Imaging for the exclusion of pulmonary embolism in pregnancy

  • Protocol
  • Diagnostic



This is a protocol for a Cochrane Review (Diagnostic test accuracy). The objectives are as follows:

To determine the diagnostic accuracy of imaging tests for the diagnosis of pulmonary embolism (PE) in pregnancy. The diagnostic accuracy of these tests or diagnostic strategies will be calculated in relation to the reference standard of pulmonary angiography (PA) or clinical follow-up for at least six weeks.

To determine the number of inconclusive test results for each diagnostic test or diagnostic strategy.

We aim to investigate the effects of a number of factors on the negative predictive value of the index tests.

Clinical factors are the following.

1. Prior testing, e.g. using clinical decision rule results and D-dimer testing.

2. Gestational age.

3. Type of reference standard.

4. Technological advances of CTPA (multi-slice versus single-slice).


Target condition being diagnosed

Pulmonary embolism (PE) is a leading cause of maternal mortality (Chang 2003; Kobayashi 2008; Lewis 2007). In the UK, approximately 30% of the maternal deaths directly related to pregnancy were caused by thromboembolism between 2003 and 2005 (Lewis 2007). The incidence approximates 0.36 to 0.48 per 1000 pregnancies and includes the postpartum period (Heit 2005; James 2006). When PE occurs anticoagulant therapy needs to be initiated to prevent further clot formation. Anticoagulant therapy is potentially harmful as it may lead to bleeding complications and can have teratogenic effects. Therefore, an objectively confirmed diagnosis of PE is vital when commencing anticoagulant therapy.

In pregnancy, a major challenge is that many symptoms of PE are similar to physical manifestations of pregnancy, for example swollen legs, dyspnoea and tachycardia. In pregnant women with suspected PE, the diagnosis is confirmed in approximately 4% of women as compared to 30% in non-pregnant patients, making accurate diagnostic tests imperative (Ginsberg 2001).

Another issue complicating the diagnosis of PE in pregnancy is that imaging tests expose the fetus to ionizing radiation. Although the level of fetal radiation exposure is far below the threshold for inducing carcinogenic effects, physicians may be hesitant to perform imaging tests in order to conform to the 'As Low As Reasonably Achievable' principle (ALARA). Given the high index of suspicion and the dangers of misdiagnosis on the one hand encouraging diagnostic efforts, and the radiation exposure and other test complications discouraging tests on the other hand, optimal diagnostic strategies are even more crucial in pregnant women compared to non-pregnant patients.

Standard pulmonary angiography (PA) is an X-ray based imaging technique using contrast agent to visualize thrombi. It is considered the reference standard for PE but is rarely used anymore because of associated complications, high radiation doses, high costs and declining availability and expertise (Rosenberg 2007; Somarouthu 2010). In particular cases, if the presence or absence of PE remains uncertain after more common tests have been performed, such as ventilation/perfusion scintigraphy (V/Q) or computed tomography (CT), PA is still an option, including in pregnant patients.

Although the risk of pregnancy-related venous thromboembolism (VTE) is relatively high compared to non-pregnant patients, the absolute numbers of pregnant women with a clinical suspicion of PE is still low. Therefore, most physicians working in the field of thrombosis infrequently encounter pregnant women with PE and thus experience is greatly lacking.

Index test(s)

Different tests and clinical decision rules are combined to provide the safest diagnostic strategies. Clinical decision rules are used to determine a patient’s pre-test probability of PE. With a low to moderate pre-test probability combined with a normal D-dimer test it is safe to rule out PE without using further diagnostic imaging in non-pregnant patients.

In pregnancy however, several items that form part of these rules comprise some of the normal physiologic manifestations of pregnancy, for example swollen legs or a heart rate above 100 beats/min, while other items are seldom seen in pregnancy, such as the presence of cancer. Secondly, during pregnancy D-dimer levels gradually rise. A normal D-dimer level is infrequently encountered in the second and third trimesters of pregnancy (Chabloz 2001; Quiroz 2005). Because of this rise in D-dimer levels during pregnancy and because the clinical decision rules are not validated in pregnant women, current practice does not comprise the diagnostic strategy of ruling out PE in pregnancy with a clinical decision rule with or without a D-dimer test. These tests are therefore not included in this review.

