A total of 2817 patients met all inclusion and no exclusion criteria, of whom 2040 (72.4%) were randomly allocated without prior consent. One participant was randomly allocated in error, and for 53 participants consent was not subsequently available. In 55 (5.4%) of 1022 participants in the conservative oxygenation group and 59 (5.8%) of 1018 in the liberal oxygenation group consent was subsequently refused or withdrawn before day 30. Primary outcome data were available for all participants who remained in the analysis. *Numbers meeting individual exclusion criteria do not add to the total as some patients met >1 criterion One patient was identified as having undergone duplicate random allocation after trial recruitment had closed. The first random allocation was retained in the analyses. Approach for consent was deemed not appropriate only in exceptional circumstances. These were where the patient was under the care of social services and an appropriate legal guardian could not be identified; due to language barriers; or an impending inquest or legal action at the participating hospital. As a result, there was not a legal basis for inclusion in the trial. Includes three patients where consent was refused or withdrawn after day 30 (one in the conservative oxygenation group and two in the liberal oxygenation group). Primary endpoint was assumed based on location of care for 28 patients in the conservative oxygenation group and 32 patients in the liberal oxygenation group.
01968-2%2Ffig1_a.webp&w=1920&q=75 "Learn How to Make Workflow for Your Research Paper")