Executive Summary

  • Pulmonary embolism is primarily a disorder that occurs in patients with identified risk factors.
  • A still-to-be-defined number of pulmonary emboli occur in patients with what appears to be apparent exacerbations of underlying chronic cardiac or pulmonary disease.
  • Diagnostic risk stratification for the pre-diagnosis probability of pulmonary embolism is very useful to guide subsequent diagnostic testing.
  • Outcome risk stratification post-diagnosis is useful to estimate the mortality and can be used to guide therapy and disposition.


Lara V. Reda, MD, Director of Quality Management, Department of Emergency Medicine, North Shore University Hospital, Assistant Professor of Emergency Medicine, Hofstra North Shore – LIJ School of Medicine, Manhasset, NY.

Andrew Choi, MD, Attending Physician, Kingwood Medical Center, Kingwood, TX.

Peer Reviewer

William J. Brady, MD, FACEP, FAAEM, Professor of Emergency Medicine and Medicine, Medical Director, Emergency Preparedness and Response, University of Virginia Operational Medical Director, Albemarle County Fire Rescue, Charlottesville, Virginia; Chief Medical Officer and Medical Director, Allianz Global Assistance.

Pulmonary Embolism

I am reading a fascinating book, The Half-Life of Facts. The author looks across several disciplines, analyzing both the growth of knowledge and the life-expectancy of individual components. One of the chapters discusses that knowledge usually grows cumulatively, building on existing understanding, generally leading to a deeper and more complete insight. Even so-called paradigm shifts rarely overturn prior accepted and useful facts. Einstein’s Theory of General Relativity did not render Newton’s Law of Gravitation obsolete; we used it to go to the moon in 1969. So it is with our understanding of the diagnostic process in pulmonary embolism. We have a reasonable well-validated approach in straightforward clinical situations. Our understanding is growing as we study alternative approaches and different patient populations. This issue covers the commonly used tools for the diagnosis of pulmonary embolism.

J. Stephan Stapczynski, MD, Editor


For emergency physicians, acute pulmonary embolism (PE) provides a particularly complex diagnostic challenge. It has been estimated that 650,000 to 900,000 individuals annually suffer a fatal or nonfatal acute pulmonary embolism.1 While the classic textbook clinical presentation is well known, it is insufficiently accurate and precise in the timely diagnosis of an acute PE. In addition, many patients presenting with seemingly typical exacerbations of their underlying cardiopulmonary disease or other chronic illness may be masking symptoms of an undiagnosed acute pulmonary embolism.2 The high acuity coupled with the unreliable clinical presentation led to the development of several clinical tools, laboratory diagnostics, and radiographical studies to increase the clinician’s diagnostic power. This article we will review the Geneva Score and Wells Criteria, as well as the Kline and PERC rules. In addition, it will discuss special patient populations and diagnostic modalities for treating pulmonary emboli.


Pulmonary emboli are caused by the blockage of the pulmonary arterial system, disrupting the ability of the lung to properly oxygenate blood from the venous circulatory system. While there are several causes of these blockages, including air, fat, and amniotic fluid, the most common cause results from a venous clot that typically travels from a deep vein thrombosis (DVT) in the lower extremity.

When a deep vein thrombosis dislodges, it travels through the venous system to the right heart. Once the clot traverses through the right heart, it enters the pulmonary circulation. The clot is forced into smaller and smaller arteries as it approaches the capillary system and eventually obstructs forward blood flow. This blockage causes dead space ventilation downstream of the obstruction. Subsequent alveolar hemorrhage and loss of surfactant can cause atelectasis, increasing pulmonary vascular resistance.

The most common site for thrombosis to initiate is the calf. Roughly 50% of these thrombi will resolve causing no clinical symptoms, but 15% will extend to more proximal veins. It is the more proximal vein thrombi that pose a risk for dislodging and causing pulmonary embolism, with some estimates as high as 45%, although pulmonary emboli are often diagnosed in the absence of peripheral DVT.3

The resultant obstruction from an acute PE causes increased afterload on the right ventricle. Massive pulmonary emboli that obstruct either the main pulmonary artery or the bifurcation can produce a significant and acute increase in afterload, causing right ventricular end diastolic distention. As the right ventricle’s pressure increases, the interventricular septum can begin to shift into the left ventricle and impede cardiac outflow. The additional distention can be compensated for with Frank-Starling preload reserve, but only to a point and at the cost of increased oxygen demand. The increased subendocardial right ventricular pressure is exerted on the coronary artery vasculature, restricting the flow of oxygenated blood to the ventricular wall in the face of increased oxygen consumption. Without sufficient reserve, the increasing stress can cause right ventricular failure.

It follows that the larger the size of the embolus, the larger the physiologic stress. Prior research has noted that two-thirds of all patients with fatal PE will die within one hour of presentation. However, only half of these deaths are attributed directly to anatomically massive emboli as defined by the occlusion of greater than 50% of the pulmonary vasculature or two or more lobar arteries. It is believed that the remaining cases are attributed to smaller or recurrent emboli.4 It appears that the clinical outcome of patients suffering PE can be stratified with both size of the embolus and with underlying cardiopulmonary status. Patients with poor baseline have a smaller reserve to compensate for the acute physiologic changes and stresses caused by a PE. It has been suggested that the term major pulmonary embolism should therefore not only reflect the size of the embolus, but also the patient’s underlying cardiopulmonary status.5

The International Cooperative Pulmonary Embolism Registry (ICOPER) demonstrated overall three-month mortality for all PE patients to be 17.4%.6 (See Tables 1 and 2.) Those patients with undiagnosed and untreated pulmonary embolism are at further risk of continued symptoms, pulmonary hypertension, and right heart failure. A related entity, known as chronic thromboembolic pulmonary hypertension (CTEPH), is thought to be the result of longstanding, non-resolving thromboemboli or from recurrent emboli. The increased pulmonary vascular resistance steadily increases chronic strain on the right heart and manifests as a clinical entity that is virtually indistinguishable from pulmonary arterial hypertension and carries a relatively poor prognosis.7

History of Diagnostics

Clinical Decision Rules. The use of clinical decision tools was popularized in the 1990s and early 2000s as a direct result of the incorporation of D-dimer assays in the diagnostic workup for venous thromboembolism. The use of d-dimer as a sensitive but nonspecific marker for the diagnosis of PE became a valuable tool for clinicians, but was only useful in low-risk patients for reducing the post-test probability of PE to less than 2%. If a patient were considered to be at high risk, a negative d-dimer would not safely exclude the diagnosis of PE and would be of little clinical yield. The development and validation of various clinical rules became important in distinguishing a category of patients who could be ruled out with a negative D-dimer.

