Summary
Coagulation dysfunction and thrombosis are major complications in patients with coronavirus disease 2019 (COVID‐19). Patients on oral anticoagulants (OAC) prior to diagnosis of COVID‐19 may therefore have better outcomes. In this multicentre observational study of 5 883 patients (≥18 years) admitted to 26 UK hospitals between 1 April 2020 and 31 July 2020, overall mortality was 29·2%. Incidences of thrombosis, major bleeding (MB) and multiorgan failure (MOF) were 5·4%, 1·7% and 3·3% respectively. The presence of thrombosis, MB, or MOF was associated with a 1·8, 4·5 or 5·9‐fold increased risk of dying, respectively. Of the 5 883 patients studied, 83·6% (n = 4 920) were not on OAC and 16·4% (n = 963) were taking OAC at the time of admission. There was no difference in mortality between patients on OAC vs no OAC prior to admission when compared in an adjusted multivariate analysis [hazard ratio (HR) 1·05, 95% confidence interval (CI) 0·93–1·19; P = 0·15] or in an adjusted propensity score analysis (HR 0·92 95% CI 0·58–1·450; P = 0·18). In multivariate and adjusted propensity score analyses, the only significant association of no anticoagulation prior to diagnosis of COVID‐19 was admission to the Intensive‐Care Unit (ICU) (HR 1·98, 95% CI 1·37–2·85). Thrombosis, MB, and MOF were associated with higher mortality. Our results indicate that patients may have benefit from prior OAC use, especially reduced admission to ICU, without any increase in bleeding.