European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229

Contents lists available at SciVerse ScienceDirect

European Journal of Obstetrics & Gynecology and

Reproductive Biology
journal homepage: www.elsevier.com/locate/ejogrb

A population-based study of venous thrombosis in pregnancy in

Scotland 1980–2005
Eleanor V. Kane a, Catherine Calderwood b, Richard Dobbie c, Carole Morris c, Eve Roman a,
Ian A. Greer d,*
a
Epidemiology and Cancer Statistics Group, Department of Health Sciences, University of York, United Kingdom
b
Royal Infirmary of Edinburgh, United Kingdom
c
Information and Statistics Division, NHS National Services, Scotland, United Kingdom
d
Faculty of Health and Life Sciences, University of Liverpool, United Kingdom

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: Data on time trends in the incidence of pregnancy-related venous thromboembolism (VTE)
Received 4 September 2012 are sparse. This report charts the incidence of pregnancy-related VTE over the period 1980–2005 in
Received in revised form 25 March 2013 Scotland, and discusses the results in relation to potential risk factors.
Accepted 29 March 2013
Study design: 1 475 301 maternity discharges from Scottish hospitals recorded on the Scottish Morbidity
Record 2 (SMR2) were included. Incidences of pregnancy-related VTE, antenatal deep venous
Keywords: thromboembolism (DVT), postnatal DVT and pulmonary embolism (PTE) were derived relative to the
Venous thromboembolism
number of deliveries, and risk factors were analysed using Poisson regression.
Pregnancy
Pulmonary embolism
Results: Over the period, VTE incidence rose from 13.7 to 18.3 per 10 000 deliveries, antenatal DVTs from
Deep vein thrombosis 8.8 to 12.2 per 10 000 deliveries and PTE from 1.5 to 3.0 per 10 000 deliveries. Postnatal DVTs, on the
other hand, declined from 4.2 to 2.7 per 10 000 deliveries. Risk factors were: age over 35 years; three or
more previous pregnancies; previous VTE; obstetric haemorrhage; and preeclampsia. Antenatal DVT risk
was highest in the most deprived areas, where events started increasing before those in less deprived
areas. Postnatal DVT risk was increased following caesarean delivery, especially when unplanned,
although after 1996, events following emergency caesarean decreased.
Conclusion: During the 26-year period, pregnancy-related VTEs increased, with the greatest rise for
antenatal DVTs. Postnatal DVTs, on the other hand, declined over the period, particularly following
emergency section. Thromboprophylaxis use following emergency delivery may have led to the
postpartum reduction. To continue to prevent events, risk assessment and intervention are required,
particularly antenatally.
ß 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction since 1995 for women at risk, particularly following caesarean

Pulmonary embolism is a major cause of maternal death in the
western world [1]. Pregnancy is a specific risk factor for venous
thromboembolism (VTE) due to changes in the haemostatic
system, reduction in venous flow and potential trauma to pelvic
vessels during the course of delivery [2]. Despite accurate data on
mortality from thromboembolism, few contemporary data on the
incidence of VTE in pregnancy have been reported, and most of
them antedate guidelines for thromboprophylaxis in pregnancy.
Throughout the UK, thromboprophylaxis has been recommended
* Corresponding author at: Faculty of Health & Life Sciences, University of
Liverpool, 3rd Floor, The Foundation Building, 765 Brownlow Hill, Liverpool L69
7ZX, United Kingdom. Tel.: +44 0151 795 0422.
E-mail address: ian.greer@liverpool.ac.uk (I.A. Greer).
delivery [3,4]. Thromboprophylaxis following delivery has not
been accepted internationally, however, due to the difficulty in
conducting randomised controlled trials among pregnant women,
as well as the relatively low absolute risk [2]. As for risk factors,
many have been extrapolated from the non-pregnant population,
although there are increasing data quantifying pregnancy-associ-
ated risk factors such as immobility, high body mass index and
medical conditions [5]. Before thromboprophylaxis guidelines, an
earlier study in the Scottish population found that women aged
over 35 years had twice the risk of antenatal DVT compared to
women under 35 years, who had a DVT rate of six per 10 000.
Following delivery, DVT risks also doubled with age over 35 years
and for emergency caesarean delivery compared to both elective
caesarean and vaginal delivery [6].
With clinical evidence lacking, examination of temporal
trends in VTE incidence may give some insight into whether

0301-2115/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejogrb.2013.03.024
224 E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229

Table 1
Incidence rates of venous thromboembolism, antenatal deep venous thrombosis, postnatal deep venous thrombosis, and pulmonary thromboembolism, Scotland 1980–2005.