Table 1. Classification of Pulmonary Embolism by Severity8,9

European Classification

American Classification

Adapted from the 2014 ESC Guidelines on the Diagnosis and Management of Acute Pulmonary Embolism8 and the AHA Guidelines for Management of Massive and Submassive Pulmonary Embolism, Iliofemoral DVT and Chronic Thromboembolic Pulmonary Hypertension.9


Hemodynamically unstable patients with suspected PE


Sustained systolic blood pressure < 90 mmHg OR

Sustained systolic pressure drop ≥ 40 mmHg

Not caused by new-onset arrhythmia, hypovolemia or sepsis


-Sustained hypotension (systolic < 90 mmHg) or requiring inotropic support not due to a cause other than PE such as arrhythmia, hypovolemia, sepsis, or LV dysfunction


-Persistent profound bradycardia (heart rate < 40 bpm)


Patients without shock or hypotension with sPESI score ≥ 1 or PESI Class ≥ III


Both RV dysfunction on echo or CT AND elevated cardiac biomarkers


-Intermediate-Risk without RV dysfunction

-PESI Class I-II or sPESI of 0 with elevated cardiac biomarkers or signs of RV dysfunction


Acute PE without systemic hypotension but with:

-RV dysfunction

RV dilation

Elevation in BNP, N-type pro-BNP

ECG changes

Elevation of troponin I or T


Patients with sPESI scores of 0 or PESI Class I-II


Acute PE and the absence of the clinical markers of adverse prognosis that define massive or submassive PE

Geneva Score. The original Geneva score, described by Wicki et al in 2001, was described in a single hospital study from Switzerland.11 Parameters, including risk factors, clinical signs, room air arterial blood gas, and chest radiograph interpretation, were used to risk stratify patients in a scoring algorithm from 0 to 16 points. Patients would then be categorized in low-, intermediate-, and high-risk groups depending on their total score. Validation studies since the initial publication have found the incidence of pulmonary embolism to be 10%, 38%, and 81% for low-, intermediate-, and high-risk groups, respectively.

Le Gal et al created a revised Geneva Score in 2006 as an adjustment to the original Geneva Score, which was independent of diagnostic testing, including blood gas and chest radiograph imaging.12 This revised scoring system ranks patients on a spectrum from 0 to 25 points primarily by using risk factors, and clinical signs and symptoms. Validations for this study found patients in low-, intermediate-, and high-risk categories to have a probability of pulmonary embolism to be 8%, 29%, and 74%, respectively.

Table 2. Pulmonary Embolism Severity Index (PESI)



Simplified PESI (sPESI)

Adapted from the 2014 ESC Guidelines on the Diagnosis and Management of Acute Pulmonary Embolism10


Age in years

1 point if > 80 years

Male gender

10 points

1 point


10 points

1 point

Chronic heart failure

10 points

1 point

Chronic pulmonary disease

20 points

1 point

Tachycardia ≥ 110 bpm

20 points

1 point

Systolic BP < 100 mmHg

30 points

1 point

Respiratory rate > 30/minute

20 points


Temperature < 36?C

20 points


Altered Mental Status

60 points


Arterial oxyhemoglobin saturation < 90%

20 points

1 point


Class I: ≤ 65 points very low 30-day mortality (0-1.6%)

Class II: 66-85 points low mortality risk (1.7-3.5%)

Class III: 86-105 points moderate mortality risk (3.2-7.1%)

Class IV: 106-125 points high mortality risk (4.0-11.4%)

Class V: ≥ 125 points very high mortality risk (10.0-24.5%)

0 points 30-day mortality risk 1.0%

≥ 1 point(s) 30-day mortality risk 10.9%

A simplified Revised Geneva Score was created by Klok et al in 2008 to improve on the weighting system of the previous two Geneva Score algorithms.13 (See Table 3.) This system used a single point for each of the elements, with the exception of tachycardia greater than 95 beats per minute, which received 2 points. The low-, intermediate-, and high-risk probabilities were found to be 8%, 29%, and 64%, respectively. Additionally, a PE unlikely and PE likely categorization scheme was used for patients with 0 to 2 points, and those who scored higher, to help guide the use of a d-dimer to safely exclude pulmonary embolism. The combined use of the simplified Revised Geneva Score with a negative d-dimer in low-risk patients safely excluded 330 patients for thromboembolic disease at presentation and at three-month follow-up.

A unique advantage to the Revised Geneva Score is that it is based entirely on objective data without any use of physician judgment on the likelihood of VTE. The Revised Geneva Score has been validated to be effective at triaging patients suspected of having PE.14 It can be used in the emergency department as an objective alternative to the more widely used Wells Score.

Table 3. Simplified Revised Geneva Score



Age > 65 years


Pervious DVT or PE


Surgery or fracture within 1 month


Active malignancy


Unilateral lower limb pain




Pain on deep vein palpation of lower limb and unilateral edema


Heart rate 75-94 bpm


Heart rate greater than
94 bpm


Score ≤ 2 unlikely to have a current PE (probability 3% with negative d-dimer)

Wells Criteria. The Wells Criteria was introduced in 1995 as a clinical decision rule based on a review of the available literature and consensus opinion.15 Patients were stratified into 10 possible outcomes based on the clinician’s suspicion and other clinical data points that were then categorized as low-, moderate-, and high-risk groups with a rate of PE of 3.4%, 27.8%, and 78.4%, respectively. The algorithmic stratified pathway of the original Wells Criteria did not lend itself to be used as an objective clinical decision tool. The physician initially assessed patients based on their signs and symptoms and divided patients into typical, atypical, and severe groups. Then the patients were further subdivided based on whether an alternative diagnosis was more likely, ultimately risk stratifying patients into low-, moderate-, and high-risk categories.

