Sometimes there’s just a question that rattles around in the back of my head. I feel like whether or not a normal D-dimer can help rule out pulmonary embolus in kids is one such question.

What is D-dimer?

D-dimer is a fibrin degradation product left over in the blood after a clot is broken down. There are two “D” monomers of fibrin that are cross linked by activated factor XIII, hence the name. D-dimers are usually only present in the blood when the clotting cascade has been activated. They are detected in serum using antigen tests such as ELISA. There are a few different tests out there, and you also need to be familiar with your lab’s reference range; <500 ng/mL for fibrin equivalent units (FEU) or <250 ng/mL for D-dimer units (DDU). D-dimer will be elevated in DVT, PE, DIC, hyperthrombotic states and also variably in liver disease, vasculitis/rheumatic conditions, malignancy, trauma, pregnancy and recent surgery.

How has it been traditionally used?

An elevated D-dimer alone is insufficient to diagnose a PE but a negative or low D-dimer can help you rule one out when there is low to intermediate probability of PE

There has been an incredible amount of work done in grown ups regarding making the diagnosis of PE relative to that in children. The PIOPED study from 1991 helped define the findings that were most often present in adults with PE. Many of you are likely familiar with the Wells score. PERC is a clinical decision rule that performs well in low prevalence settings but does require a modicum of experience in that you must assign a pretest probability. Fortunately, there is a ton of excellent FOAM content about PE, and also great evidence-focused tools available online. Per a Cochrane Review, the “estimates of sensitivity ranged from 80% to 100%, and estimates of specificity from 23% to 63%.” In general, the utility of D-dimer in grown-ups really depends on your pre-test probability.

How is PE different in children? And what’s the evidence, specifically regarding D-dimer?

As you have gathered, PE is rare in children. Many of the existing studies are case series. One large retrospective descriptive study of Pediatric Emergency Department patients in a single center between 2003-2001 saw only 105 with PE out of almost 1.2 million total patients. In the Pediatric Emergency Department, where I work, I will consider PE as a moderate to highly likely diagnostic possibility only a few times a year. Most of these kids have known risk factors, some more common than others, and which include;

  • cancer
  • congenital heart disease
  • acquired and inherited thrombophilias including rheumatologic diseases
  • indwelling central line placement
  • oral-contraceptive use
  • recent trauma or surgery with admission to the hospital
  • heavy smoker
  • obesity

The physiologic effects of PE in children can mirror those in adults. They include;

  • Infarction, where small emboli lodge in the lower lobes of the lung, leading to pain and hemoptysis.
  • Abnormal gas exchange, which can lead to dyspnea and hypoxia because larger clots obstruct pulmonary venins, leading to VQ mismatch and atelectasis
  • Shock, whereby a massive PE obstructs the vascular bed, leading to a profound increase in pulmonary vascular resistance, reflex arterial vasoconstriction due to hypoxia, which impedes right ventricular outflow, dilates the right ventricle, pushing the septum into the left ventricle and compromising cardiac output.

Fortunately, the majority of children have smaller, peripheral PE that are not as physiologically deleterious. So, with that in mind let’s take a look at a few studies. starting with the one I just alluded to, with the caveat that this is by no means an exhaustive list, since I’m focusing on D-dimer.

Agha et al. Pulmonary embolism in the pediatric emergency department. Pediatrics, 2013

This large cohort retrospectively identified PE patients based on the Wells and PERC Criteria. Only 25 of their 105 met inclusion criteria (previous diagnosis of PE at another facility being one exclusion criteria). Many of the patients had a predisposing condition.

Overall they also noted that 52% had chest pain, 44% shortness of breath, and 32% cough aligning with other reports on symptom frequency. Additionally, by Wells criteria none were in the high probability group. And the most common PERC criteria not met was tachycardia. Only 80% of the patients had a D-dimer obtained. There was no comparison with patients that didn’t have a PE and what their D-dimer was.

Rajpurkar et al. Pulmonary embolism—experience at a single children’s hospital. Thrombosis Research, 2007.

During a three year period in a single pediatric facility 14 patients were diagnosed with PE. Again, highlighting how small the population is. 10/14 had a D-dimer, and 4/10 were normal. three of these 4 patients were diagnosed on the first day of symptoms. Most (12/14) of their patients had multiple risk factors for PE. They thought that maybe those D-dimers were normal because the patients had peripheral clots – not massive ones.

Biss et al. Clinical features and outcome of pulmonary embolism in children. British Journal of Haematology, 2008.

This retrospective review of 56 patients with radiographically confirmed PE from a single center noted that D dimer was elevated at presentation in 26/30 patients (86·7%). There was no comparison with patients that did not have PE. Therefore…

Biss et al. Clinical probability score and D‐dimer estimation lack utility in the diagnosis of childhood pulmonary embolism. Journal of Thrombosis and Haemostasis, 2009.

…They did a retrospective cohort of 50 PEs and 25 normals. They compared Wells score, D-dimer and other findings. Pertinent to this post “children with PE were as likely as PE negative patients to have a D‐dimer value within the normal range (PE positive, 15%; PE negative, 25%; P=0.654). A combination of a ‘PE unlikely’ score and normal D‐dimer value occurred in 1/12 (8%) of PE negative children. Why did they feel it was a wash? Well, the authors postulated that “the similar rates of elevated D‐dimers in the PE positive and PE negative groups reflects the high rate of underlying systemic illnesses in children presenting to this institution with suspected PE.” This is because D-dimer is thought to be elevated in vasculitis and other illnesses (they cited papers on Henoch-Schonlein Purpura and Kawasaki) but listed pneumonia as something that could elevated D-dimer.