Total deliveries VTEa Antenatal DVTa Postnatal DVTa PTEa

b b b
N Rate 95% CI N Rate 95% CI N Rate 95% CI N Rateb 95% CI

Total 1 475 301 2006 13.6 13.0–14.2 1287 8.7 8.2–9.2 498 3.4 3.1–3.7 290 2.0 1.7–2.2
Age
<25 493 435 582 12.2 11.2–13.2 399 8.3 7.5–9.1 131 2.7 2.2–3.2 78 1.7 1.4–2.1
25–34 831 043 1117 13.5 12.7–14.3 706 8.5 7.9–9.2 272 3.3 2.9–3.7 170 2.0 1.7–2.3
35 150 823 307 20.7 18.2–23.3 182 11.9 10.0–13.8 95 7.1 5.5–8.6 42 2.9 2.0–3.9
Year
1980–1985 371 958 471 13.7 12.4–15.0 309 8.8 7.8–9.8 141 4.2 3.5–5.0 55 1.8 1.3–2.3
1986–1990 305 967 378 12.8 11.4–14.1 246 8.2 7.1–9.2 103 3.6 2.9–4.4 43 1.5 1.0–1.9
1991–1995 296 038 335 11.4 10.2–12.6 214 7.3 6.3–8.3 91 3.1 2.4–3.7 45 1.5 1.1–2.0
1996–2000 262 711 374 14.1 12.6–15.6 222 8.5 7.4–9.7 89 3.3 2.6–4.0 78 3.0 2.3–3.6
2001–2005 238 627 448 18.3 16.6–20.1 296 12.2 10.8–13.7 74 2.7 2.1–3.4 69 3.0 2.3–3.8
a
VTE: venous thromboembolism; DVT: deep venous thrombosis; PTE: pulmonary thromboembolism. Numbers do not add up to total number of VTE since 79 deliveries
had antenatal and postnatal DVTs, 15 deliveries had antenatal DVT and PTE, 12 deliveries had postnatal DVT and PTE, and 1 delivery had all three types of VTE; for 39
deliveries, type of VTE was not known.
b
Incidence rate per 10 000 deliveries adjusted for age and year.

thromboprophylaxis reduces the occurrence of VTE. This study
describes the pattern of VTE associated with pregnancy in Scotland
between 1980 and 2005, using routinely collected hospital
discharge data, and investigates potential risk factors.
2. Materials and methods
the log likelihood ratio test. Analyses were repeated for antenatal
DVT, postnatal DVT and PTE; for a small number of deliveries more
than one type of VTE was recorded but restricting analyses to
deliveries with only one type did not change the findings. All
analyses were conducted using Stata/IC 11.1 for Windows
(StataCorp LP).

The Scottish Morbidity Record (SMR2) is a registry containing 3. Results

clinical and demographic characteristics and outcomes for all
women discharged from Scottish maternity hospitals: regular
quality-assurance checks have confirmed that it is more than 99%
complete [7]. Obstetric histories are linked via the mother’s record,
and other national databases including SMR11 (neonatal informa-
tion up to 2002), SBR (Scottish Birth Record, from 2003) and GRO
Birth registrations (General Registry Office for Scotland) add data
relating to the baby. Hospital admissions for pregnancy-related
VTE in non-obstetric settings were obtained from the Scottish
Morbidity Record 1 (SMR1; hospital admissions data) using
probability-based matching to link maternal identifying informa-
tion to SMR2 [8]. All hospital episodes of pregnancy-related VTE,
from forty weeks before to six weeks after delivery, were identified
using ICD codes for DVT (ICD9: 325, 437.6, 451.1, 451.2, 453,
671.3–671.5; ICD10: G08, I63.6, I67.6, I80.1–I80.3, I82, O22.3,
O22.5, O87.1, O87.3, O87.9) and PTE (ICD9: 415.1, 673.2; ICD10:
I26, O88.2). Ethical approval was obtained from the Privacy
Advisory Committee of the Information Services, National Health
Service, Scotland.
The dataset comprised deliveries in Scotland to women aged
13–48 years from 1980 to 2005: data for the first decade have been
published previously [6]. There were 1 558 013 deliveries in total
but one health board was excluded due to discrepancies in their
data, leaving 1 475 301 (95%) deliveries for analysis. The incidence
of VTE in pregnancy and postpartum adjusted for age and year was
calculated relative to the total number of hospital deliveries. Age-
specific incidence rates were calculated by single year, aggregating
data for women aged 15 years or less or over 40 years. To examine
age and time trends, three-year moving averages of incidence rates
were plotted. Potential risk factors included parity, previous VTE,
ante- and postnatal haemorrhage, preeclampsia, hypertension,
mode of delivery and deprivation as well as maternal age at
delivery and year of delivery. Deprivation was based on the
Carstairs score [9], an area-based measure of socioeconomic status
calculated using 1991 Scottish census data. Incidence rate ratios
(IRR) and 95% confidence intervals (CI) were estimated using
Poisson regression [10]. Tests for interaction were conducted using
During 1980–2005, the overall incidence rate for VTE was 13.6
per 10 000 deliveries. The majority of events were antenatal (8.7
per 10 000 deliveries) rather than postnatal DVT (3.4 per 10 000
deliveries) or PTE (2.0 per 10 000 deliveries). Incidence rates are
shown by maternal age, an established risk factor in this
population [6] and by each five-year period (Table 1). Age-specific
rates of VTE in total, antenatal DVT, postnatal DVT and PTE
demonstrate an increase in all types with maternal age (Fig. 1).
Temporal trends showed that VTE increased from the mid-1990s,
mostly due to the rise in antenatal DVTs (Fig. 2). For PTE, the rate
was around 1.5 per 10 000 deliveries until 1996, when rates
increased to 3.0 per 10 000 deliveries. Postnatal DVT on the other
hand decreased over the 26 years.
Risk factors for VTE were age 35 or over (IRR = 1.73, 95% CI 1.50–
1.98), parity of 3 or more (IRR = 1.61, 95% CI 1.37–1.88), previous
VTE (IRR = 7.30, 95% CI 6.07–8.77), and haemorrhage (IRR = 1.67,
95% CI 1.43–1.93). These associations were generally present for
antenatal DVT, postnatal DVT and PTE, and most remained when
adjusted for other risk factors (Table 2). Hypertension was
associated with antenatal DVT (IRR = 1.19, 95% CI 1.01–1.39)
and preeclampsia was related to postnatal DVT (IRR = 1.66, 95% CI
1.20–2.30) and PTE (IRR = 3.14, 95% CI 2.21–4.46). In addition,
emergency and elective caesarean delivery were associated with
postnatal DVT (IRR = 2.24, 95% CI 1.79–2.81; IRR = 1.40, 95% CI
1.01–1.94 respectively).
During the study, more women gave birth in their late 30s and
early 40s and the number of caesarean deliveries increased from
around 1 in 10 deliveries in 1980–1985 to 1 in 5 in 2000–2005. The
postnatal DVT rate was lowest among women aged under 25 who
had had a vaginal delivery (2.1 per 10 000 deliveries). Compared to
this group, having a vaginal delivery at older ages or a caesarean
delivery aged under 25 increased postnatal DVT risk around three-
fold (IRR = 2.85, 95% CI 2.05–3.95; IRR = 3.43, 95% CI 2.27–5.18
respectively) while having an emergency caesarean when aged 35
and over increased the risk six-fold (IRR = 5.80, 95% CI 3.62–9.30)
(Table 3). The decrease in postnatal DVTs over time was most
 E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229 225

