YBLRE-00456; No of Pages 6

Blood Reviews xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Blood Reviews

journal homepage: www.elsevier.com/locate/blre

REVIEW

Use of clinical prediction rules and D-dimer tests in the diagnostic

management of pregnant patients with suspected acute
pulmonary embolism
L.M. Van der Pol a,b, A.T.A. Mairuhu b, C. Tromeur c, F. Couturaud c, M.V. Huisman a, F.A. Klok a,⁎
a
Department of Thrombosis and Hemostasis, Leiden University Medical Center, Leiden, The Netherlands
b
Department of Vascular Medicine, Haga hospital, The Hague, The Netherlands
c
Département de Médecine Interne et Pneumologie, Centre Hospitalo-Universitaire de Brest, Université de Bretagne Occidentale, Brest, France

a r t i c l e i n f o a b s t r a c t

Available online xxxx Because pregnant women have an increased risk of venous thromboembolism (VTE) and at the same time
normal pregnancy is associated with symptoms, mimicking those present in the setting of acute pulmonary
Keywords: embolism (PE), the latter diagnosis is frequently suspected in this patient category. Since imaging tests expose
Pulmonary embolism both mother and foetus to ionizing radiation, the ability to rule out PE based on non-radiological diagnostic
Venous thromboembolism tests is of paramount importance. However, clinical decision rules have only been scarcely evaluated in the
Pregnancy
pregnant population with suspected PE, while D-dimer levels lose diagnostic accuracy due to a physiological
D-dimer
Fibrin fibrinogen degradation products
increase during normal pregnancy. Consequently, clinical guidelines provide contradicting and weak recommen-
Decision rule dations on this subject and the optimal diagnostic strategy remains highly debated. With this systematic review,
we aimed to summarize current evidence on the safety and efficacy of clinical decision rules and biomarkers used
in the diagnostic management of suspected acute PE in pregnant patients.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction are demonstrated by the very low VTE incidences of 5.0% or less in
most studies focussing on pregnant patients with suspected PE,
Women are at an increased risk of venous thromboembolism (VTE) compared to 20–25% in the non-pregnant population [6–10]. Notably,
during pregnancy. Compared to age matched women who are not despite the frequent occurrence of suspected PE in pregnant patients,
pregnant, the risk of VTE is increased 4- to 5-fold [1]. The overall preva- large high quality studies evaluating the optimal diagnostic strategy
lence of thromboembolic events during pregnancy has been reported to for this patient category are unavailable. Moreover, the recommended
be 1.72 per 1000 deliveries with acute pulmonary embolism (PE) diagnostic management of suspected PE consisting of sequential
causing 1.1 deaths per 100,000 deliveries, accounting for 14% of all standardized clinical probability assessment, D-dimer measurement
maternal deaths in the Western World [2,3]. This higher thrombotic and computed tomography pulmonary angiography (CTPA), may not
risk is attributed to a physiological pregnancy-induced hypercoagulable be applicable to pregnant patients for three main reasons [11,12].
state as well as decreased venous outflow of the lower extremities due First, studies that were used to derive clinical prediction rules have
to mechanical venous obstruction by the uterus [4]. This hypercoagula- mostly excluded pregnant women. What's more, available prediction
bility of pregnancy may have evolved to protect women from scores have hardly been prospectively validated in pregnancy or during
haemorrhage at the time of miscarriage or childbirth, and consists of the postpartum period. Second, D-dimer levels physiologically increase
increased concentrations of factors VII, VIII and X as well as von throughout pregnancy with a peak around delivery, making this test
Willebrand factor, fibrinogen and plasminogen activator inhibitor type less useful to rule out VTE. Lastly, CTPA and ventilation perfusion (VQ)
1, whereas protein S levels are decreased [5]. lung scanning, which in the absence of other diagnostic tests, are the
In addition to the higher risk of PE, many of the common VTE symp- current cornerstone of the diagnostic management of suspected PE in
toms such as leg swelling, tachycardia, tachypnea and shortness of pregnant patients, are associated with foetal and maternal radiation ex-
breath are also associated with normal pregnancy. Consequently, in posure. Recent studies suggest that diagnostic imaging test to rule out
clinical practice physicians tend to test for PE at low thresholds, which PE in pregnant and postpartum women are ordered with a very low
threshold, such that the harms of investigation with diagnostic imaging
⁎ Corresponding author at: Department of Thrombosis and Hemostasis, LUMC,
may outweigh its benefits [13]. The ability to safely rule out acute PE in
Albinusdreef 2, 2300 RC Leiden, The Netherlands. pregnant patients without radiological tests is therefore of paramount
E-mail address: f.a.klok@LUMC.nl (F.A. Klok). importance.