CT pulmonary angiography (CTPA) has a clinical validity comparable to PA (Quiroz 2005). It is widely available and has the potential of providing an alternative diagnosis. It is performed when a clinical decision rule and D-dimer test do not concordantly rule out PE. An intraluminal filling defect is diagnostic of a PE, while the absence of a filling defect rules out the diagnosis. An inconclusive test result is present in only a small percentage of CTPAs (20 of 2199 = 0.9% (van Belle 2006)). A still unresolved issue is the clinical relevance of isolated subsegmental emboli. The rate of isolated subsegmental emboli ranges from 4% to 18% (Nijkeuter 2008; Stein 2006). Patients with isolated subsegmental PE appear to have a more benign clinical presentation than patients with a more proximal PE, and therefore the necessity to treat these patients with anticoagulants is questionable (Le Gal 2006).

Further disadvantages of CTPA include allergic reactions to the contrast agent and contrast-related nephropathy as well as exposure of the mother and fetus to radiation. The level of fetal exposure seems acceptable provided that appropriate methods of dose reduction are applied (Nijkeuter 2006). Concerns are raised that breast tissue, which receives particularly high doses of radiation (Hurwitz 2006), is especially susceptible to its carcinogenic effects, but no evidence is available on this. The iodinated contrast agents cross the placenta and carry the potential risk of neonatal thyroid function depression, though this is not supported by recent studies (Bourjeily 2010; Webb 2005).

Perfusion scintigraphy is another well-established diagnostic modality for PE. Results of both ventilation and perfusion scanning are combined and scan results can be divided into the three categories 'PE present' (high probability), 'PE absent' (very low probability or normal) or 'inconclusive' (Sostman 2008). The sensitivity of a high probability was reported to be 77% while the specificity of a very low probability or normal scan was 98% after excluding non-diagnostic readings. These non-diagnostic readings are the V/Q scan’s main drawback; the high percentage of inconclusive test results, of around 39%, then need further testing (PIOPED 1990). Moreover, the radiation exposure to the fetus is equal to or greater than in CTPA depending on the CT scanner model and imaging protocol used and the method of estimating the exposure (Doshi 2008; Hurwitz 2006a; Winer-Muram 2002). Exposure of the mother to radiation is generally less in perfusion scanning compared to CTPA (Hurwitz 2006).

Magnetic resonance (MR) angiography has the advantage of being radiation-free. However, MR angiography is not applied in daily practice because of the long acquisition times, and studies with this technique are limited (Kanal 1992; Oudkerk 2002). Another concern is the safety of gadolinium-containing contrast agents in pregnancy (Chen 2008), although progress has been made with contrast-free MR imaging (Kluge 2004).

Clinical pathway

In practice, the clinical pathway in the case of suspicion of PE in pregnant women may differ between different physicians and also between patients. Depending on the clinical presentation, a chest X-ray or electrocardiogram can be used to rule out other diseases. Clinical decision rules and D-dimer tests are not validated for excluding PE in pregnancy, however we cannot rule out that these tests are performed in daily practice. As PE and deep vein thrombosis (DVT) are both manifestations of the same disease entity (VTE), physicians sometimes opt for compression ultrasonography (CUS) of the legs in the case of suspected PE when symptoms of DVT are also present. When the diagnosis of DVT is confirmed, no further imaging of the lungs is necessary as the treatment of PE is the same as treatment of DVT.

When no leg symptoms are present, or when DVT is excluded with ultrasonography but a clinical suspicion of PE remains, or as a first-choice diagnostic test (depending on the physician), CTPA, magnetic resonance imaging (MRI) or V/Q scanning may be applied to diagnose PE. Patients with a clinical suspicion of PE who undergo CTPA, MRI or V/Q scanning are considered the focus of this review.


