PE in Pregnancy: A Complicated Diagnosis

Amal Mattu, MD


August 09, 2010


Pulmonary embolism (PE) is a leading cause of death among pregnant women in the developed world.[1] The fact that PE remains such a threat is a clear indication of the difficulty in diagnosing it. The coexistence of pregnancy makes the workup and management of PE even more problematic, for many reasons. First, there are 2 patients at risk rather than 1. Second, overdiagnosis results in unnecessary, dangerous treatments that jeopardize both patients and makes the pregnancy and delivery far more complicated. Third, the usual imaging modalities, which we use without a second thought in nonpregnant patients, suddenly become more complicated in pregnant patients.

The following article provides an excellent review of PE in pregnancy and offers helpful tips to improve our ability to diagnose and manage PE in pregnant patients.

Pulmonary Embolism in Pregnancy

Bourjeily G, Paidas M, Khalil H, Rosene-Montella K, Rodger M
Lancet. 2010;375:500-512


There is similar risk for venous thromboembolism (VTE) in all 3 trimesters of pregnancy, although the incidence of deep venous thrombosis (DVT) is approximately 3 times higher than that of PE during pregnancy.

Pregnancy-associated DVT occurs on the left side of the body in 85% of cases, the likely result of compression of the left iliac vein by the right iliac artery as well as compression by the gravid uterus. Isolated pelvic DVT is much more common in pregnancy or in the early postpartum period than in nonpregnant patients.

All 3 elements of Virchow's triad are present during pregnancy and the early postpartum period; they are hypercoagulability, vascular damage, and venous stasis.

Neither clinical judgment nor clinical decision rules that are commonly used in nonpregnant patients have proven to be effective for diagnosing pregnant patients with PE. For example, D-dimer levels rise gradually during pregnancy and then drop in the immediate postpartum period but do not return to normal until 4-6 weeks postpartum. A normal D-dimer level appears to have a high negative predictive value in patients with a low clinical suspicion for VTE, but false-positive levels are very common.

The decision to use imaging modalities that produce radiation exposure in pregnant patients is difficult because of concerns with teratogenicity. Radiation exposure of 1 Gy at any stage of pregnancy is regarded as the level above which the risk for induction of congenital abnormalities is possible. The following is a list of imaging tools commonly used in PE diagnosis.

Ventilation Perfusion Scintigraphy

The fetal radiation dose of ventilation-perfusion scintigraphy (VQ scan) is 0.00028-0.00051 Gy. The fetal radiation dose of ventilation-perfusion scintigraphy (half dose) is 0.00014-0.00025 Gy.

The negative predictive value of a normal VQ scan in pregnant patients is excellent, and in one study 70% of scans in pregnant women were interpreted as normal,[2] making this a useful first test for evaluating PE. However, 21% of women in the study had nondiagnostic scans, necessitating additional imaging tests.

Pulmonary Computed Tomographic Angiography

The fetal radiation dose of computed tomographic angiography (CTA) is 0.000003-0.000131 Gy.

Although the radiation dose to the fetus is lower for CTA than for VQ scanning, CTA produces higher levels of radiation exposure in the mother's breasts.

Breast cancer risk in 20-year-old women after a CTA is estimated to be 1 event in 143 exposures, with a lifetime attributable risk of 0.7%.[3].The risk for malignancy is inversely related to age at the time of exposure. Therefore, it is recommended that women use breast shields when getting a CTA. Breast shields can decrease the breast radiation dose from 0.05-0.08 Gy to 0.02-0.06 Gy.

A major benefit of CTA over other imaging modalities is that it often provides alternative pulmonary diagnoses.

Thus far, no studies have been published that address the accuracy or outcome of CTA in pregnant patients. However, technical limitations might produce poor vessel opacification in pregnant patients, simulating PE, and resulting in false-positive results. This appears to be more common when apparent subsegmental PEs are identified. The authors actually recommend that subsegmental PEs identified on CTA be confirmed with further testing (eg, VQ scan, lower extremity ultrasonography [US]) before making a final diagnosis.

Magnetic Resonance Imaging

The authors barely mention magnetic resonance imaging (MRI) and do not include this testing at all in their algorithms. They mention that there are insufficient accuracy and outcome data to make any conclusions about the utility of this test. Furthermore, the contrast agent gadolinium crosses the placenta, and there is insufficient safety data about this agent's use in pregnancy.

Compression Ultrasonography

Lower-extremity US is the first recommended test in pregnant patients with signs or symptoms of DVT (eg, unilateral leg swelling or pain). However, US has a very low sensitivity for detecting VTE in patients without signs or symptoms of DVT; therefore, US is not recommended as a routine first-line test in patients without signs or symptoms of DVT.

As noted earlier, pregnant patients have a higher incidence of isolated pelvic DVT, which will not be detected on lower-extremity US. False-positive US results may occur as a result of relatively slow venous flow in pregnancy.

Finally, the authors present a series of somewhat complicated diagnostic algorithms that, if nothing else, highlight the difficulty of this diagnosis and the lack of definitive data for the various imaging modalities. Initially, a US should be obtained if there are signs or symptoms of DVT; otherwise, chest imaging should be used. If the US is positive, VTE is diagnosed. Use a perfusion scan when leg symptoms are present as the first imaging test because of the lower radiation dose to the mother's breasts, if normal/near-normal VTE is excluded. Of note, the only tests in the algorithm that are considered singularly definitive are: (1) positive US (VTE diagnosed); (2) normal/near-normal perfusion scan (VTE excluded); and (3) CTA diagnosing a segmental PE (VTE diagnosed). Any other test result requires further workup (eg, ventilation scans, serial US on day 1 and 7, etc).


Low-molecular-weight heparin is the treatment of choice for PE in pregnancy, as it can be continued on an outpatient basis. Warfarin is contraindicated because of teratogenicity. Intravenous unfractionated heparin is preferred in patients with renal failure and if urgent reversal of anticoagulation is anticipated (eg, high risk for bleeding or surgery is anticipated). Thrombolytic drugs (eg, 100 mg tPA over 2 hours) can be considered in patients who are hemodynamically unstable or in patients with intractable hypoxemia. In these patients, the benefit is likely to outweigh the minor risks for bleeding and fetal loss.

The article goes on to describe longer-term treatment, use of inferior vena cava filters near term, prevention issues, and a few other topics, all of which are less relevant to emergency medicine practice and are beyond the scope of this summary.


There are only 3 things that I know for certain about PE: PE is often overdiagnosed; PE is often underdiagnosed; and over- or underdiagnosis of PE results in increased cost, morbidity, mortality, and medicolegal risk.

Almost everything else that I have read or heard is not based on hard facts but on probabilities. Unfortunately, probabilitiesare just not acceptable when we are considering the consequences of PE in a pregnant patient -- we need certainties. This manuscript doesn't solve our problems, but it does take us one step closer to certainty. The data that these authors present are very helpful in clinical decision-making. They shed light on the utility (or lack thereof) of some of the typical diagnostic tests we use, they provide more concrete information regarding radiation risks to the fetus and to the mother, and they provide sound recommendations regarding treatment. I'm going to feel a lot better about my uncertainties with this article in hand the next time I'm caring for a pregnant patient who has a possible PE.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.