Swedish obstetric thromboprophylaxis guideline: background and update

Abstract

Risk estimation concerning venous thromboembolism (VTE) and thromboprophylaxis for those at risk is routine in pregnancy. For 20 years, Swedish obstetricians have followed a weighted-risk algorithm guideline for risk estimation, based on which patient selection, timing, duration and dosage of thromboprophylaxis are determined. This article presents the latest update, the basis for the algorithm and its application for assessing moderate- to high obstetric VTE risk, defined as equal or greater absolute risk per time unit than the antepartum risk of women with one prior VTE. The risk score is based on risk factors conferring approximately fivefold increased risk of VTE or a multiple thereof. This article also presents algorithm efficacy data and describes lifestyle advice provided to patients. In our experience, the Swedish guideline for obstetric VTE risk estimation is easy to follow. It helps identify women at high risk. The risk of under- or overtreatment is thus minimised.

Keywords:

• Pregnancy
• thromboembolism
• guidelines
• lifestyle
• obstetric thromboprophylaxis

Introduction

Between 10 and 15 per 10,000 women contract pregnancy-related venous thromboembolism (VTE) and half of VTE cases occur after delivery (Lindqvist et al. 1999, Jacobsen et al. 2008a). VTE is one of the major causes of maternal mortality in Western countries (Knight 2008, Marik and Plante 2008, MacGregor et al. 2022), and women with a prior VTE have approximately 10% risk of pregnancy-related recurrent VTE if not given thromboprophylaxis (Lindqvist et al. 1999). There is consensus that women with at least 5% absolute risk of VTE during pregnancy should be recommended thromboprophylaxis. Of the many risk factors for obstetric VTE, some occur both antepartum and postpartum. These include personal or family history of VTE, thrombophilias, higher age (Lindqvist et al. 2009a) and high body mass index (BMI) (Lindqvist et al. 1999, Jacobsen et al. 2008b). Other factors confer risk mainly during the postpartum period, such as preeclampsia (Lindqvist et al. 1999, Sultan et al. 2016), placental abruption (Prochazka et al. 2003, Jacobsen et al. 2008a), caesarean section (CS) (Lindqvist et al. 1999), intrauterine foetal death (IUFD) (Sultan et al. 2016) and blood transfusion (Sultan et al. 2016, Thurn et al. 2018).

Some risk factors are associated with a risk too minor to be included in the algorithm, such as smoking (Lindqvist et al. 2009a), diabetes mellitus, hypoalbuminaemia, non-active sun exposure (Lindqvist et al. 2009b), multiple gestations, blood group other than O (Jacobsen et al. 2008a), varicose veins, sedentary lifestyle (Lindqvist et al. 2009a), emergency vs. elective CS (Macklon and Greer 1996) and profuse blood loss without transfusion (Thurn et al. 2018).

Obstetric thromboprophylaxis consists of daily subcutaneous injections of low-molecular-weight heparin (LMWH), recommended due to the absence of placental transfer, thus avoiding effect on the foetus. Lifestyle advice is provided and compression stockings are recommended (Chauleur et al. 2008, Dargaud et al. 2009, Lindqvist and Hellgren 2011, Nelson-Piercy et al. 2015, Bates et al. 2018). Several thromboprophylaxis guidelines have been published (Chauleur et al. 2008, Dargaud et al. 2009, Lindqvist and Hellgren 2011, Nelson-Piercy et al. 2015, Sultan et al. 2016, Bates et al. 2018). Although some divide pregnant women into different risk groups (Nelson-Piercy et al. 2015, Bates et al. 2018), others apply weighted risk scores (Chauleur et al. 2008, Dargaud et al. 2009, Thrombosis Canada 2023). Thus, obstetric thromboprophylaxis regimens differ in different countries, and the extent of implementation also depends on the size of the population at risk (5% to 40%). There is an increasing risk of VTE throughout pregnancy, and the risk peaks during the first weeks postpartum. In this article, we present the basis for the Swedish algorithm for VTE risk estimation and management in women diagnosed as at moderate to high risk of VTE, as well as the most recent guideline update. This guideline is not designed for women with very high risk of VTE such as antithrombin (AT) deficiency, mechanical heart valves, antiphospholipid syndrome (APS) or those on chronic anticoagulant therapy.

