S62 The prediction of pulmonary embolism & CTPA findings in the COVID-19 crisis

ObjectiveTo determine if the key features used to predict and diagnose pulmonary embolism (PE) in the hospital setting were affected by the COVID-19 crisis Our hospital protocol utilises a two-tier Wells probability score with values of 4 or less requiring a positive D-dimer test to determine if imaging is necessary The components of a Wells score and D-dimer levels may be affected by coincident coronavirus infection MethodsObservational data has been collected for patients presenting to acute services with a possible PE at our trust as part of ongoing pathway development A representative month (April 2019) was used to provide a cohort for comparison with the COVID-19 patients who were investigated for PE during April 2020 ResultsDuring April 2020, 126 patients had a CTPA with 30 diagnosed PEs (23 8%) compared with 2019 when only seven PEs were diagnosed from 59 scans (11 8%) The calculated Wells score for the 2020 cohort had a mean of 4 5 and median of 4 5, in 2019 the mean Wells was 4 0 with a median of 4 5 The most common components of the Wells score seen in 2020 were: PE most likely diagnosis (n=89), heart rate >100bpm (n=88), immobile for >3 days (n=61) The available D-dimer results indicate they were significantly higher in the COVID cohort (12097 n=72) than the 2019 group (3367 n=28, p<0 05 t-test) During COVID D-dimer levels were significantly higher in patients with a PE (25207)) than those without PE (7000, p<0 01) From the 126 CTPAs 50 cases had CT features consistent with COVID-19 disease, ten of whom also had a PE 75 cases had COVID proven on viral swab PCR or CT criteria COVID proven patients had higher platelet count, ferritin and CRP with lower lymphocyte count (all p<0 05 t-test) compared with the rest of the 2020 cohort The COVID proven patients with a PE also had a significantly higher D-dimer than without a PE (38156 vs 3855 p<0 01) and a trend towards a prolonged INR (1 52 vs 1 11 p=0 09) ConclusionThe COVID-19 crisis was associated with an increase in PE diagnoses and diagnostic rates on imaging with higher D-dimer levels Thresholds and predictionmodels may need to be re-evaluated