http://dx.doi.org/10.1016/j.blre.2016.09.003
0268-960X/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003
2 L.M. Van der Pol et al. / Blood Reviews xxx (2016) xxx–xxx

Most literature that has been published on the optimal diagnostic trimester and 0.61 mg/L in the postpartum period. This corresponds to
management of pregnant patients with suspected PE targets the 79%, 50%, 22%, 0% and 69% of women with a normal D-dimer respective-
discussion whether CTPA or VQ scintigraphy should be preferred as ly, using the established threshold of 0.5 mg/L (Fig. 1).
imaging test. However, the greatest improvement lies in excluding PE A second study applied a comparable design: D-dimer levels were
without imaging tests, a subject that remains scarcely studied. The measured in all trimesters as well as postpartum in 89 healthy pregnant
aim of this review was therefore to generate a complete overview of subjects using a quantitative latex immunoagglutination D-dimer assay
all literature on excluding PE without radiological imaging in pregnant (HemosIL D-dimer HS assay, Instrumentation Laboratory, Milan, Italy)
patients. To do so, we have evaluated current evidence on the safety [18]. The included women were between 18 and 40 years old and had
and efficacy of clinical decision rules and biomarkers used in the no history of VTE. Women with a prior history of diabetes, SLE, chronic
diagnostic management of suspected acute PE in pregnant patients, by hypertension, hepatic or renal disease were excluded from the study. As
performing a systematic search for relevant articles and abstracts in with the previous study, D-dimer levels increased in the course of
MEDLINE (via PubMed), EMBASE, Web of Science, CINAHL, the pregnancy with a normalization in the postpartum period. Using the
Cochrane Database of Systematic Reviews, Academic Search Premier, reference value of 0.230 mg/L, 84%, 33%, 1% and 88% of patients would
Science Direct and the Clinical trials registry that were published in have had a normal D-dimer tests in the three trimesters and postpartum
the last 10 years. The search string is detailed in Appendix A. respectively.
The third study was a large cross sectional study including 1343
2. D-dimer tests in pregnant patients with suspected PE pregnant females in whom D-dimer levels were measured at one
point during pregnancy or post-partum with the latex-based
2.1. Natural course of D-dimer levels during pregnancy and post-partum immunoturbidimetry on the STA-R evolution coagulation analyser
(Diagnostica Stago) [19]. Again, the D-dimer levels increased signifi-
In the 1990s several studies have shown that D-dimer levels rise cantly during the pregnancy trimesters. During the first trimester, 85%
steadily throughout pregnancy [14–16]. In the past 10 years, five studies of pregnant women had a D-dimer below the threshold of 0.5 mg/L,
further contributed to this subject. The first one enrolled 50 healthy which decreased to 29% and 4.1% in the second and third trimester
women who were seeking medical advice to become pregnant, in respectively. The D-dimer levels returned to normal after 42 days post-
whom sequential D-dimer tests were performed during the course of partum. D-dimer concentrations of women who delivered by caesarean
their pregnancy [17]. The mean age of these women was 31 years, 82% sections were significantly higher than those delivering vaginally on the
were Caucasian and 44% had never been pregnant before. Extensive 2nd and 3rd postpartum days, but this difference disappeared on day
exclusion criteria of the study were applied to rule out any alternative 42. Normalization of D-dimer levels occurred on average 3 weeks post-
cause of increased D-dimer, such as renal disease or malignancy. partum in the fourth observational study [20].
Blood samples were drawn at preconception, during each trimester of The fifth study showed that the amount of D-dimer increase during
pregnancy (at 12, 24 and 36 weeks), and four weeks postpartum. pregnancy was more pronounced in twin pregnancies than in singleton
Successful pregnancy was achieved in 32 patients and complete data pregnancies [21]. This study evaluated 1106 patients with a singleton
were available for 18 study subjects. D-dimer levels as measured with pregnancy and 25 patients with a twin pregnancy. In singleton pregnan-
a quantitative latex immunoagglutination D-dimer assay (MDA cies, the mean values of D-dimer in the first and third trimesters were
immunoturbidimetric Assay; Organon Teknika) increased throughout 1.1 ± 1.0 and 2.2 ± 1.6 mg/L, and for twin pregnancies, the mean values
the pregnancy with a mean of 0.43 mg/L at baseline, 0.58 mg/L in the of D-dimer were 1.1 mg/L ± 0.7 and 3.7 mg/L ± 2.5 respectively. In this
first trimester, 0.83 mg/L in the second trimester, 1.2 mg/L in the third study, the increase of D-dimer in twin pregnancies was higher than for

Fig. 1. Overview of pregnant women (%) with D-dimer levels below non pregnancy threshold during pregnancy.

