ABSTRACT
Background : Diagnostic strategies for suspected pulmonary embolism (PE) have not been prospectively evaluated in COVID-19 patients. Aims : Evaluate diagnostic strategies for suspected PE in patients with COVID-19. Methods : Prospective, multicenter, cohort study in 708 patients with suspected PE and (suspected) COVID-19, without anticoagulant therapy, from 14 hospitals (March 1 -October 29, 2020). The study was approved by Institutional Review Boards of these hospitals and informed consent was obtained by opt-out approach. Different management protocols for suspected PE were used in the participating hospitals: YEARS algorithm, computed tomography pulmonary angiography (CTPA) only, or Wells rule with variable D-dimer threshold. We evaluated the safety (3-month failure rate) and efficiency (number of CTPAs avoided) of the three strategies. Results : Baseline characteristics are presented in Table 1;PE prevalence was 28%. YEARS was applied in 36%, Wells rule in 4.2% and CTPA only in 52%;7.3% was not tested because of hemodynamic instability. Within YEARS, PE was considered excluded without CTPA in 29%, of which one patient developed non-fatal PE during followup (failure rate 1.4%;95%CI 0.03-7.4;Table 2). 117 patients had a negative CTPA within YEARS (46%), of whom 10 were diagnosed with non-fatal venous thromboembolism (VTE) during follow-up (failure rate 8.5%;95%CI 4.2-15;Table 2). In patients imaged with CTPA only, 65% had an initial negative CTPA, of whom 10 patients were diagnosed with non-fatal VTE during follow-up (failure rate 4.1%;95%CI 2.0-7.5). CTPA was avoided in 2 patients (6.7%) managed with Wells rule. Conclusions : CTPA could be avoided in 29% of patients managed by YEARS, with a low failure rate (1.4%). The failure rate of a negative CTPA, used as a sole test or within YEARS, was higher than in a non-COVID-19 population and warrants ongoing suspicion in these patients. Our results underline the applicability of the YEARS algorithm in COVID-19 patients with suspected PE.