SEARCH

SEARCH BY CITATION

Keywords:

  • Antenatal;
  • case–control study;
  • pregnancy;
  • pulmonary embolism

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

Objectives  To estimate the incidence of antenatal pulmonary embolism and describe the risk factors, management and outcomes.

Design  A national matched case–control study using the UK Obstetric Surveillance System (UKOSS).

Setting  All hospitals with consultant-led maternity units in the UK.

Participants  A total of 143 women who had an antenatal pulmonary embolism between February 2005 and August 2006. Two hundred and fifty nine matched control women.

Methods  Prospective case and control identification through the UKOSS monthly mailing.

Main outcome measures  Incidence and case fatality rates with 95% CIs. Adjusted odds ratio estimates.

Results  One hundred per cent of UK consultant-led obstetric units contributed data to UKOSS. A total of 143 antenatal pulmonary embolisms were reported, representing an estimated incidence of 1.3 per 10 000 maternities (95% CI 1.1–1.5). Seventy per cent of women had identifiable classical risk factors for thromboembolic disease. The main risk factors for pulmonary embolism were multiparity (adjusted odds ratio [aOR] 4.03, 95% CI 1.60–9.84) and body mass index ≥ 30 kg/m2 (aOR 2.65, 95% CI 1.09–6.45). Nine women who had a pulmonary embolism should have received antenatal thromboprophylaxis with low-molecular-weight heparin (LMWH) according to national guidelines; only three (33%) of them did. Six women (4%) had a pulmonary embolism following antenatal prophylaxis with LMWH; three of these women (50%) were receiving lower than recommended doses. Two women had recurrent pulmonary emboli (1.4%, 95% CI 0.2–5.1%). Five women died (case fatality 3.5%, 95% CI 1.1–8.0%).

Conclusions  Significant severe morbidity from thromboembolic disease underlies the maternal deaths from pulmonary embolism in the UK. This study has shown some cases where thromboprophylaxis was not provided according to national guidelines, and there may be scope for further work on guideline implementation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

Pulmonary embolism has been identified as the most important cause of direct maternal mortality in the UK today and has been comprehensively studied as part of the continuing Confidential Enquiry into Maternal Deaths (now the Confidential Enquiry into Maternal and Child Health [CEMACH]).1 In contrast, nonfatal pulmonary embolism in pregnancy has not been studied extensively. Data do not exist in the UK on the incidence of this disorder, its sequelae or management.

Nonfatal antenatal pulmonary embolism is now widely regarded as an obstetric ‘near-miss’ event.2 A near-miss may be defined as ‘a severe life-threatening obstetric complication necessitating an urgent medical intervention to prevent likely death of the mother’.3 In developed countries such as the UK where maternal deaths are rare, the factors surrounding individual deaths may be peculiar to the event and not, therefore, generalisable. Analysis of near-miss events such as antenatal pulmonary embolism has several advantages over death-only audits. It allows for more rapid reporting and more widely applicable conclusions than death-only reports. It also allows the comparison of maternal deaths with near-misses and examination of the quality of care. Further advantages include the identification of additional risk factors that may be amenable to change and the provision of information to improve service planning.2 A series of risk factors for fatal pulmonary embolism have been identified,4 and UK guidelines exist for the use of prophylaxis.5 However, it is not known to what extent these risk factors are present in women with nonfatal pulmonary embolism or how appropriately prophylaxis is used in antenatal pulmonary embolism. This information is important to improve management of this serious life-threatening condition.

The aim of this study was to carry out a population-based national case–control study using the UK Obstetric Surveillance System (UKOSS)6 to estimate the incidence of antenatal pulmonary embolism in the UK, to investigate and quantify the risk factors, management and outcomes and to explore the extent to which preventive and management guidelines are followed.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

A matched case–control study was undertaken. Over the study period, we anticipated identifying 90 cases and 180 controls with an estimated power of 80% to detect odds ratios between 1.9 and 4.0 at the 5% level of statistical significance, for a prevalence range for the potential risk factors of between 5 and 40%. For example, a study of this size would be able to detect an odds ratio of 2.5 for a risk factor that occurred in 15% of the controls, such as maternal obesity (body mass index [BMI] ≥ 30 kg/m2).