Wells reinterpreted the initial data retrospectively and developed a simpler scoring system that could be used in conjunction with d-dimer assays to exclude pulmonary embolism. Using this scoring system, which included the clinician’s suspicion of pulmonary embolism, the patients could be categorized using various cutoffs into a low-, moderate-, and high-risk stratification or into a PE unlikely and PE likely scheme. (See Table 4.) The latter scheme was created with the specific intention of being used with d-dimer to exclude PE. This was later validated in a multicenter prospective trial showing a negative predictive value for the use of the low clinical probability scoring system and a negative d-dimer to be 99.5%.16

Table 4. Wells Score



Traditional interpretation

Score > 6.0 – High (probability 59%)

Score 2.0 – 6.0 – Moderate (probability 29%)

Score < 2 – Low (probability 15%)

Alternative Interpretation

Score > 4 – PE likely. Consider imaging.

Score 4 or less – PE unlikely. Consider d-dimer.

Clinically suspected DVT


Alternate diagnosis is less likely than PE


Tachycardia (heart rate > 100 bpm)


Immobilization (≥ 3 d) or surgery in previous four weeks


History of DVT or PE




Malignancy (with treatment within 6 months) or palliative


While the Wells Criteria is widely used, the varying cutoffs used to stratify patients complicate its use. Additionally, the inclusion of “an alternative diagnosis is less likely than PE” as being worth 3 points by itself places a patient in an intermediate-risk group. As this relies intrinsically on clinician judgment, the Wells Criteria cannot be truly considered an objective clinical decision rule. Studies have shown that the use of clinical judgment allowed for poor inter-observer reproducibility.17 Interestingly, it appears that it is this criterion that has provided for much of the predictive value of the Wells Score.18 Therefore, as long as the emergency physician is comfortable with the use of a subjective judgment of probability in the implementation of a clinical tool, the Wells Score has been validated in the risk stratification of patients with suspected VTE.

Table 5. Kline Score

Is the patient 50 years old or younger with a heart rate less than or equal to the systolic blood pressure (shock index ≤ 1.0)?

If yes, the patient is eligible for d-dimer testing

For patients older than 50 or with a shock index higher than 1.0:

Does the patient have unexplained hypoxemia?

Does the patient have unilateral leg swelling?

Has the patient had surgery requiring general anesthesia in the past 4 weeks?

Does the patient have hemoptysis?

If no to all, the patient is still eligible for d-dimer testing

Kline Rule. The Kline Rule was described in 2002 in a multicenter prospective study of ED patients in urban hospitals. The study focused on criteria for the safe use of d-dimer testing in emergency department patients with suspected pulmonary embolism.19 The authors found that in patients younger than 50 years and with a heart rate less than or equal to the systolic blood pressure (shock index ≤ 1.0) would be eligible. Those who were not would have to go through 4 sequential questions involving hypoxemia, unilateral leg swelling, recent surgery, and a history of hemoptysis. The authors found that the use of this clinical decision tool to find patients eligible for d-dimer study would yield a posttest probability of PE to 1.1%. (See Table 5.)

Table 6. Pulmonary


  • Age ≥ 50 years?

  • Heart Rate ≥ 100 bpm?

  • O2 Saturation on Room Air < 95%?

  • Prior History of DVT/PE?

  • Recent Trauma or Surgery?

  • Hemoptysis?

  • Exogenous Estrogen?

  • Unilateral Leg Swelling?

    If no to all criteria and with clinician’s pre-test probability < 15%, PERC Rule criteria are satisfied and no further workup needed with < 2% probability of PE

    PERC Rule. While the Kline Rule is used to determine an optimal patient population for excluding the diagnosis of pulmonary embolism with d-dimer testing, a separate rule called the Pulmonary Embolism Rule-out Criteria (PERC) was used to find patients who could be ruled out without diagnostic testing. Created by the same author as the Kline Rule, the PERC algorithm uses a series of eight questions based on the patient’s vital signs, demographics, and history, in addition to the clinician’s suspicion, to classify patients as being ruled out without further workup.20 The original prevalence in the low-risk group was calculated using logistic regression to be around 1.8%. (See Table 6.) This was considered to be the equipoise between the benefits and risks of further investigation including potential adverse outcomes due to CT-PA studies.

    Several validation studies of the PERC rule have been done and have found similar results since the original publication.21,22 However, although the data were collected prospectively, the PERC analysis was done retrospectively, limiting the recommendations for its use. Additionally, the patients were selected into a low-risk group based on subjective analysis of their presentation. One prospective study of the PERC rule found the prevalence in the PERC-negative group to be 5.4% and when combined with the revised Geneva score to be 6.4%.23 The study concluded that the PERC rule, used alone or when combined with the Geneva score, was unacceptable in safely excluding the diagnosis of PE without further testing. While the PERC rule is widely used, more research needs to be done to validate an acceptable low-risk group, and clinicians should be very careful in applying it in the emergency department.

    Laboratory Studies

    D-Dimer. Formation. During clot formation, fibrinogen, a soluble plasma glycoprotein, is cleaved by thrombin into highly self-adhesive fibrin monomers. Activated factor XIII then covalently cross links fibrin monomers, forming protofibrils and an insoluble fibrin gel initiating the clot formation. When plasmin begins degrading the fibrin gel, it does so by cleaving the polymers at several sites. One such degradation point yields a product of a cross-linked fibrin polymer with two D fragments, thusly named d-dimer. These soluble d-dimer molecules can be found from fibrin degradation of both soluble fibrin polymers and from insoluble fibrin gels in clots.24

    Detections. The detection of these soluble d-dimer molecules requires monoclonal antibodies with varying specificities. Depending on the epitope detected, detection mechanism, instrument calibration, and laboratory standards, the cut-offs for positive results differ. There are several ways of testing d-dimer levels; however, the two most studied and validated means of quantifying serum d-dimer concentrations are the enzyme-linked immunosorbent assay (ELISA) and the second-generation latex agglutination tests (immunoturbidimetric tests) with a commonly used cutoff of 500 µg/L.25 It is therefore imperative for the emergency physician to confirm that their laboratory uses the validated ELISA or latex agglutination tests prior to interpreting the results.