As anticoagulant therapy is potentially harmful, but untreated PE can be fatal, false positive and false negative diagnostic results can have major consequences. Diagnostic tests for PE have been extensively studied in non-pregnant patients (Bounameaux 2010). Pregnant women have a five-fold increased risk of VTE compared to age-matched controls (Pomp 2008). Yet pregnant women are usually excluded from diagnostic and management studies and so a strong evidence base for the management of pregnancy-related PE is missing (Middeldorp 2011). Extrapolating results from studies in, and guidelines for, non-pregnant patients is not advisable since pregnant women form a clinically distinct subgroup for several reasons outlined below.

Physiological changes in pregnancy may mimic the symptomatology of PE, and therefore PE can more often be suspected in pregnant women. This complicates the diagnostic process since it is undesirable to unnecessarily expose patients to a diagnostic path and to radiation. On the other hand, PE is a leading cause of maternal mortality and the diagnosis should not be missed.

Clinical decision rules available for non-pregnant patients cannot be used for pregnant women because several items of these rules comprise some of the normal physiologic manifestations of pregnancy, for example swollen legs or a heart rate above 100 beats/min, while other items are seldom seen in pregnancy, such as the presence of cancer.

Another physiological change during pregnancy is the progressive rise in D-dimer levels. Without adjusting the cut-off value for the D-dimer test to detect PE the test will yield more false positive results, that is it will have a lower specificity. Raising the cut-off would theoretically decrease its sensitivity.

The main issue following in utero exposure to radiation at typical diagnostic levels is induction of malignancies. The number of excess malignancy cases up to age 15 years following irradiation in utero is considered to be 1 in 16,000 per mSv (Streffer 2003). A CTPA exposes the fetus to 0.013 mSv of radiation and a perfusion scan to 0.11 to 0.20 mSv. Given the difficulty of measuring radiation and the variations in protocols used, these numbers should be interpreted as equally low and they justify the use of radiation in order to prevent a potentially life-threatening PE. Of course radiation exposure should be kept as low as possible, in line with the ALARA principle (Hendee 1986), but at the same time clinicians need to be aware that risks of radiation need to be weighed against the risks of untreated PE or complications of unnecessary anticoagulant therapy (Nijkeuter 2004).

Finally, the average age of pregnant women is lower than the age of the participants in the studies from which diagnostic accuracy data are derived. They are therefore less likely to have cardio-respiratory comorbidity which can result in abnormal V/Q scans (Matthews 2006), though in pregnancy a raised diaphragm might result in abnormal perfusion scan results.

For these reasons, it is important to regard pregnant women as a distinct subgroup in which tests might perform differently and additional risks need to be taken into consideration. Other research groups have come to the same conclusion and from this point of view reviewed the available literature on PE diagnostics specifically for pregnant women (Bourjeily 2010a; Brown 2010; Duran-Mendicuti 2011; Nijkeuter 2006; Rodger 2010; Rosenberg 2007; Tan 2011). Although these narrative reviews are useful as guides in the clinical dilemma of suspected PE in pregnancy, they lack the rigorous and explicit methodology of a Cochrane review with its transparent guidelines, minimisation of bias and regular updates. There is a need for such a systematic review in this field where there are still many uncertainties and controversies.


To determine the diagnostic accuracy of imaging tests for the diagnosis of pulmonary embolism (PE) in pregnancy. The diagnostic accuracy of these tests or diagnostic strategies will be calculated in relation to the reference standard of pulmonary angiography (PA) or clinical follow-up for at least six weeks.

Secondary objectives

To determine the number of inconclusive test results for each diagnostic test or diagnostic strategy.

We aim to investigate the effects of a number of factors on the negative predictive value of the index tests.

Clinical factors are the following.

1. Prior testing, e.g. using clinical decision rule results and D-dimer testing.

2. Gestational age.

3. Type of reference standard.

4. Technological advances of CTPA (multi-slice versus single-slice).


Criteria for considering studies for this review

Types of studies

Study designs eligible for inclusion will be consecutive patient series.


Participants will be pregnant women with a clinical suspicion of PE. The conditions for suspicion of PE are left to the discretion of the authors of the included studies. Only data on symptomatic patients will be included in the review.