Case definition

Cases were defined as any woman in whom either (1) pulmonary embolism was confirmed using suitable imaging (angiography, computed tomography, magnetic resonance imaging or ventilation–perfusion scan showing a high probability of pulmonary embolism) or (2) pulmonary embolism was confirmed at surgery or postmortem or (3) a clinician had made a diagnosis of pulmonary embolism with signs and symptoms consistent with pulmonary embolism present, and the woman had received a course of anticoagulation therapy of greater than 1 week of duration.

Control selection

Controls were defined as any woman delivering a fetus or infant who had not suffered from an antenatal pulmonary embolism. Reporting clinicians were asked to select two controls, identified as the two women delivering immediately before the case in the same hospital. If the controls identified did not deliver before the case, they were deemed to have been selected incorrectly, and clinicians were asked to reselect appropriate control women.

Data collection

Cases were identified through the UKOSS monthly mailings between February 2005 and August 2006. The UKOSS methods have been described in detail elsewhere.6 Briefly, nominated clinicians in each consultant-led maternity unit in the UK were sent a card each month and asked to report all cases of antenatal pulmonary embolism, thus covering the entire cohort of UK births in this period. Clinicians who reported a case were then asked to complete data collection forms for the case and controls, detailing demographic and other potential risk factors, management and outcomes. Up to five reminders were sent if completed forms were not returned. All data requested were anonymous. On receipt of data collection forms, cases were checked to confirm that they met the case definition. Duplicate reports were identified by comparing the woman’s year of birth, hospital and expected date of delivery.

Additional case ascertainment

To ensure all cases were identified, we independently contacted all radiology departments, who were asked to report any cases of pulmonary embolism in pregnant women, reporting only their year of birth and date of diagnosis. Where a case was identified which had not been reported through UKOSS, the relevant UKOSS reporting clinician was contacted and asked to complete case and control data collection forms. In addition, the CEMACH was contacted at the end of the study and asked to identify any maternal deaths from antenatal pulmonary embolism occurring during the study period, again providing only their year of birth and date of diagnosis. These were compared with maternal deaths reported through UKOSS.

Ethics committee approval

The UKOSS general methodology (04/MRE02/45) and this study (04/MRE02/74) were approved by the London Multicentre Research Ethics Committee.

Statistical analyses

Incidence rates with 95% CIs were calculated using the available national maternity data (2005) as a proxy denominator for 2005 and 2006.7 This study covered the entire cohort of UK births, and therefore, risks with 95% CIs were calculated using the population proportions derived from the control women.

In this matched study, odds ratios were estimated throughout using conditional logistic regression. A full regression model was developed by including both explanatory and potential confounding factors in a core model if there was a pre-existing hypothesis or evidence to suggest that they were causally related to pulmonary embolism, for example pre-existing thrombophilia and smoking. Continuous variables were tested for departure from linearity by the addition of quadratic terms to the model and interaction tested by the addition of interaction terms and subsequent likelihood ratio testing. A correlation matrix of the covariates was constructed to check for collinearity. To estimate the proportion of cases attributable to specific causes, we calculated population proportional attributable risks (aetiologic fractions) using the adjusted odds ratios from the multivariate analysis and the proportion of controls exposed as an estimate of population exposure.8 All analyses were carried out using STATA v9 software.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

All 229 eligible UK hospitals participated in UKOSS, representing coverage of 100% of all births in the UK. Data collection was complete for 93% of notified cases of pulmonary embolism (Figure 1). One hundred and thirty eight confirmed cases of pulmonary embolism were reported through UKOSS, and an additional three through radiology departments and two through CEMACH over the 19 months of the study. There were thus a total of 143 antenatal pulmonary embolisms in an estimated 1 132 964 maternities,7 giving an incidence of 1.3 per 10 000 maternities (95% CI 1.1–1.5). No further data were obtained about the two women identified through CEMACH alone. No controls were identified for nine women, and only one control for a further five women, thus subsequent analyses were performed on 141 women and 259 correctly selected controls.

image

Figure 1. Case reporting and completeness of data collection.