Fig. 1. Age-specific incidence rates of venous thromboembolism, antenatal deep venous thrombosis, postnatal deep venous thrombosis, and pulmonary thromboembolism,
Scotland 1980–2005. Three-year sliding average rates are plotted, e.g. the rate plotted at age 20 is the average of rates at ages 19, 20 and 21. Rates are adjusted for year of
delivery.

Fig. 2. Year-specific incidence rates of venous thromboembolism, antenatal deep venous thrombosis, postnatal deep venous thrombosis, and pulmonary thromboembolism,
Scotland 1980–2005. Three-year sliding average rates are plotted, e.g. the rate plotted for 1985 is the average of rates across 1984, 1985 and 1986. Rates are adjusted for age at
delivery.

notable for emergency caesarean, where incidence rates fell from
12.3 per 10 000 deliveries (95% CI 7.4–17.1) in 1980–1985 to 3.9
per 10 000 deliveries (95% CI 1.8–6.0) in 2001–2005 (Table 4). In
1991–1995, immediately before prophylaxis guidelines, the rate
was 7.4 per 10 000 deliveries (95% CI 4.2–10.6) and following their
introduction, it fell to 4.9 (95% CI 2.4–7.5) in 1996–2000.
Compared to 1980–1985 rates, the incidence of postnatal DVTs
was significantly lower from 1996 onwards (1996–2000:
IRR = 0.43, 95% CI 0.23–0.81; 2001–2005: IRR = 0.34, 95% CI
0.18–0.66). This pattern remained when adjusted for parity,
deprivation, antenatal haemorrhage, preeclampsia, hypertension
and previous VTE.
While postnatal DVT rates decreased, antenatal DVT rates rose.
The increasing trend occurred at all maternal ages and not just
among women aged 35 and over (data not shown). This VTE was
increased amongst women living in the most deprived areas of
Scotland (Table 2) and the temporal trends differed from those
living in less deprived areas. In the 1980s, antenatal DVT rates were
similar for women living in the most and less deprived areas (9.3
per 10 000 deliveries, 95% CI 7.2–11.4; 8.2 per 10 000 deliveries,
95% CI 7.2–9.3, respectively). Rates began to increase in the 1990s
in the most deprived areas (1996–2000: 13.0 per 10 000 deliveries,
95% CI 9.9–16.0; 2000–2005: 16.3 per 10 000 deliveries, 95% CI
12.8–19.9), but not until the millennium in the less deprived areas
 226
Table 2
Associations between antenatal deep venous thrombosis, postnatal deep venous thrombosis, and pulmonary thromboembolism; and age, year of delivery, deprivation, parity, hypertension, previous venous thrombosis embolism,
haemorrhage, preeclampsia and delivery mode.

