Introduction Patients with COVID‐19 are known to have a coagulopathy with a thrombosis risk. It is unknown whether this is due to a generalized humoral prothrombotic state or endothelial factors such as inflammation and dysfunction. The aim was to further characterize thrombin generation using a novel analyser (ST Genesia, Diagnostica Stago, Asnières, France) and a panel of haematological analytes in patients with COVID‐19. Methods Platelet poor plasma of 34 patients with noncritical COVID‐19 was compared with 75 patients with critical COVID‐19 (as defined by WHO criteria) in a retrospective study by calibrated automated thrombography and ELISA. Patients were matched for baseline characteristics of age and gender. Results Critical patients had significantly increased fibrinogen, CRP, interleukin‐6 and D‐dimer compared to noncritical patients. Thrombin generation, in critical patients, was right shifted without significant differences in peak, velocity index or endogenous thrombin potential. Tissue plasminogen activator (tPA), tissue factor pathway inhibitor (TFPI) and vascular endothelial growth factor (VEGF) were significantly increased in the critical versus noncritical patients. Critically ill patients were on haemodiafiltration (31%; heparin used in the circuit) or often received escalated prophylactic low‐molecular weight heparin. Conclusion These results confirm increased fibrinogen and D‐dimer in critical COVID‐19‐infected patients. Importantly, disease severity did not increase thrombin generation (including thrombin‐antithrombin complexes and prothrombin fragment 1 + 2) when comparing both cohorts; counter‐intuitively critical patients were hypocoaguable. tPA, TFPI and VEGF were increased in critical patients, which are hypothesized to reflect endothelial dysfunction and/or contribution of heparin (which may cause endothelial TFPI/tPA release).

Summary This study describes the use of a simple charcoal product (DOAC‐RemoveTM) to allow haemostasis assays on patients taking direct oral anticoagulants (DOAC). In the proposed algorithm, patients taking DOAC are screened using the dilute thrombin time (dTT) and anti‐Xa assay. If either are positive then DOAC‐Remove is utilised. In a validation, DOAC‐Remove did not interfere with coagulation testing in normal plasma or in patients on DOAC with a known lupus anticoagulant (LA). Of 1566 routine patient samples tested, 125 (8%) had evidence of anti‐Xa activity (>0·1 iu/ml) or prolonged dTT suggestive of either a direct/indirect Xa inhibitor or direct thrombin inhibitor. All of these 125 patients had a prolonged dilute Russell viper venom time (dRVVT) screening test and 106 had a LA detected by dRVVT after phospholipid correction. After DOAC‐Remove, 91 patients (73%) had a negative dRVVT screen. After further investigation only 9 (7%) had a positive LA. DOAC‐Remove prevented 5% of patients having a LA inappropriately detected. DOAC did not significantly affect the LA activated partial thromboplastin time (aPTT) ratio, protein S antigen or protein C activity. DOAC cause low intrinsic factor assays, high prothrombin time/aPTT and high activated protein C sensitivity ratio, which DOAC‐Remove reversed (P < 0·05). Despite recommendations, haemostasis testing for patients on DOAC continues; this algorithm aids diagnostic accuracy. Further validation and research are warranted.

No abstract