If not all women in the study meet the inclusion criteria of the present review, data will need to be extractable for the relevant participants. If the relevant data are not extractable and cannot be supplied by the author, the study will be excluded. Study populations from secondary and tertiary care settings, comprising inpatients or outpatients, can be included. We will exclude primary care patients.

Index tests

Index tests for PE will be CTPA, perfusion scintigraphy and MRI. The number of women who have a negative or normal index test result will be calculated and the number of false negatives will be determined by PA or clinical follow-up.

Target conditions

The target condition of this review is symptomatic PE.

Reference standards

We will accept both PA and clinical follow-up for at least six weeks as reference standards, though it is anticipated that few or no studies will use PA as the reference standard.

The diagnosis of PE is confirmed on PA in the case of a new intraluminal filling defect, cut-off of contrast material in a vessel more than 2.5 mm in diameter on PA.

If clinical follow-up is applied as the reference standard, objective diagnostic testing needs to be performed in the case of symptoms of PE. The objective criterion for the diagnosis of PE is a (new) intraluminal filling defect on CTPA; cut-off of contrast material in a vessel more than 2.5 mm in diameter; a new perfusion defect involving at least 75% of a segment, with corresponding normal ventilation (that is a high-probability lung scan); a new non-diagnostic lung scan accompanied by documentation of DVT by ultrasonography or venography; or confirmation of a new PE at autopsy. For clinical follow-up, PE will also be diagnosed in the case of symptoms of PE combined with diagnosis of DVT. The objective criterion for the diagnosis of DVT is a (new) non-compressible venous segment or a substantial increase (4 mm or more) in the diameter of the thrombus during full compression in a previously abnormal segment on ultrasonography, or a new intraluminal filling defect on venography.

In case both PA and clinical follow-up are performed, PA will be the reference standard.

Search methods for identification of studies

There will not be any language restrictions.

Electronic searches

We will search MEDLINE and EMBASE (OvidSP) systematically for potentially eligible studies. The search strategies are specified in Appendix 1 and Appendix 2. They contain three sets of terms reflecting the research question of the review: the tests (index tests), the target condition (PE) and the patient (pregnant women). We will execute the searches with venous thrombosis or thromboembolism as the target condition to increase sensitivity. We will not use a methodological search filter or search term such as 'diagnostic' since doing so has been shown to decrease the sensitivity of the search without significantly benefiting the specificity (Leeflang 2006; Ritchie 2007).

Searching other resources

We will use the studies identified by the electronic search as eligible for inclusion in the review as 'seeds' for the 'related article' feature in PubMed and 'find similar' feature in EMBASE to identify more potentially relevant articles. We will perform a citation search for these studies in Web of Science to identify articles that have cited them. We will handsearch the reference lists of the included studies and of previous reviews on the subject for potentially relevant studies. Finally, we will contact experts in the field to enquire about unpublished studies or studies that were not included in EMBASE or MEDLINE.

Data collection and analysis

Selection of studies

The authors will retrieve the articles from the search and review all abstracts of potentially relevant studies, judged first by title with a very low threshold. Publications considered potentially relevant from the abstract will then have a full text review to determine whether they meet the inclusion criteria defined above. Two authors (MN and PJ) will independently complete this process. They will resolve disagreement by discussion or involvement of a third author (SM), when necessary, whose judgement will then be decisive.

Data extraction and management

Two authors will use standardized, piloted data extraction forms to independently extract data from all the selected studies. They will resolve any disagreement by discussion or involvement of a third author, when necessary, whose judgement will then be decisive. The authors will collect the following data from each study for each test being researched: general information (title, authors, year of publication, status of publication, period of recruitment, study design); descriptives of the relevant patient (pregnant patients with suspected symptomatic PE) population (age, sex, ethnicity, country of recruitment, inpatients or outpatients, healthcare setting, PE symptoms); information on the test under study; reference test used (clinical follow-up or PA, or both); items for the study quality assessment (see below); data on intra- and inter-observer variability; number of inconclusive test results in the diseased and non-diseased group, and numbers of true and false positives and true and false negatives for the relevant women. In the case of missing data, we will make an attempt to contact the original authors.