Total deliveries Antenatal DVTa Postnatal DVTa PTEa

(N = 1 475 301)
N = 1287 IRRb 95% CI Adj. IRRb Adj. 95% CI N = 498 IRRb 95% CI Adj. IRRb Adj. 95% CI N = 290 IRRb 95% CI Adj. IRRb Adj. 95% CI

Age (mean, sd) 27.2 (5.55) 27.6 (5.84) 1.08c 1.03–1.13 1.06c 1.01–1.11 28.7 (6.15) 1.25c 1.16–1.35 1.25c 1.16–1.36 28.6 (5.79) 1.21c 1.10–1.34 1.14c 1.03–1.26
<25 493 435 399 1 Ref 1 Ref 131 1 Ref 1 Ref 78 1 Ref 1 Ref
25–34 831 043 706 1.03 0.91–1.16 1.06 0.93–1.20 272 1.27 1.03–1.57 1.27 1.03–1.58 170 1.22 0.93–1.59 1.22 0.93–1.62
35 150 823 182 1.40 1.17–1.68 1.33 1.10–1.60 95 2.58 1.97–3.38 2.34 1.76–3.10 42 1.49 1.01–2.18 1.34 0.90–2.00

E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229
Year 1.09d 1.05–1.13 1.09d 1.05–1.14 0.96d 0.90–1.02 0.91d 0.85–0.97 1.24d 1.14–1.34 1.20d 1.10–1.30
1980–1985 371 958 309 1 Ref 1 Ref 141 1 Ref 1 Ref 55 1 Ref 1 Ref
1986–1990 305 967 246 0.96 0.81–1.14 0.97 0.82–1.15 103 0.86 0.67–1.11 0.87 0.67–1.13 43 0.93 0.63–1.39 0.95 0.63–1.42
1991–1995 296 038 214 0.86 0.72–1.02 0.86 0.72–1.03 91 0.75 0.57–0.97 0.75 0.57–0.99 45 0.97 0.66–1.45 0.98 0.65–1.47
1996–2000 262 711 222 0.99 0.83–1.18 1.00 0.84–1.20 89 0.78 0.60–1.02 0.77 0.59–1.02 78 1.84 1.30–2.61 1.87 1.31–2.68
2001–2005 238 627 296 1.45 1.23–1.70 1.49 1.26–1.76 74 0.69 0.52–0.92 0.66 0.49–0.89 69 1.76 1.23–2.52 1.76 1.21–2.56
Deprivation quintile
1st (most affluent) 264 660 234 1 Ref 1 Ref 82 1 Ref 1 Ref 48 1 Ref 1 Ref
2nd 267 301 223 0.94 0.79–1.13 0.98 0.81–1.17 94 1.14 0.84–1.53 1.20 0.89–1.62 57 1.18 0.80–1.73 1.26 0.86–1.85
3rd 269 384 202 0.85 0.70–1.02 0.89 0.74–1.07 83 0.99 0.73–1.35 1.09 0.80–1.48 58 1.19 0.81–1.74 1.28 0.87–1.88
4th 290 056 247 0.96 0.81–1.15 1.02 0.85–1.22 93 1.03 0.77–1.39 1.16 0.86–1.56 53 1.01 0.68–1.49 1.12 0.76–1.66
5th (least affluent) 345 362 354 1.16 0.98–1.37 1.26 1.06–1.49 133 1.24 0.94–1.64 1.43 1.08–1.89 69 1.10 0.76–1.59 1.29 0.89–1.88
Not Recorded 38 538 27 0.79 0.53–1.18 0.90 0.60–1.35 13 1.09 0.61–1.95 1.18 0.65–2.14 5 0.72 0.28–1.80 1.08 0.42–2.76
Paritye
0 725 228 650 1 Ref 1 Ref 253 1 Ref 1 Ref 144 1 Ref 1 Ref
1–2 667 418 524 0.83 0.74–0.93 0.80 0.71–0.90 199 0.80 0.66–0.97 0.83 0.68–1.01 116 0.75 0.59–0.97 0.84 0.65–1.08
3+ 82 648 113 1.40 1.14–1.72 1.19 0.96–1.46 46 1.38 1.00–1.90 1.32 0.95–1.84 30 1.43 0.95–2.14 1.50 0.99–2.27
Previous VTEa
No 1 462 434 1207 1 Ref 1 Ref 472 1 Ref 1 Ref 276 1 Ref 1 Ref
Yes 12 867 80 7.42 5.91–9.33 7.97 6.30–10.1 26 5.37 3.61–7.99 6.06 4.03–9.12 14 5.28 3.07–9.07 5.71 3.28–9.94
Hypertension
No 1 298 755 1110 1 Ref 1 Ref 424 1 Ref 1 Ref 234 1 Ref 1 Ref
Yes 176 546 177 1.19 1.01–1.39 1.18 0.96–1.44 74 1.24 0.97–1.59 0.92 0.65–1.31 56 1.81 1.35–2.42 1.00 0.63–1.59
Preeclampsia
No 1 403 448 1216 1 Ref 1 Ref 458 1 Ref 1 Ref 254 1 Ref 1 Ref
Yes 71 853 71 1.19 0.94–1.52 1.03 0.76–1.39 40 1.66 1.20–2.30 1.60 1.01–2.53 36 3.14 2.21–4.46 2.98 1.72–5.17
Antenatal haemorrhage
No 1 338 579 1186 1 Ref 1 Ref 450 1 Ref 1 Ref 247 1 Ref 1 Ref
Yes 86 722 101 1.36 1.11–1.66 1.34 1.09–1.64 48 1.75 1.30–2.36 1.54 1.14–2.08 43 2.77 2.01–3.83 2.64 1.90–3.65
Postnatal haemorrhage
No – – – – – – 438 1 Ref 1 Ref 231 1 Ref 1 Ref
Yes – – – – – – 60 1.47 1.12–1.93 1.20 0.90–1.60 59 2.33 1.74–3.12 2.19 1.63–2.93
Delivery mode
Vaginal 1 233 136 – – – – – 358 1 Ref 1 Ref – – – – –
Elective caesarean 91 621 – – – – – 41 1.40 1.01–1.94 1.39 1.00–1.94 – – – – –
Emergency caesarean 150 544 – – – – – 99 2.24 1.79–2.81 2.06 1.63–2.61 – – – – –
a
VTE: venous thromboembolism; DVT: deep venous thrombosis; PTE: pulmonary thromboembolism.
b
Incidence rate ratios (IRR) and 95% confidence intervals (CI) estimated using Poisson regression; adjusted incidence rate ratios (Adj. IRR) were adjusted for age at delivery, year of delivery, deprivation, parity, hypertension,
previous VTE, haemorrhage and preeclampsia, and for postnatal DVT, mode of delivery.
c
IRR for 5-year increase in age compared to age below 20.
d
IRR for 5-year increase compared to 1980–1985 period.
e
Seven deliveries had no information on parity.
 E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229 227