Hennelly et al. Detection of Pulmonary Embolism in High-Risk Children. Journal of Pediatrics, 2016.

This was a retrospective review of 561 patients that underwent PE workup. They applied the Wells criteria, PERC and an internally derived PE prediction rule that was established via recursive partitioning. This rule included the presence of oral contraceptive use, tachycardia, and oxygen saturation <95%. The test characteristics were;

Again, clinical suspicion/pre-test probability is very important when determining the utility of a positive D-dimer.

Kanis et al. Clinical characteristics of children evaluated for suspected pulmonary embolism with D-dimer testing. Arch Dis Child, 2018.

This was a, descriptive univariate, retrospective analysis of patients who had D-dimer obtained during a workup for PE. D-dimer was ordered in 526 children, and 34/526 had PE (6.4%). Most were seen in the ED (80%) and 88% were teenagers with a 2:1 female predominance. Per the authors “children with PE had higher mean heart and higher respiratory rate and a lower pulse oximetry and haemoglobin concentration.” Their univariate analysis established that 5 conditions were more frequently seen in conjunction with PE as opposed to no PE; surgery, central line, limb immobility, prior PE or deep vein thrombosis and cancer. Ultimately they argued that their findings could serve as the basis for a decision rule for children.

Kanis et al. Diagnostic accuracy of the D-dimer in children. Arch Dis Child, 2018.

This was actually published in the same issue as the previous study and used the same 526 patients. The main goal of the study was to determine the test characteristics of D-dimer in diagnosing PE. In all of the PE positive patients (34/526) the D-dimer was elevated (positive). Here is what they found – though I’d encourage you to read the article itself

  • The mean D-dimer value was 2104±1394 ng/mL in the 34 PE+ children and 586±962 ng/mL in 492 PE− children
  • 236/526 (45%) had a positive D-dimer including all 34 patients who were positive for PE
  • The sensitivity was 100% (95% CI 89% to 100%)
  • The specificity was 290/492 (58%, 95% CI 54% to 63%)
  • The area under the receiver operating characteristic curve was 0.90 (95% CI 0.87 to 0.94)

This means that all children with PE in their sample had an elevated D-dimer, so “ruling in” was not an issue. The specificity was 58%, so D-dimer alone is insufficient to rule out. And the authors are clear that you need clinical suspicion and a history of risk factors to enhance the usefulness of the D-dimer.

Bringing it all together

OK, so it’s clear that there is a paucity of evidence into the utility of D-dimer in children, especially relative to the knowledge base for adults. Here is what I know to be true;

  • There are far, far fewer PEs in kids than adults and most children with PE with have an underlying risk factor
  • Kids will not be as sick by and large
  • The number of patients with an elevated D-diver, who do NOT have a PE is pretty high
  • CTPA is not without risk, but is, in my practice the best and most available test for children
  • An externally validated decision rule for PE in children is needed, and given the scarcity of the diagnosis, it will likely have to rely on multi center data

Final thoughts and recommendations

Here are some closing thought as, grounded in my review of the (admittedly incomplete) evidence:

  • If PE is your most likely diagnosis, or if it is the most likely serious thing you are trying to rule out, especially if the patient has a known risk factor and/or they are tachypneic, tachycardic and hypoxic go ahead and order a CTPA and skip the D-dimer.
  • If you think that PE might be the diagnosis, but you have low to intermediate pre-test probability, AND you’d be OK with either an alternative diagnosis or not having a definitive diagnosis other than “non-specific chest pain” go ahead and get a D-dimer. You can also use the Wells criteria or PERC, to help serve as scaffold for your risk assessment, but know that the criteria are NOT entirely relevant to all kids. Alternatively, per one study the following 3 factors are associated with increased likelihood of PE if the D-dimer is positive – oral contraceptive use, tachycardia, and oxygen saturation <95%
    • If that D-dimer is elevated, get a CTPA, knowing that there is still a greater likelihood it WON’T be a PE
    • If the D-dimer is negative, you haven’t completely ruled out a PE, but you can be reassured
  • If you don’t think that it could be a PE at all then skip the D-dimer. Don’t open that door unless you want to hang out in that room for awhile.


Agha et al. Pulmonary embolism in the pediatric emergency department. Pediatrics. 2013 Oct;132(4):663-7. doi: 10.1542/peds.2013-0126. Epub 2013 Sep 2.

Crawford et al. D-dimer test for excluding the diagnosis of pulmonary embolism. Cochrane Database Syst Rev. 2016 Aug 5;(8):CD010864. doi: 10.1002/14651858.CD010864.pub2.

Hennelly et al. Detection of Pulmonary Embolism in High-Risk Children. Journal of Pediatrics, 2016.

Kanis et al. Clinical characteristics of children evaluated for suspected pulmonary embolism with D-dimer testing. Arch Dis Child, 2018.

Kanis et al. Diagnostic accuracy of the D-dimer in children. Arch Dis Child, 2018.

MD Calc: PERC Rule

Raja et al. Evaluation of Patients With Suspected Acute Pulmonary Embolism: Best Practice Advice From the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2015;163(9):701.

Singh et al. Pulmonary embolism rule-out criteria (PERC) in pulmonary embolism–revisited: a systematic review and meta-analysis. Emerg Med J. 2013;30(9):701. Epub 2012 Oct 4. 

Stein et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest. 1991 Sep; 100(3):598-603.

Wells et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost. 2000 Mar;83(3):416-20.