Methods

In 2002, we presented two different models of obstetric VTE risk estimation. One yielded an individually based, absolute VTE risk and was suitable for computerised assessment or inclusion in computerised medical records (Lindqvist, Kublikas, et al. 2002, Lindqvist, Olofsson, et al. 2002). The other was a weighted risk assessment based on several major risk factors associated with a fivefold increased risk or a multiple thereof (Lindqvist, Olofsson, et al. 2002, Lindqvist and Hellgren 2011). The reference and working group on haemostatic disorders in obstetrics and gynaecology (Hem-ARG) of the Swedish Society of Obstetrics and Gynecology (SFOG) chose the latter assessment model (Lindqvist and Hellgren 2011). The working group’s guideline was published and implemented in 2004 (Hem-ARG 2004). SFOG’s national guideline was published in 2009. The algorithm has since been revised several times; the 2023 version is presented in this article.

As mentioned earlier, the relative risk of VTE is calculated based on risk factors conferring approximately fivefold increased risk of VTE. Antepartum and postpartum risk factors include personal or family history of VTE, thrombophilias, higher age (Lindqvist et al. 2009a), high BMI (Lindqvist et al. 1999, Jacobsen et al. 2008b), immobilisation (Jacobsen et al. 2008b), inflammatory bowel disease (Bernstein et al. 2001, Grainge et al. 2010) and major infectious disease (James et al. 2006, Liu et al. 2009). Ovarian hyperstimulation syndrome (OHSS) after in vitro fertilisation is a major risk factor (100-fold increased risk) during the whole first trimester (see Table 1) (Rova et al. 2012, Henriksson et al. 2013).

Table 1. Risk points for venous thromboembolism (VTE), compiled to total risk score as basis for prescribing thromboprophylaxis.

Risk points	1p	2p	3p	> 4p^a
				High risk
Risk factors
	Heter FV Leiden mut	Prot S deficiency	Hom FV Leiden mut	Prior VTE
	Heter prothrombin mut	Prot C deficiency	Hom prothrombin mut	APS without prior VTE
	Age > 40 years	Immobilisation^b	Double mutation	OHSS^c
	BMI^d 30–40	BMI^d > 40
	Heredity for VTE^e
	Inflammatory bowel disease
	Hyperhomocysteinaemia^f		Very high risk^g
	Other major risk factor^h		Chronic anticoagulation^i
	Postpartum risk factors		Antithrombin deficiency
	Preeclampsia		Repeated VTEs
	Placental abruption		APS with VTE
	Caesarean section		Covid-19 pneumonia, required oxygen
	Blood transfusion
	Stillbirth
	Severe infection, sepsis

Heter: heterozygous; Hom: homozygous; mut: mutation; APS: anti-phospholipid syndrome; Prot: protein, FV: Factor V; DOAC: direct oral anticoagulants; ASA: acetylsalicylic acid

^a Prior VTE or APS without VTE yields risk score 4.

^b Fracture cast, strict bed rest or Covid 19 infection, short-term thromboprophylaxis.

^c OHSS, thromboprophylaxis during whole first trimester.

^d BMI early pregnancy.

^e VTE among 1st-grade relative aged < 50 years.

^f Homocysteine >8 µmol/L during pregnancy, risk of sinus thrombosis.

^g Women in this group are classified as very high risk, independent of other risk factors.

^h For example cancer, cancer treatment, SLE, heart disease, sickle cell trait or essential thrombocytosis.

ⁱ Warfarin, DOAC, LMWH, but not low-dose ASA.

Risk factors that confer risk mainly during the postpartum period include preeclampsia (Lindqvist et al. 1999, Sultan et al. 2016), placental abruption (Prochazka et al. 2003, Jacobsen et al. 2008b), CS (Lindqvist et al. 1999), IUFD (Sultan et al. 2016) and blood transfusion (see Table 1) (Sultan et al. 2016, Thurn et al. 2018). Rare risk factors, such as cancer, ongoing cancer treatment, systemic lupus erythematosus (SLE), heart disease, sickle cell trait and essential thrombocytosis, are not taken into account in the algorithm, but may be added at the discretion of the clinician, under ‘Other major risk factor’ (Table 1).