Download figure to PowerPoint

Risk factors

Only 99 women (70%) had any identifiable classical risk factors for thromboembolic disease5 (Table 1); the majority had only one risk factor (75% of those with risk factors). Classical risk factors were identified in 49% of controls (OR 2.35, 95% CI 1.51–3.68), with maternal age and obesity making the largest contribution to both groups. The factors included in the final analysis were: age, ethnicity, socio-economic group, marital status, smoking status, BMI, known thrombophilia, history of thrombosis, family history of thrombosis, recent bedrest, recent long haul travel, deep venous thrombosis (DVT) diagnosed in this pregnancy, surgery in pregnancy, parity and twin pregnancy.

Table 1.  Classical risk factors for thromboembolic disease identified in study subjects5
FactorsCases, n (%*) (n = 141)Controls, n (%*) (n = 259)
  • *

    Percentages of those with data. Note that subjects may have more than one individual factor and percentages will not add up to 100%.

Previous VTE6 (4)0 (0)
Thrombophilia5 (4)1 (0.5)
Age older than 35 years43 (30)67 (26)
Obesity (BMI ≥ 30 kg/m2)34 (28)35 (16)
Parity > 42 (1)8 (3)
Varicose veins1 (1)0 (0)
Sickle cell trait1 (1)2 (1)
Immobility (>4 days bedrest)12 (9)3 (1)
Proteinuric pre-eclampsia4 (3)5 (2)
Long haul travel6 (4)5 (2)
Surgery in pregnancy3 (2)1 (0.4)
At least one factor99 (70)126 (49)
Any one factor74 (52)94 (36)
Any two factors15 (11)22 (8)
Any three factors9 (6)8 (3)
Any four factors1 (1)2 (1)

After adjustment, only two factors were associated with a significantly raised odds ratio for antenatal pulmonary embolism (Table 2). Women with a pulmonary embolism were more than twice as likely to have a BMI of 30 kg/m2 or greater (OR 2.65, 95% CI 1.09–6.45) compared with a BMI of 20–24.9 kg/m2. There was a suggestion of a gradient of increasing risk with higher BMI (BMI < 20 kg/m2: OR 0.69, 95% CI 0.18–2.59; BMI 20–24.9 kg/m2: OR 1.0 (reference); BMI 25–29.9 kg/m2: OR 2.51, 95% CI 1.02–6.19; BMI ≥ 30 kg/m2: OR 2.65, 95% CI 1.09–6.45), but we were unable to investigate subgroups of women with BMI greater than 30 kg/m2 because of small numbers in these groups. Cases were more likely to be multiparous than primiparous (OR 4.03, 95% CI 1.60–9.84). The odds ratio did not differ significantly between women who had only had one previous delivery (OR 5.67, 95% CI 2.01–15.9) and those who had had two or more (OR 3.12, 95% CI 1.19–8.19). Ten women had had a DVT diagnosed before their pulmonary embolism, with an adjusted odds ratio of 5.94 (0.60–58.9).

Table 2.  Maternal risk factors for antenatal pulmonary embolism
Risk factorsNumber (%*) of cases (n = 141)Number (%*) of controls (n = 259)Unadjusted OR (95% CI)Adjusted OR (95% CI)
  • *

    Percentages of those with complete data.

  • **

    Baseline comparison group.

  • ***

    Stable estimate of risk not possible due to small number of controls with the relevant factor.