Assessment of methodological quality

Two authors will independently assess the methodological quality of the selected studies using the QUADAS-2 instrument (Table 1). This is an evidence-based tool for the assessment of the methodological quality of diagnostic studies included in systematic reviews that was developed in 2011 as a revision of the QUADAS-1 tool (Whiting 2011). The authors will score each item 'yes', 'no' or 'unclear' according to the criteria outlined in Table 1. They will resolve disagreements by discussion or involvement of a third author, when necessary, whose judgement will then be decisive. We will present the results in a narrative summary as well as in a methodological quality graph and table (Whiting 2011).

Table 1. Quadas 2 checklist
  Item plus signalling questions Criteria for scoring 'yes', 'no' and 'unclear'


Was a consecutive or random sample of patients enrolled?

Was a case-control design avoided?

Did the study avoid inappropriate exclusions?

We will score this item 'yes' when: subjects were consecutively or randomly selected; 90% or more of the subjects were evaluated at the hospital; 5% or less of subjects had received anticoagulant therapy within 24 hours prior to testing; 30% or less is diagnosed with comorbidities such as chronic obstructive pulmonary disease or other pulmonary diseases, malignancy or pregnancy complications (pre-eclampsia, syndrome of haemolysis, elevated liver enzymes and low platelets or eclampsia), and 10% or less of the subjects had prior testing for this episode of suspected PE. We will score 'No' if one of these criteria was not met.


Were the index test results interpreted without knowledge of the results of the reference standard?

If a threshold was used, was it pre-specified?

We will score this item 'yes' in the following cases: if the authors state that the index test interpreter was unaware of the result of the reference test or if the order of testing was index test before reference test for every subject. Even if clinical follow-up was the reference test, the order of testing still has to be stated for the item be scored 'yes'. We will score the item ‘no’ for studies in which it is stated that the test interpreter of the index test was aware of the result of the reference test. In other cases we will score this 'unclear'.

In case of studies directly comparing diagnostic accuracy of 2 index tests against the reference standard, these test results had to be interpreted without knowledge of the results of the comparator index test, and this item will be scored similarly to the approach described above.

No threshold was used.



Is the reference standard likely to correctly classify the target condition?

Were the reference standard results interpreted without knowledge of the results of the index test?

PA and clinical follow-up of at least six weeks are both considered to be able to correctly classify the target condition, the latter only if objective diagnostic tests are used in case of suspected venous thromboembolism.

We will score this item 'yes' if the authors state that the reference tests were interpreted without knowledge of the results of the index test. Furthermore, in case of clinical follow-up as a reference standard, any clinical suspicion of venous thrombosis during follow-up needs to be followed by objective diagnostic testing, i.e. CUS or venography for a suspicion of DVT and scintigraphy, CTPA, or PA for a clinical suspicion of PE. In case a patient died during follow-up, death was classified as caused by PE in case of confirmation by autopsy, in case of an objective test positive for PE prior to death or if PE could not be confidently excluded as the cause of death.



Was there an appropriate interval between index test(s) and reference standard?

Did all patients receive a reference standard?

Did all patients receive the same reference standard?

Were all patients included in the analysis?

With PA, a time period of less than 24 hours between the index and the reference test will be considered short enough to be sure that the target condition did not change between the tests, either because of natural progression of the disease or because of therapeutic intervention. For studies using PA as the reference test, we will score this item 'yes' if the time between index and reference test was less than 24 hours.

Similarly in case of studies directly comparing diagnostic accuracy of index tests, a time period of less than 24 hours between the index tests as well as the reference test will be considered short enough.

During clinical follow-up there is the possibility that the disease diminishes, through natural progression or through intervention. Or there is even the possibility that the condition arises during follow-up where it was not present at the time of the index test. Studies using clinical follow-up will therefore be scored 'no' for this item.

We will score this item 'no' if less than 90% or a non-random selection of participants underwent the reference test.

We will score this item 'no' if less than 90% of subjects that had an index test result either underwent PA or had clinical follow-up as the reference test.

