Determining if the strategy use in the evaluation of PE is ready for the spotlight.
The evaluation of dyspneic pregnant patients at risk of pulmonary embolism (PE) presents a significant challenge to most emergency physicians.
Background
Pregnant women are usually young and otherwise healthy, but the consequences of a missed PE can be disastrous, and the risk of over evaluation and over imaging with the resultant exposure to excess medical radiation is something we would like to avoid/reduce. The additional complicating feature of these presentations is that the expected physiologic changes of pregnancy may cause women in their second and third trimesters to experience leg swelling, tachycardia and sensation of dyspnea at baseline; common signs/ symptoms of venous thromboembolism (VTE).
The estimated incidence of PE in pregnancy is 1.72 cases per 1000 deliveries, with a death rate of 1:100,000 deliveries. [1]
Previous Findings
D-dimer use as a strategy to reduce the need for advanced imaging in the evaluation of VTE is a well-established practice in the non-pregnant population, and most studies that evaluate the use of D-dimer diagnostic strategies have excluded pregnant patients in the past.
Historically, D-dimer use in pregnant patients has not been recommended. There is good evidence that baseline D-dimer levels in pregnant women will progressively increase as their pregnancy progresses. This reduces the likelihood of a negative result below the established threshold of 500 ng/ml.
In 2011, the American Thoracic Society clinical practice guideline recommended against the use of D-dimer in the evaluation of VTE in pregnancy. [2] This recommendation was based on the admittedly weak evidence of one retrospective study of 37 pregnant patients with suspected PE who had V/Q scan and D-dimer testing. Sensitivity of D-dimer for PE in this study was 73%. Additional indirect evidence used in the recommendation was three prospective studies evaluating a total of 389 patients for DVT. D-dimer was 100% sensitive for DVT, however there was a very low rate of DVT in this population.
D-dimer Adjustments
Jeff Kline, MD, proposed a strategy of trimester specific pregnancy adjusted D-dimer cutoffs in a 2013 interview on Rob Orman’s ERCast podcast [4] Kline is an emergency physician, and an expert in venous thromboembolism.
The proposed strategy was to perform D-dimer testing on pregnant women with a pretest probability less than 40% (Wells score less than or equal to 4), and a negative PERC rule with an adjusted heart rate cutoff of 105 bpm. Adjusted D-dimer thresholds of 750/1000/1250 ng/ml for first, second and third trimester respectively were recommended. This strategy has physiologic plausibility (in the same way that age adjusted D-dimer cutoffs for the elderly have physiologic plausibility) however, this strategy has not been prospectively validated.
Another study to deliberately evaluate the utility and accuracy of D-dimer testing as a diagnostic strategy in the evaluation of VTE in pregnancy was the DiPEP study (Diagnosis of PE in Pregnancy) published in 2018. [5] This was a prospective observational cohort study of 310 pregnant or postpartum patients in the UK who were suspected of having a VTE.
The authors measured multiple biomarkers, including D-dimer, that was compared between patients with confirmed VTE and those where VTE was excluded. A group of women with known DVT were added to the study population to increase the prevalence of VTE in the study, and to improve the estimates of biomarker sensitivity. This study did not identify a threshold for any biomarker (including D-dimer) that would optimize sensitivity in the upper 90% range while maintaining reasonable specificity.
One specific weakness of the study is that a large proportion of patients were already on anticoagulation at the time of D-dimer measurement. This can impact the accuracy of this test. An additional weakness of the study was that there was no incorporation of pretest probability included in the diagnostic pathway employed. For more on the statistics, strengths, and weaknesses of this study check out the rebelem.com blog post by Rick Pescatore DO. [5,6]
In late 2018, Righini and colleagues sought to incorporate pretest probability into a testing strategy using a D-dimer threshold of 500 ng/ml, compression ultrasonography, and CT pulmonary angiography. [7] This was a multicenter, multinational prospective diagnostic outcome study of 395 pregnant women who were clinically suspected of PE with signs of acute onset of new or worsening shortness of breath or chest pain without another obvious cause.
Patients were classified as low, intermediate or high risk based on their Revised Geneva Score (see table below). Those with low or intermediate pretest probability had D-dimer screening performed and had no further imaging if their D-dimer was <500 ng/ml. High risk, or D-dimer > 500 ng/ml patients had bilateral compression ultrasonography performed. If the ultrasound was positive, they were treated empirically for PE. If the ultrasound was negative, CT Pulmonary Angiography was performed. All patients were followed for three-months and the primary outcome was VTE event during the follow up period in women who did not get anticoagulation based on negative results in their initial work-up.
Table One: Revised Geneva Score
| Variable | Points |
| Risk Factors | |
| Age > 65 | 1 |
| Previous DVT or PE | 3 |
| Surgery (under GA, or lower ext. fracture within 1 month) | 2 |
| Active malignant condition (or cured <1 year) | 2 |
| Symptoms | |
| Unilateral lower-limb pain | 3 |
| Hemoptysis | 2 |
| Clinical Signs | |
| Heart rate 75-94 beats/min | 3 |
| Heart rate >95 beats/min | 5 |
| Pain on lower limb deep venous palpation and unilateral edema | 4 |
| Clinical Probability | |
| Low | 0-3 total |
| Intermediate | 4-10 total |
| High | >/= 11 total |
Pretest probability was low in 48.6% of patients, intermediate in 50.6%, and high in 0.8%. D-dimer testing was negative in 11.7% of low/intermediate pretest probability patients. The proportion of negative D-dimer testing decreased with increased gestational age: 25.3% first trimester, 11.1% second trimester and 4.2% third trimester. Overall PE rate was 7.1%. (positive compression ultrasound, CTPA, or V/Q scan). CT pulmonary angiography was performed in 84% of patients.
