Podcast: Play in new window

Subscribe: RSS

Reference: Bannelier et al. Failure rate of D-dimer testing in patients with high clinical probability of pulmonary embolism: Ancillary analysis of three European studies. AEM Feb 2025

Date: February 27, 2025

Guest Skeptic: Dr. Lauren Westafer an Assistant Professor in the Department of Emergency Medicine at the University of Massachusetts Medical School – Baystate. She is the cofounder of FOAMcast and a pulmonary embolism and implementation science researcher. Dr. Westafer serves as the Social Media Editor and a research methodology editor for Annals of Emergency Medicine.

Case: A 57-year-old woman presents to the emergency department (ED) with pleuritic posterior chest/back pain, shortness of breath, and left leg swelling. Her oxygen saturation is 95% on room air, blood pressure is 125/70, and her heart rate is 106 beats per minute. She has a swollen left leg that is tender but neurovascularly intact, without signs of cellulitis. She is on oral hormone replacement therapy (estrogen), and has a history of hypertension (on lisinopril). Chest x-ray shows no pneumonia or pneumothorax and bedside ultrasound reveals no B-lines or effusion.

Background: Pulmonary embolism (PE) is a potentially life-threatening diagnosis that ED physicians must always have on their radar. The challenge, however, lies in balancing the risks of missing a PE with the harms of unnecessary imaging. D-dimer testing has changed how we workup suspected PE patients by serving as a highly sensitive yet non-specific biomarker for venous thromboembolism (VTE). This test has become a crucial component of modern PE diagnostic pathways.

There is a need to “right size” testing such that we do not miss clinically important PEs without exposing very low risk patients to risks of imaging (ionizing radiation, contrast, allergic reaction, cost, and misdiagnosis). The principle behind D-dimer testing is its high negative predictive value (NPV) in ruling out PE, particularly in patients with low to moderate pretest probability. Current guidelines suggest that in these groups, a negative D-dimer result (below the defined threshold) can reliably exclude PE without the need for imaging. The Wells Score and Revised Geneva Score (RGS) and YEARS are widely used clinical prediction rules to stratify risk and guide appropriate use of D-dimer testing.

However, in high-risk patients, the utility of D-dimer is less clear. The positive predictive value (PPV) of the test is low due to the presence of multiple potential causes of elevated D-dimer. Some of the causes of elevated D-dimers include cancer, infection, trauma, and post-surgical states. Given that a negative D-dimer result is uncommon in high-risk patients and the high prevalence of PE in this group, traditional teaching has recommended skipping the test and proceeding directly to CT pulmonary angiography (CTPA). The CT scanner has been called the “donut of truth” by some physicians. 

Despite these long-standing recommendations, recent studies have questioned whether a D-dimer–based approach could still be safe in selected high-risk patients. With age-adjusted D-dimer thresholds and Bayesian approaches refining risk stratification, a re-evaluation of the test’s performance in high-probability patients is warranted.

Clinical Question: Can an age-adjusted D-dimer strategy safely exclude PE in patients with a high clinical probability of PE?

Reference: Bannelier et al. Failure rate of D-dimer testing in patients with high clinical probability of pulmonary embolism: Ancillary analysis of three European studies. AEM Feb 2025

  • Population: Patients with a high clinical probability of PE (Wells >6 or Revised Geneva Score >10) and underwent D-dimer testing in the ED.
    • Excluded: Patients with missing D-dimer values, missing data elements for the RGS or Wells score, inconclusive CTPA.
  • Intervention: Age-adjusted or standard (<500) cut off for d-dimer
  • Comparison: None
  • Outcome:
    • Primary Outcome: Failure rate of the age-adjusted D-dimer (i.e. a VTE in a patient with a negative D-dimer) at index visit or during 3-month follow up
    • Secondary Outcomes: Failure rate using a fixed 500 ng/mL D-dimer threshold and the failure rate of CTPA (i.e. a VTE in a patient with a negative CTPA) during 3-month follow up
  • Type of Study: This was a post hoc analysis of three European studies (two cluster randomized trials: PROPER, MODIGLIANI and one retrospective study: TRYSPEED)

This is an SGEMHOP but we were unable to get in contact with the lead and corresponding author. We hope they will see or hear this episode and respond to our nerdy questions.

