Does D-dimer rule out PE? YES, but…

Written by Zach Adams, OSU EM resident // Edited by Michael Barrie EM Assistant Professor.

Bottom line: D-dimer reliably excludes PE in low/moderate risk patients, however use clinical history/exam to guide pursuing advanced imaging.

An otherwise healthy 23 year-old presents with worsening shortness of breath for the past 7 days.  Shortness of breath began suddenly while at rest, has been continuous, and with associated reduced exercise tolerance.  Five days ago she was seen in the ED with a negative d-dimer and CXR and discharged with return precautions.  Since then, she states that the symptoms have progressed. She says she cannot lie flat.  ROS reveals a history of antecedent URI 10 days ago.  She is a non-smoker, takes no birth control pills, does not have a personal or family history of DVT or PE, and denies recent prolonged travel.  She appears uncomfortable and takes deep inspirations every 3-5 seconds.  Her physical exam reveals tachypnea with otherwise normal exam.  ECG shows normal sinus rhythm.  CXR is normal and repeat d-dimer is negative.

The patient has bounced back with worsening symptoms.  I struggled with this clinical question: should we obtain a CT PE to rule out PE despite a negative d-dimer in this low risk patient?

Pulmonary emboli can present in a variety of ways, from no symptoms at all to sudden death.  Symptoms are classically described sudden onset (seconds to minutes) of dyspnea followed by pleuritic chest pain and cough. The PIOPED II study reported the most common symptoms of PE among their population as (1):

  • Dyspnea at rest or with exertion (73%)
  • Pleuritic chest pain (44%)
  • Cough (37%)
  • Orthopnea (28%)
  • Calf or thigh pain and/or swelling (44%)
  • Wheezing (21%)
  • Hemoptysis (13%)

According to the PIOPED II study, commonly presenting sings are (1):

  • Tachypnea (54%)
  • Calf or thigh swelling, erythema, edema, tenderness, palpable cord (47%)
  • Tachycardia (24%)
  • Rales (18%)
  • Decreased breath sounds (17%)
  • Accentuated pulmonic component of the second heart sound (15%)
  • JVD (14%)
  • Fever mimicking pneumonia (3%)

Our patient has the most common presenting sign, so what risk factors does she have then?  We can divide these into patient-associated risk factors (permanent) and circumstantial (transient).  From a risk standpoint, they can be described as (2):

Major (RR 5-20):

  • Postoperative major surgery – abdominal/pelvic, hip/knee joint replacement
  • Late pregnancy, caesarian section, puerperium
  • Lower limb fractures or extensive varicosities
  • Malignancies – abdominal/pelvic and advance/metastatic stage
  • Immobilization
  • History of previous thromboembolism

Minor (RR 2-4)

  • Cardiovascular – congenital heart disease, heart failure, hypertension, central venous catheter
  • Estrogen use, including oral contraceptives and hormone replacement therapy
  • Misc. – COPD, neurological impairment, thrombotic defects, long distance travel, obesity
  • Other – IBD, nephrotic syndrome, dialysis, myeloproliferative disease, PNH

Based on risk factors, the case above describes a low pre-test probability for PE.  But does a negative d-dimer is excludes her from having a PE in this instance? Fortunately, at least in this case – YES!

When working up individuals for PE, we begin by determining the pre-test probability using some well known clinical scores, most commonly with Well’s and PERC criteria.  Studies have shown, however, that gestalt estimates and these probability scores have comparable sensitivity when combined with a d-dimer (3).  It should be noted that these scores do not apply to the pregnant patient population, and are less likely accurate in the older population (4).  That said, we should apply them when appropriate.

Using the Well’s score in the non-pregnant patient population with suspected PE, we can classify patients into categories of unlikely (score <4) or likely (score >4) using the following:

  • Clinical symptoms of deep vein thrombosis (DVT) (3 points)
  • Other diagnoses are less likely than PE (3 points)
  • Heart rate >100 (1.5 points)
  • Immobilization three or more days or surgery in previous four weeks (1.5 points)
  • Previous DVT/PE (1.5 points)
  • Hemoptysis (1 point)
  • Malignancy (1 point)

We use the PE rule-out criteria (PERC) to rule out low risk patients with NO further testing. In the low and moderate risk category, a d-dimer is used to further use to evaluate risk.