Small groups comprising special cases are not managed according to this guideline. These include women with AT deficiency, prior multiple VTEs, mechanical heart valves or APS with prior VTE, as well as those on chronic anticoagulant therapy. These women are assessed as having ≥15% absolute VTE risk (‘very high risk’) in relation to pregnancy and they are managed individually with ‘high-dose prophylaxis’. Moreover, pregnant women with Covid-19 infection requiring hospitalisation and oxygen therapy are at very high risk. All women at ‘very high risk’ are recommended twice-daily thromboprophylaxis, aiming at measurable anti-factor X activity (anti-FXa) 24 h per day.

Weighted risk factor-based algorithm

The algorithm for those at moderate to high risk, thus not applicable for ‘very high risk’ cases, is presented below. Risk points are compiled to yield a total risk score (see Table 1). Women with risk score 1 are estimated to have fivefold increased risk; risk score 2 (two variables with 1 point or one with 2 points) indicates 25-fold increased risk; risk score 3 indicates 125-fold increased risk and risk score ≥ 4 indicates 5% absolute antepartum risk.

The absence of a specific risk factor generates 0 points. Thus, the baseline risk is defined as a condition in which all included risk factors are negative. Compared to the total mean risk of 10 to 15/10,000 pregnancies, the baseline absolute risk has been estimated at 1.2/10,000 antepartum and 0.93/10,000 postpartum in 2002 (Lindqvist, Olofsson, et al. 2002). When all personal absolute risk estimates were summed up, the total corresponded to the number of VTEs in the study population in the same study (Lindqvist, Kublikas, et al. 2002). Women with a prior VTE or APS diagnosis receive a risk score of 4, regardless of other risk factors. All women with previous VTE, except central retinal vein thrombosis, are given thromboprophylaxis (Epstein et al. 2015). The model also takes into account the fivefold higher risk per time unit postpartum (Salonen Ros et al. 2001, Jacobsen et al. 2008b). Postpartum thromboprophylaxis of varying duration is recommended, e.g. one week for risk score 2 and at least six weeks for risk score 3 (Table 2).

Table 2. Management, based on risk score in Table 1.

Risk score = 1	No treatment. Lifestyle advice
Risk score = 2	Postpartum normal-dose prophylaxis at least 7 days and short-term, normal-dose prophylaxis during temporary risk factor
Risk score = 3	Postpartum normal-dose profylax for at least 6 weeks
Risk score > 4	Antepartum normal-dose prophylaxis beginning as soon as pregnancy is known, and until at least 6 weeks postpartum

Moderate- or high-dose prophylaxis.

Recommendations, ‘normal-dose’ thromboprophylaxis

The total risk score determines patient selection, timing, duration and dosage of thromboprophylaxis (Table 2). The basis of the algorithm is a recommendation of LMWH thromboprophylaxis at a risk level corresponding to the antepartum risk of women with one previous VTE, i.e. 0.4% of the pregnant population (Lindqvist, Olofsson et al. 2002). In the following, we will present ‘normal-dose’ thromboprophylaxis during pregnancy and postpartum in women at moderate to high risk of VTE (see Table 3 for doses).

Table 3. Dosage, low-molecular-weight heparin (LMWH).

	Weight^a (kg)	Dalteparin s.c. IE/24 h	Tinzaparin s.c. IE/24 h	Enoxaparin s.c. mg/24 h
Risk score = 2–4
Moderate to high risk	< 90	5,000	4,500	40
‘Normal-dose prophylaxis’	> 90	7,500^b	8,000^b	60^b
‘Very high risk’ of VTE^c
‘High-dose prophylaxis’ (twice daily)	<50	2,500 × 2^b	2,500 × 2^b	20 × 2^b
	50–90	5,000 × 2^b	4,500 × 2^b	40 × 2^b
	>90	7,500 × 2^b	4,500 + 8,000^b	60 × 2^b

^a Weight in early pregnancy.