    It has been noted that physiologic d-dimer concentrations increase with age.26 In the ADJUST-PE project, a large multicenter prospective study in Europe, researchers validated the use of an age-dependent d-dimer cutoff for patients older than the age of 50.27 By using the patient’s age multiplied by 10 in conjunction with a Wells Score, the researchers were able to safely exclude PE without the inclusion of any false negatives. Patients younger than the age of 50 were subject to the standard 500 µg/L cutoff. This is a particularly useful tool when considering the use of a CT with IV contrast in elderly patients.

    Utility. A meta-analysis by Stein et al. found that the ELISA and quantitative rapid ELISA had negative likelihood ratios that performed more favorably than a normal to near-normal lung scan and was comparable to a negative result with a lower-extremity duplex ultrasonography. Given its high negative predictive value and lack of specificity, it has been used as a unidirectional diagnostic tool where a negative result can be used to safely exclude acute venous thrombosis in low- and moderate-risk patients, while a positive result’s relevance is unclear. In high-risk patients, the clinical utility of a normal d-dimer concentration is lost, as a false-negative rate becomes too high and further diagnostic testing is needed.

    Diagnostic Imaging

    V/Q Scan. The ventilation/perfusion or V/Q scan has historically been used as a first-line diagnostic imaging test for pulmonary embolism. The exam is performed in conjunction with a chest X-ray in two phases. The ventilation phase involves the inhalation of a gaseous radionucleotide that is administered to the patient during inhalation. The perfusion phase includes an intravenous injection of radioactive technetium macro aggregated albumin (Tc99m-MAA) and a comparison is made to find areas with ventilation but lacking in perfusion. This test may be performed in conjunction with a 3-dimentional SPECT image. The result is then categorized as normal, low probability, intermediate probability, or high probability based on the degree of suspected mismatch. (See Table 7.)

    Unfortunately, while high probability scans in the context of high pretest suspicion are found to be helpful in the diagnosis of PE, the interpretation of intermediate probability scans and high probability scans in intermediate risk patients remains more problematic. Patients with a low probability of PE with a high-probability scan have demonstrated a documented PE via direct angiography of about 50%. Similarly, patients with an intermediate probability scan with a high clinical likelihood had a probability of 66%. It was found that only 4% of patients with a low or normal probability scan in the setting of low-clinical suspicion had a pulmonary embolism.28 In other words, only when the clinical suspicion and the V/Q test result are concordant, are the results of a V/Q scan useful. Additionally, shortages in availability of Tc-99m limit the usage of V/Q scans. In the late 2000s, a sharp decrease in the supply of molybdenum-99, a precursor to technetium-99m, caused a worldwide shortage of the isotope.29

    Pulmonary Angiography. Historically, conventional catheter-directed pulmonary angiography was considered the gold standard for the diagnosis of pulmonary embolism. Although direct visualization of a filling defect under fluoroscopy was considered definitive, limitations in technique and interpreter agreement limited its utility with subsegmental arteries.30 Additionally, this test is invasive, time-consuming, and has been outperformed by more modern techniques.

    CT Pulmonary Angiography. First introduced in 1992, the single detector spiral CT pulmonary angiogram (CT-PA) was used in the detection of acute pulmonary embolism.31 Diagnosis was based on direct visualization of intraluminal clots by detecting complete or partial filling defects, “railway track” signs, and mural defects. These results were then compared to direct pulmonary angiography via catheterization. The initial study noted some false positives from intersegmental lymph nodes and asymmetric pulmonary filling, but found that the CT-PA was sufficiently reliable to detect thromboembolism in second- to fourth-generation pulmonary vessels. Additionally, the CT-PA allows for clinicians to detect alternate diagnoses, making it an attractive diagnostic tool and of greater use over the V/Q scan.

    The Prospective Investigation of Pulmonary Embolism Diagnosis II Trial (PIOPED II) was designed to determine the reliability of a multidetector CT-PA and whether the additional use of the Wells score improved the ability to detect or rule out pulmonary embolism.32 The study found that using CTA in conjunction with lower extremity CTV maximized sensitivity over CTA alone when used correctly.

    Since the publication of the PIOPED II, adherence to the correct selection of patients who would benefit most from a CT-PA has been found to be deficient. One study found that nearly half of CT exams meant for assessing pulmonary embolism were not ordered in concordance with accepted guidelines and the initial PIOPED II study.33 The selection of patients without consideration of pretest probability with clinical tools or d-dimer studies increases the chance of a false positive, leading to inappropriate anticoagulation. In other words, by haphazardly ordering CT-PA studies too broadly, the emergency physician risks both the harms of the diagnostic tool with radiation and nephrotoxic dye and an unacceptably high false-positive rate that can lead to the inappropriate use of anticoagulation.

    Magnetic Resonance Angiography. The use of MRA has been suggested in the diagnosis of acute pulmonary embolism. The lack of ionizing radiation is of particular interest for pregnant and pediatric patients, but the diagnostic yield is still questionable. In addition, the limited availability in most emergency departments and the requirement that the patient hold their breath for 15-20 seconds add further limitations to its use. The IRM-EP study found that while MRA had good diagnostic yield for proximal pulmonary embolism, the sensitivity dropped significantly for segmental and subsegmental pulmonary emboli.34 A recently published, single site study found the outcomes for patients with negative MRA to be excellent, with a negative predictive value of 97% at 3 months and 96% at 1 year. The authors hypothesized that while the loss in sensitivity from the IRM-EP study was attributed to subsegmental emboli, they may be of questionable clinical significance.35 Further research needs to be done before MRA is used as a first test for the exclusion of pulmonary embolism in the emergency department.

