In our last two pieces on pulmonary embolism (PE), 1 and 2, we discussed five studies that attempted to describe the prevalence of pulmonary embolism in patients presenting to the ED with syncope. In this post, we will review some common methodologies for working up PE in the ED.
PE is one of those usually considered not-to-be-missed diagnoses. It has traditionally been thought that a missed PE can be devastating, although recent studies have started to challenge this idea.[1] So, there should be no surprise that this is an area ripe for research.
How does one work up a patient with a suspected pulmonary embolism?
Well, let’s start with Wells criteria. The Wells criteria are used to objectively generate a pre-test risk for a particular patient.[2-4] The Wells score is comprised of seven questions with point values from 1 to 3:
1. Clinical signs and symptoms of DVT: 3 pts
2. PE is #1 diagnosis OR equally likely: 3 pts
3. Heart rate > 100: 1.5 pts
4. Immobilization for at least 3 days OR surgery in the previous 4 weeks: 1.5 pts
5. Previous objectively diagnosed PE or DVT: 1.5 pts
6. Hemoptysis: 1 pt
7. Malignancy w/ treatment within 6 months or palliative: 1 pt
There are then two ways to risk-stratify patients and generate a pre-test probability from the Wells score. There is the 3-category approach where a score of <2 predicts a risk of 1.3%, 2-6: 16.2%, and >6: 37.5%. And, there is the 2-category approach where a score of ≤4 predicts risk of 12.1% and >4: 37.1%.
A d-dimer with a threshold of 500ng/mL FEUs has a negative likelihood ratio (nLR) of 0.05. The Fagan nomogram of the 3-category approach utilizing an FEU d-dimer cut off of 500ng/mL is in Figure 1.
Fagan Nomogram for Well’s Criteria and 500ng/ml FEU as cut-off
Figure 1: Red line represents low-risk patients, Blue Line represents moderate-risk patients, and Green line represents high-risk patients.
As you can see from the Fagan nomogram, a high-risk individual with a d-dimer < 500 has a post-test probability of >1%, which is generally considered too high to rule out PE. As such, it is generally not recommended to use a d-dimer in high-risk individuals. Another way to look at this is that the high-risk population has such a high prevalence that the negative predictive value (NPV) of d-dimer testing is lowered, making the d-dimer test unsuitable for ruling out PE. (If you remember back to your biostatistics, NPV is changed by prevalence, where specificity and sensitivity are not.) So, high-risk patients should undergo a more definitive study, namely a CTPA or a VQ scan.
To reduce the number of tests and d-dimers ordered, many clinicians then apply the PERC (PE Rule-out Criteria) to generate a more acceptable post-test probability.[5] PERC has a nLR of 0.17. Any positive criterion means that a patient is “PERC positive” for PE:
1. Age ≥ 50
2. SpO2 < 95% (room air)
3. HR <100
4. No prior PE or DVT
5. No recent surgery or trauma (within prior 4wk)
6. No hemoptysis
7. No estrogen use
8. No unilateral leg swelling
These eight criteria were initially studied on patients determined to have a low pre-test probability based on clinician gestalt and identified as having a pre-test probability of <15%. However, in the validation study, they showed that a pre-test probability of 6% and PERC neg was required to rule out PE.[5] Figure 2 shows PERC with the results from the validation study with a nLR of 0.17 and a VTE rate of 7%[5]:
Fagan Diagram for PERC
Figure 2. Fagan nomogram of PERC criteria for a patient with a pre-test risk of 7%, the red line represents the nLR of the test, and the blue line represents the positive likelihood ratio (pLR) of the test.
Some argue that a more appropriate way to establish low pre-test probability is to use a Wells score of <2. For the experienced clinician who can reliably assign a pre-test probability of PE, PERC can be used without a Wells score to determine a pre-test risk.
