A UK trial involving 16 500 mechanically ventilated intensive care unit (ICU) patients found no 90-day survival benefit for conservative supplemental oxygen over usual oxygen therapy. Nevertheless, the study, published in JAMA, did demonstrate the accuracy and cost-effectiveness of conducting a large trial with a simple intervention.
Oxygen is one of the most commonly administered treatments to patients in ICUs, but liberal oxygen therapy to avoid the risks of hypoxaemia may lead to harm, so finding the right level could optimise outcomes. Trials to date have shown mixed results.
For COVID patients admitted to the ICU with severe hypoxaemia, survival without life support was extended with conservative oxygen therapy. In a paediatric ICU study, conservative oxygen therapy resulted in a reduction in a composite of organ support at 30 days or death. A meta-analysis of 13 trials showed no differences between liberal and conservative oxygen therapy.
Even with just a small difference in survival benefit, with tens of millions of patients mechanically ventilated in the ICU would still mean significant numbers of lives saved. Other tests of new drugs and procedures in the ICU are hampered by high cost, as Seitz et al. noted in an accompanying editorial, so this sort of trial comparing two approaches to a common therapy is much more affordable.
The UK Intensive Care Unit Randomised Trial Comparing Two Approaches to Oxygen Therapy (UK-ROX) trial was initiated to determine if there was a difference between conservative and usual oxygen therapy.
The trial randomised 16 500 patients across 97 ICUs in the UK to either conservative oxygen therapy or usual oxygen therapy, in adults receiving mechanical ventilation and supplemental oxygen in the ICU. The primary outcome was mortality at 90 days. Conservative oxygen therapy targeted a peripheral oxygen saturation (Spo2) of 90% (range, 88%-92%), while usual oxygen therapy was at the discretion of the treating clinician.
Patients were early in mechanical ventilation (median, 5 hours), were severely ill (median predicted mortality risk, 35%), had a range of critical illnesses (eg, > 5000 patients with sepsis and > 1500 patients with hypoxic-ischaemic encephalopathy) and with significant hypoxaemia (eg, > 11 000 patients with a Pao2:Fio2 ratio, consistent with acute respiratory distress syndrome). Obtaining informed consent from the patients was, of course, largely not feasible, so this requirement was waived for the study.
Exposure to supplemental oxygen was 29% lower for those in the conservative oxygen therapy group compared with the usual oxygen therapy group. Of the patients randomised to conservative oxygen therapy, 35.4% died by 90 days compared with 34.9% of patients receiving usual oxygen therapy.
No differences were seen for secondary outcomes, including ICU stay, days free of life support and mortality at various time points. No interactions for confirmed or suspected COVID, ethnicity or other illnesses were observed.
Post hoc analysis showed weak evidence of increased harm from conservative oxygen therapy among the first 10 patients in each site but no difference for the random enhanced data collection sample compared with standard data collection.
Seitz et al. pointed out that the high level of adherence to the conservative target resulted in a mean oxygen saturation of 93.3%, versus 95.1% for usual care. The differences in oxygen saturation (1.9%) and Fio2 (0.04) between the trial groups in UK-ROX were about half the magnitude of some prior trials, due to not aiming for widely separated targets, and usual care varies considerably depending on location and clinical considerations.
Therefore, the researchers concluded that the findings do not support an approach of reducing oxygen exposure by targeting an Spo2 of 90% in mechanically ventilated adults receiving oxygen in the ICU. They suggest that future research may involve using AI to determine specific situations where conservative or liberal oxygen therapy may have beneficial outcomes.
References:
Martin DS, Gould DW, Shahid T, et al. Conservative Oxygen Therapy in Mechanically Ventilated Critically Ill Adult Patients: The UK-ROX Randomized Clinical Trial. JAMA. 2025;334(5):398–408. doi:10.1001/jama.2025.9663
Seitz KP, Casey JD, Semler MW. Patient, Treatment, Outcome—Large Simple Trials of Common Therapies. JAMA. 2025;334(5):395–397. doi:10.1001/jama.2025.9657