Sociodemographic factors
Age (years)
 <3598 (70)190 (74)1**1**
 35 or over43 (31)67 (26)1.29 (0.82–2.06)0.90 (0.42–1.90)
Ethnicity
 White107 (76)202 (80)1**1**
 Non-White33 (24)50 (20)1.25 (0.67–2.34)0.77 (0.23–2.61)
Socio-economic group
 Managerial and professional occupations29 (23)63 (28)1**1**
 Intermediate occupations29 (22)55 (24)1.12 (0.59–2.17)1.00 (0.39–2.59)
 Routine and manual occupations/unemployed71 (55)107 (48)1.43 (0.83–2.48)1.27 (0.49–3.30)
Marital status
 Single33 (24)52 (20)1.23 (0.70–2.15)1.88 (0.71–5.02)
 Married/cohabiting106 (76)202 (80)1**1**
Smoking status
 Never/Ex smoker149 (59)84 (61)1**1**
 Smoked during pregnancy102 (41)54 (39)0.94 (0.59–1.49)0.89 (0.40–1.95)
Booking body mass index (BMI) (kg/m2)
 <207 (6)24 (11)0.63 (0.24–1.63)0.69 (0.18–2.59)
 <2547 (38)114 (52)1**1**
 25–29.935 (28)47 (21)1.68 (0.92–3.09)2.51 (1.02–6.19)
 30 or over34 (28)35 (16)2.14 (1.15–3.99)2.65 (1.09–6.45)
Medical and prothrombotic factors
Known thrombophilia
 No135 (97)254 (100)1*****
 Yes4 (3)0 (0)∞ (1.95–∞)***
History of thrombosis
 No135 (96)254 (100)1*****
 Yes6 (4)0 (0)∞ (2.91–∞)***
Thrombosis in first-degree relatives
 No119 (88)239 (97)1**1**
 Yes17 (13)7 (3)4.78 (1.86–12.3)2.30 (0.61–8.68)
Recent long haul travel
 No132 (96)229 (98)1**1**
 Yes6 (4)5 (2)1.37 (0.36–5.17)4.79 (0.44–52.3)
Recent bedrest (>4 days duration)
 No128 (91)247 (99)1*****
 Yes12 (9)3 (1)20.1 (2.59–156.1)***
DVT this pregnancy
 No131 (93)252 (99)1**1**
 Yes10 (7)2 (1)8.55 (1.84–39.7)5.94 (0.60–58.9)
Surgery in pregnancy
 137 (98)255 (100)1**1**
 3 (2)1 (0.4)6.00 (0.62–57.7)0.61 (0.03–11.1)
Factors related to pregnancy and reproduction
Parity
 040 (28)119 (46)1**1**
 1+101 (72)138 (54)2.41 (1.51–3.84)4.03 (1.60–9.84)
Twin pregnancy
 No134 (96)253 (98)1*****
 Yes6 (4)4 (2)3.00 (0.84–10.6)***

Two accepted risk factors for pulmonary embolism were associated with nonsignificantly raised odds; long haul travel, defined as a journey of 4 hours or more within the previous 2 weeks (OR 4.79, 95% CI 0.44–52.3) and a history of thrombosis in first-degree relatives (OR 2.30, 95% CI 0.61–8.68). The power of the study to detect significantly raised odds for these factors was limited by low frequencies of each risk factor in the control group. A reliable estimate of the odds could not be reached for twin pregnancy, a history of thrombosis, a history of bedrest or known thrombophilia due to the small numbers of women in both groups reported to have these risk factors. Age older than 35 years (OR 0.88, 95% CI 0.42–1.86) and surgery in pregnancy (OR 0.61, 95% CI 0.03–11.1) were not independently associated with antenatal pulmonary embolism in this study.

Clinicians were asked to report whether women had been diagnosed with anticardiolipin antibodies, antiphospholipid syndrome, antithrombin deficiency, Factor V Leiden, lupus anticoagulant, protein C deficiency, protein S deficiency, prothrombin gene variant or any other condition with associated thrombophilia. In total, ten women (7%) had a heritable or acquired thrombophilia identified either before or after their pulmonary embolism. The only heritable thrombophilia identified in these women was Factor V Leiden, in four women (4%). Of the four women with a known prior diagnosis of thrombophilia, three were heterozygous for Factor V Leiden and one had antiphospholipid antibody syndrome. Seventy-six per cent of women were reported to have been tested for thrombophilia following their pulmonary embolism either during pregnancy or postdelivery (104 of 137 not diagnosed prior to pregnancy). Only a further six women were reported to have had a thrombophilia identified after their pulmonary embolism; three of these women (50%) had antiphospholipid antibody syndrome, one was homozygous for Factor V Leiden and in two the exact diagnosis was not documented.

Assuming causality, population proportional attributable risks indicate that 62% of pulmonary emboli were attributable to multiparity and 21% to BMI of 30 kg/m2 or over.