Authors’ Conclusions: “In this study, ruling out pulmonary embolism in high-risk patients based on D-dimer below the age-adjusted threshold was safe, with no missed pulmonary embolism. However, the sample size was not large enough to draw a definitive conclusion on the safety of this strategy.”

Quality Checklist for Observational Study:

  1. Did the study address a clearly focused issue? Yes
  2. Did the authors use an appropriate method to answer their question? No
  3. Was the cohort recruited in an acceptable way? No
  4. Was the exposure accurately measured to minimize bias? Yes
  5. Was the outcome accurately measured to minimize bias? Yes
  6. Have the authors identified all-important confounding factors? No
  7. Was the follow up of subjects complete enough? No
  8. How precise are the results? Confidence intervals are wide (failure rate 0.0%–6.5%), reducing certainty.
  9. Do you believe the results? Unsure
  10. Can the results be applied to the local population? No
  11. Do the results of this study fit with other available evidence? No
  12. Funding of the Study and Conflicts of Interest? Funding was not specified and the authors declare no COIs.

Results: The analysis included 651 patients who met inclusion criteria. Study subjects were 60% female and had a median age of 68 years. Overall prevalence of PE in this cohort was 31% (n=204).

Key Result: No patients with a D-dimer below the age-adjusted threshold were found to have a PE.

  • Primary Outcome: 70 patients had a negative D-dimer (age-adjusted below the threshold of age x 10). The failure rate was 0.0% (95% CI 0.0% to 6.5%). The Bayesian analysis estimated a 76.2% probability that the failure rate was below 2%.
  • Secondary Outcomes:
    • Failure Rate Using a Fixed 500 ng/mL D-dimer Threshold. 48 patients had D-dimer levels below 500 ng/mL. None of these patients had PE after follow-up. Failure rate: 0.0% (95% CI 0.0%–7.4%). The Bayesian analysis estimated a 62.8% posterior probability that the failure rate was below 2%.
    • Failure Rate of CTPA (Exploratory Analysis). 845 patients underwent CTPA in the ED and six were subsequently diagnosed with PE at 3-month follow up giving a failure rate of 0.7% (95% CI 0.3 to 1.6%).

1. Post Hoc Design: This type of study design increases the risk of several biases. Post hoc analyses involve examining data that were not originally intended to answer a specific research question. This means the study design and data collection were not prospectively planned to assess the safety of D-dimer testing in high-risk PE patients. Because of this, certain methodological safeguards, such as pre-specified inclusion criteria, standardized follow-up protocols, and predefined statistical analyses, may not have been in place. These limitations can lead to potential biases in data interpretation. The main sources of bias arise from the inability to control for confounding variables that were not accounted for in the original studies.

2. Selection Bias: This can occur when the individuals included in a study do not accurately represent the broader population to which the results are meant to apply. In this study, patients were included if they had both a D-dimer and a CTPA order. Based on professional society recommendations, these patients should theoretically undergo diagnostic imaging rather than D-dimer evaluation. It is possible that these patients were characteristically different than those who did not have a CTPA or a D-dimer.

3. Information Bias: This is a concern because patients with missing D-dimer values or incomplete follow-up were excluded. The incomplete data set may have skewed the results if these excluded patients had different clinical outcomes compared to those who remained in the analysis. This type of bias could lead to an overestimation of the safety of a D-dimer-based rule-out strategy in high-risk patients. Ultimately, the study findings might not accurately reflect real-world risk.

4. Study Asymmetry: Nearly 90% of patients came from one cohort (TRYSPEED) that appears categorically different than the PROPER and MODIGLIANI. Average D-dimer values were nearly triple in TRYSPEED than the other two trials combined and few cases had a D-dimer under the 500 ng/mL threshold. This raises the question of how useful a standard D-dimer threshold would be in this type of population.

