ImportanceGiven the high risk of thrombosis and anticoagulation-related bleeding in patients with hypoxemic COVID-19 pneumonia, identifying the lowest effective dose of anticoagulation therapy for these patients is imperative.ObjectivesTo determine whether therapeutic anticoagulation (TA) or high-dose prophylactic anticoagulation (HD-PA) decreases mortality and/or disease duration compared with standard-dose prophylactic anticoagulation (SD-PA), and whether TA outperforms HD-PA; and to compare the net clinical outcomes among the 3 strategies.Design, Settings, and ParticipantsThe ANTICOVID randomized clinical open-label trial included patients with hypoxemic COVID-19 pneumonia requiring supplemental oxygen and having no initial thrombosis on chest computer tomography with pulmonary angiogram at 23 health centers in France from April 14 to December 13, 2021. Of 339 patients randomized, 334 were included in the primary analysis—114 patients in the SD-PA group, 110 in the HD-PA, and 110 in the TA. At randomization, 90% of the patients were in the intensive care unit. Data analyses were performed from April 13, 2022, to January 3, 2023.InterventionsPatients were randomly assigned (1:1:1) to receive either SD-PA, HD-PA, or TA with low-molecular-weight or unfractionated heparin for 14 days.Main Outcomes and MeasuresA hierarchical criterion of all-cause mortality followed by time to clinical improvement at day 28. Main secondary outcome was net clinical outcome at day 28 (composite of thrombosis, major bleeding, and all-cause death).ResultsAmong the study population of 334 individuals (mean [SD] age, 58.3 [13.0] years; 226 [67.7%] men and 108 [32.3%] women), use of HD-PA and SD-PA had similar probabilities of favorable outcome (47.3% [95% CI, 39.9% to 54.8%] vs 52.7% [95% CI, 45.2% to 60.1%]; P = .48), as did TA compared with SD-PA (50.9% [95% CI, 43.4% to 58.3%] vs 49.1% [95% CI, 41.7% to 56.6%]; P = .82) and TA compared with HD-PA (53.5% [95% CI 45.8% to 60.9%] vs 46.5% [95% CI, 39.1% to 54.2%]; P = .37). Net clinical outcome was met in 29.8% of patients receiving SD-PA (20.2% thrombosis, 2.6% bleeding, 14.0% death), 16.4% receiving HD-PA (5.5% thrombosis, 3.6% bleeding, 11.8% death), and 20.0% receiving TA (5.5% thrombosis, 3.6% bleeding, 12.7% death). Moreover, HD-PA and TA use significantly reduced thrombosis compared with SD-PA (absolute difference, −14.7 [95% CI −6.2 to −23.2] and −14.7 [95% CI −6.2 to −23.2], respectively). Use of HD-PA significantly reduced net clinical outcome compared with SD-PA (absolute difference, −13.5; 95% CI −2.6 to −24.3).Conclusions and RelevanceThis randomized clinical trial found that compared with SD-PA, neither HD-PA nor TA use improved the primary hierarchical outcome of all-cause mortality or time to clinical improvement in patients with hypoxemic COVID-19 pneumonia; however, HD-PA resulted in significantly better net clinical outcome by decreasing the risk of de novo thrombosis.Trial RegistrationClinicalTrials.gov Identifier: NCT04808882