K. Factor V Q506
mutation (activated protein C resistance) associated with reduced
intrapartum blood loss—a possible evolutionary selection mechanism.
Thromb Haemost 1998; 79: 69–73.
THE LANCET • Vol 353 • April 10, 1999
45 Kupferminc MJ, Amiram E, Steinman N, et al. Increased frequency of
genetic thrombophilia in women with complications of pregnancy.
N Engl J Med 1999; 1: 9–13.
46 Macklon NS, Greer IA, Reid AW, Walker ID. Thrombocytopenia
antithrombin deficiency and extensive thromboembolism in
pregnancy—treatment with low molecular weight heparin. Blood
Coag Fibrinol 1995; 6: 672–75.
47 Nelson-Piercy C. Hazards of heparin: allergy, heparin-induced
thrombocytopenia and osteoporosis. In: Greer IA, ed. Bailliere’s
clinical obstetrics and gynaecology—thromboembolic disease in
obstetrics and gynaecology. London: Bailliere Tindall, 1997: 489–509.
48 Bates SM, Ginsberg JS. Anticoagulants in pregnancy: fetal defects. In:
Greer IA, ed. Bailliere’s clinical obstetrics and gynaecology—
thromboembolic disease in obstetrics and gynaecology. London:
Bailliere Tindall, 1997: 479–88.
49 Warkentin TE, Levine MN, Hisch J, et al. Heparin induced
thrombocytopenia in patients treated wtih low molecular weight
heparin or unfractionated heparin. N Engl J Med 1995; 332: 1330–35.
50 Letsky E. Peripartum prophylaxis of thromboembolism. In: Greer IA,
ed. Bailliere’s clinical obstetrics and gynaecology—thromboembolic
disease in obstetrics and gynaecology. London: Bailliere Tindall, 1997:
23–43.
51 Greer IA. Epidemiology, risk factors, and prophylaxis of venous
thromboembolism in obstetrics and gynaecology. In: Greer IA, ed.
Bailliere’s clinical obstetrics and gynaecology—thromboembolic
disease in obstetrics and gynaecology. London: Bailliere Tindall, 1997:
403–30.
52 Macklon NS, Greer IA. Technical note: compression stockings and
posture—a comparative study of their effects on the proximal deep
veins in the leg at rest. Br J Radiol 1995; 68: 515–18.
Essay
Helen
Peter Reagan
She pointed to her left cheek and said, “You can kiss me
now. Right here”. Taken by surprise, and extremely
moved, I kissed her lightly and sat back. She smiled and
looked up brightly. Afterwards she asked me for some
brandy. Beth had diluted it fifty-fifty with a soft drink.
She wasn’t used to drinking much alcohol, and she
choked on it just a little as it went down. Beth offered her
a mint, but she waved it away and said she was sleepy.
Douglas held her right hand and arm, as I held her left.
Beth massaged her feet. Her breathing was slow, even,
and relaxed. About a minute later I asked her how she
was doing and she mumbled, “Tired”. We took two of
the pillows from behind her head so she could sit back
more easily, and Helen fell into a deep sleep.
The three of us sat around her bed talking quietly
about the emotional struggle we’d each been through. 3
weeks earlier, I had received a phone call from a retired
colleague about a patient who wanted help ending her life
under the new Oregon Death With Dignity Law (see
Lancet 1994; 344: 1493–94). Her previous physicians had
not been prepared to offer a life-ending prescription.
Could I see her soon?
5 days later she was sitting in a wheelchair in an
examination room with her son Douglas and her
daughter Beth. She was elderly, frail, and hooked to
Lancet 1999; 353: 1265–67
2406 NE 19th Avenue, Portland, OR 97212, USA (P Reagan MD)
THE LANCET • Vol 353 • April 10, 1999
53 Thrombombolic risk factors (THRIFT) Consensus Group. Risk of
and prophylaxis for venous thromboembolism in hospital patients.
BMJ 1992; 305: 567–74.
54 Macklon NS. Diagnosis of deep venous thrombosis and pulmonary
embolism. In: Greer IA, eds. Bailliere’s clinical obstetrics and
gynaecology—thromboembolic disease in obstetrics and gynaecology.
London: Bailliere Tindall, 1997: 463º77.
55 Lowe GDO. Treatment of venous thromboembolism. In: Greer IA,
ed. Bailliere’s clinical obstetrics and gynaecology—thromboembolic
disease in obstetrics and gynaecology. London: Bailliere Tindall, 1997:
511–21.
56 Thomson AJ, Walker ID, Greer IA. Low molecular weight heparin for
immediate management of thromboembolic disease in pregnancy.
Lancet 1998; 352: 1904.
57 Brennand JE, Walker ID, Greer IA. Anti-activated factor X profiles in
pregnant women receiving antenatal thromboprophylaxis with
Enoxaparin. Acta Haematol 1999; 101: 53–55.
58 Blomback M, Bremme K, Hellgren M, Lindberg H. A
pharmacokinetic study of dalteparin (Fragmin) during late pregnancy.
Blood Coag Fibrinol 1998; 9: 343–50.
59 Gibson JL, Ekevall K, Walker ID, Greer IA. Puerperal
thromboprophylaxis: comparison of the anti-Xa activity of enoxaparin
and unfractionated heparin. Br J Obstet Gynaecol 1998; 105: 795–97.
60 Horlocker TT, Heit JA. Low molecular weight heparin:
biochemistry, pharmacology, perioperative prophylaxis regimens and
guidelines for regional anesthetic management. Anesth Analg 1997; 85:
874–85.
61 Report of the RCOG Working Party on Prophylaxis Against
Thromboemolism in Gynaecology and Obstetrics. London: Royal
College of Obstetricians and Gynaecologists, 1995.
oxygen, but completely astute, with a delightful twinkle; I
liked her immediately. She and her family carefully
detailed her history of breast cancer, her years since her
first mastectomy, the recurrence on the other side, the
metastases, and the downhill course. The second tumour
had developed when she was 82; she had refused all
treatment for it after the lumpectomy. Her husband had
suffered a lingering death a decade earlier; she was opting
for life while it was good and then a clean exit.
I discussed the requirements of the law with her. When
I got to the part about a written request, she handed me a
notarised statement signed by Douglas and a neighbour. I
looked into her eyes and shivered as I contemplated the
enormity of what I was facing. I took a deep breath and
we started discussing her attempts to get help with dying.
She had asked her original physician, but he was about to
relocate his practice and didn’t want to be involved. He
had recommended hospice care and had referred her to a
second primary physician who, he thought, might have
been open to aid in dying. The second physician had seen
her twice, ordered a chest radiograph, and also
encouraged hospice care. He agreed that she had a short
life-expectancy but recorded in her notes that she was
probably depressed. Helen and her family felt he was not
ready to provide assistance and was buying time. She
said, “He made me wait 24 days, and he never intended
to help me. I think he still expected me to come back to
him. He should have told me at the beginning”.
1265