What may be considered the most commonly accepted way to improve the pLR of d-dimer tests and decrease inappropriate testing is the age-adjusted d-dimer.[6] ACEP has actually already given its support to the use of age-adjusted d-dimers.[6] The idea of the age-adjusted d-dimer is that as we age, we expect to have higher baseline d-dimer levels. To adjust a patient’s d-dimer for patients over the age of 50, you take their age and multiply it by 10, and this will be the upper limit of normal in FEUs ng/mL. For example, a 75-year-old person would have an age-adjusted d-dimer cutoff of 750ng/mL in FEUs. The nLR of the test is 0.08. As patients get older, age-adjusted d-dimer loses some of its sensitivity but has improved specificity. Figure 3 is the Fagan nomogram for a patient with a pre-test probability of 10% and shows a post-test probability of <1%. Again, age-adjusted d-dimer alone is not a good test for “non-low-risk” patients.
Figure 3. Age-adjusted d-dimer with prevalence of 10%. Red line represents the nLR and Blue line represents the pLR
One of the new kids on the block is the YEARS algorithm.[8] The YEARS algorithm is made up of three criteria: clinical signs of a DVT, Hemoptysis, and PE is the “most likely diagnosis”. YEARS is meant to dichotomously risk-stratify low to moderate risk patients, where a low to moderate risk patient who meets no positive YEARS criteria allows the clinician to use a1000 ng/mL FEU d-dimer cutoff. However, if any one YEARS criteria is positive, then the clinician uses a 500ng/ml FEU d-dimer cutoff. This has been validated in the United States and has also been analyzed (but not yet validated) in pregnant women,[9] where negative YEARS, negative bilateral lower extremity dopplers, and d-dimer <1000ng/ml FEUs safely rules out PE. The negative likelihood ratio is 0.1. For patients with a pre-test probability of 10%, there will be a post-test probability of 1% . Figure 4 shows the Fagan nomogram for the YEARS algorithm for patients with a pre-test risk of 5%.
Figure 4. YEARS criteria Fagan nomogram. Red line represents a pre-test probability of 5% (PE incidence in the study) and a negative test yielding a post-test probability of 0.5%, while blue line represents the pLR for the test.
The newest kid on the block is PEGED (Pulmonary Embolism Graduated d-dimer),[10] which is similar to YEARS as it utilizes different d-dimer levels for patients based on their risk. They broke patients into three levels of risk using Wells: low risk 0-4, moderate risk 4.5-6, and high risk >6. Patients with low risk had a cutoff d-dimer of 1000ng/mL FEUs; moderate-risk patients had a 500ng/mL FEUs cutoff, and high-risk patients had a CTPA or VQ scan. It should be noted that although they had no false negatives, they lost 9 patients to follow-up in their low-risk group. I performed a worst-case scenario analysis assuming all 9 of those patients had a PE to generate a nLR of 0.08, similar to age-adjusted d-dimer, which can be seen in figure 5 below.
Figure 5. PEGED study. Of note nLR is calculated based on worst-case analysis where all 9 patients lost to follow up had PE. In their study, no PEs were missed based on any rule-out criteria. the red line represents the nLR and the blue lines represents the pLR
You might have been reading this post and been asking yourself “what the hell is an FEU?”. FEU stands for Fibrinogen Equivalent Units. This is what many labs use to measure d-dimers, however, there is another unit that you need to be aware of, DDU, which stands for d-dimer Units. One DDU is equivalent to about two FEUs, so if your lab reports DDUs, instead of the FEUs, you need to divide all of the above-discussed values by two. So for PEGED and YEARS, you would use 500ng/ml FEUs and 250ng/ml DDUs. For age-adjusted you would multiply by 5 instead of 10, so for a 750-year-old, your upper limit would be 375ng/ml DDUs.