Diagnosis and management

The median gestational age at diagnosis was 28 weeks (range 2–40 weeks) (Figure 2). A range of diagnostic techniques were used. Ninety-one women (65%) had a ventilation–perfusion scan, 42 (30%) had computed tomography, 22 (16%) had echocardiography and 16 (11%) were diagnosed by angiography. Forty women (28%) were diagnosed using a combination of techniques. Sixty-five women (46%) had their pulmonary emboli diagnosed solely by ventilation–perfusion scanning. Four women (3%) were diagnosed and treated on the basis of clinical symptoms alone without any confirmatory diagnostic test; one of these women had previously been diagnosed with a DVT.

image

Figure 2. Gestational age at pulmonary embolism (completed weeks). Figures above bars show numbers of women.

Download figure to PowerPoint

Prophylaxis

Nine women who had a pulmonary embolism should have received antenatal thromboprophylaxis with low-molecular-weight heparin (LMWH) according to national guidelines;5 only three (33%) of them did. One of these three received a lower dose than that recommended. In total, six women (4%) had a pulmonary embolism following antenatal prophylaxis with LMWH. All these women were treated with a once daily dose of LMWH, and all were of normal booking body weight (range 56–89 kg). Three of the women (50%) in whom LMWH prophylaxis failed were receiving lower than recommended doses; two received dalteparin 2500 units once daily and one enoxaparin 20 mg once daily. The other three women were receiving prophylaxis with enoxaparin 40 mg once daily. A further ten women used graduated elastic compression stockings alone as prophylaxis against venous thromboembolism (VTE).

Treatment

One hundred and thirty-four women (97%) were treated for pulmonary embolism with LMWH (Table 3); 49% with a once-daily dosage schedule. Two women died before receiving therapeutic anticoagulation, and the management of one woman is unknown. The remainder were managed with unfractionated heparin. One woman received thrombolysis, and five women had an inferior vena caval filter inserted.

Table 3.  LMWH treatment regimens used in women with antenatal pulmonary embolism
TherapyOnce daily administration, n (%*)Twice daily administration, n (%*)Total, n (%**)
  • *

    Percentage of women treated with the same drug.

  • **

    Percentage of all women.

Enoxaparin32 (39)51 (61)83 (62)
Dalteparin11 (44)14 (56)25 (19)
Tinzaparin23 (88)3 (12)26 (19)

Outcomes following pulmonary embolism

There were three maternal deaths from pulmonary embolism reported through UKOSS, one following a pulmonary embolism in the first trimester of pregnancy and two in the third trimester. All three cases were also identified through CEMACH. An additional two deaths were identified through CEMACH, both occurring early in the first trimester (case fatality 3.5%, 95% CI 1.1–8.0%; mortality rate 4.4 per million maternities, 95% CI 1.4–10.3/million). Two women had recurrent pulmonary emboli (recurrence rate 1.4%, 95% CI 0.2–5.1%). One woman was receiving enoxaparin once daily in the recommended dose for treatment of pulmonary embolism outside pregnancy. The second woman was receiving enoxaparin twice daily but also at a dose suitable for pulmonary embolism outside pregnancy. Eight women were reported to have additional severe morbidity subsequent to their pulmonary embolism (5.7%, 95% CI 2.5–10.9%), including two women who required ventilation, two who had septicaemia and one women with haemolysis, elevated liver enzymes and low platelet count syndrome (HELLP). No women had haemorrhagic complications, and no women was reported to have developed thrombocytopenia or osteoporotic fractures during pregnancy.

Pregnancy outcomes are known for 140 women with pulmonary embolism. Five of these women had their pregnancies terminated. The median gestation at delivery of those with continuing pregnancies was 38 weeks (range 26–42). Twenty-seven per cent of women with a pulmonary embolism and continuing pregnancies (n= 35) delivered prior to 37 weeks of gestation compared with 10% of control women (n= 25) (unadjusted OR 3.32, 95% CI 1.84–6.00). Women with a pulmonary embolism were more likely to have labour induced than control women (unadjusted OR 2.05, 95% CI 1.24–3.39). Maternal pulmonary embolism and requirement to plan anticoagulation at delivery were cited as the reason for induction in 26 of the 42 women who had labour induced (62%). Forty-one per cent of women with a pulmonary embolism (n= 56) were delivered by caesarean section (unadjusted OR 2.22, 95% CI 1.39–3.55). Half were planned caesarean deliveries and half were carried out as an emergency, with maternal compromise cited as the most common indication for emergency delivery.