Table 3
Associations between postnatal deep venous thrombosis, age and mode of delivery.

Age Vaginal delivery Elective caesarean Emergency caesarean

a a b b a a b b
Total N Rate 95% CI IRR 95% CI Total N Rate 95% CI IRR 95% CI Total N Ratea 95% CIa IRRb 95% CIb

<25 433 707 91 2.1 1.7–2.6 1 Ref 16 816 10 6.3 2.4–10.3 2.86 1.49–5.49 42 912 30 7.1 4.5–9.6 3.43 2.27–5.18
25–34 687 471 206 3.0 2.6–3.4 1.48 1.15–1.89 56 618 19 3.4 1.9–5.0 1.70 1.04–2.80 86 954 47 6.1 4.3–8.0 2.79 1.96–3.98
35 111 958 61 5.8 4.3–7.4 2.85 2.05–3.95 18 187 12 7.4 2.6–12.2 3.56 1.94–6.52 20 678 22 15.7 8.2–23.2 5.80 3.62–9.30
a
Incidence rate per 10 000 deliveries.
b
Incidence rate ratios (IRR) and 95% confidence intervals (CI) estimated using Poisson regression adjusting for year of admission. Test for interaction between age at
delivery and mode of delivery: x2 = 8.61, p = 0.07.

Table 4
Associations between postnatal deep venous thrombosis and mode of delivery by year of delivery.

Year Total Vaginal delivery Elective caesarean Emergency caesarean

N Ratea 95% CIa IRRb 95% CIb Total N Ratea 95% CIa IRRb 95% CIb Total N Ratea 95% CIa IRRb 95% CIb

1980–1985 325 485 106 3.5 2.8–4.2 1 Ref 18 625 7 3.9 0.9–6.9 1 Ref 27 848 28 12.3 7.4–17.1 1 Ref
1986–1990 262 919 70 2.9 2.2–3.5 0.79 0.59–1.07 16 053 12 8.1 3.4–12.8 1.97 0.77–5.00 26 995 21 8.3 4.6–12.0 0.76 0.43–1.34
1991–1995 249 867 65 2.6 2.0–3.2 0.73 0.54–1.00 17 520 5 2.9 0.3–5.4 0.73 0.23–2.32 28 651 21 7.4 4.2–10.6 0.68 0.39–1.20
1996–2000 212 906 67 3.0 2.2–3.7 0.84 0.62–1.15 18 418 7 4.4 1.0–7.8 0.95 0.33–2.75 31 387 15 4.9 2.4–7.5 0.43 0.23–0.81
2001–2005 181 959 50 2.3 1.6–3.0 0.71 0.51–1.01 21 005 10 4.7 1.6–7.8 1.17 0.44–3.15 35 663 14 3.9 1.8–6.0 0.34 0.18–0.66
a
Incidence rate per 10 000 deliveries.
b
Incidence rate ratios (IRR) and 95% confidence intervals (CI) estimated using Poisson regression adjusting for age at delivery. Test for interaction between year of delivery
and mode of delivery: x2 = 14.1, p = 0.08.

Table 5
Associations between antenatal deep venous thrombosis and deprivation by year of delivery.

Year 1st to 4th quintiles of deprivation 5th quintile of deprivation (least affluent)

Total N Ratea 95% CIa IRRb 95% CIb Total N Ratea 95% CIa IRRb 95% CIb

1980–1985 278 418 228 8.2 7.2–9.3 1 Ref 93 540 81 9.3 7.2–11.4 1 Ref
1986–1990 231 672 195 8.3 7.2–9.5 1.02 0.84–1.24 74 295 51 6.6 4.7–8.5 0.78 0.55–1.11
1991–1995 228 714 152 6.7 5.6–7.7 0.80 0.65–0.98 67 324 62 9.6 7.2–12.1 1.03 0.74–1.43
1996–2000 204 959 148 7.2 6.1–8.4 0.86 0.70–1.06 57 752 74 13.0 9.9–16.0 1.40 1.02–1.92
2001–2005 186 176 210 11.2 9.7–12.8 1.34 1.11–1.62 52 451 86 16.3 12.8–19.9 1.77 1.30–2.41
a
Incidence rate per 10 000 deliveries.
b
Incidence rate ratios (IRR) and 95% confidence intervals (CI) estimated using Poisson regression adjusting for age at delivery. Test for interaction between deprivation and
year of delivery: x2 = 17.4, p = 0.002.