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003
 L.M. Van der Pol et al. / Blood Reviews xxx (2016) xxx–xxx 3

singleton pregnancies, although statistical substantiation was not evidence for recommending a certain threshold depending on pregnan-
provided. According to the reported information, this is the only study cy duration is poor at best.
that evaluated twin pregnancies. The other selected studies in this
described paragraph do not specify the pregnancy into a single- or 2.3. Studies using D-dimer to rule out PE
twin pregnancy.
The final study is a cross-sectional study which determined D-dimer Although many of the studies identified in our search described the
levels in 416 pregnant women at one random time point during their D-dimer test results in incident cases of VTE in general and in acute PE
pregnancy, and in 32 age-matched healthy non-pregnant women [22]. specifically, we did not identify any study that ruled out PE based on a
The reason for inclusion of the 32 non-pregnant controls is not entirely normal D-dimer level alone. In the studies discussed in this review,
clear. The authors have described that these controls are added after we could extract 45 patients with confirmed VTE with known D-
recommendation of the “International Federation of Clinical Chemistry dimer levels [6,18,23,25–28]. These levels exceeded the assay-specific
and Clinical and Laboratory Standards Institute”. Exclusion criteria predefined threshold of 0.5 mg/L in 44 patients, and were normal in
were designed to minimize any possible influence on D-dimer levels: one patient, for a sensitivity of 98% (95% confidence interval (CI) 88–
any history of thromboembolic disease, auto-immune diseases and 99.9). Of note, we anticipate the issue of publication bias on this matter.
morbid obesity. D-dimer analysis was performed using the AMAX From this, it may be hypothesised that this 98% sensitivity could be
AUTO D-dimer kit, based on an immunoturbidimetric method. Median sufficient to allow for a sufficiently high negative predictive value of a
age of the study subjects was 27 years old. normal D-dimer to rule out acute PE, especially because of the known
As compared to non-pregnant women, D-dimer and fibrinogen low disease frequency in cohorts of pregnant patients with suspected
levels were found to be elevated in each trimester. PE. The above described adjusted trimester-specific D-dimer thresholds
In summary, the specificity of the D-dimer test decreases may – if properly validated – increase the specificity and with that the
considerably during pregnancy (Fig. 1). Notably most PE suspicions usefulness of the test, although prospective data to confirm that
occur in the third trimester, in which the specificity of D-dimer test hypothesis are lacking.
approximates 0% when applying diagnostic thresholds established in
non-pregnant patients [6,8]. 3. Other biomarkers in pregnant patients with suspected PE

2.2. Alternative D-dimer thresholds A few alternative biomarkers to replace D-dimer tests or to be used
in combination with D-dimer test have been suggested in recent years.
There are six studies that evaluated alternative D-dimer thresholds In the above described study by Ercan and colleagues, fibrinogen levels
in pregnant patients with suspected VTE (Table 1). were assessed along with D-dimer levels [22]. Fibrinogen levels were
The above described study from Ercan and colleagues also calculates higher in the pregnant patients then in the non-pregnant controls, but
different D-dimer thresholds for each trimester [22]. Reference intervals remained stable during the first 2 trimesters, after which a statistically
for normal D-dimer levels were determined as 0.11–0.40 mg/L, 0.14– significant but absolute small increase was shown. Because of the
0.75 mg/L and 0.16–1.3 mg/L for the first, second and third trimester wide range in fibrinogen levels as well as the lack of both a standardized
respectively [22]. The second study was of the same design [19]. This normal value and a relevant threshold for PE, fibrinogen seems not to
large cross sectional study included 1343 pregnant Chinese women. have potential to replace D-dimer tests in the diagnostic work-up of
The adjusted thresholds for normal D-dimer level were determined at suspected PE in pregnant patients.
0.66 mg/L, 2.29 mg/L and 3.12 mg/L during the first, second and third In a case report, it was suggested to increase the specificity of D-
trimester [19]. dimer tests by combining that with N-terminal pro-B-type-natriuretic
Two above described studies and 2 additional small studies peptide (NT-proBNP) measurement [26]. NT-proBNP is secreted by
measured D-dimer levels during different time points of the pregnancy the cardiac ventricles in response to dilatation or increased intraventric-
and used the range of identified D-dimer values to suggest new ular pressure. With that, NT-proBNP has been established as a prognos-
modified normal values (Table 1) [17–19,22–24]. All 6 studies used tic serum marker for acute PE but not as a diagnostic test [29]. Moreover,
different methods of determining the alternative thresholds, i.e. based it has been shown that NT-proBNP levels increase during normal
on the extreme values or on certain percentiles. Also, and importantly, pregnancy as well, making it an unlikely useful diagnostic tool in the
the identification of a ‘normal range’ is in fact very different from setting of suspected PE [30].
identifying a safe exclusion threshold for PE. Together with the wide Lastly, the diagnostic accuracy of a protein C sensitivity test for PE in
variety in suggested alternative thresholds, this indicates that the pregnant women was evaluated in a small Japanese study [25]. This

Table 1
Identified studies that evaluated adjusted thresholds for normal D-dimer levels in pregnant patients during the trimesters of pregnancy.