Outcomes are known for 135 infants. There were six deaths (two stillbirths in association with maternal deaths, one unexplained stillbirth, two neonatal deaths in infants with major congenital anomalies unrelated to anticoagulation therapy and one neonatal death following premature delivery), giving a perinatal mortality rate of 44 per 1000 (95% CI 16–94/1000) in comparison with the national rate of 8.2 per 1000 (95% CI 8.0–8.4/1000).9

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

The UK national incidence of antenatal pulmonary embolism as estimated by this study is 1.3 per 10 000 or approximately 1 per 7700 maternities. There are very few comparable studies documenting the incidence of both fatal and nonfatal antenatal pulmonary embolism, and none conducted prospectively in a similar manner. All the other studies have been undertaken using retrospective review of routinely coded hospital admissions data. Two national studies, from Sweden 1990–9310 and the USA 2000–01,11 document similar incidence rates of 0.9 per 10 000 (95% CI 0.7–1.3/10 000) and 1.2 per 10 000 deliveries (95% CI 1.2–1.3), respectively. However, a study in the North West Thames region of London between 1988 and 1997 documented a significantly lower incidence (0.6/10 000, 95% CI 0.4–0.9),12 and a lower incidence was also documented in an 11-year retrospective study from Glasgow (0.7/10 000, 95% CI 0.1–1.3/10 000),13 although the numbers in the latter study were small. Routinely coded data in the UK are known to be relatively unreliable,14 and it is likely that these observed differences are simply due to differences in case ascertainment. We included four women in our study who were diagnosed with pulmonary embolism clinically without any confirmatory diagnostic tests, which may potentially account for a higher incidence. However, exclusion of these women does not significantly alter our incidence estimate and is therefore unlikely to explain the observed differences. Both the UK-based studies were conducted more than 8 years prior to the UKOSS study, and we cannot rule out the possibility that the incidence of antenatal pulmonary embolism in the UK has changed over time due to a changing pattern of risk factors.

Just over two-thirds of the women with antenatal pulmonary embolism had a documented risk factor for the condition, as identified from current UK thromboprophylaxis guidelines; only 16% had more than one risk factor. This has been observed before; only one-third of women in a US study of VTE15 had a documented risk factor, and 72% of women with VTE in the Scottish study had clinical risk factors or an identified thrombophilic abnormality.13 This suggests that to impact further on prevention of the condition, additional consideration needs to be given as to how these risk factors are used to guide prophylaxis. In addition, further work is needed to determine whether certain individual factors may be associated with a sufficient level of risk to warrant prophylaxis in the absence of other risk factors, for example further investigation of the potential gradient in risk identified in obese women.

The main risk factors for antenatal pulmonary embolism identified in this cohort of women were multiparity and obesity. Even this large national study, including 143 affected women in a cohort of more than 1.1 million maternities, had insufficient power to show any other associations as statistically significant. We were also unable to investigate a potential gradient in risk associated with rising obesity. This highlights the need for very large, multinational studies of rare conditions such as antenatal pulmonary embolism to accurately identify and quantify significant risks.

Other researchers have investigated risk factors for VTE in pregnancy, but not for antenatal pulmonary embolism alone. Lindqvist et al.,10 in common with the UKOSS study, did not show a significantly increased risk of VTE associated with age or smoking status. There was a borderline increase in risk in women with multiple pregnancies. The pattern of risk observed with parity in their study was, however, very different to that which we identified. In this Swedish study,10 women who had had one previous delivery at more than 24 weeks of gestation were at the lowest risk; a significantly raised odds ratio was only observed in primiparous women and in women with a parity of three or greater. In our study, we observed a stepped increase in risk between primiparous and multiparous women; we did not observe any gradient in risk with increasing parity, although we have limited power to investigate this in women with very high parity. These differences may be due to the different populations studied; the Swedish study10 also included women with DVT.

The Swedish study10 did not investigate the role of obesity in the aetiology of VTE. This was investigated in the North West Thames study,12 but the investigators were unable to show a significant association, possibly due to small numbers. Obesity (BMI categorisation unspecified) was strongly associated with VTE in pregnancy and the puerperium in the study using the US National Inpatient Sample,11 with an odds ratio of 4.4 (95% CI 3.4–5.7). Our estimate of risk is compatible with this, although it should be noted that obesity may be a stronger risk factor for postnatal VTE, particularly in the context of operative delivery.