If we choose to use a d-dimer, we have the following:

  • PE unlikely, d-dimer <500 ng/mL – no further testing
  • PE unlikely, d-dimer >500 ng/mL – further testing is indicated (CTPA or VQ scan)

If we choose to use the PERC rule:

  • PE unlikely, PERC negative – no further testing
  • PE unlikely, PERC positive – obtain d-dimer or further imaging per above

Our patient is low-risk and has a negative d-dimer, but how certain can we be that we’re not missing something still given her symptoms?

In a study by den Exter, et. Al, they compared 4,044 patients with a suspected PE presenting more than 7 days from the onset of symptoms to those presenting within 7 days to assess the safety of excluding a PE utilizing a clinical decision rule and d-dimer testing.  They found that the failure rate of an unlikely clinical probability and a normal d-dimer was 0.5% in each of the groups (5).  That is impressively low, and they concluded that even with a delayed presentation (as in our case) the combination of the above algorithms is adequately sensitive.

A negative d-dimer in high risk patient’s unfortunately does not always exclude PE.  In a study by Gibson, et. Al, 1,722 patients were evaluated for PE.  563 of these had a normal d-dimer, with 477 being low probability and 86 being high probability patients by Well’s criteria.  Only 5 (1.1%) of the low probability patients with a negative d-dimer were diagnosed with a PE by CTPA.  However, 9.3% of the high probability group were found to have PE by imaging, indicating that even with a negative d-dimer, clinical suspicion should determine if further imaging is indicated.

So, if a patient bounces back with worsening symptoms but has a negative d-dimer, is advanced imaging indicated?

In this case, we diagnosed pneumomediastinum!  Air was seen both in the mediastinum and dissecting up through the proximal soft tissues of the neck on the study.  Though physical exam did not pick this up, we did have our answer and she was admitted for further imaging, including a contrast esophogram.  For a quick read on benign pneumomediastinum, check out

Ultimately we had our diagnosis, which would have otherwise been missed if we didn’t have concern that something was wrong.  In the words of OSUEM attending Dr. Martin:

“When a patient presents with the same complaint, do more.”

That said, the final diagnosis in this case was not PE and so I still feel confident using the well established risk criteria and d-dimer for ruling out pulmonary embolism.

For more reading on some novel concepts including the use of age adjusted d-dimers, check out:


  1. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Stein PD, Beemath A, Matta F, Weg JG, Yusen RD, Hales CA, Hull RD, Leeper KV Jr, Sostman HD, Tapson VF, Buckley JD, Gottschalk A, Goodman LR, Wakefied TW, Woodard PK. Am J Med. 2007;120(10):871.
  2. Exp Clin Cardiol. 2013 Spring; 18(2): 129–138. Pulmonary embolism, part I: Epidemiology, risk factors and risk stratification, pathophysiology, clinical presentation, diagnosis and nonthrombotic pulmonary embolism. Jan Bĕlohlávek, MD PhD, Vladimír Dytrych, MD, and Aleš Linhart, MD PhD
  3. Clinical decision rules for excluding pulmonary embolism: a meta-analysis. Lucassen W, Geersing GJ, Erkens PM, Reitsma JB, Moons KG, Büller H, van Weert HC. Ann Intern Med. 2011;155(7):448.
  4. TIAccuracy of the Wells clinical prediction rule for pulmonary embolism in older ambulatory adults. Schouten HJ, Geersing GJ, Oudega R, van Delden JJ, Moons KG, Koek HL. J Am Geriatr Soc. 2014 Nov;62(11):2136-41. Epub 2014 Nov 3.
  5. Impact of delay in clinical presentation on the diagnostic management and prognosis of patients with suspected pulmonary embolism. AUden Exter PL, van Es J, Erkens PM, van Roosmalen MJ, van den Hoven P, Hovens MM, Kamphuisen PW, Klok FA, Huisman MV. Am J Respir Crit Care Med. 2013 Jun;187(12):1369-73.
  6. Gibson NS, et al. The importance of clinical probability assessment in interpreting a normal d-dimer in patients with suspected pulmonary embolism. Chest. October 2008;134(4)