5. Wide Confidence Intervals: The study only had a small number of high-risk patients with negative D-dimer results (70 out of 651) below the age-adjusted threshold. This resulted in a wide 95% confidence interval (CI) around the point estimate and leads to uncertainty about the true failure rate of a D-dimer–based strategy in these high-risk PE patients. While the failure rate was 0%, the upper limit of the 95% CI extended to 6.5%. This level of uncertainty makes it difficult to confidently recommend using D-dimer as a rule-out strategy in high-risk patients without additional evidence from larger studies. While it is reassuring the Bayesian analysis estimated a 76% probability that the failure rate was below 2%, it still leaves room for doubt, especially in clinical decision-making where missing even a small number of true PE cases could have severe consequences.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: There may be a role for D-dimer in high-risk patients, but at present the certainty of this role and optimal D-dimer threshold is unknown.

SGEM Bottom Line: We cannot recommend using a D-dimer strategy in ruling out PE in high-risk patients.

Case Resolution: Based on the patient’s clinical risk, you send the patient for CTPA. She has bilateral segmental pulmonary emboli. You risk stratify the patient and she meets criteria for outpatient treatment. You give her an oral anticoagulant and discuss expectant management.

Dr. Lauren Westafer

Clinical Application: At this time, I will await more definitive evidence before routinely using D-dimer in patients with high clinical probability of PE.

What Do I Tell the Patient? I suspect you may have a blood clot in your leg. Part of this clot has likely broken off and travelled to your lungs, giving you the pain and shortness of breath. We are going to get a CT scan to confirm if you have a pulmonary embolism (blood clot in your lung). 

Keener Kontest: There was no winner last week. The answer we were looking for is that MDW measures monocyte distribution width. This represents the variability in the size of circulating monocytes. During the early stages of infection and sepsis, monocytes undergo activation in response to inflammatory stimuli, leading to morphological changes and increased heterogeneity in their size. This variability is detected by advanced hematology analyzers

Listen to the SGEM podcast for this week’s question. If you think you know the answer, then send an email to thesgem@gmail.com with “keener” in the subject line. The first correct answer will receive a shoutout on the next episode.

SGEMHOP: Now it is your turn, SGEMers. What do you think of this episode on using D-dimers to rule out PE in high-risk patients? Post your comments to social media using #SGEMHOP. The best social media feedback will be published in AEM.

Other SGEM Episodes on Pulmonary Embolism:

  • SGEM#51: Home (Discharging Patients with Acute Pulmonary Emboli Home from the Emergency Department)
  • SGEM#118: I Hope you Had a Negative D-dimer (ADJUST PE Study)
  • SGEM#126: Take me to the Rivaroxaban – Outpatient treatment of VTE
  • SGEM#163: Shuffle off to Buffalo to Talk Thrombolysis for Acute Pulmonary Embolism
  • SGEM#184: We Weren’t Born to Follow-Up – The PEITHO Long-Term Follow-up Study
  • SGEM#219: Shout, Shout, PERC Rule Them Out
  • SGEM#276: FOCUS on PE in Patients with Abnormal Vital Signs
  • SGEM#277: In the Pregnant YEARS – Diagnosing Pulmonary Embolism
  • SGEM#282: It’s All ‘bout that Bayes, ‘Bout that Bayes- No Trouble – In Diagnosing Pulmonary Embolism
  • SGEM#295: Teacher, Teacher – Tell Me How to Do It (Diagnose a PE)
  • SGEM#323: Mama I’m Comin’ Home – For Outpatient Treatment of a Pulmonary Embolism
  • SGEM#416: She’s Always A Woman, Query PE?

REMEMBER TO BE SKEPTICAL OF ANYTHING YOU LEARN, EVEN IF YOU HEARD IT ON THE SKEPTICS’ GUIDE TO EMERGENCY MEDICINE