Skip to content

PE in pregnancy: which diagnostic tests do you use?


Pulmonary embolism (PE) can be a deadly disease and one of the most challenging diagnosis to make in a pregnant patient. Patients may present with signs and symptoms that might also be present in a normal uncomplicated pregnancy. Even in nonpregnant patients, the diagnosis of venous thromboembolism (VTE) such as PE can be quite challenging.

One challenge is that pregnant patients already have an increased risk of venous and arterial thromboembolism compared to nonpregnant patients due to the physiologic and structural changes during the intra- and postpartum period. Another challenge is that clinical decision aids such as Wells’ criteria were not validated to rule out PE in pregnancy. This issue is problematic in the assessment of the pregnant patient whom some would consider, at least, to be at a moderate risk level for having PE.

According to a retrospective chart review1 of 125 patients, a modified Wells’ criteria was used retroactively to identify patients without PE via CT pulmonary angiogram (CTPA). From the patients identified the negative predictive value was 100% with a modified Wells’ criteria ≤0.001).


What about using a D-dimer test to risk-stratify? A BestBets review addressed the question of using a negative D-dimer to rule out PE in pregnancy. The research performed by BestBets in 2004 and last modified in 2011 selected 5 studies out of 151 studies which addressed the question of interest. These studies had a small number of patients, and consisted of 2 prospective, a retrospective, a systematic review, and a case report study. Several studies showed normal elevation of D-dimers as pregnancy progressed, which is something that should be considered for future research on D-dimer cutoff values.

BestBets’ conclusion (2011): There is NO current evidence to support the use of a single isolated negative D-dimer result to rule out PE in the pregnant patient.

Why don’t we just get a CTPA to rule-out all pregnant patients who may have a PE?

Typically for non-pregnant patients, one starts the PE work up by obtaining a chest xray and determining a pre-test probability. Obtain a D-dimer for low pretest probability patients. If the D-dimer comes back elevated or if the pre-test probability is already moderate-high to start with, perform an imaging study such as a CTPA (or a V/Q scan).

For pregnant patients, however, we just discussed that D-dimers aren’t as helpful. Should we get a CTPA (or V/Q scan)? Such imaging studies place both the fetus and mother at risk for future malignancies. For this reason, if we order these tests, we should keep radiation exposure to “As Low As Reasonably Achievable” (ALARA) for both. We can provide a shield to protect the fetus and/or also place a urinary catheter to drain the radioactive isotope. Interestingly, the high cardiac output in pregnancy might not allow for appropriate vascular opacification in CTPA. Furthermore, the dye used in the V/Q scan has the potential of causing hypothyroidism in the unborn fetus.

Using ultrasound

An alternative approach before obtaining a CTPA or V/Q scan is to start a PE workup with a chest x-ray and a lower extremity (LE) ultrasound to assess for deep venous thrombosis (DVT). If the LE ultrasound is positive, the treatment threshold is surpassed and the patient should be started on anticoagulants for a likely PE. If the ultrasound study shows no DVT, a more extensive work up involving irradiation (CTPA or V/Q scan) should now be considered in the decision process.

This still leaves many questions unanswered unfortunately

Ultimately in the end, your clinical decision should be a multifactorial process, which considers the uncertainties of diagnostic findings, risk-benefit justification in obtaining or not obtaining a CTPA, patient’s values, and your overall clinical judgment.
Since there is no convincing literature on how to work up PEs for pregnant patients, I would love to hear comments on your thought process.
  • Should we use elevated D-dimer cutoffs to rule out pulmonary embolism in pregnancy?
  • Would you skip D-dimers altogether since pregnant patients were excluded from the Wells derivation studies (making Wells scores not externally valid for pregnant patients)?
  • Would you skip D-dimers altogether since pregnant patients are at greater baseline risk for VTE and thus automatically fall into the moderate-to- high risk category?
  • Do you start with a lower extremity ultrasound? Would you feel ok with a one time negative lower extremity ultrasound to rule out PE?


Additional Reading
  1. Cutts, BA et al. New directions in the diagnosis and treatment of pulmonary embolism in pregnancy. Am J Obstet Gynecol. 2013 Feb;208(2):102-8. [PMID 22840412]
  2. Heit, JA et al. Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med. 2005 Nov 15;143(10):697-706. [PMID 16287790]
  3. James, AH. Venous Thromboembolism in Pregnancy. Arterioscler Thromb Vasc Biol. 2009 Mar;29(3):326-31. [PMID 19228606]
  4. Le Gal G. et al. Diagnostic value of single complete compression ultrasonography in pregnant and postpartum women with suspected deep vein thrombosis: prospective study. BMJ. 2012 Apr 24;344:e2635. [PMID 22531869]
  5. Shahir, K. et al. Pulmonary embolism in pregnancy: CT pulmonary angiography versus perfusion scanning. AJR Am J Roentgenol. 2010 Sep;195(3):W214-20. [PMID 20729418]