There were no confirmed cases of VTE during the three-month follow up period in patients that were low/intermediate probability with negative D-dimer tests. For more on the statistics, strengths, and weaknesses of this study, see the rebelem.com blog by Rick Pescatore, DO. [7,8]
Findings
Most of the major PE decision rule studies have deliberately excluded pregnant patients. The YEARS study group did not deliberately exclude pregnant patients, but only enrolled a very low number. They prolonged their study period to deliberately enroll more pregnant patients to assess a pregnancy adapted YEARS algorithm. [9] This was a prospective study that screened 510 pregnant women suspected of PE. Three criteria from the original YEARS algorithm were screened for: Clinical signs of DVT, hemoptysis and PE as the most likely diagnosis. Patients with 0/3 criteria and a D-dimer less than 1000 ng/ml were discharged without further imaging. Patients with 1 or more criteria and a D-dimer less that 500 ng/ml were also discharged. Patients with clinical signs of DVT had bilateral lower extremity compression ultrasonography performed. If positive, they were presumed to have a PE and placed on anticoagulation without confirmatory CT pulmonary angiogram. If negative, and D-dimer was less than 500 ng/ml, they were discharged without further imaging.
The primary outcome was incidence of VTE during the three-month follow up period.
PE was diagnosed in 4% of patients. Sixty-one percent of patients in this study had a CT pulmonary angiogram performed. CT pulmonary angiogram was avoided in 65% of first trimester patients, 46% of second trimester patients, and 32% of third trimester patients. The three-month incidence of symptomatic VTE was 0.21%. One patient was diagnosed with a proximal DVT and there were no diagnoses of PE in the group that had negative D-dimer testing at the pre-specified thresholds of 500/1000 ng/ml depending on the number of YEARS criteria present.
Conclusion
The new data from 2018 and 2019 (Righini pretest probability study, and pregnancy adapted YEARS algorithm) provide clinicians challenged with safely ruling out PE and avoiding the harms of medical radiation with an option of using D-dimer in their diagnostic evaluation.
In the first trimester there is utility of a D-dimer testing strategy that can reduce the number of CT scans performed, reducing the risk of fetal teratogenicity. As D-dimer sensitivities for VTE drop in the second and third trimester D-dimer use becomes more of a challenge, however there is still demonstrated reduction in CT use in this gestational age group (YEARS reduced CT use by 46% in second, and 32% in third trimester patients). Reduction in maternal breast radiation exposure in the second and third trimester is an important goal, as is decreased resource utilization, and decreased cost of care.
Anytime you consider the incorporation of an evidence-based practice improvement that is a significant departure/progression from your usual practice it is reasonable to attempt to incorporate these changes on a departmental or even institutional level.
Bring the evidence to your department chair and your thrombosis team. Convincing the ‘needle experts’ that you have a better way of sifting through the ‘haystack of dyspnea’ in your emergency department, and then making an institutionally supported change may be better than blindsiding them with new evidence.
References:
[1] James AH. ET AL. Venous thromboembolism during pregnancy and the postpartum period: incidence, risk factors, and mortality. Am J Obstet Gynecol 2006;194:1311-5 PMID: 16647915
[2] Leung AN et al. An Official American Thoracic Society/Society of Thoracic Radiology Clinical Practice Guideline: Evaluation of Suspected Pulmonary Embolism in Pregnancy. Am J Respir Crit Care Med 2011. Nov 15;184(10):1200-8 PMID: 22086989
[3] Rob Orman, ‘Pulmonary embolism in pregnancy with Jeff Kline’. ERCAST podcast, April 24, 2013. Available at: http://ercast.libsyn.com/pulmonary-embolism-in-pregnancy-with-jeff-kline
[4] Kline J. et al. D-dimer concentrations in normal pregnancy: new diagnostic thresholds are needed. Clin Chem. 2005 May;51(5):825-9. PMID: 15764641
[5] Hunt, Beverly J., et al. The DiPEP (Diagnosis of PE in Pregnancy) biomarker study; An observational cohort study augmented with additional cases to determine the diagnostic utility of biomarkers for suspected venous thromboembolism during pregnancy and puerperium. Br J Haematol 2018 Mar;180(5):694-704 PMID: 29359796
[6] Rick Pescatore. D-dimer and Pregnancy: The DiPEP Study, REBEL EM blog, March 19, 2018. Available at: https://rebelem.com/d-dimer-and-pregnancy-the-dipep-study/
[7] Righini, M., et al. Diagnosis of Pulmonary Embolism During Pregnancy. A Multicenter Prospective Management Outcome Study. Ann Intern Med. 2018 Dec 4;169(11):766-773 PMID: 30357273
[8] Rick Pescatore. D-dimer in Pregnancy: Limiting Radiation with Pre-test Probability, REBEL EM blog, November 29, 2018, Available at: https://rebelem.com/d-dimer-in-pregnancy-limiting-radiation-with-pre-test-probability/
[9] van der Pol, L. M., et al. Pregnancy-Adapted YEARS Algorithm for Diagnosis of Suspected Pulmonary Embolism. N Engl J Med. 2019 Mar 21;380(12):1139-1149 PMID: 30893534
