Pulmonary Embolism Rule-Out Strategy Safely Reduces Chest Imaging

A pulmonary embolism (PE) diagnostic strategy using a higher D-dimer threshold under the so-called YEARS rule combined with age adjustment for selected patients who couldn't be ruled out by PE rule-out criteria (PERC) appeared safe in a randomized trial.

For emergency department patients with suspected PE, the strategy wasn't inferior for thromboembolic events compared with a conventional diagnostic strategy of subjective estimation of pretest probability, D-dimer testing if clinical probability isn't high, and then chest imaging if D-dimer exceeds the threshold.

The one case of venous thromboembolism diagnosed by 3 months compared with five in the usual care group (0.15% vs 0.80%) had a one-sided 97.5% CI no greater than 0.21% for the difference, which fell within the noninferiority margin.

And as hoped, the intervention cut down on chest imaging (30.4% vs 40.0%, adjusted difference 95% CI −13.8% to −3.5%), Yonathan Freund, MD, PhD, of Sorbonne Université in Paris, and colleagues reported in JAMA.

That finding "emphasizes the value of combining the two criteria sets," the researchers concluded.

Of the other six secondary endpoints, only median stay in the emergency department showed a significant difference between strategies, an adjusted 1.6-hour advantage (95% CI −2.3 to −0.9) with the intervention applying the YEARS criteria in PERC-positive patients.

"Taken together, there is a role for efficient diagnostic algorithms to decrease the use of CTPA [computed tomography pulmonary angiography] in patients with clinically suspected pulmonary embolism," according to an accompanying editorial.

Editorialists Marcel Levi, MD, PhD, of Amsterdam University Medical Center, and Nick van Es, MD, PhD, of University College London Hospitals NHS Foundation Trust, called the study elegant in refining the diagnostic algorithm.

"However, in the acute clinical setting of emergency medicine, which is usually busy or even hectic, such a rather complex approach toward triage for CTPA is not only time consuming, but may also prove challenging," Levi and van Es wrote. "A simple diagnostic approach based on the YEARS algorithm combined with age-adjusted D-dimer testing in all patients might have been just as efficient and safe overall, while less burdensome."

And, they added, "from a patient perspective, a negative diagnostic algorithm for pulmonary embolism does not diminish the physician's obligation to consider other diagnoses that explain the symptoms, for which chest CT scans may still be needed and helpful."

The trial included two adult emergency departments in Spain and 16 in France in a cluster-randomized design with crossover. Enrollment was suspended during the early part of the COVID-19 pandemic while chest imaging was routine for suspected COVID-19.

The semi-structured diagnostic strategies for PE started with the clinician's gestalt, with patients only enrolled upon clinical suspicion of a PE due to acute onset of chest pain, worsening acute dyspnea, syncope, or a combination thereof and either less than 15% subjective probability plus at least one PERC score element or a subjective probability in the 16%-50% range. PERC criteria were:

• Age ≥50 years

• Pulse rate ≥100/min

• Arterial oxygen saturation ≤94%

• Unilateral leg swelling

• Hemoptysis

• Recent trauma or surgery

• Prior PE or deep venous thrombosis

• Exogenous estrogen use

The intervention then involved assessing the YEARS algorithm, with its three clinical criteria (clinical signs of deep vein thrombosis, hemoptysis, and PE as the most likely diagnosis), together with D-dimer. Those with no YEARS criteria were ruled out for PE if D-dimer was below 1,000 ng/mL. Those with YEARS criteria were ruled out if D-dimer was below the threshold for their age (age×10 ng/mL for age 50+).

The usual care group simply got chest imaging if their D-dimer was above the age-adjusted threshold.

Notably, 80% of participants in the intervention group had zero YEARS criteria, such that the change in strategy (chest imaging) was triggered only if the D-dimer level was above 1,000 ng/mL. No missed PEs occurred in these patients, for a failure rate of 0.00% (95% CI 0.00%-0.71%). "Consequently, these data demonstrating the safety of the intervention strategy are particularly robust," the researchers wrote.

However, that left the study somewhat underpowered for the patients with a non-zero YEARS score and a D-dimer level above the age-adjusted threshold but below 1,000 ng/mL, Freund's group noted. "No missed PEs were found in this subgroup, but the upper bound of the 95% CI of the failure rate was 5.36%, which was above the predefined safety threshold."

Comment