Although several reports recommend the use of systemic anticoagulation therapy in patients with severe coronavirus disease 2019 (COVID-19) pneumonia, appropriate target population and timing of administration are unknown. We assessed association between therapeutic anticoagulation administration with unfractionated heparin and outcomes in patients with severe COVID-19 pneumonia, assuming that anticoagulant administration effects are influenced by therapy timing.

This retrospective observational study included severe COVID-19 patients requiring mechanical ventilation in a tertiary emergency critical care hospital intensive care unit (ICU) in Japan from May 1, 2020 to September 30, 2021. All included patients were divided into early and late-phase administration groups based on therapeutic anticoagulant administration timing (≤5 and >5 days, respectively, after commencing oxygen therapy). Primary outcomes (in-hospital mortality and adverse events related to anticoagulation therapy) and secondary outcomes [veno-venous extracorporeal membrane oxygenation (ECMO), ventilator-free days (VFD), and ICU-free days] were compared between groups using univariate and multivariate models.

Of 198 included patients 104 (52.5%) and 94 (47.5%) were in early-phase and late-phase administration groups, respectively. Although background characteristics were similar between the groups, the early-phase administration group had a significantly lower in-hospital mortality rate (3.8% vs. 27.7%; p < 0.001), lower adverse event rates (1.9% vs. 12.8%; p < 0.001), significantly longer VFD and ICU-free days, and lower ECMO rates, than the late-phase administration group, in the multivariate model.

Late administration of therapeutic-dose anticoagulation in patients with severe COVID-19 pneumonia was significantly associated with worse outcomes than early administration.