O’Connor C, Moriarty J, Walsh J, Murray J, Coulter-Smith S, Boyd W. The application of a clinical risk stratification score may reduce unnecessary investigations for pulmonary embolism in pregnancy. J Matern Fetal Neonatal Med. 2011;24(12):1461-1464. [PubMed]
  • jeff tabas

    Nice summary. I would say that many symptoms of pregnancy mimic PE, so can be a difficult diagnosis. In a patient that I intend to image with CTPA, I would send a d-dimer first and if negative I would stop. I see no drawback. You use standard cutoffs (500) for 1st trimester and post partum, but can use cutoffs of 150% for 2nd trimester (750) and 200% (1000) for 3rd trimester based on work of Jeff Kline. I would obtain bilateral LE dopplers first – in the off chance they are positive, would stop. If you really want to avoid the radiation I would consider MRA/MRV instead of CTPA, which is sensitive but has high rate of technically inadequate imaging.

    • Hi Jeff, thanks for reading and for your response. This is quite a challenge, good thing is not that common. The conversation definitely helps to have different options in mind.

  • Thanks for the points, Jeff. I agree with your assessment and approach.

    I saw the cutoff values for d-dimers published by Jeff Kline with the numbers you provided. They were tests run on normal pregnant patients. To be nitpicky, they weren’t quite validated for risk-stratification purposes in patients with PE-like symptoms during pregnancy, although it makes sense that one could apply this data

    Kline JA, Williams GW, Hernandez-Nino J. D-dimer concentrations in normal pregnancy: new diagnostic thresholds are needed. Clin Chem. 2005 May;51(5):825-9. PubMed PMID: 15764641.

    MRA/MRV’s do indeed eliminate the radiation perspective but I agree with you that it’s so low yield since IV gadolinium contrast isn’t used (crosses placental barrier). Note, it’s way down on the list of recommended imaging choices per the ACR Appropriateness Criteria (Radiology guidelines):

    Sounds like validating the higher cutoffs for pregnancy, by trimester, will be a useful study to conduct, so that we don’t have think about irradiating so many pregnant patients. Any takers anyone?

  • I would say that by and large, pregnant women don’t have PE. If as a group they did (on a par with knee surgery, for example) Kline et al. would have had the numbers to publish real guidelines already. When the rare case occurs that they do, it seems to be peripartum. We know that Dimer rises progressively thru each trimester. Does it follow that thrombosis risk rises incrementally as well? I don’t know but I would surmise not- more of an asymptotic curve as we approach delivery. Certainly I practice as if a first trimester pt. is no different risk than a non-pregnant one.
    That makes me like the Dimer approach (low risk grp) but as pointed out by Dr. Lin, not validated yet. Most helpful to me in 2nd trimester when physiology mimics PE. Alternative dx helpful as well. I have avoided imaging with aggressive albuterol therapy and cured SOB sx to the satisfaction of the referring OB. 3rd trimester cutoffs worry me as the rising risk of PE may make the “Dimer r/o” a losing proposition.

    • Thanks Pik. Agree, most of the PE occur during peri/post-partum, and uncommon during pregnancy. I didn’t go into it, but considering alternative diagnoses and involving OB are important aspects in the management.

    • Yes, but main point is that a 3rd trimester pt. is in auch higher risk category: so should you apply a Dimer at all? We’re taught that these are mainly useful in low risk and equivocal pts. So even if the “Kline cutoffs” gain footing, I will likely not use them in late preg. at all.

  • Anonymous

    Interesting discussion.

    I think there are three different type of patients:
    1) A sick pt: A pt that is hypoxic, tachycardic, complaining of chest pain and with a swollen leg. This pt needs imaging for a definitive diagnosis once they are stable enough unless there is a clear alternate diagnosis like a big pneumonia or an echo showing pregnancy associated cardiomyopathy or something- though an abnormal echo will probably won’t remove PE in the differential and if a pneumonia is big enough to make you that sick you will probably still need a CT chest.
    2) A patient with mild subjective SOB without any specific cause but also without any objective signs of PE. No swollen legs, no hypoxia, no chest pain, no tachypnea. This is the pt you don’t think has a PE but you want to cover your butt a little. We all understand that this pt is at higher risk for PE than the general population but there is nothing to push me into a workup here. I might ultrasound the legs, get an EKG looking for strain, but that’s pretty much as far as I’ll go. I’ll have a conversation about “there’s this thing called a PE…” and talk to them about why I don’t want to get a CT. We are paid to weigh the risks and benefits. To me the definite risk of radiation (and dye) outweigh the benefit (small chance of finding a probably clinically insignificant PE).
    3) The ones that are not screaming PE, but there’s something that makes you think they might legitimately have a PE. I might beat around the bush a little: get ultrasounds, maybe get a dimer (though the “Kline cut-offs” probably aren’t ready for prime time). At the end of the day, I think you just have to bite the bullet and get the CT. There is really nothing else that will rule out this potentially lethal disease that you think they might acutally have. I probably order less than 15 CTPE’s on pregnant patients per year. I kinda feel bad when I do it, but if you are doing it to protect the pt and baby and not to cover your butt, you will sleep well.