    Special Populations

    It is important to consider special populations when dealing with the diagnosis of pulmonary embolism. Pregnant, pediatric, and oncologic patients, among others, pose specific challenges that complicate an already intricate workup.

    Table 7. V/Q Scan Interpretation




    No perfusion deficit

    Low probability

    Perfusion deficit with matched ventilation deficit

    Intermediate probability

    Perfusion deficit that corresponds to parenchymal abnormality on chest X-ray

    High probability

    Multiple segmental perfusion deficits with normal ventilation

    Pregnancy. Pregnant patients are at particular risk for the development of VTE and PE. Pregnant patients are physiologically hypercoagulable in preparation for hemorrhage during delivery, making them seven to 10 times more likely to suffer a thrombosis than age-matched controls.36 Interestingly, 85% of lower extremity deep vein thromboses in pregnant patients are on the left side due to compression of the left iliac vein by the right iliac artery and the uterus.37 Pulmonary embolism has been cited as the leading cause of maternal mortality during pregnancy and the six-week postpartum stage. The risk of VTE has been estimated to be around 5-12 per 10,000 during pregnancy and 3-7 per 10,000 in the postpartum period.38

    The classic signs and symptoms of PE seen in the general population are seen normally during pregnancy. For this reason, many of the clinical tools used for risk stratifying patients do not apply to pregnant patients. Furthermore, the diagnosis of pulmonary embolism relies on radiologic studies that expose the mother and developing fetus to potentially harmful ionizing radiation.

    D-dimer levels fluctuate during normal pregnancies. Although the d-dimer concentration level is expected to be normal during the first trimester, it begins to rise during the second trimester before dropping off after delivery and reaching a normal level after 4-6 weeks postpartum.39 These normally elevated levels can lead clinicians to expose pregnant patients to unneeded radiographic studies. It remains to be seen if a gestational age-based cutoff can be developed and used reliably, but given the difficulty in establishing a safe cutoff in the general population. However, given the normally elevated serum d-dimer levels during pregnancy, a negative result has been shown to be useful in ruling out a pulmonary embolus in pregnant and post-partum patients.40 One meta-analysis of five studies showed that a negative d-dimer in the presence of clinical suspicion could be used to rule out PE in pregnant patients. It was recommended that because the false-positive rate was so high using traditional cutoffs, that its use be limited to the first two trimesters.41 More research clearly needs to be done before the emergency physician uses D-dimers in pregnancy to safely rule out PE during pregnancy.

    Table 8. Rates of DVT/PE by Malignancy


    Rate of DVT/PE per 10,000 patients

    Based on an analysis of more than 1.2 million Medicare patients admitted with a malignancy (Levitan et al. 1999)61



































    There is much trepidation about the use of advanced imaging studies in pregnant patients to avoid the potential teratogenic and oncogenic effects on the developing fetus. One approach to avoid ionizing radiation exposure in the pregnant patient is to image the legs with duplex doppler ultrasonography. Identifying a proximal DVT indicates the same treatment as a PE. If the ultrasound is negative, then advanced imaging is considered.

    The radiation dose to the fetus during V/Q scans has been estimated to be 100-370 µGy and 280 µGy to the mother’s breast.42 This is compared to the generally accepted safe teratogenic threshold of 0.1 Gy that has been extrapolated by limited human data and mouse and rat models.43 The oncogenic risk of radiation to the developing fetus is considered dangerous at 0.01 Gy above background radiation, which represents a 0.01% increase in risk of cancer before the age of 20 years.44 This can be compared to a combined radiation dose to the fetus of 0.004 Gy for a chest radiograph, ventilation perfusion scan and conventional pulmonary angiogram.45 Given these data, it seems that the mortality associated with a missed diagnosis of a pulmonary embolism outweighs the potential teratogenic and oncogenic risks. If the emergency physician suspects pulmonary embolism in a pregnant patient, the risks and benefits should be communicated with the patient and the harms of a missed diagnosis of a PE should be emphasized. An often overly exaggerated and poorly understood risk of radiation to the developing fetus should not cloud the physician’s judgment.

    The V/Q scan has been used widely in the pregnant population with some studies showing that it is the most frequently used diagnostic tool for this group.46 Given the use of radioactive isotopes in both the ventilation and perfusion phase of this test, it is of some concern. It has been suggested that performing the perfusion scan with a reduced dose should be performed first with a subsequent ventilation phase only in the event that a defect is noted to potentially limit the radioactive exposure.47

    The negative predictive value of a normal study in pregnant patients has been found to be acceptable. Two small retrospective studies have shown a lack of VTE events in pregnant patients with normal V/Q scans at follow-up, making a V/Q study a reasonable first test for ruling out PE.48 Unfortunately, non-diagnostic results would require further imaging and radiation.

    There are very limited data on the accuracy of CT-PA in pregnancy. It has been found that pulmonary artery opacification is limited in pregnant patients, which may obscure results and necessitate the need for adjusted protocols.49

    Table 9. Recommended Diagnostic Strategy

    Clinical Situation

    Recommended Approach

    Previously healthy stable patient

    • Risk stratify using one of the validated tools
    • Measure d-dimer in low and intermediate risk patients using one of the validated high-sensitivity assays
    • Image with CT-PA for high-risk patients and patients with elevated age-adjusted d-dimer

    Patient with severe IV contrast allergy

    • Risk stratify using one of the validated tools
    • Measure d-dimer in low and intermediate risk patients using one of the validated high-sensitivity assays
    • Image with radionuclide ventilation-perfusion lung scan for high-risk patients and patients with elevated age-adjusted d-dimer

    Unstable patient

    • Echocardiography
    • Treat with unfractionated heparin if signs of RV overload and consider fibrinolytic therapy
    • When patient stabilized, perform CT-PA