(2000–2005: 11.2 per 10 000 deliveries, 95% CI 9.7–12.8 per 10 000
deliveries) (Table 5).
4. Comments
Over the 26-year period, the rate of venous thrombosis
associated with pregnancy increased from 14 in 1980–1985 to
18 of every 10 000 deliveries in 2001–2005. The rise in pregnancy-
associated VTE was due to increasing numbers of antenatal DVTs.
In the most deprived areas of Scotland, the rise started before, and
incidence of antenatal DVTs remained above, that in less deprived
areas. PTE rates were constant across the first 16 years of the study
until a slight increase around 1996–1997 and have remained at
three per 10 000 deliveries since. Not all types of VTE were on the
increase. The incidence of postnatal DVTs declined from over four
to under three per 10 000 deliveries. Postnatal DVTs have
decreased the most amongst women delivering by emergency
caesarean, particularly after 1995.
Our study, based on almost 1.5 million deliveries in Scotland, is
one of the largest population-based studies of pregnancy-related
VTE, and covers the period when thromboprophylaxis guidelines
in pregnancy were introduced. This population-based study using
routinely collected hospital data has some limitations. Changes in
diagnostic techniques and in practice during the study period may
have impacted our findings. For instance, increased epidural rates
and antibiotic prophylaxis for caesarean delivery may have
reduced the VTE risk. Conversely, in the early 1980s before the
advent of ultrasound venography [11], VTE may have been
diagnosed clinically rather than by X-ray venography due to
concerns of radiation exposure to the fetus. Such clinical diagnosis
may have overdiagnosed the problem in the earlier years of this
study in contrast to objectively confirmed diagnosis at later times.
Pregnancy-related VTE may not necessarily present to obstetric
departments and 2% of cases were identified by non-obstetric ICD
codes. Coding of VTE could be an issue, and since validation of
codes against medical records elsewhere suggests positive
predictive values of more than 80% for the broad groupings of
VTE, DVT and PTE [12,13], it is possible that rates are over-
estimated. Changes in diagnoses and coding quality are unlikely to
explain the differing trends in antenatal and postnatal events, and
by social deprivation. One health board, comprising 5% of
deliveries, had higher than typical numbers of PTEs in two
consecutive years, and its exclusion did not change the findings for
other types of VTE. We were unable to explain the increase in
antenatal DVTs in terms of risk factors, such as obesity,
immobilisation during pregnancy, smoking in pregnancy, and
multiple pregnancies for example, due to a lack of data. Our data
could not be linked to prescribing and therefore we cannot
establish which women received thromboprophylaxis.
The incidence of pregnancy-related VTE in Scotland is amongst
the highest reported in the last decade. Elsewhere, rates per 10 000
deliveries have ranged from five in South Korea, ten in England and
Australia, 12 in Denmark to 17 or more in North America and Hong
Kong [14–21]. Temporal trends described in North America,
228 E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229
Scandinavia and England mostly relate to data up to the 1990s
[20,22–25]. One study described pregnancy-related VTE incidence
from 1991 through 2005 using hospital inpatient records [20].
These Canadian data found a decline in DVTs during pregnancy and
postpartum. During the study, there was a change in practice from
treating women with DVTs as outpatients rather than inpatients
and this may have influenced their observation. This same study
and another conducted in the US both reported increases in PTE
incidence from the early 1990s to the mid-2000s [20,26] while an
earlier US study reported a decline from 1966 to 1995 [25]. Our PTE
data showed a constant rate until 1996/1997, and a higher but
constant rate since.
In the UK, postpartum prophylaxis for women at risk has been
advocated for almost 20 years [27]. In September 1995, the first
Scottish Intercollegiate Guidelines Network (SIGN) document on
prophylaxis of VTE specifically published a section on pregnancy
giving clear national guidelines applicable to the Scottish Health
Service [4]. The guideline highlighted pregnancy as a risk factor
particularly for the puerperium, and where there was a history of
previous thromboembolism or thrombophilia, age over 35, obesity,
emergency caesarean or major surgery or concurrent medical
illness in pregnancy. Pharmacological prophylaxis with heparin
with or without graduated elastic compression stockings was
recommended for patients at risk. In the same year, the Royal
College of Obstetricians and Gynaecologists (RCOG) published
their guidelines for thromboprophylaxis in obstetrics [3], and in
particular following caesarean delivery, with consideration of
similar risk factors to the SIGN recommendations. This was
disseminated to all Fellows and Members of the College in late
1995. The reduction in postnatal VTE following the introduction of