^b Initial dose.

^c measurable anti-FXa activity before next dose, i.e. >0.05–0.1 IE/mL.

VTE, venous thromboembolism.

Women with risk score ≥ 4 usually initiate ‘normal-dose’ thromboprophylaxis as soon as they know they are pregnant. Women with risk scores 2 and 3 are started on thromboprophylaxis 4 h after delivery; no further laboratory testing is needed. LMWH injections are always combined with lifestyle advice and recommendations for compression stockings. Knee-high stockings are usually recommended, beginning in early pregnancy and until at least 12 weeks after delivery. Women with post-thrombotic syndrome are advised to use grade 2 compression stockings.

Upon initiation of LMWH treatment, platelet count, activated prothrombin time (APTT) and international normalized ratio (INR) are tested. Platelet count is repeated after two weeks in women previously treated with unfractionated heparin, for early identification of heparin-induced thrombocytopenia (HIT). Women weighing ≤90 kg (in early pregnancy) are administered a ‘normal dose’ once daily throughout pregnancy, without further laboratory test monitoring. For those weighing >90 kg, a higher initial dose is advised (see Table 3), and anti-FXa testing is recommended about two weeks after initiation. Anti-FXa is targeted at 0.20–0.45 U/mL 3 h post-injection. If indicated, the dose is increased or decreased by half a ‘normal dose’. If the first anti-FXa level is within range, no further testing is required, provided there are no complications and no abnormal weight gain. The recommended different doses are guided by maternal weight, based on experience and pharmacokinetic studies (Blomback et al. 1998, Norris et al. 2004). Women on antepartum thromboprophylaxis follow the course of routine maternal healthcare, supplemented by one extra visit scheduled at approximately 34 gestational weeks for planning delivery and puerperium care. The issue of epidural and spinal analgesia requires especial attention. Lifestyle advice, if not already given, is provided during this visit and postpartum.

In case of temporary risk situations, such as strict bed rest, fracture casts, severe infections or post-surgery, short-term thromboprophylaxis is advised. Women with OHSS are recommended thromboprophylaxis for the whole first trimester (Rova et al. 2012).

Thromboprophylaxis during delivery and postpartum

On arrival at the delivery ward, platelet count, APTT and INR are tested and the time of the last LMWH injection is noted. We aim for vaginal delivery with spontaneous onset. LMWH is discontinued in active labour and the next dose is scheduled at 4 h postpartum. To avoid abdominal wall haematomas during the first seven days after CS, we recommend that LMWH injections be given in the thigh. Thromboprophylaxis postpartum is administered according to Table 3, i.e. 1 week for risk score 2 and 6 weeks for risk score ≥ 3. In the event of complications, individualised management is advised. Heavy blood loss has priority over thromboprophylaxis, but prophylaxis should be (re-)initiated after stabilisation. If a change to warfarin is planned, it is usually initiated one to two weeks after delivery. Dosing of warfarin postpartum is more complicated than in other clinical situations and higher doses may be needed (Brooks et al. 2002). It is recommended that women with protein S or C deficiency be given 1 week of simultaneous warfarin and LMWH, to avoid paradoxical thrombosis. Thromboprophylaxis is also recommended 4 weeks after miscarriage, abortion or ectopic pregnancy, as well as 1 week after surgery such as appendectomy.

Discussion

Comments on individual risk factors

Obesity, BMI ≥ 40

Women with BMI ≥ 40 were overrepresented among fatal pulmonary embolism (PE) cases in the UK and Ireland audit ‘Saving lives, improving mothers’ care’ (8/37) (Tuffnell et al. 2018). Since the risk of PE is increased 15-fold among this population (OR 14.9), BMI > 40 is allotted 2 risk points (Larsen et al. 2007). It is recommended that these women be given, in early pregnancy, lifestyle counselling and information about VTE and PE symptoms (Tuffnell et al. 2018).