Statistical analysis and data synthesis

We will extract from each study the available data on true and false negatives and positives for the relevant participants in order to calculate specificity, sensitivity and post-test probabilities. If a sufficient number of studies is available, the data will be pooled in a meta-analysis for each diagnostic test. Results will be presented in tables organised by index test as well as in a forest plot.

The forest plot will display for each study: the sensitivity and specificity values, both with the 95% confidence interval; the percentage of inconclusive results in the diseased group and the non-diseased group; the data from the two by two table and sensitivity and specificity as point estimates with 95% confidence intervals. We will plot the sensitivity and specificity of the individual studies in a summary receiver operator curve (SROC). We will use a bivariate random-effects model approach for calculating summary estimates of sensitivity and specificity (Reitsma 2005). Summary estimates will also be translated into likelihood ratios. To illustrate the findings, the post-test probabilities will be presented for a range of plausible values of the pre-test probabilities of PE. The software used for this part of the analysis will be SAS (SAS Institute Inc, Cary, NC, USA).

Besides the possibility of being positive or negative, a test might be inconclusive, for instance with V/Q scans. In practice, these inconclusive results may either be considered positive or negative. The conventional approach to calculating sensitivity and specificity, ignoring intermediate results, does not provide a way of incorporating this third outcome category. Ignoring the inconclusive category would lead to overestimations of the sensitivity and specificity. Calculating the sensitivity as a percentage by dividing the true positives by the total of participants with the disease suggests that 100% minus that percentage is the proportion of diseased participants missed by the test. This is not the case since this proportion also comprises the participants that did not have a conclusive test result. Even though they did not have a conclusive test result, these participants are not by definition falsely classified. Calculating specificity for a test with an inconclusive result category is similarly thwarted. For index tests with inconclusive results, we will explicitly state the proportion of participants with an inconclusive test result in the diseased group and in the healthy group, respectively. In our primary analyses, we will calculate sensitivity and specificity while considering the intermediate results as 'negatives'. In sensitivity analyses, we will re-analyse the data treating inconclusive results as being 'positives'.

Clinicians may be more interested in the proportion of PE patients they will miss after a negative index test, or after an inconclusive index test. Therefore, we will also estimate summary post-test probabilities of having PE for the three index test result categories. This can be done by extending the bivariate model for multiple index test categories, providing sufficient data are available and providing that the models will converge (Bipat 2007; Leeflang 2012).

Investigations of heterogeneity

To explore anticipated sources of heterogeneity we will add predefined covariates in the meta-regression. This will only be undertaken for index tests or strategies that are investigated in more than three studies that report on the particular covariate (Higgins 2009; Lambert 2002; Morton 2004; Thompson 2002). The authors expect that some of the anticipated factors listed below may not be introduced in meta-regression because of a too limited number of studies. It is, nevertheless, useful to predefine these covariates in case there are sufficient data, perhaps in a later stage, and to be able to informally explore their effects.

Clinical factors are the following.

1. Prior testing and clinical decision rule results. Patients with suspected PE are often stratified into high and low risk groups with the use of clinical decision rules or empirical clinical likelihood. The prevalence of PE differs between these groups. Results of prior diagnostic tests also influence the likelihood that the target condition is present. Therefore, any clinical decision rule or other prior diagnostic test that was not the index test of the particular study will be recorded and will be introduced into the analysis of heterogeneity. The same applies for chest X-ray as a prior test. Chest X-ray can be used as a triage to determine which patients are likely to have a conclusive perfusion scintigraphy. Perfusion scintigraphy in patients with abnormal findings on chest X-ray is more often inconclusive compared to scintigraphy in patients with normal chest X-rays. Therefore, in case of abnormal findings on chest X-ray CTPA might be the preferred test, whereas in patients with normal chest X-ray results perfusion scintigraphy may be more often considered.

2. Stage of pregnancy. For a number of reasons, pregnant women should be considered a distinct subgroup with regard to the diagnosis of PE. The importance of these reasons can vary during the course of pregnancy, for example the progressive rise of D-dimer levels or signs of pregnancy that mimic symptoms of PE. Trimester will therefore be added as a covariate to the analysis.