Study D-dimer assay Study design Number of pregnant study Suggested threshold for normal
participants D-dimer (mg/L)

1st 2nd 3rd

trimester trimester trimester

Kline, 2005 [17] MDA immune-turbidometric assay D-dimer measured preconceptionally, 50 0.76 1.09 1.48
(Organon Teknika) each trimester and 4 weeks post-partum
Kovac, 2010 [18] HemosIL D-dimer HS (IL) D-dimer tested each trimester and 6–8 weeks 89 0.27 0.46 0.64
post-partum
Wang, 2013 [19] Latex-based immunoturbidimetry single D-dimer measurement at random 1343 0.66 2.29 3.12
(Diagnostica Stago) moment during pregnancy
Ercan, 2015 [22] AMAX AUTO D-dimer Kit single D-dimer measurement at random 416 0.40 0.75 1.30
(Trinity Biotec Plc) moment during pregnancy
Parilla, 2016 [23] Unknown single D-dimer testing and diagnostic 45 0.95 1.29 1.70
imaging (spiral CT or VQ scan)
Kappert, 2009 [38] HemosIL D-Dimer HS 500 test D-dimer tested in each trimester 50 n.p. n.p. n.p.
Morse, 2004 [24] D-dimer assay (IL) D-dimer tested each trimester 48 0.28 0.47 0.64

Note: n.p. = not provided.

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003
4 L.M. Van der Pol et al. / Blood Reviews xxx (2016) xxx–xxx

