Author: Dr. Shivon Manchanda
“Pregnancy-Adapted YEARS Algorithm for Diagnosis of Suspected Pulmonary Embolism.”
Background:
The diagnosis of PE can be challenging. We do not want to scan everyone, yet even with the creation of decision-making tools, our use of CTPA is going up, while our yield of actual PE is going down. And pregnancy makes the diagnostic process even harder! This is because most decision-making tools leave out pregnant patients in their studied populations. The Wells Score gets hard to use when pregnant patients are tachycardic and short of breath at baseline, and because of that you can’t even use the PERC rule. An OBGYN resident I’m friends with also recently said “We laugh at people that order D-dimers in our pregnant patients.” Great.
Problematically, PE is still a leading cause of maternal mortality worldwide. While the prevalence of PE isn’t that high compared to nonpregnant patients, pregnant women are still more likely to get DVTs. So you have a muddled clinical picture because of normal physiologic changes, no good clinical decision making rules, and even though they aren’t supposed to get PE’s, they get so many DVTs there’s no way you aren’t going to be worried that your patient is one of the few whose DVT broke off and went up to their lungs.
So what do we do? There was a YEARS study in 2017 that established a promising PE algorithm, AND it included pregnant patients. But not enough were in the original study…so what if we just got more?
Study Question:
○ Can a pregnancy-adapted YEARS algorithm safely avoid diagnostic imaging in pregnant women with suspected pulmonary embolism?
Design:
- Continuation of the YEARS study – ran for 5 more years
- Multicenter, international (France and the Netherlands), and prospective
- 510 pregnant women over the age of 18 were sent from the ED or obstetric ward for clinical suspicion of PE – new onset or worsening dyspnea +/- tachycardia +/- hemoptysis
- Went through this algorithm:
Picture from MDCalc
- Only change from original study: two-point ultrasound if signs of DVT
- If positive, they were anticoagulated, and that’s it, no need to look for a PE
- Everyone was followed for 3 months
- Independent committee assessed suspected VTE and death during follow-up
Results:
○ 20/498 of the included patients found to have a VTE (4% 95% CI 2.6 – 61)
- Primary outcome: Incidence of VTE during the 3 month follow up in the group that was not indicated for anticoagulation
- 477 (96%) with PE ruled out at baseline
- Only one with popliteal DVT found at follow-up (0.21%)
- Secondary outcome: Proportion of patients where CTPA was not indicated
- 195 patients were ruled out
- CTPA was safely avoided in 39% (95% CI 35-44)
- They even looked at efficiency by trimester!
- 1st trimester: 65% avoided
- 2nd trimester: 46% avoided
- 3rd trimester: 32% avoided
Discussion:
The Pregnancy-Adapted YEARS study is a one of the largest studies conducted on the diagnostic approach to pregnant patients and shows promising results. It has several strengths including having patients from all three trimesters, almost complete follow up with the patients, and its use of an independent committee. While it is good that this is an international study, it needs to be externally validated since patients were only in select centers in the Netherlands and France.
The algorithm does also have some of the same sticking points of other PE tools though: heavy use of subjectivity. 89% of the people that met at least one YEARs criteria met it because “PE was the most likely diagnosis.” Plus, physicians could know the dimer results prior to evaluation, biasing their assessment and elevating the power of the test.
At the same time, physician gestalt can also be the most important factors in the diagnosis of PE! So, including this is both good and bad, it just depends on the provider which makes its usefulness quite variable. Also, the d-dimer being known early is something that realistically happens with the Wells calculation, so at least for comparison purposes this is not the biggest deal. For a global assessment as an “objective” tool, well we still need to find ways to do better.
Anyway, overall the pregnancy adapted years study did save many pregnant patients from getting a CT. Looking at the efficiency breakdown though, it’s way more efficient in the first trimester. By the third, it’s better than no rules but still not great. Thus, I’d be more inclined to use this tool the earlier the patient is in her pregnancy. The study does note different average d-dimer levels during each trimester, so maybe we need to stratify our cutoffs by trimester?
Take away:
The Pregnancy-Adapted YEARS algorithm is the largest and best study we have to date on the pregnant population and can safely rule out PE and decrease CT studies, particularly in the 1st trimester.
Reference:
- Pol, Liselotte M. van der, Cecile Tromeur, Ingrid M. Bistervels, Fionnuala Ni Ainle, Thomas van Bemmel, Laurent Bertoletti, Francis Couturaud, et al. “Pregnancy-Adapted YEARS Algorithm for Diagnosis of Suspected Pulmonary Embolism.” New England Journal of Medicine 380, no. 12 (March 21, 2019): 1139–49. https://doi.org/10.1056/NEJMoa1813865.