    Patient with active cancer

    • Measure d-dimer using one of the validated high-sensitivity assays
    • Image with CT-PA for patients with elevated age-adjusted d-dimer

    Pregnant patient

    • Measure d-dimer using one of the validated high-sensitivity assays during the first two trimesters
    • Image legs with duplex doppler ultrasound, treat with weight-adjusted low molecular weight heparin if proximal DVT identified
    • If ultrasound negative, image with radionuclide perfusion lung scan for patients with elevated d-dimer

    Patient with apparent exacerbation of underlying cardiopulmonary disease

    • Measure d-dimer using one of the validated high-sensitivity assays
    • Image with CT-PA for patients with elevated age-adjusted d-dimer

    The fetal radiation for CT-PA dose appears to be equal to or lower than VQ scans, although there is some variability with how the precise protocol.50 However, CT-PA exposes the mother’s breasts to a significant amount of radiation that has been attributed to an increased risk of breast cancer, with an estimated 0.7% lifetime risk, which can be attenuated with radioprotective breast shielding.51 Therefore, while the risk of radiation is not insignificant, it should not overwhelm the harm of a missed diagnosis. These risks can be communicated to the patient and it is important for the emergency physician to explain these studies as clearly as possible.

    Oncology. Cancer has been identified as an independent risk factor for the development of VTE and PE. The tumor cell procoagulants, tumor associated inflammatory procoagulants, and mediators of platelet adhesion generated by tumor cells have been attributed to this hypercoagulable state.52 A single site retrospective outpatient study found a 2.87% incidence of PE in oncology patients, as opposed to an estimated 0.11% incidence in the general population.53 It has been estimated that cancer is responsible for 18% of all VTE. Patients with cancer are thought to be at a seven-fold increased risk of VTE.54

    However, it appears that not all cancers represent an equivalent elevation of risk of VTE. Several retrospective studies have found that certain cancers including CNS, pancreatic, upper GI, and lung malignancies have a particularly high incidence of PE, possibly attributed to immobility. (See Table 8.) The effect was most pronounced with CNS neoplasms with an estimated odds ratio of 27.05. This was compared to a relatively low risk of PE in patients suffering from hematologic and breast cancers.55 It also appears that among cancer patients, the risk is highest in the first three months after diagnosis and in the presence of distant metastases.56

    To make matters worse, frequently used chemotherapy agents, erythropoiesis-stimulating agents, and indwelling catheters all increase the risk of VTE significantly. Blom et al. found a three-fold increased risk for VTE attributed to chemotherapy over the already elevated risk for oncologic patients. Additionally, hormonal therapies and erythropoietin-stimulating agents have been associated with additional risks for VTE.57

    It is believed that there are several mechanisms responsible for the increased risk of VTE in cancer patients. It has been found that a majority of patients with cancer had increased levels of coagulation factors V, VIII, IX, and XI, along with deficiencies in ADAMTS13, causing the accumulation of large von Willebrand multimers.58,59 Additionally, many tumors have been found to cause abnormal activation of tissue factor, pathologically triggering the extrinsic pathway.60 These pathophysiologic changes in cancer are compounded by other factors frequently seen in cancer patients, including immobility, the use of long-standing catheters, and increased age.

    Although a patient’s history of active cancer or treatment is included in many clinical decision rules regardless of type of cancer, it may be helpful for the emergency physician to be cognizant of the fact that not all cancers carry an equivalent risk. Additionally, it is important to understand that the debility imparted by these cancers and their treatments carry independent risks for the development of VTE.

    Pediatrics. Pulmonary embolism is a rare diagnosis in pediatric patients, with an estimated yearly incidence of about 0.9 per 100,000.62 The lack of validated clinical tools like the Wells Score and PERC for children limits the clinician. However, similar risk factors for PE and VTE apply to these patients, including obesity, immobility, estrogen therapy, central venous catheter, malignancy, prothrombotic diseases, surgery, and trauma.63 It should be noted that the pretest probabilities for a 12-year-old female with no medical history and a 17-year-old smoker on oral contraceptives are radically different despite both being categorized as pediatric.

    The risks of cumulative ionizing radiation doses to pediatric patients undergoing CT scans are of particular concern to pediatric patients. There are unique considerations when utilizing CT scans in pediatric patients. The thyroid gland, breast, and gonads, are structures that are particularly sensitive to radiation and have a higher potential for developing into cancer.64 It is thought that the combination of smaller organs receiving radiation and a longer lifetime post-exposure to manifest potential oncogenic effects culminate in a significantly higher risk to these patients.65

    It is difficult to precisely quantify and communicate the risks and benefits to patients and family, especially given the lack of precise science in the literature. A position statement by the American Associations of Physicists in Medicine states that the risks of medical imaging below 50 mSv (significantly higher than a CT scan) are too low to be detectible and may be nonexistent and goes on to state that any predictions about the hypothetically attributable cancer risks would be inherently speculative.66

    Obesity. Obesity has been noted as an independent risk factor for the development of DVT and, given the growing epidemic of obesity in this country, it is important to keep this in mind when risk stratifying patients. Given the risks of other comorbidities frequently seen in obese patients, it is important to keep pulmonary embolism in the differential diagnosis of shortness of breath in these patients.67

    Renal Patients. Patients who require chronic renal dialysis are at increased risk of VTE. While platelet dysfunction and coagulopathy are known to occur in this population, it may be counterintuitive that these patients are at higher risk for VTE. It has been hypothesized that this may be due in part to other traditional risk factors, including immobility, advanced age, indwelling catheters, and frequent surgical procedures. These patients are also known to have a higher prevalence of systemic inflammation and endothelial damage, both of which may contribute to the development of thrombi.68


    The diagnosis of pulmonary embolism is particularly challenging to the emergency physician, as it requires careful case-specific considerations, an involved, and sometimes time-consuming diagnostic workup, and is treated best by quick, timely, and accurate diagnosis and initiation of treatment. Given the imperfect tools available for diagnosis, it is imperative to maximize the diagnostic accuracy. Evidence-based literature supports the use of decision rules to estimate the pretest probability of PE before going on to diagnostic imaging. (See Table 9.) Two decision rules are most commonly used in clinical practice: the Wells Criteria and the Revised Geneva Score. The Wells Criteria takes into consideration physician clinical judgment in the scoring, whereas the Revised Geneva Score is based on objective information only. Even after using the clinical decision rules to determine a patient’s pre-test probability of a PE, failure to select the right diagnostic study or failure to accurately interpret the results can lead to a missed diagnosis or an inappropriate and potentially dangerous initiation of anticoagulation. As further academic studies are conducted and as technologies continue to improve in the area of diagnosis of pulmonary embolism, our understanding and accuracy will continue to evolve.