3. Reference standard. It is desirable to allow only one reference standard for all studies. However, PA used to be the gold standard but in more recent literature it has been replaced by clinical follow-up. Therefore, we decided to include both as acceptable reference standards. This might lead to differential verification and cause heterogeneity.

4. Technological advances of index tests: multi-slice versus single-slice CTPA. The newer generation CTPA has been shown to diagnose more subsegmental PE. This will be added as a covariate to take into account the technological advances throughout the years.

We will also explore heterogeneity more informally by way of visual inspection of forest plots stratified by the potential factors of heterogeneity.

Analyses will be performed using SAS statistical software (SAS Institute Inc, Cary, NC, USA).

Sensitivity analyses

We will perform sensitivity analysis to explore the effect on the overall results of historical versus current inclusion of participants. Diagnostic accuracy studies are by definition cross-sectional because two tests are compared on their ability to make a diagnosis at the same moment in the course of disease. Inclusion of patients can however be historical, that is retrospective. This design is more prone to selection bias and is therefore an anticipated source of heterogeneity.

We will also investigate the effect of considering inconclusive index test results as being 'positive' and compare these results with when they were considered to be 'negative'.

Assessment of reporting bias

We will not explore publication bias since adequate methods of detecting publication bias have not yet been developed for diagnostic accuracy studies (Deeks 2005).


Appendix 1. MEDLINE search strategy


1. Thromboembolism/

2. Venous Thromboembolism/

3. Thrombosis/

4. Venous Thrombosis/

5. Upper Extremity Deep Vein Thrombosis/

6. (thromboemboli$ or microthrombus or thrombus or thrombo$ or thrombilic or thrombotic).ti,ab.

7. (DVT or VTE).ti,ab.

8. Pulmonary Embolism/

9. (pulmonary adj embol$).ti,ab.

10. (pulmonary adj thrombo$).ti,ab.

11. (lung adj embol$).ti,ab.

12. (lung adj thrombo$).ti,ab.

13. PE.ti,ab.

14. or/1-13

15. exp Pregnancy/

16. pregnan$.ti,ab.

17. prepartum.ti,ab.

18. antepartum.ti,ab.

19. prenatal.ti,ab.

20. antenatal.ti,ab.

21. exp Pregnancy Trimesters/

22. trimester?.ti,ab.

23. or/15-22

24. diagnostic imaging/

25. exp Lung/ra, ri, us [Radiography, Radionuclide Imaging, Ultrasonography]

26. Pulmonary Artery/ra, ri, us [Radiography, Radionuclide Imaging, Ultrasonography]

27. Angiography/

28. (pulmonary adj3 angiogr$).ti,ab.

29. (lung adj3 angiogr$).ti,ab.

30. magnetic resonance imaging/

31. magnetic resonance angiography/

32. Diffusion Magnetic Resonance Imaging/

33. ((magnetic resonance or MR or MRI or NMR) adj5 (angiogra$ or arteriogra$)).ti,ab.

34. MRA.ti,ab.

35. tomography/

36. exp tomography, emission-computed/

37. exp tomography, x-ray/

38. tomography scanners, x-ray computed/

39. ((CT or CAT) adj5 (angiogra$ or arteriogra$ or tomograph$)).ti,ab.

40. (comput$ adj3 tomogra$).ti,ab.

41. tomodensitometry.ti,ab.

42. (cat adj4 scan$).ti,ab.

43. (ct adj4 scan$).ti,ab.

44. (CTA or CTPA).ti,ab.

45. three dimensional-ct.ti,ab.

46. (MDCT or MSCT).ti,ab.

47. multislice.ti,ab.

48. multi-slice.ti,ab.

49. (multi-row or multirow).ti,ab.

50. (single-slice or singleslice).ti,ab.

51. 3d-cta.ti,ab.

52. exp ultrasonography, doppler/

53. ultrasonography/

54. exp ultrasonography, prenatal/

55. ultrasonics/

56. ultrasound.ti,ab.

57. ultrasonogra$.ti,ab.

58. ultrasonic$.ti,ab.

59. echograph$.ti,ab.

