Prevalence of venous thromboembolism in patients undergoing diagnostic venous ultrasound during the first SARS-CoV-2 pandemic.

ABSTRACT

Background: Venous thromboembolism appears to be associated with severe COVID-19 infection than in those without it. However, this varies considerably depending on the cohort studied. The aims of this single-centre, multi-site retrospective cross-sectional study were to assess the number of all venous scans performed in the first month of pandemic in a large university teaching hospital, to evaluate the incidence of deep venous thrombosis (DVT), and assess the predictive ability of the clinical information available on the electronic patient record in planning work-up for DVT and prioritising ultrasound scans. Patients and methods: All consecutive patients undergoing venous ultrasound for suspected acute DVT between 1st of March and 30th of April 2020 were considered. Primary outcome was the proportion of scans positive for DVT; the secondary outcomes included association of a positive SARS-CoV-2 PCR test, demographic, clinical factors, and Wells scores. Results: 819 ultrasound scans were performed on 762 patients across the Trust in March and April 2020. This number was comparable to the corresponding pre-pandemic cohort from 2019. The overall prevalence of DVT in the studied cohort was 16.1% and was higher than before the pandemic (11.5%, p=.047). Clinical symptoms consistent with COVID-19, irrespective of the SARS-CoV-2 PCR test result (positive_COVID_PCR OR 4.97, 95%CI 2.31-10.62, p<.001; negative_COVID_PCR OR 1.97, 95%CI 1.12-3.39, p=.016), a history of AF (OR 0.20, 95%CI 0.03-0.73, p=.037), and personal history of venous thromboembolism (VTE) (OR 1.95, 95%CI 1.13-3.31, p=.014), were independently associated with the diagnosis of DVT on ultrasound scan. Wells score was not associated with the incidence of DVT. Conclusions: Amongst those referred for the DVT scan, SARS-CoV-2 PCR test was associated with an increased risk of VTE and should be taken into consideration when planning DVT work-up and prioritising diagnostic imaging. We postulate that the threshold for imaging should possibly be lower.