    1. Bell WRT. The clinical features of submassive and massive pulmonary emboli. Am J Med 1977;62:355-360.
    2. Shapira-Rootman M. The prevalence of pulmonary embolism among patients suffering from acute exacerbations of chronic obstructive pulmonary disease. Emerg Radiol 2014 Oct 26. [Epub ahead of print]
    3. Rahimtoola A. Acute pulmonary embolism: An update on diagnosis and management. Curr Probl Cardiol 2005;30:61-114.
    4. Wood KE. Major pulmonary embolism: Review of a pathophysiologic approach to the golden hour of hemodynamically significant pulmonary embolism. Chest 2002;121:3:877.
    5. Hoagland PM. Massive pulmonary embolism: Pulmonary embolism and deep vein thrombosis. Philadelphia, PA: WB Saunders, 1985: 179-208.
    6. Goldhaber SZ. Acute pulmonary embolism: Clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999;353:1396-1389.
    7. Pepke-Zaba J. Chronic thromboembolic pulmonary hypertension (CTEPH): Results from an international prospective registry. Circulation 2011;124(18):1973-1981.
    8. Konstantinides SV. 2014 ESC Guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J 2014 Aug 29. [Epub ahead of print]
    9. Jaff MR. AHA. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis and chronic thromboembolic pulmonary hypertension. Circulation 2011;123:1788-1830.
    10. Konstantinides SV. 2014 ESC Guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J 2014 Aug 29. [Epub ahead of print]
    11. Wicki J, Perneger TV, Junod AF, et al. Assessing clinical probability of pulmonary embolism in the emergency ward. Arch Intern Med 2001;161:92-97.
    12. Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmonary embolism in the emergency department: The revised Geneva score. Ann Intern Med 2006;144:165-171.
    13. Klok FA, Mos IC, Nijkeute M, et al. Simplification of the revised Geneva score for assessing clinical probability of pulmonary embolism. Arch Intern Med 2008;169:2131-2136.
    14. Le GG et al. Prediction of pulmonary embolism in the emergency department: The revised Geneva Score. Ann Intern Med 2006;144:165-171.
    15. Wells PS, Ginsberg JS, Anderson DR, et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998;129:997-1005.
    16. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001;135:98-107.
    17. Wolf SJ, et al. Prospective validation of Wells Criteria in the evaluation of patients with suspected pulmonary embolism. Ann Emerg Med 2004;44:503-510.
    18. Klok FA. Alternative diagnosis other than pulmonary embolism as a subjective variable in the Wells clinical decision rule: Not so bad after all. J Thromb Haemost 2007;5:1079-1080.
    19. Kline JA, Nelson RD, Jackson RE, Courtney DM. Criteria for the safe use of d-dimer testing in emergency department patients with suspected pulmonary embolism: A multicenter US study. Ann Emerg Med 2002;39(2):144-152.
    20. Kline JA. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6:772-780.
    21. Wolf SJ, et al. Assessment of the pulmonary embolism rule-out criteria rule for evaluation of suspected pulmonary embolism in the emergency department. Am J Emerg Med 2008;26:181-185.
    22. Kline JA. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6:772-780.
    23. Hugli O, et al. The pulmonary embolism rule-out criteria rule does not safely exclude pulmonary embolism. J Thromb Haemost 2011;9:300-304.
    24. Adam SS, et al. D-dimer antigen: Current concepts and future prospects. Blood 2009;113(13):2878-2887.
    25. Stein PD, et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: A systematic review. Ann Intern Med 2004;140(8):589-602.
    26. Righini M. Effects of age on the performance of common diagnostic tests for pulmonary embolism. Am J Med 2000;109(5):357-361.
    27. Righini M. Age-adjusted d-dimer cutoff levels to rule out pulmonary embolism: The ADJUST-PE study. JAMA 2014;311(11)1117-1124.
    28. Gray HW. Pulmonary embolism exclusion: A practical approach to low probability using the PIOPED data. Prospective Investigation of Pulmonary Embolism Diagnosis. Eur J Nucl Med 1998;25:271-276.
    29. Wald ML. Radioactive drug for tests is in short supply. The New York Times July 23, 2009.
    30. Stein PD. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism. Radiology 1999;210:689-691.
    31. Remy-Jardin M, Remy J, Wattinne L, et al. Central pulmonary thromboembolism: Diagnosis with spiral volumetric CT with the single-breath-hold technique — comparison with pulmonary angiography. Radiology 1992;185:381-387.
    32. Stein PD, et al. Multidetector computed tomography for acute pulmonary embolism. Curr Opin Pulm Med 2007;13(5):384-388.
    33. Adams DM. Adherence to PIOPED II investigators’ recommendations for computed tomography pulmonary angiography. Am J Med 2013;126;36-42.
    34. Revel MP. Diagnostic accuracy of magnetic resonance imaging for an acute pulmonary embolism: Results of the ‘IRM-EP’ study. J Thromb Haemost 2012;10(5):743-750.
    35. Schiebler ML. Effectiveness of MR angiography for the primary diagnosis of acute pulmonary embolism: Clinical outcomes at 3 months and 1 year. J Magn Reson Imaging 2013;38(4):914-925.
    36. Ray JG. Deep vein thrombosis during pregnancy and the puerperium: A meta analysis of the period of risk and the leg of presentation. Obstet Gynecol Surv 1999;54:265-271.
    37. Goldhaber SZ. A prospective registry of 5451 patients with ultrasound-confirmed deep vein thrombosis. Am J Cardiol 2004;93:259-262.