these guidelines suggests the possibility of their impact in practice.
This decrease may be greater than our data reflect since during the
study period, older maternal age, caesarean delivery and obesity
have become more common. It is worth noting that while specific
pharmacological thromboprophylaxis after caesarean delivery is
commonly practised in Europe, it is much less common in North
America.
While cause and effect cannot be demonstrated here, the
temporal relationship for postnatal DVTs is suggestive of, and
consistent with, benefits from thromboprophylaxis seen in other
settings [28]. Outwith pregnancy, there are randomised trials of
thromboprophylaxis in patients at risk; in general surgery,
orthopaedic surgery and in medical patients, for instance [28].
There are, however, no large-scale randomised trials of thrombo-
prophylaxis in pregnancy or the puerperium. While there have
been attempts to conduct such studies, any sizeable trial of an anti-
thrombotic agent in pregnancy or the puerperium is unlikely as
this is perceived as a difficult area to study in view of the
relationship with pregnancy and low absolute event rate. Trials in
this area have been summarised in a recent Cochrane Review [29]
and the lack of direct evidence has also been highlighted by the
RCOG and the National Institute for Clinical Excellence [30,31]. In
the absence of randomised controlled trials, the presented data
may be of value in assessing potential benefits from thrombopro-
phylaxis in pregnancy and the puerperium.
While postnatal DVTs have decreased, the rise in antenatal
DVTs is of concern. A reduction in incidence of this VTE may also
be achieved through thromboprophylaxis. Women with 3 or
more risk factors have been recommended to receive the
intervention since the 1995 guidelines [4], but the impact of
these guidelines on antenatal DVT occurrence could not be
assessed due to insufficient information. It is interesting to note
that the trend with age was not present in recent data from
Denmark where anticoagulation prophylaxis has been recom-
mended for older women [17]. Our data show that, although
older women consistently had the highest incidence, antenatal
DVTs rose in all age groups. Some DVTs in pregnancy may be
preventable if risk factors such as obesity, smoking in pregnancy,
immobilisation and deprivation could be addressed. Previous
VTE and the other risk factors examined here explained 5% or less
of antenatal DVTs, although a higher proportion could be
explained if there is non-differential misclassification of VTE
resulting in underestimation of the risk estimates. As in the
Scottish population as a whole, obesity has become more
common among pregnant women in recent years [32]. Using
data from another study where a third of women were
overweight and a BMI of 25 kg m 2 or more increased antenatal
DVT risk to 1.6, being overweight may account for over 15% of
antenatal DVTs [5]. Attributable fractions depend on the
prevalence of the exposure in a population but nevertheless,
this example gives some indication of the impact being
overweight may have on the occurrence of antenatal DVTs.
Whatever underlying factors are causing the rise, it is clear that
women from the most deprived areas of Scotland have a greater
incidence than those from less deprived areas. Nevertheless, the
increase in antenatal DVTs has occurred among all women in our
study, rising by almost 50% over the last ten years. These data
show a decrease in postnatal DVTs following the advent of
guidelines for thromboprophylaxis but on the other hand, an
increase in antenatal DVTs that needs addressing.
Conflict of interest
The authors have no competing interests.
Acknowledgements
The Executive Health Department, Chief Scientists Office
funded data extraction and processing conducted by the Informa-
tion and Statistics Division, NHS National Services. The study was
approved by the Privacy Advisory Committee of the Information
Services, National Health Service, Scotland.
References
[1] Centre for Maternal and Child Enquiries (CMACE). Saving Mothers’ Lives:
Reviewing maternal deaths to make motherhood safer: 2006–2008. BJOG
2011;118:1–203.
[2] Bates SM, Greer IA, Pabinger I, Sofaer S, Hirsh J. Venous thromboembolism,
thrombophilia, antithrombotic therapy, and pregnancy: American College of
Chest Physicians Evidence-Based Clinical Practice Guidelines (8th ed.). Chest
2008;133(June (6 Suppl.)):844S–86S.
[3] RCOG. Report of the RCOG working party on prophylaxis against thromboem-
bolism in gynaecology and obstetrics. RCOG Press; 1995.
[4] Scottish Intercollegiate Guidelines Network. Prophylaxis of venous thrombo-
embolism a national clinical guideline recommended for use in Scotland by the
Scottish Intercollegiate Guidelines Network. Edinburgh: Scottish Intercolle-
giate Guidelines Network; 1995.
[5] Jacobsen AF, Skjeldestad FE, Sandset PM. Ante- and postnatal risk factors of
venous thrombosis: a hospital-based case-control study. J Thromb Haemost
2008;6:905–12.