Only 10 of the 141 women with antenatal pulmonary embolism had a DVT diagnosed before their pulmonary embolism. Other studies suggest there are more than three times as many DVTs occurring antenatally than pulmonary embolisms,13,15 suggesting that, in the UK, treatment of the majority of DVTs is adequate to prevent subsequent pulmonary embolism. However, our observation emphasises the need for a high index of suspicion for pulmonary embolism in the presence of suggestive symptoms even in the absence of clinical evidence of DVT.

Studies document very variable rates of thrombophilia in pregnant women with VTE,13,16 although the rates are generally higher than documented in our study. The only heritable thrombophilia we noted was Factor V Leiden, and the only acquired thrombophilia was antiphospholipid syndrome. A recent report from the RIETE registry documented a comparable rate of thrombophilia in women with VTE during pregnancy,17 with thrombophilias detected in 17 of 72, six with Factor V Leiden and three with antiphospholipid syndrome. However, it is still likely that this important risk factor is being underdiagnosed in the UK. Tests for thrombophilias are commonly postponed until several weeks or months postpartum. The most likely explanation for underdiagnosis is that complete testing was not undertaken during pregnancy when we collected the UKOSS data.

Comprehensive guidelines exist in the UK for both prophylaxis5 and treatment18 of VTE. Only two of the nine women with an antenatal pulmonary embolism who should have received LMWH prophylaxis according to these guidelines received prophylaxis at the appropriate dose. Half of the women who were given antenatal LMWH were given lower than the recommended dose, possibly due to a misinterpretation of the guidelines concerning dose adjustment with bodyweight. There is thus scope for further work on guideline implementation. The absolute number of women managed inappropriately is, however, small. By applying the guidelines to the risk-factor profile of the controls, we estimate that at least 31 000 women in the UK (3% of maternities) were successfully managed with antenatal LMWH prophylaxis over the period of this study and did not subsequently develop a pulmonary embolism. Only two women in this study can be regarded as ‘failures’ of recommended prophylaxis, in that they were receiving the recommended dose of LMWH according to guidelines and nevertheless developed an antenatal pulmonary embolism. All of the six women who received LMWH prophylaxis and went on to develop an antenatal pulmonary embolism were given prophylaxis using a once daily dosage schedule, which suggests that the efficacy of a once daily regimen needs further investigation.

Nearly 94% of the women who had an antenatal pulmonary embolism were not eligible for antenatal thromboprophylaxis according to current guidelines, suggesting that there is scope for revision of these guidelines. If the guidelines were changed to recommend prophylaxis on the basis of two or more risk factors instead of the current three or more, an additional 13 women would have been eligible for prophylaxis, but so would an estimated additional 102 000 normal pregnant women (9% of maternities). The results of our case–control study showed only one different risk factor that might be incorporated into the Royal College of Obstetricians and Gynaecologists prophylaxis guidelines and that is parity of one or greater. Currently, only a parity of five or greater is included as a risk factor. We reassessed the guidelines including parity of one or greater on our group of cases; an additional eight women who had a pulmonary embolism would have been eligible for LMWH prophylaxis, with an estimated additional 62 000 normal pregnant women eligible (5.5% of maternities). Either of these changes will therefore have major cost implications, and in addition, LMWH prophylaxis is not without risk of adverse effects. Thus, until further studies are undertaken to assess the cost benefit, we cannot make any recommendations for change to current prophylaxis guidelines on the basis of our results.

Five women died from antenatal pulmonary embolism in our study. We were thus unable to undertake a comparison of risk factors for fatal and nonfatal antenatal pulmonary embolism. Two women who died from pulmonary embolism in the first trimester of pregnancy were identified through CEMACH alone. The UKOSS system relies on case reporting by obstetricians and midwives, but not physicians or gynaecologists. It is possible, therefore, that some women with nonfatal antenatal pulmonary embolism occurring in the first trimester, whose pregnancy was lost or terminated, were also not reported to UKOSS. Our estimate of incidence may therefore be an underestimate. However, we did attempt to identify cases through radiologists and identified only three additional cases. We do not feel, therefore, that the under ascertainment is likely to be large. We plan to undertake further comparison between our nonfatal cases and the fatal cases reported through CEMACH following the publication of maternal deaths data for 2003–05.