Previous VTE

In a prior version of the guideline, women with VTE after surgery or other temporary risk factors started thromboprophylaxis at 20 weeks of gestation. However, after several VTEs occurred during the first half of pregnancy among these women, we assessed the difference in risk level between permanent and temporary risk factors as too minor to warrant different clinical managements. Thus, antepartum thromboprophylaxis is advised, beginning as soon as pregnancy is recognised, for all women with a history of VTE, making the guideline easier to follow.

Higher maternal age

Higher age is a well-documented risk factor for VTE. However, maternal age > 40 is the weakest risk factor included in the algorithm, which is why maternal age was not set at >35, a common limit (Jacobsen et al. 2008a, Dargaud et al. 2009, Nelson-Piercy et al. 2015).

CS

The magnitude of CS as a risk factor differs between studies. We consider it to confer approximately fivefold VTE risk increase (Macklon and Greer 1996, Lindqvist et al. 1999). A lower risk is reported in studies in which a thromboprophylaxis programme was implemented, but LMWH use was not taken into consideration (Sultan et al. 2016). We consider the disparity in risk between emergency CS and elective CS too small to warrant different managements (Macklon and Greer 1996).

IUFD

Both UK and Swedish data indicate that IUFD is linked to fivefold increased risk of VTE (Sultan et al. 2016). In previous versions of the algorithm, we assumed this risk was a result of immobilisation due to grief and might best be addressed by lifestyle counselling. However, many women are not given appropriate lifestyle advice and we have thus added IUFD as a major risk factor.

Blood transfusion

Both blood transfusion and profuse postpartum haemorrhage (PPH) have been reported as independent VTE risk factors (Sultan et al. 2016). A recent study showed that blood transfusion is an independent major risk factor, but that PPH without blood transfusion is not (Thurn et al. 2018). The plausible explanation is that bleeding consists of whole blood, which is replaced by concentrated blood cells without plasma which would have included anticoagulant proteins (Thurn et al. 2018).

When to prescribe thromboprophylaxis

The guideline’s risk score is a clinical aid in avoiding under- and overtreatment with thromboprophylaxis. The percentage of women with risk score 2 based on anamnestic risk factors is 2.6% during pregnancy and 6.1% postpartum (Lindqvist, Olofsson, et al. 2002). The reason for this discrepancy is that several of the risk factors mainly increase the risk of VTE after delivery (Table 4). Based on the algorithm, 0.4% of all women will be given antepartum thromboprophylaxis. Thromboprophylaxis will be given to approximately 6% one week postpartum and to 0.9% six weeks postpartum (Table 3).

Table 4. Absolute risk of venous thromboembolism at different risk scores.

VTE risk score	Risk increase	Prevalence		Pregnancy, absolute risk	Postpartum, absolute risk
				VTE/10,000	VTE/10,000
		AP	PP
0p	1	73%	66%	≈1.2	≈0.9
1p	5	17%	24%	≈6	≈5
2p	25	2.6%	6.1%	≈30	≈23
3p	125	0.2%	0.9%	≈160	≈120
4p, ‘High risk’	≈500	0.4%	0.4%	≈500	≈500
‘Very high risk’	> 750	<0.1%		> 750	> 700

AP: antepartum; PP: postpartum; p: points;

VTE : venous thromboembolism.

Efficacy of LMWH thromboprophylaxis

According to the algorithm, both antepartum and postpartum thromboprophylaxis prevent ∼88% of VTE cases in women with one prior VTE (Lindqvist et al. 2010). The highest risk for these women occurs 43 to 100 days after delivery, i.e. soon after termination of thromboprophylaxis (Lindqvist et al. 2010). After delivery, a considerable proportion of women without prior VTE have a risk corresponding to the antepartum risk among women with one prior VTE. Following the guideline has been calculated to reduce the risk of postpartum PE by two-thirds (Lindqvist et al. 2008). Since one-sixth of maternal mortality is caused by PE and two-thirds of all lethal PE occur after delivery, mortality might theoretically be reduced by one-twelfth (Lindqvist et al. 2008).