study was based on the observation that functional sensitivity to of 59 pregnant patients in total. All women with proven PE had D-
activated protein C (APC) decreases during pregnancy and especially dimer level N 0.5 mg/L and/or a likely clinical probability, compared
during pregnancy-associated VTE. To test their hypothesis, the authors with only 11 (26%) of the women in whom PE was ruled out. A total of
measured the normalized APC sensitivity ratio in 111 randomly selected 55 patients (55/59, 93%) had an unlikely clinical probability, of whom
healthy Japanese pregnant females and compared that with those in 31 also had a normal D-dimer test result (31/59, 53%). None of these
200 non-pregnant females (selection criteria not provided) and 7 preg- latter patients was diagnosed with PE at baseline, for a sensitivity of
nant patients with established VTE using an endogenous thrombin 100% (95%CI 80–100). Notably, follow-up data were not available.
potential-based assay. Indeed, the sensitivity to APC in patients with Although these findings support the hypothesis that a normal D-dimer
VTE was reduced in comparison to the control groups, although the in combination with an unlikely clinical probability safely rules out
measured values largely overlapped. A sensitive diagnostic threshold acute PE, it needs to be stressed that this was not an outcome study
could not be extracted from the acquired data. and concerned a limited number of patients. Hence, this study should
Because the biomarkers seem unsuitable for clinical practice and be regarded as hypothesis generating.
validation studies in larger patient numbers as well as prospective
outcome trials for all three suggested biomarkers are lacking, it seems 5. Current studies
that no valuable options to replace the D-dimer test for this purpose
are available in the near future. Based on a search in international trial registries, two large prospec-
tive studies on the subject of optimizing the diagnostic management of
4. Clinical decision rules in pregnant patients with suspected PE suspected PE in pregnant patients are currently running. The Swiss
“Ruling Out Pulmonary Embolism During Pregnancy: a Multicenter
4.1. Diagnostic accuracy of clinical decision rules Outcome Study” evaluates the safety and efficacy of a diagnostic
strategy of sequential clinical probability assessment, D-dimer mea-
Current diagnostic workup of patients with suspected acute PE surement, lower limb compression ultrasonography and multi-slice
usually starts with the assessment of clinical pre-test probability using computed tomography (NCT00771303). The study aims to include
standardized and validated clinical prediction rule. The best validated, 300 pregnant patients with a clinical suspicion of acute PE. Major exclu-
and therefore most widely used, clinical decision rules are the original sion criteria include age less than 18 years, absence of informed consent,
and simplified versions of the Wells rule and revised Geneva rule [9, allergy to contrast medium, impaired renal function (creatinine
31–33]. In our literature review, we identified three articles that clearance less than 30 mL/min as estimated by the Cockcroft-Gould
evaluated the Wells rule in pregnant patients, of which two reported formula) and geographic inaccessibility for follow-up. The results of
on the same patient cohort [6–8,23]. Other clinical prediction rules this study are expected in 2016.
have not yet been tested in pregnant or postpartum women with The Artemis study is a Dutch-French initiative, that prospectively
suspected PE. Since there is great paucity of studies on clinical decision evaluates the safety and efficacy of the YEARS diagnostic algorithm in
rules, detailed information on discriminatory factors is missing as well. pregnant patients (NTR 5913). The YEARS criteria form a simple
The first study was a retrospective evaluation of 81 pregnant and 22 algorithm consisting of the only three Wells items that significantly
post-partum patients who were referred for CTPA imaging because of added incremental value to the D-dimer test – haemoptysis, signs of
clinically suspected PE [6,7]. The Wells score was calculated post-hoc deep vein thrombosis and ‘PE most likely diagnosis’ - in combination
from the medical charts. The majority of patients (60%) were in the with CTPA if necessary [35]. In patients without any of the 3 Wells
third trimester of their pregnancy and 4.8% of patients were diagnosed items and a D-dimer level b 1.0 mg/L, PE is considered excluded without
with PE at baseline. Of the 14% of patients with a high clinical probability further testing. In patients with 1 or more of the items, a D-dimer
(Wells score of 6 or more points), 35% were diagnosed with PE, whereas level b 0.5 mg/L is required to rule out PE. In all patients with either a
PE was ruled out in all of the patients with a Wells score of less than D-dimer level of ≥1.0 mg/L or ≥0.5 mg/L in combination with at least
6 points, suggesting an maximal sensitivity of 100% and a specificity of one YEARS criterion, CTPA is indicated (NTR 4193). The derivation
90% with this diagnostic threshold. Notably, follow-up data were not study of this algorithm suggested that this approach is safe -incidence
available. of symptomatic VTE in untreated patients during three months
The second study involved 183 pregnant patients with suspected PE, follow-up was 1.9% (24/1295, 95%CI 1.2–2.7%) - and more efficient -
of whom 58% were in the third trimester [8]. All were referred for VQ- absolute reduction in CT-scans of 11%- when compared to the current
scanning and the Wells score was calculated by two independent asses- standard strategy. The sample size of this study is determined to be
sors retrospectively by examining medical records of the initial presen- 445 patients. Major exclusion criteria are i) age b 18 years, ii) treatment
tation. PE was confirmed in six patients (3.3%). A total of 107 patients with therapeutic low molecular weight heparin, unfractionated heparin
(58%) had an unlikely pre-test probability (Wells score ≤ 4 points), of or other therapeutic anticoagulants, initiated 24 h or more prior to
whom none were diagnosed with PE (0%). Of the patients with a likely eligibility assessment, iii) unable to give consent, iv) unable to complete
clinical probability (Wells score N 4 points), 7.9% were diagnosed with follow-up or life-expectancy b3 months and v) contraindication to
PE for a sensitivity of 100% and a specificity of 60%. As for the previous CTPA because of iodine allergy. The study objectives include validation
study, follow-up data were not available. of the clinical utility and safety of the Years-diagnostic algorithm in
The third study prospectively evaluated whether trimester specific D- pregnant patients and the assessment of the percentage of pregnant
dimer level and an unlikely clinical probability by the Wells score would patients with suspected PE in whom a CTPA is required when adhering
be a safe criterion to rule out PE [23]. The trimester specific D-dimer to the YEARS-diagnostic algorithm. The results of the Artemis study are
thresholds applied were b 0.95 mg/L, b1.29 mg/L and b 1.70 mg/L in expected in 2018.
the first, second and third trimester respectively, derived from unclear
historical data [34]. The Wells score was calculated at baseline, after 6. What do the guidelines recommend?
which all patients were subjected to diagnostic imaging (either CTPA
or VQ-scan), independent of the D-dimer level or pre-test probability. Guidelines provide contradicting and low grade recommendations
The final diagnosis was based on the imaging test alone. A total of 45 with regard to the role of D-dimer testing and clinical pre-test probabil-
women with suspected PE were included. The median age was ity assessment in the diagnostic management of acute PE in pregnancy.
30 years and the majority of them were multiparous. In addition to Guidelines from the 2014 European Society of Cardiology and the 2016
these, 14 prior patients with an established PE diagnosis of whom D- German Society of Thrombosis and Haemostasis recommend measuring
dimer levels were available were included as well, resulting in a cohort D-dimer levels, stating that normal D-dimer levels do exclude PE in

