Aims/Objectives/Background The diagnosis and management of COVID-19 has presented a novel challenge in the emergency department (ED). Early and sensitive predictors of outcome are needed to improve management of COVID-19 patients. Recent evidence has suggested a role for a COVID-19 associated pro-thrombotic coagulopathy as part of the underlying pathology. The aim was to evaluate the prognostic utility of D-dimer as a biomarker predictive of outcome in COVID-19 patients.

Methods/Design We retrospectively analysed data for 326 cases of confirmed COVID-19 presenting to our ED at the Royal Sussex County Hospital in Brighton between 13th March and 17th June 2020. During this period 2687 attendances were triaged to the ‘red’ COVID-19 zone with symptoms suspicious for COVID-19, amongst whom 326 admissions were confirmed to have COVID-19 by CT, chest x-radiograph or PCR swab. D-dimers were measured in ED for 265/326. Peak D-dimer measured during admission was collected to evaluate deteriorations subsequent to admission. Receiver-operating characteristic curves were used to determine an optimal cutoff for discrimination.

Results/Conclusions D-Dimer elevation >0.5µg/mL was seen in 93.5% of admitted patients with confirmed COVID-19. Multivariable logistic regression suggested that age >75 (OR=3.01 95% CI 1.65–5.49 p=0.0003) and D-dimer (measured in ED) >1.25µg/mL (OR=2.06 95% CI 1.08–3.93 p=0.0276) were associated with increased mortality. D-dimer measured in ED predicted mortality with sensitivity of 76.5% and specificity of 41.3%. The D-dimer rose by >1.00 for 30/265 patients subsequent to admission of whom 8/30 (26.7%) died (all mortality 16%) and 11/30 (36.7%) were escalated to intensive care. Peak D-dimer measured during admission >3.2µg/mL predicted hospital mortality with 50% sensitivity and 72.4% specificity (OR=2.16 95% CI 1.16–3.99 p=0.0145).

The results of this study support the growing argument that a raised D-dimer may play an important role as a prognostic marker in patients with COVID-19, perhaps indicative of a pro-thrombotic coagulopathy within the underlying pathology.