So which algorithm should you use? Well PEGED isn’t yet validated, so that might dissuade you. YEARS algorithm is validated and may even work in pregnant patients. Personally, I use a mixture. I use Wells, and then for the very low-risk (0-1.5 points), use PERC. If the patient is low to moderate-risk, I apply age-adjusted and YEARS criteria, and if patient is high-risk I go straight to imaging. If the patient is pregnant, I utilize shared decision-making and consider applying YEARS for pregnant patients. Of note, this methodology is not itself validated, but I believe supported by the evidence. If PEGED is successfully validated, then I think a similar algorithm could be used instead of YEARS.
References:
1. Kline JA, Kabrhel C. Emergency Evaluation for Pulmonary Embolism, Part 1: Clinical Factors that Increase Risk. J Emerg Med [Internet] 2015;48(6):771–80. Available from: http://www.sciencedirect.com/science/article/pii/S0736467914014280
2. Wells PS, Anderson DR, Rodger M, et al. Excluding Pulmonary Embolism at the Bedside without Diagnostic Imaging: Management of Patients with Suspected Pulmonary Embolism Presenting to the Emergency Department by Using a Simple Clinical Model and d-dimer. Ann Intern Med. 2001;135:98–107. doi: https://doi.org/10.7326/0003-4819-135-2-200107170-00010
3. Prospective validation of wells criteria in the evaluation of patients with suspected pulmonary embolism, Wolf, Stephen J. et al., Annals of Emergency Medicine, Volume 44, Issue 5, 503 – 510
4. Writing Group for the Christopher Study Investigators*. Effectiveness of Managing Suspected Pulmonary Embolism Using an Algorithm Combining Clinical Probability, D-Dimer Testing, and Computed Tomography. JAMA. 2006;295(2):172–179. doi:10.1001/jama.295.2.172
5. KLINE, J.A., COURTNEY, D.M., KABRHEL, C., MOORE, C.L., SMITHLINE, H.A., PLEWA, M.C., RICHMAN, P.B., O’NEIL, B.J. and NORDENHOLZ, K. (2008), Prospective multicenter evaluation of the pulmonary embolism rule‐out criteria. Journal of Thrombosis and Haemostasis, 6: 772-780. doi:10.1111/j.1538-7836.2008.02944.x
6. PENALOZA, A., ROY, P.‐M., KLINE, J., VERSCHUREN, F., LE GAL, G., QUENTIN‐GEORGET, S., DELVAU, N. and THYS, F. (2012), Performance of age‐adjusted D‐dimer cut‐off to rule out pulmonary embolism. Journal of Thrombosis and Haemostasis, 10: 1291-1296. doi:10.1111/j.1538-7836.2012.04769.x
7. Wolf, S., Hahn, S., Nentwich, L., Raja, A., Silvers, S. and Brown, M. (2018). Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department With Suspected Acute Venous Thromboembolic Disease. Annals of Emergency Medicine, 71(5), pp.e59-e109. doi: https://doi.org/10.1016/j.annemergmed.2018.03.006
8. Multicenter Evaluation of the YEARS Criteria in Emergency Department Patients Evaluated for Pulmonary Embolism, Kabrhel, Christopher et. al., Academic Emergency Medicine 2018; 25: 987– 994, https://doi.org/10.1111/acem.13417
9. van der Pol LM, Tromeur C, Bistervels IM, et al. Pregnancy-Adapted YEARS Algorithm for Diagnosis of Suspected Pulmonary Embolism. N Engl J Med [Internet] 2019;380(12):1139–49. Available from: https://doi.org/10.1056/NEJMoa1813865
10. Kearon C, Wit KD, Parpia S, et al. Diagnosis of Pulmonary Embolism with d-Dimer Adjusted to Clinical Probability. New England Journal of Medicine. 2019;381(22):2125-2134. doi:10.1056/nejmoa1909159.
Noah Berland
Latest posts by Noah Berland (see all)
- Teaching Shifts: “Sore Throat” - November 12, 2020
- Teaching Shifts - October 20, 2020
- Social EM Journal Club: Food Insecurity - October 16, 2020
0 Comments