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003
 L.M. Van der Pol et al. / Blood Reviews xxx (2016) xxx–xxx 5

pregnancy (Class IIb, level C recommendation) whereas guidelines from Appendix A. Search strategy for systemic review
the American Thoracic Society/Society of Thoracic Radiology (2011) and
the Royal college of Obstetricians and Gynaecologists (2015) recom- (“Pulmonary Embolism”[Mesh] OR “Pulmonary Embolism”[tw] OR
mend that D-dimer should not be used to exclude PE in pregnancy Pulmonary Embol*[tw] OR “lung embolism”[tw] OR lung embol*[tw] OR
and physicians should refrain from clinical probability assessment “Pulmonary Thromboembolism”[tw] OR Pulmonary Thromboembol*[tw]
(Class III/IV, level C recommendation) [13,36,37]. OR “Lung Thromboembolism”[tw] OR Lung Thromboembol*[tw] OR “Pul-
monary Thrombo-embolism”[tw] OR Pulmonary Thrombo-embol*[tw]
OR Lung Thrombo-embol*[tw]) AND (“Pregnancy”[Mesh:noexp] OR
7. Conclusion
“Pregnancy”[tw] OR “Pregnant Women”[Mesh] OR Pregnan*[tw] OR
“Gestation”[tw] OR “Gravidity”[mesh] OR “Gravidity”[tw] OR “Pregnancy
In the setting of suspected PE in pregnant women, a frequent
in Adolescence”[mesh] OR “Pregnancy, High-Risk”[mesh] OR “Pregnancy,
situation, more than ever the diagnosis of PE should be ruled-in or
Multiple”[mesh] OR “Superfetation”[tw] OR “Pregnancy, Unplanned”
ruled-out. However, the role of clinical probability assessment in the
[mesh] OR “Pregnancy, Unwanted”[mesh] OR “Pregnancy Complications,
diagnostic management in pregnant patients is uncertain. There is
Cardiovascular”[mesh:noexp] OR “Pregnancy Complications,
great paucity of studies on clinical decision rules and detailed informa-
Hematologic”[mesh]) AND (“decision rule”[tw] OR “decision rules”[tw]
tion on discriminatory factors is missing as well. D-dimer tests with
OR “Decision Making”[mesh] OR “Decision Support Techniques”[mesh]
the conventional threshold lack the specificity to be of incremental
OR “Decision Support Systems, Clinical”[Mesh] OR “d-dimer”[tw] OR
diagnostic value. Alternative thresholds have been suggested but lack
ddimer*[tw] OR “fibrin fragment D” [Supplementary Concept] OR “Fibrin
adequate prospective validation in high quality studies with adequate
Fibrinogen Degradation Products”[Mesh] OR “fibrin fragment D1
sample sizes using standardized contemporary high-sensitive D-dimer
dimer”[tw] OR “fibrin fragment DD”[tw]).
assays. The issues are not only the accuracy of the diagnosis manage-
ment, primary based on clinical probability and optimized threshold
D-dimer, of suspected PE in pregnant women, but also to reduce References
recourse to irradiative tests. Discrepancies in leading international
[1] Heit JA, Kobbervig CE, James AH, Petterson TM, Bailey KR, Melton III LJ. Trends in the
guidelines emphasize the great need for more studies on this subject,
incidence of venous thromboembolism during pregnancy or postpartum: a 30-year
of which two are currently under way. For now, imaging tests remain population-based study. Ann Intern Med 2005;143:697–706.
the cornerstone of evidence based diagnostic management of suspected [2] James AH, Jamison MG, Brancazio LR, Myers ER. Venous thromboembolism during
PE in pregnancy and we recommend not to make use of D-dimer tests or pregnancy and the postpartum period: incidence, risk factors, and mortality. Am J
Obstet Gynecol 2006;194:1311–5.
clinical decision rules in this patient category until this approach has [3] Guimicheva B, Czuprynska J, Arya R. The prevention of pregnancy-related venous
been proven safe. thromboembolism. Br J Haematol 2015;168:163–74.
[4] James AH. Venous thromboembolism in pregnancy. Arterioscler Thromb Vasc Biol
2009;29:326–31.
8. Practice points [5] James AH. Pregnancy-associated thrombosis. Hematology Am Soc Hematol Educ
Program 2009:277–85.
[6] O'Connor C, Moriarty J, Walsh J, Murray J, Coulter-Smith S, Boyd W. The application
of a clinical risk stratification score may reduce unnecessary investigations for pul-
− Pregnant women are at higher risk of VTE, but many of the common
monary embolism in pregnancy. J Matern Fetal Neonatal Med 2011;24:1461–4.