Thrombophilias

Women with protein C or protein S deficiency are allotted 2 points. Of these women, most who do not have a personal history of VTE do have a first-degree relative with VTE diagnosed at age < 50. Thus, they usually have a risk score of 3 (2 points for thrombophilia and 1 point for heredity). According to the algorithm, six weeks of thromboprophylaxis is recommended postpartum. In the absence of other risk factors, a heterozygous carrier of the Factor V Leiden (FVL) mutation (= APC resistance) is given risk score 1 and thromboprophylaxis is not recommended. However, in combination with first-degree heredity for VTE at age < 50, the same woman would be given risk score 2 and thromboprophylaxis would be recommended for one week after delivery. If she was also delivered by CS, the risk score would increase to 3 (FVL, heredity and CS), and six weeks of postpartum thromboprophylaxis would be recommended. Most women with homozygous FVL mutation (3 points) also have first-degree heredity for VTE at age < 50 (1 point), yielding a risk score of 4. In this case, thromboprophylaxis is recommended both antepartum and for six weeks postpartum.

For obstetric purposes, carriers of the prothrombin gene mutation can be treated the same way as those with an FVL mutation. Hyperhomocysteinaemia is not a thrombophilia but is often considered to be one because several polymorphisms in methionine metabolism may increase homocysteine levels. Hyperhomocysteinaemia is mainly a risk factor for sinus thrombosis, conferring a four- to tenfold risk increase (1 point) (Nagaraja et al. 2008, Saposnik et al. 2011), and is usually treated with folic acid 5 mg/day or with folic acid combined with vitamins B12 and B6.

Cost-effectiveness

An assessment has shown that the guideline is cost-effective, especially postpartum among women with risk scores 2 or 3. If the recommendations are followed, the cost of LMWH is calculated at less than half of that of treating thrombotic complications (Lindqvist et al. 2008).

Lifestyle advice

All women at risk should be given lifestyle counselling and be advised to wear compression stockings. Fertile-age women with a history of one previous VTE have ≈ 1% annual risk of VTE at age 30 years, rising to 5% per year at age 50 (Christiansen et al. 2010, Lindqvist et al. 2010). Half of all VTEs will occur during high-risk, and half during low-risk, situations. In low-risk situations, only lifestyle changes are possible. Exercise halves the VTE risk (Lindqvist et al. 2009a). A brisk daily walk of about 30 min is recommended. Water aerobics and swimming are other suitable alternatives that do not strain the pelvis. Active sun exposure habits seem to reduce the risk of VTE, in comparison to low sun-exposure habits (Lindqvist et al. 2009b). This may be an effect of higher levels of nitric oxide or higher vitamin D sufficiency among the latter (Lindqvist et al. 2022). The Swedish MISS study showed that the VTE risk is increased by 50% during winter, compared to other seasons (Lindqvist et al. 2009b). Smokers are also at higher risk of VTE and should be encouraged to quit (Lindqvist et al. 1999, Jacobsen et al. 2008b). Women with BMI 25–30 are at 30% increased risk of VTE, compared to those with BMI in the normal range (BMI 20–25). Obese women (BMI ≥ 30) are at approximately fivefold increased risk of VTE (Lindqvist et al. 1999, Jacobsen et al. 2008a). Women at increased risk of VTE have medical reasons to maintain a normal weight, preferably through exercise (Lindqvist et al. 1999).

We conclude that the Swedish obstetric venous thromboprophylaxis guideline is user-friendly, effective and economical. Its implementation results in thromboprophylaxis recommendations that help avoid both under- and overtreatment of women at increased risk of VTE.

Ethical approval

No ethical permission is needed to publish guidelines

Authors’ contributions

PL and MH contributed to conception and design. PL prepared the draft and all three authors participated in writing and critical review, as well as approving the final version of the article.

Registration number

Not applicable

Disclosure statement

Pelle Lindqvist has received lecture and education honoraria from Pfizer Sweden and Leo Pharma. Margareta Hellgren has received lecture honoraria from Pfizer Sweden, Leo Pharma, CSL Behring and Octapharma Nordica. She has been a consultant to Pfizer Sweden and has received research funding from Pfizer Sweden, Octapharma Nordica and CSL Behring.

Data availability statement

Data sharing not applicable – no new data generated

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.