    BTW, love your blog.

    • Thanks, interestingly the literature I looked at (not sure if I might have missed it) did not include electrocardiograms as part of the workup. That makes a lot of sense.

  • Some comments written in the Real-Time Peer Review Demographics form:

    Still waiting for the perfect approach.
    I hear that there may be a role for increasing the D Dimer cutoff, but haven’t read up yet.
    Chan et al in 2002 suggested that v/Q normal or non-diagnositc -> treat as normal and no missed PE.
    WIner-Muram et al [and others] suggest CT has equivalent [and maybe even less] radiation exposure than V/Q.
    Read somewhere that Kline suggested PERC Rule + Leg Dopplers?
    — Nadim

    I would do the lower leg DVT study and if positive treat, if not and I have no other explanation for the pt. presentation I would discuss the benefit and the risk with the patient and decide where to go ?..
    — Anonymous EM PGY-2

  • Another great response from the Demographics form:

    Great piece on PE in pregnancy. Just some comments and then I was going to attempt to answer the questions listed:

    -Although pregnant patients are at a greater risk for VTE, I don’t think all of them are automatically moderate PTP for disease. I agree that Well’s was not validated in pregnant patients and I find it less useful in this group. My opinion (and it is purely opinion) is to use gestalt or PERC. If your PERC negative (with elevated hr cut-off of 105) and you were low pre-test probability for VTE, you’re done–no d dimer, no imaging. As mentioned, the symptoms of pregnancy often overlap with PE and I feel that your gestalt means alot. Also, the highest risk for VTE is 3rd Trimester and post partum period, so I believe you can risk stratify certain pregnant patients to minimal/low risk of disease based on their symptoms, trimester, and your gestalt.

    In response to d-dimer use to r/o VTE in pregnancy, my thought is why not use it. The first decision is if you are going to embark on a work-up at all. If you are (pt is moderate PTP or even higher), send the d-dimer (if you have a high sensitivity d-dimer). B/c if it is negative, I think you are done. You are certainly done if you had moderate PTP of disease. It is debatable if you had high pre-test probability. If you were committed to a work up and the d-dimer is positive, then image.

    -In terms of imaging, there is data to support if CXR is negative, V/Q scan gives you more bang for your buck. If CXR abnormal, CT is the way to go. So if you have ability to do both, follow the above algorithm. If VQ unavailable, CT is the only choice you have for definitive imaging and radiation doses are within safe range for fetus. I worry more about radiation to maternal breasts than to fetus. I agree with ALARA principle regardless of what imaging procedure you do.

    Other ways to cut radiation for VQ is to do perfusion only scan and if perfusion normal, ventilation portion can probably be excluded. Also radiology can also use smaller dose of Technetium. I was unaware Technetium had potential to cause fetal hypothyroidism, I was under the impression it was the iodinated contrasts. Even though this risk is present, it is a small risk due to short exposure time of fetus to the contrast.

    On to the questions:
    -We can use elevated d-dimer cut-offs based on trimester. There is data to support these cut-offs in healthy pregnant volunteers although I don’t believe this has been looked at in pregnant patients where VTE is being considered.

    -I would not skip d-dimers altogether. Your first step is are you going to embark on work up at all. If you are, I think send the d-dimer if you would i b/c if negative and you have high sensitivity d dimer, you may be able to terminate work up. If positive, you can proceed with work up.

    -Although pregnant patients are at greater baseline risk for PE, I do think you can risk stratify certain patients to a low enough risk to not even work them up. Use PERC with elevated hr cut-off.

    -As far as ultrasound, I think it should be done. Again, decide if you want to work patient up at all. If the answer is yes, send the d dimer. If negative, done. If positive, get Ultrasound. If Ultrasound, positive, treat and done. If Ultrasound is negative, I feel you must move to imaging. I would not be comfortable with one time negative ultrasound.

    sorry for this long response, I have an interest in this topic.

    Nilesh Patel

  • Anonymous

    I am confused
    My impression was that d-dimers rule out pulmonary embolus only in the low risk patient. I thought that a negative d-dimer in moderate and high probability patients was not of real value.

    Can someone clarify?
    Thank you!