VTE symptoms are also associated with normal pregnancy, which [7] O'Sullivan C, Moriarty J, Walsh J, Coulter-Smith S, Boyd W. Application of a clinical
makes it necessary to assess clear diagnostic management recom- risk stratification score in pregnancy and the puerperium - can unnecessary investi-
gations for pulmonary embolism be avoided? Am J Obstet Gynecol 2009:S68
mendations for pregnant patients with suspected PE.
[Conference].
− Diagnostic management of pulmonary embolism in pregnancy is [8] Cutts BA, Tran HA, Merriman E, Nandurkar D, Soo G, Dasgupta D, et al. The utility of
still a challenge due to great paucity of prospective validation studies the Wells clinical prediction model and ventilation-perfusion scanning for pulmo-
concerning D-dimer thresholds in pregnancy as well as clinical nary embolism diagnosis in pregnancy. Blood Coagul Fibrinolysis 2014;25:375–8.
[9] Douma RA, Mos IC, Erkens PM, Nizet TA, Durian MF, Hovens MM, et al. Performance
probability assessment. of 4 clinical decision rules in the diagnostic management of acute pulmonary embo-
− Different D-dimer thresholds have been suggested, but a validated lism: a prospective cohort study. Ann Intern Med 2011;154:709–18.
cut-off value for D-dimer in pregnancy to exclude PE safely is [10] van Es N, van der Hulle T, van Es J, den Exter PL, Douma RA, Goekoop RJ, et al. Wells
rule and d-dimer testing to rule out pulmonary embolism: a systematic review and
lacking. individual-patient data meta-analysis. Ann Intern Med 2016;165:253–61.
− For now, due to lack of validation of alternative D-dimer thresholds [11] Huisman MV, Klok FA. Magnetic resonance imaging for diagnosis of acute pulmo-
for pregnant patients, imaging tests remain the cornerstone of evi- nary embolism: not yet a suitable alternative to CT-PA. J Thromb Haemost 2012;
10:741–2.
dence based diagnostic management of suspected PE in pregnancy. [12] Huisman MV, Klok FA. Current challenges in diagnostic imaging of venous thrombo-
embolism. Blood 2015;126:2376–82.
[13] Leung AN, Bull TM, Jaeschke R, Lockwood CJ, Boiselle PM, Hurwitz LM, et al. An offi-
cial American Thoracic Society/Society of Thoracic Radiology clinical practice guide-
line: evaluation of suspected pulmonary embolism in pregnancy. Am J Respir Crit
9. Research agenda Care Med 2011;184:1200–8.
[14] Nolan TE, Smith RP, Devoe LD. Maternal plasma D-dimer levels in normal and com-
plicated pregnancies. Obstet Gynecol 1993;81:235–8.
− Large registries of pregnant patients with regular D-dimer measure- [15] Giavarina D, Mezzena G, Dorizzi RM, Soffiati G. Reference interval of D-dimer in
pregnant women. Clin Biochem 2001;34:331–3.
ments are necessary to accurate assess the natural course of D-dimer
[16] Francalanci I, Comeglio P, Alessandrello Liotta A, Cellai AP, Fedi S, Parretti E, et al. D-
levels throughout pregnancy. dimer plasma levels during normal pregnancy measured by specific ELISA. Int J Clin
− Prospective outcome studies are necessary to establish the safety of Lab Res 1997;27:65–7.
[17] Kline JA, Williams GW, Hernandez-Nino J. D-dimer concentrations in normal preg-
ruling out acute PE based on a normal D-dimer level.
nancy: new diagnostic thresholds are needed. Clin Chem 2005;51:825–9.
[18] Kovac M, Mikovic Z, Rakicevic L, Srzentic S, Mandic V, Djordjevic V, et al. The use of
D-dimer with new cutoff can be useful in diagnosis of venous thromboembolism in
pregnancy. Eur J Obstet Gynecol Reprod Biol 2010;148:27–30.
[19] Wang M, Lu SM, Li S, Shen FX. Reference intervals of D-dimer during the pregnancy
Conflict of interest and puerperium period on the STA-R evolution coagulation analyzer. Clin Chim Acta
2013;425:176–80.
None. [20] Padgett J, Gough J, DeVente J, Dixon A, Brewer K. The effects of childbirth on D-dimer
levels. Acad Emerg Med 2015:S314 [Conference].
This research did not receive any specific grant from funding [21] Nishii A, Noda Y, Nemoto R, Ushiro K, Ohno T, Mochizuki Y, et al. Evaluation of D-
agencies in the public, commercial, or not-for-profit sectors. dimer during pregnancy. J Obstet Gynaecol Res 2009;35:689–93.