60. (USS or DUS or CDUS or CEUS).ti,ab.

61. (doppler or duplex).ti,ab.

62. sonograph$.ti,ab.

63. sonogram$.ti,ab.

64. (contrast adj4 US).ti,ab.

65. Phlebography/

66. radionuclide imaging/ or perfusion imaging/

67. Ventilation-Perfusion Ratio/

68. (VQ or V?Q or V?P).ti,ab.

69. SPECT.ti,ab.

70. perfusion.ti,ab.

71. scintigraph$.ti,ab.

72. or/24-71

73. 14 and 23 and 72

Appendix 2. EMBASE search strategy

1. thromboembolism/

2. thrombosis/

3. vein thrombosis/



6. (thrombus or microthrombus or thrombotic or thrombilic or thromboemboli$ or thrombos$).ti,ab.


8. deep vein$ thrombo$.ti,ab.

9. deep venous thrombo$.ti,ab.

10. (dvt or VTE).ti,ab.

11. exp lung/

12. pulmonary artery/

13. lung embolism/

14. (pulmonary adj embol$).ti,ab.

15. (pulmonary adj thrombo$).ti,ab.

16. (lung adj embol$).ti,ab.

17. (lung adj thrombo$).ti,ab.

18. PE.ti,ab.

19. or/1-18

20. exp pregnancy/

21. pregnan$.ti,ab.

22. prepartum.ti,ab.

23. trimester?.ti,ab.

24. antepartum.ti,ab.

25. prenatal.ti,ab.

26. antenatal.ti,ab.

27. or/20-26

28. exp diagnostic imaging/

29. exp nuclear magnetic resonance imaging/

30. angiography/

31. (pulmonary adj3 angiogr$).ti,ab.

32. (lung adj3 angiogr$).ti,ab.

33. ((magnetic resonance or MR or MRI or NMR) adj5 (angiogra$ or arteriogra$)).ti,ab.

34. MRA.ti,ab.

35. Radiodensitometry/

36. exp tomography/

37. ((CT or CAT) adj5 (angiogra$ or arteriogra$ or tomograph$)).ti,ab.

38. (comput$ adj3 tomogra$).ti,ab.

39. tomodensitometry.ti,ab.

40. (cat adj4 scan$).ti,ab.

41. (ct adj4 scan$).ti,ab.

42. (CTA or CTPA).ti,ab.

43. three dimensional-ct.ti,ab.

44. (MDCT or MSCT).ti,ab.

45. (multi-slice or multislice or multi-row or multirow).ti,ab.

46. (single-slice or singleslice).ti,ab.

47. 3d-cta.ti,ab.

48. exp echography/

49. computed tomography scanner/

50. ultrasound/

51. ultrasound.ti,ab.

52. ultrasonogra$.ti,ab.

53. ultrasonic$.ti,ab.

54. echograph$.ti,ab.

55. (USS or DUS or CDUS or CEUS).ti,ab.

56. (doppler or duplex).ti,ab.

57. sonograph$.ti,ab.

58. sonogram$.ti,ab.

59. (contrast adj4 US).ti,ab.

60. exp phlebography/

61. scintiangiography/ or scintiphlebography/

62. computer assisted scintigraphy/

63. phlebogra$.ti,ab.

64. lung ventilation perfusion ratio/

65. lung perfusion/

66. (VQ or V?Q or V?P).ti,ab.

67. imag$.ti,ab.

68. perfusion.ti,ab.

69. scintigraph$.ti,ab.

70. SPECT.ti,ab.

71. lung angiography/

72. or/28-71

73. 19 and 27 and 72

Contributions of authors

The protocol was written by Thijs van Mens and Paulien de Jong, with comments from Mathilde Nijkeuter and Saskia Middeldorp. Mariska Leeflang provided support with the methodological aspects of the protocol.

Mathilde Nijkeuter is contact author and guarantor of the review.

Declarations of interest

Authors have no conflicts of interest to declare.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • Chief Scientist Office, Scottish Government Health Directorates, Scottish Government, UK.

    The PVD Group editorial base is supported by the Chief Scientist Office