[6] Macklon NS, Greer IA. Venous thromboembolic disease in obstetrics and
gynaecology: the Scottish experience. Scott Med J 1996;41:83–6.
[7] Information Services Division Scotland. Data quality assurance: assessment of
maternity data (SMR02) 2008–2009. Edinburgh: NHS National Services Scot-
land; 2010.
[8] Kendrick S, Clarke J. The Scottish record linkage system. Health Bull (Edinb)
1993;51:72–9.
[9] Carstairs V, Morris R. Deprivation and health in Scotland. Health Bull (Edinb)
1990;48:162–75.
[10] Breslow NE, Day NE. Statistical methods in cancer research. Volume II—the
design and analysis of cohort studies. IARC Sci Publ 1987;(82):1–406.
[11] Greer IA, Barry J, Mackon N, Allan PL. Diagnosis of deep venous thrombosis in
pregnancy: a new role for diagnostic ultrasound. Br J Obstet Gynaecol
1990;97:53–7.
[12] White RH, Brickner LA, Scannell KA. ICD-9-CM codes poorly indentified venous
thromboembolism during pregnancy. J Clin Epidemiol 2004;57:985–8.
[13] Larsen TB, Johnsen SP, Møller CI, Larsen H, Sørensen HT. A review of medical
records and discharge summary data found moderate to high predictive values
 E.V. Kane et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 169 (2013) 223–229
of discharge diagnoses of venous thromboembolism during pregnancy and
postpartum. J Clin Epidemiol 2005;58:316–9.
[14] Won HS, Kim DY, Yang MS, Lee SJ, Shin H-H, Park JB. Pregnancy-induced
hypertension, but not gestational diabetes mellitus, is a risk factor for venous
thromboembolism in pregnancy. Korean Circ J 2011;41:23–7.
[15] Lyall H, Myers B. Incidence and management of 82 cases of pregnancy-
associated venous thromboembolism occurring at a single centre—compari-
son with Voke et al. Br J Haematol 2008;142:309–11 [author reply 311–2].
[16] Sharma S, Monga D. Venous thromboembolism during pregnancy and the
post-partum period: incidence and risk factors in a large Victorian health
service. Aust N Z J Obstet Gynaecol 2008;48:44–9.
[17] Virkus RA, Løkkegaard ECL, Bergholt T, Mogensen U, Langhoff-Roos J, Lidegaard
Ø. Venous thromboembolism in pregnant and puerperal women in Denmark
1995–2005. A national cohort study. Thromb Haemost 2011;106:304–9.
[18] Chan LY, Tam WH, Lau TK. Venous thromboembolism in pregnant Chinese
women. Obstet Gynecol 2001;98:471–5.
[19] James AH, Jamison MG, Brancazio LR, Myers ER. Venous thromboembolism
during pregnancy and the postpartum period: incidence, risk factors, and
mortality. Am J Obstet Gynecol 2006;194:1311–5.
[20] Liu S, Rouleau J, Joseph KS, et al. Epidemiology of pregnancy-associated venous
thromboembolism: a population-based study in Canada. J Obstet Gynaecol
Can 2009;31:611–20.
[21] O’Connor DJ, Scher LA, Gargiulo 3rd NJ, Jang J, Suggs WD, Lipsitz EC. Incidence
and characteristics of venous thromboembolic disease during pregnancy
and the postnatal period: a contemporary series. Ann Vasc Surg 2011;25:
9–14.
[22] Andersen BS, Steffensen FH, Sørensen HT, Nielsen GL, Olsen J. The cumulative
incidence of venous thromboembolism during pregnancy and puerperium—an
11 year Danish population-based study of 63,300 pregnancies. Acta Obstet
Gynecol Scand 1998;77:170–3.
229
[23] Simpson EL, Lawrenson RA, Nightingale AL, Farmer RD. Venous thromboem-
bolism in pregnancy and the puerperium: incidence and additional risk factors
from a London perinatal database. BJOG 2001;108:56–60.
[24] Stein PD, Hull RD, Kayali F, et al. Venous thromboembolism in pregnancy: 21-
year trends. Am J Med 2004;117:121–5.
[25] Heit JA, Kobbervig CE, James AH, Petterson TM, Bailey KR, Melton 3rd LJ. Trends
in the incidence of venous thromboembolism during pregnancy or postpartum:
a 30-year population-based study. Ann Intern Med 2005;143:697–706.
[26] Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the
United States: evidence of overdiagnosis. Arch Intern Med 2011;171:831–7.
[27] Thromboembolic Risk Factors (THRIFT) Consensus Group. Risk of and prophy-
laxis for venous thromboembolism in hospital patients. BMJ 1992;305:
567–74.
[28] Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembo-
lism: American College of Chest Physicians Evidence-Based Clinical Practice
Guidelines (8th ed.). Chest 2008;133(6 Suppl.):381S–453S.
[29] Tooher R, Gates S, Dowswell T, Davis L-J. Prophylaxis for venous thromboem-
bolic disease in pregnancy and the early postnatal period. Cochrane Database
Syst Rev 2010;(5):CD001689.
[30] Greer IA, Thomson A. Thrombosis and embolism during pregnancy and the
puerperium, the acute management of (green-top 37b) [Internet]. Available
from: http://www.rcog.org.uk/womens-health/clinical-guidance/thrombo-
embolic-disease-pregnancy-and-puerperium-acute-management-gre.
[31] NICE. Reducing the risk of venous thromboembolism (deep vein thrombosis
and pulmonary embolism) in patients admitted to hospital. Clinical Guide-
lines, CG92. London: National Clinical Guideline Centre – Acute and Chronic
Conditions; 2010. Report No.: CG92.
[32] Kanagalingam MG, Forouhi NG, Greer IA, Sattar N. Changes in booking body
mass index over a decade: retrospective analysis from a Glasgow Maternity
Hospital. BJOG 2005;112:1431–3.