Diagnosing PE in Pregnant Women


Venous thromboembolism (VTE) occurs in pregnant women at 7-10 times the incidence in age-matched controls. Pregnant women who get VTE have deep venous thrombosis (DVT) approximately 3 times as often as pulmonary embolism (PE). The increased risk is similar during all three trimesters, starts to diminish after delivery, and returns to baseline by six weeks post-partum.

The left leg is affected in 85% of pregnancy-associated DVT, possibly from compression of the left iliac vein. Isolated pelvic DVT is also more common in pregnancy. In the developed world, pregnancy-associated VTE is the leading cause of maternal mortality.

Why the increased risk?

Pregnant women are in a hypercoagulable state in preparation to staunch the bleeding associated with delivery. Venous stasis is likely due to compression of the pelvic veins by the uterus and progesterone-induced venodilation. Protein S is an anticoagulant which develops reduced activity, and activated protein C resistance increases. There is an increase in fibrinogen and multiple coagulation factors. At the same time, there is less fibrinolysis due to increases in plasminogen activator inhibitors, and subsequently reduced activity of endogenous tPA. Delivery (either vaginal or operative) can damage blood vessels, also leading to increased risk of post-partum VTE.

Risk stratification

While clinical decision rules (eg PERC or Wells criteria) are validated and routinely used for non-pregnant patients, no tools exist for pregnant women. Pregnancy-associated VTE has unique risk factors, has a clear left leg predominance, and complicating matters further, the symptoms of pulmonary embolism often mimic expected changes in pregnancy (eg dyspnea, tachycardia, leg edema). Currently, clinical judgement is the best approach for risk stratification in these patients.

It is important to remember that potential maternal mortality from untreated PE will generally outweigh the theoretical fetal risk from radiation exposure. If you have clinical concern for PE in a pregnant patient, you should perform the appropriate diagnostic tests based on shared-decision making.


D-dimer is routinely used in non-pregnant patients to exclude VTE, but there is a physiologic increase in d-dimer levels during normal pregnancy, and the thresholds for when to suspect VTE have not been validated. A low d-dimer may potentially exclude VTE in pregnancy (further research needed), but due to the low specificity of an elevated level this test should generally not be used to exclude or risk stratify pregnant patients. This is consistent with recommendations by the American Thoracic Society.

NON-VALIDATED: Dr. Jeff Kline (of VTE fame) has suggested possibly using d-dimer cutoffs 50% higher than normal for each trimester:

·         Normal Lifespan d-dimer test cutoff: 230

·         First trimester: 345

·         Second trimester: 460

·         Third trimester: 575

* Lifespan uses a d-dimer cutoff of 230 in the ED and 300 for inpatients

NON-VALIDATED: Dr. Jeff Kline has also suggested in first trimester that negative d-dimer + PERC negative (with allowable HR up to 105) makes PE unlikely, however this approach is not validated.

Ultrasound for DVT

In pregnant women with signs or symptoms of both PE and DVT, you should start with compression ultrasonography. If the test is positive, anticoagulation can generally be started without advanced chest imaging. It is not recommended to perform DVT ultrasound without signs or symptoms of DVT.

Chest x-ray

If you are concerned for PE but your pregnant patient has no signs or symptoms of DVT, the next step is a chest x-ray, which can evaluate for alternative pathology and is a branch point in the decision for advanced chest imaging. Both VQ scans and CT pulmonary angiograms have a risk of being non-diagnostic.

VQ Scan

Ventilation perfusion scintigraphy (VQ scan) results are based on probability of a ventilation-perfusion mismatch, typically categorized as normal/low, moderate, or high. However, this is based on pre-test probability, and given the difference in PE prevalence between pregnant and non-pregnant patients, interpretation is open to more variability. VQ scan can theoretically expose a fetus to increased radiation as the radionuclide is renally excreted and collects in the urinary bladder. Consider temporarily placing a Foley catheter to minimize this.

A retrospective study by Cahill et al. in 304 patients found that if the chest x-ray is normal, VQ scan will more likely be diagnostic compared with CTPA (94% versus 70%). If you had enough clinical concern to obtain a VQ scan and it is non-diagnostic, you should proceed to a CT Pulmonary Angiogram.

CT Pulmonary Angiogram

CT Pulmonary Angiogram (CTPA) generally exposes a fetus to similar or lower levels of ionizing radiation compared with VQ scan (though much higher radiation exposure to the mother’s breasts). CTPA can also reveal alternate diagnoses while VQ scans typically cannot. If the patient’s chest xray is abnormal but you remain clinically concerned for PE, you should proceed to CTPA rather than VQ scan. Subsegmental PE is often over-diagnosed with the high resolution of newer CT machines, and you should consider this possibility prior to starting anticoagulation.

Shared Decision Making

Ultimately, balancing the possibility of VTE with the risk of radiation exposure to the mother and fetus should be decided with shared decision making. Remember that PE is potentially lethal, and you’re balancing that with a risk of fetal radiation exposure which is not well described. Whichever option is chosen, be sure to document your shared decision making discussion with the patient. As always, include that the patient has capacity, specific risks/benefits were explained, and that she was able to verbalize back her understanding.

Faculty Reviewer: Dr. Kristina McAteer

References and More FOAM

Bourjeily, Ghada, et al. "Pulmonary embolism in pregnancy." The Lancet 375.9713 (2010): 500-512.