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003
6 L.M. Van der Pol et al. / Blood Reviews xxx (2016) xxx–xxx

[22] Ercan S, Ozkan S, Yucel N, Orcun A. Establishing reference intervals for D-dimer to [31] Wells PS, Anderson DR, Rodger M, Ginsberg JS, Kearon C, Gent M, et al. Derivation of
trimesters. J Matern Fetal Neonatal Med 2015;28:983–7. a simple clinical model to categorize patients probability of pulmonary embolism:
[23] Parilla BV, Fournogerakis R, Archer A, Sulo S, Laurent L, Lee P, et al. Diagnosing pul- increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000;
monary embolism in pregnancy: are biomarkers and clinical predictive models use- 83:416–20.
ful? AJP Rep 2016;6:e160–4. [32] Le Gal G, Righini M, Roy PM, Sanchez O, Aujesky D, Bounameaux H, et al. Prediction
[24] Morse M. Establishing a normal range for D-dimer levels through pregnancy to aid of pulmonary embolism in the emergency department: the revised Geneva score.
in the diagnosis of pulmonary embolism and deep vein thrombosis. J Thromb Ann Intern Med 2006;144:165–71.
Haemost 2004;2:1202–4. [33] Klok FA, Mos IC, Nijkeuter M, Righini M, Perrier A, Le Gal G, et al. Simplification of the
[25] Hirai K, Sugimura M, Ohashi R, Suzuki K, Itoh H, Sugihara K, et al. A rapid activated revised Geneva score for assessing clinical probability of pulmonary embolism. Arch
protein C sensitivity test as a diagnostic marker for a suspected venous thromboem- Intern Med 2008;168:2131–6.
bolism in pregnancy and puerperium. Gynecol Obstet Invest 2011;72:55–62. [34] Abbassi-Ghanavati M, Greer LG, Cunningham FG. Pregnancy and laboratory studies:
[26] Franz MB, Guenthner-Biller MM, Kainer F, Schiessl B. Pro-B-type natriuretic peptide a reference table for clinicians. Obstet Gynecol 2009;114:1326–31.
as a marker of pulmonary embolism in pregnancy: a case report. Arch Gynecol [35] Van Es J, Beenen LFM, Douma RA, Den Exter PL, Mos ICM, Kaasjager HAH, et al. A
Obstet 2008;277:439–40. simple decision rule including D-dimer to reduce the need for computed tomogra-
[27] To MS, Hunt BJ, Nelson-Piercy C. A negative D-dimer does not exclude venous phy scanning in patients with suspected pulmonary embolism. J Thromb Haemost
thromboembolism (VTE) in pregnancy. J Obstet Gynaecol 2008;28:222–3. 2015;13:1428–35.
[28] Chan WS, Lee A, Spencer FA, Chunilal S, Crowther M, Wu W, et al. D-dimer testing in [36] Linnemann B, Bauersachs R, Rott H, Halimeh S, Zotz R, Gerhardt A, et al. Diagnosis of
pregnant patients: towards determining the next ‘level’ in the diagnosis of deep vein pregnancy-associated venous thromboembolism - position paper of the working
thrombosis. J Thromb Haemost 2010;8:1004–11. Group in Women's Health of the Society of Thrombosis and Haemostasis (GTH).
[29] Klok FA, Mos IC, Huisman MV. Brain-type natriuretic peptide levels in the prediction Vasa 2016;45:87–101.
of adverse outcome in patients with pulmonary embolism: a systematic review and [37] Konstantinides SV, Torbicki A, Agnelli G, Danchin N, Fitzmaurice D, Galie N, et al.
meta-analysis. Am J Respir Crit Care Med 2008;178:425–30. 2014 ESC guidelines on the diagnosis and management of acute pulmonary embo-
[30] Yoshimura T, Yoshimura M, Yasue H, Ito M, Okamura H, Mukoyama M, et al. Plasma lism. Eur Heart J 2014;35:3033-69k.
concentration of atrial natriuretic peptide and brain natriuretic peptide during nor- [38] Kappert G, Halimeh S, Rott H. Reference ranges for the determination of D-dimers in
mal human pregnancy and the postpartum period. J Endocrinol 1994;140:393–7. pregnancy. J Thromb Haemost 2009:750 [Conference].

Please cite this article as: Van der Pol LM, et al, Use of clinical prediction rules and D-dimer tests in the diagnostic management of pregnant
patients with suspected acute pulmonar..., Blood Rev (2016), http://dx.doi.org/10.1016/j.blre.2016.09.003