Systematic screening for pulmonary embolism in patients with suspected COVID-19 in the emergency department: Aprospective cohort study

Background : Pulmonary embolism (PE) seems to be a prevalent complication in patients with Coronavirus disease 2019 (COVID-19). However, all studies thus far were retrospective analyses without systematic PE-screening, thus highly susceptible to selection bias. Therefore, the true prevalence of PE and optimal D-dimer cut-off cannot be derived from such studies. Aims : We aimed to evaluate the true prevalence of PE in patients with suspected COVID-19 and the diagnostic yield of systematic PE screening. Methods : We evaluated all patients with suspected COVID-19 who were admitted to our hospital via the Emergency Department between April 7th and May 31st 2020. Patients were assessed according to a prespecified clinical protocol and were evaluated for PE using the YEARS-algorithm. CT-pulmonary angiography (CTPA) was performed in case D-dimer levels were ≥1000 ng/mL, or ≥500 ng/ mL if patients had ≥1 YEARS item (hemoptysis, clinical signs for deep vein thrombosis, and/or PE as the most likely diagnosis). Patients using anticoagulant drugs were excluded. Results : 666 patients with suspected COVID-19 were included in the present analyses (Figure 1). CTPA was performed in 393 patients, revealing PE in 51 patients. Thus, 1 out of 7.7 CTPAs was positive for PE. Interestingly, the PE-prevalence did not differ between patients with COVID-19 ( n = 190) and those without (13.5% vs. 12.8%, P = 0.87). D-dimer levels were higher in patients with PE (Figure 2). The lowest D-dimer value ruling out PE was 1258 μg/L. In multivariate regression analysis, only D-dimer was associated with PE (OR 1.07 for every 1000 μg/L increase, 95%CI 1.02-1.13, P = 0.007). Conclusions : In the first study on systematic PE screening in ED patients with suspected COVID-19, we found an overall PE prevalence of 7.7%. We believe this is a feasible approach for early PE-detection. Our analysis does not support the use of higher D-dimer thresholds to rule out PE in this population.