In conclusion, significant severe morbidity from thromboembolic disease underlies the maternal deaths from antenatal pulmonary embolism reported in the UK, with approximately 30 women diagnosed with the condition for each woman who died. This study has shown some cases where thromboprophylaxis was not undertaken according to national guidelines, and there may thus be scope for further work on guideline implementation. The major risk factors identified were multiparity and obesity; however, without additional studies of cost-effectiveness, we are unable to recommend any changes to current guidelines, which would aid prevention of this serious condition. Nearly one-third of women with antenatal pulmonary embolism had no classical risk factors, and only nine women with classical risk factors were eligible for prophylaxis under current guidelines. Further work is needed to assess how information about these risk factors may be used to guide prophylaxis.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

M.K. is funded by the National Coordinating Centre for Research Capacity Development of the Department of Health. J.J.K. was partially funded by a National Public Health Career Scientist Award from the Department of Health and NHS R&D (PHCS 022). The National Perinatal Epidemiology Unit is funded by the Department of Health in England. The views expressed in this paper are those of the authors and do not necessarily reflect those of the Department of Health.

Writing committee

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

M.K., National Perinatal Epidemiology Unit; Jennifer J Kurinczuk, National Perinatal Epidemiology Unit; Catherine Nelson-Piercy, Guys and St Thomas’ Hospital; Ian Greer, University of York; Patsy Spark, National Perinatal Epidemiology Unit; Peter Brocklehurst, National Perinatal Epidemiology Unit.

UKOSS Steering Committee

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

Catherine Nelson-Piercy (Chair), Guys and St Thomas’ Hospital; Jenny Furniss (Vice-chair), Lay Member; Sabaratnam Arulkumaran, Royal College of Obstetricians and Gynaecologists; Jean Chapple, Faculty of Public Health; Cynthia Clarkson, National Childbirth Trust; Natasha Crowcroft, Health Protection Agency; Andrew Dawson, Nevill Hall Hospital; James Dornan, Royal College of Obstetricians and Gynaecologists; Shona Golightly, CEMACH; Ian Greer, University of York; Mervi Jokinen, Royal College of Midwives; Gwyneth Lewis, Department of Health; Richard Lilford, Department of Public Health and Epidemiology, University of Birmingham; Margaret McGuire, Scottish Executive Health Department; Richard Pebody, Health Protection Agency; Derek Tuffnell, Bradford Hospitals NHS Trust; James Walker, National Patient Safety Agency; Steve Yentis, Chelsea and Westminster Hospital; Carole Harris, National Perinatal Epidemiology Unit; Marian Knight, National Perinatal Epidemiology Unit; Jennifer Kurinczuk, National Perinatal Epidemiology Unit; Peter Brocklehurst, National Perinatal Epidemiology Unit.

Contribution to authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

M.K. designed the study, coordinated data collection, coded the data, carried out the analysis and wrote the first draft of the paper. J.J.K. assisted with the design of the study and supervised the data collection and analysis and contributed to writing the paper. P.S. assisted with data coding, conducted validation of the data and some analysis. C.N.P. and I.G. provided clinical input to the study and contributed to the analysis and writing of the paper. P.B. had the original idea for the surveillance system, provided advice at every stage of the study and contributed to the writing and editing of the paper. P.B. will act as study guarantor.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References

This study would not have been possible without the contribution and enthusiasm of the UKOSS reporting clinicians who notified cases and completed the data collection forms. We would also like to acknowledge the members of the UKOSS Steering Committee who provided advice throughout the study and supplied useful comments on earlier drafts of the paper. The support of the Royal College of Obstetricians and Gynaecologists, Royal College of Midwives, Obstetric Anaesthetists Association, Faculty of Public Health, National Childbirth Trust, and the CEMACH contributed greatly to the success of the study.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Funding
  8. Writing committee
  9. UKOSS Steering Committee
  10. Contribution to authorship
  11. Acknowledgements
  12. References