    38. Simpson EL. Venous thromboembolism in pregnancy and the puerperium: Incidence and additional risk factors from a London perinatal database. Br J Obstet Gynaecol 2001;108:56-60.
    39. Morse M. Establishing a normal range for d-dimer levels through pregnancy to aid in the diagnosis of pulmonary embolism and deep vein thrombosis. J Thromb Haemost 2004;2:1202-1204.
    40. Choi H. The diagnostic utility of d-dimer, and other clinical variables in pregnant and postpartum patients with suspected acute pulmonary embolism. Emerg Med J 2014;31(0):783-784.
    41. BET2: Current evidence does support the use of negative d-dimer to rule out suspected pulmonary embolism in pregnancy. Emerg Med J 2014;31:946-947.
    42. Cook JV. UK agency reports slight increase in radiation exposure. BMJ 2005;330:1229.
    43. International Commission on Radiation Protection. Pregnancy and medical radiation. ICRP Publication 84. Ann ICRP 2000: 3-41.
    44. Valentin J. Biological effects after prenatal irradiation-embryo and fetus (International Commission on Radiologic Protection publican 90). Ann ICRP 2003;33:1-206.
    45. Bourjeily G. Pulmonary embolism in pregnancy. Lancet 2010;375:500-512.
    46. Boiselle PM. Pulmonary embolus in pregnant patients: Survey of ventilation-perfusion imaging policies and practice. Radiology 1998;207:201-206.
    47. Chan W-S. Diagnosis of deep vein thrombosis and pulmonary embolism in pregnancy. Thromb Res 2002;107:85-91.
    48. Chan WS. Suspected pulmonary embolism in pregnancy: Clinical presentation, results of lung scanning, and subsequent maternal and pediatric outcomes. Arch Intern Med 2002;162:1170-1175.
    49. Andreou AK. Does pregnancy affect vascular enhancement in patients undergoing CT pulmonary angiography? Eur Radiol 2008;18(12):2715-2722.
    50. Winer-Muram HT. Pulmonary embolism in pregnant patients: Fetal radiation dose with helical CT. Radiology 2002;224:487-492.
    51. Hopper KD. The breast: In-plane X-ray protection during diagnostic thoracic CT shielding with bismuth radioprotective garments. Radiology 1997;205:853-858.
    52. Rickles FR. Hemostatic alterations in cancer patients. Cancer Metastasis Rev 1992;11:237-248.
    53. Shinagare AB. Incidence of pulmonary embolism in oncologic outpatients at a tertiary cancer center. Cancer 2011;117:3860-3866.
    54. Blom JW. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005;293:715-722.
    55. Shinagare AB. Incidence of pulmonary embolism in oncologic outpatients at a tertiary cancer center. Cancer 2011;117(16):3860-3866.
    56. Blom JW. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005;293:715-722.
    57. Noble S. Epidemiology and pathophysiology of cancer-associated thrombosis. Br J Cancer 2010;102 Suppl 1:S2-9.
    58. Hoffman R. Cancer and thrombosis revisited. Blood Rev 2001;15:61-67.
    59. Oleskowicz L. Deficient activity of von Willebrand’s factor-cleaving protease in patients with disseminated malignancies. Cancer Res 1999;59:2244-2250.
    60. Kakkar AK. Extrinsic-pathway activation in cancer with high factor VIIa and tissue factor. Lancet 1995;346:1004-1005.
    61. Levitan N. Rates of initial and recurrent thromboembolic disease among patients with malignancy vs those without malignancy. Risk analysis using Medicare claims data. Medicine 1999;78:285-291.
    62. Stein PD. Incidence of venous thromboembolism in infants and children: Data from the National Hospital Discharge Survey. J Pediatr 2004;145(4)563-565.
    63. Andrew M. Venous thromboembolic complications (VTE) in children: First analysis of the Canadian Registry of VTE. Blood 1994;83(5):1251-1257.
    64. Brenner DJ. Estimating cancer risks from pediatric CT: Going from the qualitative to the quantitative. Pediatr Radiol 2002;32:228-231.
    65. Huda W. Radiation dose and image quality. Imaging Children Edinburgh, United Kingdom: Elsevier Science; 2003.
    66. AAPM position statement on radiation risks from medical imaging procedures. The American Association of Physicists in Medicine, 2011. http://wwww.aapm.org/org/policies/details.asp?id=318 (accessed September 2014).
    67. Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000;160(22):3415-3420.
    68. Tveit DP. Chronic dialysis patients have high risk for pulmonary embolism. Am J Kidney Dis 2002;39(5):1011-1017.
    69. Jaff MR. Management of Massive and Submassive Pulmonary Embolism, Iliofemoral Deep Vein Thrombosis, and Chronic Thromboembolic Pulmonary Hypertension: A Scientific Statement From the American Heart Association. Circulation 2001;123:1788-1830.
    70. Ghignome M. Volume expansion versus norepinephrine in treatment of a low cardiac output complicated an acute increase in right ventricular afterload in dogs. Anesthesiology 1984;6:132-135.
    71. Mercat A. Hemodynamic effects of fluid loading in acute massive pulmonary embolism. Crit Care Med 1999;27:540-544.
    72. Layish DT. Pharmacologic hemodynamic support in massive pulmonary embolism. Chest 1997;111:218-224.
    73. Aujesky D. Outpatient versus inpatient treatment for patients with acute pulmonary embolism: An international, open-label, randomized, non-inferiority trial. Lancet 2011;378:41-48.
    74. Piran S. Outpatient treatment of symptomatic pulmonary embolism: A systematic review and meta-analysis. Thrombosis Research 2013;132:515-519.
    75. Chatterjee S. Thrombolysis for pulmonary embolism and risk of all-cause mortality, major bleeding, and intracranial hemorrhage: A meta-analysis. JAMA 2014;311(23):2414-2421.